Chinese herbal medicine for diabetic peripheral neuropathy

  • Review
  • Intervention

Authors


Abstract

Background

Chinese herbal medicine is frequently used for treating diabetic peripheral neuropathy in China. Many controlled trials have been undertaken to investigate its efficacy.

This is an update of a Cochrane review that was first published in the year 2011.

Objectives

To assess the beneficial effects and harms of Chinese herbal medicine for people with diabetic peripheral neuropathy.

Search methods

On 14 May 2012, we searched the Cochrane Neuromuscular Disease Group Specialized Register CENTRAL (2012, Issue 4 in The Cochrane Library), MEDLINE (January 1966 to May 2012), EMBASE (January 1980 to May 2012), AMED (January 1985 to May 2012) and in October 2012, the Chinese Biomedical Database (CBM) (1979 to October 2012), Chinese National Knowledge Infrastructure Database (CNKI) (1979 to October 2012), and VIP Chinese Science and Technique Journals Database (1989 to October 2012). We searched for unpublished literature in the Chinese Conference Papers Database, and Chinese Dissertation Database (from inception to October 2012). There were no language or publication restrictions.

Selection criteria

We included randomised controlled trials of Chinese herbal medicine (with a minimum of four weeks treatment duration) for people with diabetic peripheral neuropathy compared with placebo, no intervention, or conventional interventions. Trials of herbal medicine plus a conventional drug versus the drug alone were also included.

Data collection and analysis

Two authors independently extracted data and evaluated trial quality. We contacted study authors for additional information.

Main results

Forty-nine randomised trials involving 3639 participants were included. All trials were conducted and published in China. Thirty-eight different herbal medicines were tested in these trials, including four single herbs (extracts from a single herb), eight traditional Chinese patent medicines, and 26 self concocted Chinese herbal compound prescriptions. The trials reported on global symptom improvement (including improvement in numbness or pain) and changes in nerve conduction velocity. The positive results described from the 49 studies of low quality are of questionable significance. There was inadequate reporting on adverse events in the included trials. Eighteen trials found no adverse events. Two trials reported adverse events: adverse events occurred in the control group in one trial, and in the other it was unclear in which group the adverse events occurred. 29 trials did not mention whether they monitored adverse events. Conclusions cannot be drawn from this review about the safety of herbal medicines, due to inadequate reporting. Most of the trials were of very low methodological quality and therefore the interpretation of any positive findings for the efficacy of the included Chinese herbal medicines for treating diabetic peripheral neuropathy should be made with caution.

Authors' conclusions

Based on this systematic review, there is no evidence to support the objective effectiveness and safety of Chinese herbal medicines for diabetic peripheral neuropathy. No well-designed, randomised, placebo controlled trial with objective outcome measures has been conducted.

Résumé scientifique

Les plantes médicinales chinoises contre la neuropathie diabétique périphérique

Contexte

Les plantes médicinales chinoises sont souvent utilisés en Chine dans le traitement de la neuropathie diabétique périphérique. De nombreux essais contrôlés ont été réalisés pour évaluer leur efficacité.

Ceci est une mise à jour d'une revue Cochrane publiée pour la première fois en 2011.

Objectifs

Évaluer les effets bénéfiques et délétères des plantes médicinales chinoises pour les patients atteints de neuropathie diabétique périphérique.

Stratégie de recherche documentaire

Le 14 mai 2012, nous avons effectué des recherches dans le registre spécialisé CENTRAL du groupe Cochrane sur les affections neuromusculaires (La Bibliothèque Cochrane, numéro 4, 2012), dans MEDLINE (de janvier 1966 à mai 2012), EMBASE (de janvier 1980 à mai 2012), AMED (de janvier 1985 à mai 2012) et, en octobre 2012, nous avons consulté la Chinese Biomedical Database (CBM) (de 1979 à octobre 2012), la Chinese National Knowledge Infrastructure Database (CNKI) (de 1979 à octobre 2012) et la VIP Chinese Science and Technique Journals Database (de 1989 à octobre 2012). Nous avons recherché des articles non publiés dans les bases de données Chinese Conference Papers Database et Chinese Dissertation Database (depuis leur création jusqu'en octobre 2012). Aucune restriction de langue ou de publication n'a été appliquée.

Critères de sélection

Nous avons inclus des essais contrôlés randomisés comparant l'utilisation de plantes médicinales chinoises (avec une durée minimale de traitement de quatre semaines) chez des patients atteints de neuropathie diabétique périphérique à l'utilisation d'un placebo, à l'absence d'intervention, ou aux interventions conventionnelles. Les essais comparant l'association des plantes médicinales et d'un médicament traditionnel au traitement par médicament seul ont également été inclus.

Recueil et analyse des données

Deux auteurs ont travaillé indépendamment pour extraire les données et évaluer la qualité des essais. Nous avons contacté les auteurs des études pour obtenir des informations supplémentaires.

Résultats principaux

Quarante-neuf essais randomisés impliquant 3 639 participants ont été inclus. Tous les essais ont été menés et publiés en Chine. Trente-huit plantes médicinales différentes ont été testées dans ces essais, y compris quatre plantes individuelles (extraits d'une seule plante), huit remèdes traditionnels chinois brevetés et 26 préparations personnalisées à base de plantes médicinales chinoises. Les essais signalaient une amélioration globale des symptômes (y compris une amélioration des symptômes d'engourdissement ou de douleur) et des changements dans la vitesse de conduction nerveuse. Le caractère significatif des résultats positifs décrits dans ces 49 études de faible qualité est discutable. Les événements indésirables étaient mal rapportés dans les essais inclus. Dans dix-huit essais, aucun événement indésirable n'a été observé. Deux essais rapportaient des événements indésirables : dans un essai, les événements indésirables étaient survenus dans le groupe témoin, l'autre essai ne spécifiait pas dans quel groupe les événements indésirables étaient survenus. Dans 29 essais, aucune mention n'était faite au sujet d'une surveillance des événements indésirables. Aucune conclusion ne peut être tirée de cette revue sur l'innocuité des plantes médicinales, en raison de l'inadéquation des rapports. La plupart des essais étaient de très faible qualité méthodologique, et les résultats positifs concernant l'efficacité des plantes médicinales chinoises dans le traitement de la neuropathie diabétique périphérique doivent être interprétés avec prudence.

Conclusions des auteurs

D'après cette revue systématique, il n'existe aucune preuve permettant d'affirmer l'efficacité et l'innocuité objectives des plantes médicinales chinoises dans le traitement de la neuropathie diabétique périphérique. Aucun essai randomisé bien conçu, contrôlé contre placebo et avec des critères de jugement objectifs n'a été réalisé.

초록

당뇨 말초신경병증에 대한 중약의 치료 효과

배경

중국에서는 당뇨 말초 신경병증의 치료에 중약을 자주 이용한다. 그 효능을 조사하기 위해 많은 대조 임상시험이 이루어져 왔다.

이 리뷰는 2011년에 첫 출판된 코크란 리뷰를 업데이트 한 것이다.

목적

당뇨 말초 신경병증 환자에 대한 중약 치료의 효과와 위해를 평가하기 위함.

검색 전략

2012년 5월 14일, 코크란 신경근육질환 그룹 임상연구 등록부, 센트럴 (CENTRAL, 2012년 4호의 코크란 라이브러리까지), 메드라인 (1966년 1월 ~ 2012년 5월), 엠베이스 (1980년 1월 ~ 2012년 5월), 아메드 (1985년 1월 ~ 2012년 5월) 을 검색했다. 중국 생의학 문헌 데이터베이스 및 중국 국가 지식 기반 데이터베이스 (1979년 ~ 2012년 10월) 과 VIP 중국 과학 기술 저널 데이터베이스 (1989년 ~ 2012년 10월) 역시 검색했다. 미출간 자료의 검색은 중국 컨퍼런스 자료 데이터베이스 및 중국 학위논문 데이터베이스 (자료 제공 시작일부터 2012년 10월까지) 를 이용했다. 분석 대상 연구의 선정은 언어 및 출판 형태에 구애받지 않았다.

선정 기준

당뇨 말초 신경병증 환자에게 (적어도 4주의 치료 기간을 설정하여) 중약 치료와 대조 치료 (위약, 무처치, 서양의학적 치료) 를 비교한 무작위 대조 연구를 분석에 포함시켰다. 서양의학적 약물과 중약 병행 요법을 서양의학적 약물 단독 사용과 비교한 임상시험 역시 분석에 포함시켰다.

자료 수집 및 분석

두 저자가 각각 독립적으로 자료를 추출하고 임상시험의 품질을 평가하였다. 일차 연구의 저자들에게도 연락하여 추가 정보를 요청하였다.

주요 결과

무작위 대조 연구 41 건 (연구참가자 3639명) 이 분석되었다. 모든 연구는 중국에서 수행되고 중국어로 출판되었다. 4 종의 단방 (단일 중약의 추출물) 및 특허 제조된 8 종의 전통 중약, 26 종의 자체 탕전 중약 처방 등 총 38 종의 다양한 중약이 연구에 사용되었다. 전반적 증상 호전 (저림 또는 통증의 개선) 과 신경 전도 속도 상 변화가 임상시험에서 보고되었다. 품질이 낮은 49개 연구에서 긍정적 결과를 보고했지만, 유의한 결과인지는 의문스럽다. 이상반응에 대한 보고의 질은 부적절했다. 18개 연구에서는 이상반응이 없다고 보고하였다. 다른 두 건의 연구 중, 한 건에서는 이상반응이 대조군에서 나타났다고 보고하였고, 다른 한 건에서는 이상반응이 나타났지만 치료군 또는 대조군 중 어디에서 생긴 것인지 불분명하게 보고했다. 29개 임상연구는 아예 이상반응의 발생 여부조차 보고하지 않았다. 보고 품질의 부적절성 때문에, 중약의 안전성에 대한 결론은 이 리뷰를 통해 내릴 수 없다. 대부분 임상시험의 방법론적 질은 매우 낮았으므로, 당뇨 말초 신경병증에 대해 이 리뷰에서 분석된 중약의 긍정적 효과는 해석 시 주의를 기울여야 한다.

연구진 결론

리뷰 결과에 비추어 볼 때, 당뇨 말초 신경병증에 대한 중약 치료의 객관적 효과와 안전성을 지지하는 근거는 없다. 객관적 평가 도구를 이용한 양질의 무작위 위약 대조 연구는 지금껏 이루어지지 않았다.

역주

이 리뷰는 김건형 (부산대학교 한의학전문대학원) 이 번역하였습니다. 번역 내용과 관련한 궁금점은 김건형 (pdchrist@gmail.com) 으로 연락주십시오.

アブストラクト

糖尿病性末梢神経障害に対する漢方薬

背景

中国では糖尿病性末梢神経障害に対し、漢方薬がよく用いられる。その有効性を調査するため、多くのコントロール試験が実施されてきた。

2011年に初版発表されたコクランレビューをアップデートした。

目的

糖尿病性末梢神経障害の患者を対象に漢方薬の有効性と有害性を評価すること。

検索戦略

2012年5月14日に、Cochrane Neuromuscular Disease Group Specialized Register CENTRAL (2012年 第 4 号コクランライブラリ)、MEDLINE ( 1966年1月~2012年5月)、EMBASE ( 1980年1月~ 2012年5月)、 AMED (1985年1月~ 2012年5月) 2012年10月に、 Chinese Biomedical Database (CBM) (1979年~2012年10月)、 Chinese National Knowledge Infrastructure Database (CNKI) (1979 年~2012年10月)、VIP Chinese Science and Technique Journals Database (1989年~October 2012年10月)を検索した。 Chinese Conference Papers Database, and Chinese Dissertation Database (初版~2012年10月) から未発表文献を検索した。 言語、公表文献の制限はなかった。

選択基準

糖尿病性末梢神経障害を有する患者に対する漢方薬を(治療期間最低4週間)、プラセボ、無介入、従来型介入と比較したランダム化比較試験を対象とした。従来の薬剤に漢方薬を併用した群と従来の薬剤単独群を比較した試験も含まれた。

データ収集と分析

2名の著者が独立してデータを抽出し、試験の質を評価した。その後追加された情報については、研究著者に問い合わせた。

主な結果

3639例の参加者を対象とした49件のランダム化試験が含まれた。試験はすべて中国で実施され発表された。4種の単味製剤(生薬1種からの抽出物)、8種の伝統的な中国特許医薬品、26種の漢方薬自家製剤を含む、38種の異なる漢方薬が試験された。試験では広汎な症状の改善(麻痺や疼痛を含む)と神経伝達速度の変化について報告された。質の低い試験49件により良好な結果が報告されたが、有意性は疑問である。不適切な有害事象報告が含まれた。 18件の試験では有害事象が認められなかった。 2件の試験では有害事象が報告された。1件は対照群で発現し、もう1件は発現した群が不明であった。29件の試験では有害事象の有無について記載がなかった。不適切な報告であるため、漢方薬の安全性についてこのレビューから結論は得られない。試験のほとんどは、方法論的に質が低かった。従って糖尿病性末梢神経障害に対する、漢方薬を含む治療の有効性に関する結果は、慎重に判断しなければならない。

著者の結論

このシステマティック・レビューによると、糖尿病性末梢神経障害に対する漢方薬の客観的な効果と安全性を支持するエビデンスは認められなかった。客観的アウトカム判定を伴う、適切にデザインされたランダム化プラセボ比較試験は、実施されていない。

訳注

《実施組織》厚生労働省「「統合医療」に係る情報発信等推進事業」(eJIM:http://www.ejim.ncgg.go.jp/)[2016.1.5]
《注意》この日本語訳は、臨床医、疫学研究者などによる翻訳のチェックを受けて公開していますが、訳語の間違いなどお気づきの点がございましたら、eJIM事務局までご連絡ください。なお、2013年6月からコクラン・ライブラリーのNew review, Updated reviewとも日単位で更新されています。eJIMでは最新版の日本語訳を掲載するよう努めておりますが、タイム・ラグが生じている場合もあります。ご利用に際しては、最新版(英語版)の内容をご確認ください。

Plain language summary

Chinese herbal medicine for treatment of diabetic peripheral neuropathy

Diabetic peripheral neuropathy (DPN) is one of the most common complications of diabetes. It is characterised by a progressive loss of nerve fibres that predisposes the person to painful or insensitive extremities, ulceration and amputation, and results in a large disease burden in terms of incapacity for work, quality of life and use of healthcare resources. This systematic review identified a total of 49 trials that included 3639 participants with DPN. Ten of the trials were new at this first update of the review. We evaluated the effects of various herbal formulations (including single herbs, Chinese proprietary medicines and mixtures of different herbs) for treating people with DPN. All the identified clinical trials were performed and published in China. The trials reported on global symptom improvement (including improvement in numbness or pain) and changes in nerve conduction velocity. The positive results described from the 49 studies of low quality are of questionable significance. There was inadequate reporting on adverse events in the included trials. Most of the trials did not mention whether they monitored for adverse effects. Only two trials reported adverse events but an adverse event occurred in the control group in one trial and it was unclear in which group they occurred in the other trial. Conclusions about the safety of herbal medicines cannot therefore be drawn from this review due to inadequate reporting. Most of the trials were of very low methodological quality and the interpretation of any positive findings for the efficacy of the included Chinese herbal medicines for treating DPN should be made with caution. Based on this systematic review, there is no evidence to support the objective effectiveness and safety of Chinese herbal medicines for DPN. No well-designed, randomised placebo controlled trial with objective outcome measures has been conducted.

Résumé simplifié

Les plantes médicinales chinoises dans le traitement de la neuropathie diabétique périphérique

La neuropathie diabétique périphérique diabétique (NDP) est l'une des complications les plus courantes du diabète. Elle se caractérise par une perte progressive de fibres nerveuses qui prédispose à des douleurs ou à une insensibilité aux extrémités, à une ulcération et à une amputation, et constitue un fardeau important en termes d'incapacité de travail, de qualité de vie et d'utilisation des ressources sanitaires. La présente revue systématique a relevé un total de 49 essais qui incluaient 3 639 participants atteints de NDP. Dix de ces essais étaient nouveaux pour cette première mise à jour de la revue. Nous avons évalué les effets de différentes préparations à base de plantes (y compris des plantes seules, des remèdes chinois brevetés et des mélanges de plantes) pour le traitement des patients atteints de NDP. Tous les essais cliniques analysés ont été effectués et publiés en Chine. Les essais signalaient une amélioration globale des symptômes (y compris l'amélioration des problèmes d'engourdissement ou de douleur) et des changements dans la vitesse de conduction nerveuse. Le caractère significatif des résultats positifs décrits dans ces 49 études de faible qualité est discutable. Les informations concernant les événements indésirables étaient mal rapportées dans les essais inclus. La plupart des essais n'indiquaient pas si les effets indésirables avaient fait l'objet d'un suivi ou non. Seuls deux essais rapportaient l'occurrence d'événements indésirables, mais l'un de ces événements était signalé dans le groupe témoin et, pour l'autre, dans un autre essai, le groupe dans lequel l'événement était survenu était mal indiqué. Aucune conclusions sur l'innocuité des plantes médicinales ne peut donc être tirée de cette revue en raison du caractère inadapté des rapports. La plupart des essais étaient de très faible qualité méthodologique et les résultats positifs concernant l'efficacité des plantes médicinales chinoises dans le traitement de la NDP doivent être interprétés avec prudence. D'après cette revue systématique, il n'existe aucune preuve permettant d'affirmer l'efficacité et l'innocuité objectives des plantes médicinales chinoises dans le traitement de la NDP. Aucun essai randomisé contrôlé par placebo, bien conçu et avec des critères de jugement objectifs n'a été réalisé.

Notes de traduction

Traduit par: French Cochrane Centre 14th January, 2014
Traduction financée par: Pour la France : Minist�re de la Sant�. Pour le Canada : Instituts de recherche en sant� du Canada, minist�re de la Sant� du Qu�bec, Fonds de recherche de Qu�bec-Sant� et Institut national d'excellence en sant� et en services sociaux.

쉬운 말 요약

당뇨 말초 신경병증에 대한 중약의 치료 효과

당뇨 말초 신경병증은 가장 흔한 당뇨 합병증 중 하나이다. 환자는 점점 신경 섬유 손상이 진행됨에 따라 말초 감각이 떨어지거나 통증을 느끼게 되며, 궤양이나 사지부 절단 등으로 끝내 업무 능력 상실, 삶의 질 저하, 보건의료 자원 이용 증가 등의 큰 질환 부담을 안게 된다. 이 체계적 문헌고찰은 총 3639 명의 당뇨 말초 신경병증 환자가 참여한 49개의 임상시험을 분석하였다. 리뷰 업데이트를 통해 10개의 새로운 연구를 포함시켰다. 우리는 당뇨 말초 신경병증 환자 치료를 위한 다양한 중약 처방 (단일 중약, 특허를 받은 제조 중약, 여러 가지가 혼합된 중약 등) 의 효과를 평가하였다. 분석된 모든 임삼시험들은 중국에서 수행되고 중국어로 출판되었다. 임상시험에서는 전반적 증상 호전 (저림 또는 통증 개선) 및 신경 전도 검사 상의 변화 등을 보고하였다. 품질이 낮은 49개 연구에서 얻은 긍정적 결과는 신뢰하기 어렵다. 분석한 임상시험의 이상반응 보고는 부적절했다. 대부분의 임상시험들은 이상반응 측정 여부조차 언급하지 않았다. 단 두 건의 임상시험에서 이상반응 발생을 보고했다. 대조군에서 발생한 연구가 1건 있었지만, 어느 군에서 발생했는지 불명확하게 보고한 연구도 1건 있었다. 따라서 이 리뷰에서는 부적절한 보고 품질 때문에 중약 안전성에 대한 결론을 내리기 어렵다. 대부분 임상시험의 방법론적 질은 매우 낮았으며, 당뇨 말초 신경병증 치료에 대한 중약 치료의 긍정적 결과는 주의깊게 해석해야 한다. 리뷰 결과에 비추어 볼 때, 당뇨 말초 신경병증에 대한 중약 치료의 객관적 효과와 안전성을 지지하는 근거는 없다. 객관적 평가도구를 이용한 양질의 무작위 위약 대조 연구는지금껏 이루어지지 않았다.

역주

이 리뷰는 김건형 (부산대학교 한의학전문대학원) 이 번역하였습니다. 번역 내용과 관련한 궁금점은 김건형 (pdchrist@gmail.com) 으로 연락주십시오.

平易な要約

糖尿病性末梢神経障害の治療に対する漢方薬

糖尿病性末梢神経障害(DPN)は糖尿病の合併症として最もよくみられる症状のひとつである。これは、進行性の神経線維損失を特徴とするため、疼痛、四肢の感覚鈍麻、潰瘍を生じやすく患部切断し、その結果、就業能力、生活の質、医療費において多大な負担が発生する。今回のシステマティック・レビューでは3639例のDPN患者を対象とする総計49件の試験を同定した。そのうち10件がこのレビューで新たにアップデートされた。DPN患者の治療に対し、種々の漢方製剤の効果を評価した(漢方単味製剤、中国特許医薬品、異なる漢方の混合製剤を含む)。同定された試験はすべて、中国で実施、発表された。試験は幅広い症状の改善と(麻痺や疼痛の改善を含む)神経伝達速度の変化について報告された。質の低い試験49件により良好な結果が報告されたが、有意性は疑問である。不適切な有害事象報告が含まれた。ほとんどの試験では、有害事象を観察したかどうかについて触れられていなかった。有害事象の報告された試験が2件のみあったが、1件は対照群に発現しており、もう1件は有害事象の発現した群が不明であった。したがって、報告が不適切であることから、漢方薬の安全性についてこのレビューからは結論に至らなかった。試験のほとんどは、方法論的に非常に質が低く、DPNの治療に対し漢方薬が含まれる場合の有効性に関する良好な結果は慎重に判断されるべきである。このシステマティック・レビューからは、DPNに対する漢方薬の客観的効果と安全性を支持するエビデンスは認められなかった。客観的アウトカム指標の設定された適切なデザインのランダム化プラセボ対照比較試験は実施されていない。

訳注

《実施組織》厚生労働省「「統合医療」に係る情報発信等推進事業」(eJIM:http://www.ejim.ncgg.go.jp/)[2016.1.5]
《注意》この日本語訳は、臨床医、疫学研究者などによる翻訳のチェックを受けて公開していますが、訳語の間違いなどお気づきの点がございましたら、eJIM事務局までご連絡ください。なお、2013年6月からコクラン・ライブラリーのNew review, Updated reviewとも日単位で更新されています。eJIMでは最新版の日本語訳を掲載するよう努めておりますが、タイム・ラグが生じている場合もあります。ご利用に際しては、最新版(英語版)の内容をご確認ください。

Background

Diabetic peripheral neuropathy (DPN) is one of the most common complications of both insulin dependent (type 1) and non-insulin dependent (type 2) diabetes mellitus. DPN is characterised by a progressive loss of nerve fibres that predisposes the person to painful or insensitive extremities, neuropathic ulceration and amputation. Clinical symptoms and signs that are determined from a full neurological assessment provide the principal method for detecting DPN. The clinical evaluation may be supplemented by objective tests of nerve function such as nerve conduction studies (NCS), quantitative sensory testing (QST), quantitative autonomic testing and nerve biopsy (Vinik 2000; Perkins 2003).

DPN results in a large disease burden in terms of incapacity for work, quality of life and use of healthcare resources. The prevalence of DPN varies considerably due to the variation in criteria for diagnosis and patient selection (O'Hare 1994; Tesfaye 1996). In the EURODIAB IDDM Complications Study (Tesfaye 1996), 28% of people with type 1 diabetes had diabetic neuropathy. Fedele 1997 reported a similar prevalence rate (32.3%) in people with type 1 or type 2 diabetes. The prevalence of neuropathy was 34% in the Pittsburgh Epidemiology of Diabetes Complications Study (Maser 1989) and, using similar criteria to Fedele 1997, clinically detectable peripheral neuropathy was present in 39% of participants in the Diabetes Control and Complications Trial (DCCT) (DCCT 1988). In the Seattle Prospective Diabetic Foot Study (Adler 1997), 50% of the study participants for whom neuropathy testing was available were found to have peripheral sensory neuropathy at baseline.

The pathophysiology of DPN is multifactorial with genetic, environmental, behavioural, metabolic, neurotrophic, and vascular factors (Vinik 1999; Oates 2001; Vincent 2002). Long-term elevation of plasma glucose levels has been established as contributing to its development. The complex and interrelated effects of hyperglycaemia include increased metabolic flux through the polyol pathway with consequent sorbitol and fructose accumulation and reduced sodium-potassium ATPase levels, altered fatty acid metabolism, alterations in the redox state, reduced myoinositol, accumulation of advanced glycated end products, accelerated neuronal apoptosis, immunological alterations, changes in blood flow, and increased oxidative stress (Perkins 2003). The exact steps linking these pathophysiological mechanisms to the clinical evolution of DPN are uncertain, but a recent study demonstrated that the morphological severity of DPN is strongly linked to poor glycaemic control in type 1 and type 2 diabetic patients (Perkins 2001), emphasising the importance of hyperglycaemia as a fundamental risk factor in the development of DPN.

Data regarding the risk factors for the development and progression of DPN are far fewer, possibly due to greater difficulties in diagnosing and classifying the disease. Available evidence implicates significant correlations between the presence of DPN and increasing age, duration of diabetes, glycated haemoglobin (HbA1c) concentration, body height, the presence of background or proliferative diabetic retinopathy, smoking, high-density lipoprotein cholesterol, and the presence of cardiovascular disease (Girach 2006). In addition, new associations have been identified, including elevated diastolic blood pressure, the presence of severe ketoacidosis, increased fasting triglycerides, and the presence of microalbuminuria (Tesfaye 1996).

The only established intervention to treat DPN is strict glycaemic control, as demonstrated in type 1 diabetes by the results of the DCCT Research Group (DCCT 1993), and in type 2 diabetes by the UK Prospective Diabetes Study (UKPDS) Group (UKPDS 1998), and Kumamoto studies (Shichiri 2000). These studies have shown that intensive glycaemic control that is maintained for many years confers a 60% risk reduction in DPN for people with type 1 diabetes and limits the progressive sensory loss in people with type 2 diabetes. One Cochrane review examined the evidence for enhanced glucose control in the prevention of distal symmetric polyneuropathy in people with type 1 and type 2 diabetes (Callaghan 2012). The results show that enhanced glucose control significantly prevents the development of clinical neuropathy and reduces nerve conduction and vibration threshold abnormalities in type 1 diabetes mellitus. In type 2 diabetes mellitus, enhanced glucose control reduces the incidence of clinical neuropathy, although this was not formally statistically significant (P = 0.06). However, enhanced glucose control does significantly reduce nerve conduction and vibration threshold abnormalities. Importantly, enhanced glucose control significantly increases the risk of severe hypoglycemic episodes, which needs to be taken into account when evaluating its risk benefit ratio.Other tested therapeutic strategies include antioxidant therapy, inhibition of protein kinase C, fatty acid supplementation, immunotherapy, and inhibition of advanced glycation end-products (AGEs), or glutamate carboxypeptidase II (Siemionow 2004). While many of these therapies are successful in animal models of diabetes, these treatments have not yet been proved effective for human patients (Siemionow 2004). In addition, there are symptomatic therapies for control of the painful symptoms of DPN, including the tricyclic antidepressants; selective serotonin-reuptake inhibitors; antiepileptics such as gabapentin, phenytoin, lamotrigine, and zonisamide; opioids and tramadol; and non-steroidal anti-inflammatory drugs (Duby 2004).

Traditional Chinese medicine (TCM) has been used widely in China for thousands of years. TCM has unique theories of aetiology, systems of diagnosis and treatment that are vital to its practice. Generally speaking, the aim of TCM is "to consider the person and the environment as a whole and to treat different patients with different methods that suits him the best" (Tu 2006). Chinese herbal medicine (CHM) is a component of TCM. CHM has been used for a long time in China, orally or topically, and alone or in combination with Western conventional medicine to treat diabetes (Liu 2004), including to treat DPN. Clinical studies have suggested that CHM has therapeutic potential for DPN (Halat 2003; Chen 2006). Several randomised clinical trials have suggested that CHM can reduce symptoms and blood glucose levels, improve motor nerve conduction velocity (MCV) and sensory nerve conduction velocity (SCV), and delay the progression of DPN (Fan 1999; Tong 2003). In vivo, compound Chinese herbal medicines are reported to improve sciatic nerve function and ameliorate morphological changes in rats with DPN, probably through depleting free radical production, improving endothelial cell function, and ameliorating impaired blood flow (Zhang 2006). In the first version of this review (Chen 2011), we included 39 RCTs of CHM for the treatment of DPN and concluded that there was no evidence to support the objective effectiveness and safety of CHM for DPN due to the low quality of RCTs included. After two years, new clinical trials have been conducted and published, and this update became necessary.

Objectives

To assess the beneficial effects and harms of Chinese herbal medicine for people with diabetic peripheral neuropathy.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) describing a correct randomisation procedure. We excluded quasi-randomised trials. We took trials with a significant skew distribution of participants in groups that could not be explained by the randomisation principle to be non-RCTs and excluded them (Pocock 1990) unless they prespecified the use of unequal randomisation.

Types of participants

We included participants of any age or sex with diabetic peripheral neuropathy (DPN). The definition of DPN that was used in the studies had to conform to the following diagnostic criteria:

  1. the patient had diabetes mellitus by clearly defined or internationally recognised criteria, such as the World Health Organization (WHO) criteria (WHO 1999);

  2. the patient had a predominantly distal symmetrical sensorimotor polyneuropathy of the limbs, including subjective complaints of pain, tingling, numbness, weakness, and reduced functioning of the peripheral nerves demonstrated by a nerve conduction test; and

  3. other causes of sensorimotor polyneuropathy were excluded.

Types of interventions

The Chinese herbal medicine (CHM) interventions included single herbs (including extracts from a single herb), a Chinese proprietary medicine, or a compound of several herbs irrespective of preparation (for example decoction, oral liquid, tablet, capsule, pill, powder, or injection). The mode of delivery (for example oral, intramuscular or intravenous), dosage, and regimen of herbs were not restricted. We included trials if the treatment was given for a minimum of four weeks.

The control interventions were: no intervention, placebo, or another conventional medicine used with the intention of glycaemic control or symptomatic care. We also included trials of medicinal herb plus conventional medicine versus conventional medicine alone. We permitted co-intervention(s) as long as they were the same in all arms of the randomised trial.

The following comparisons were tabulated, where data were available:
(1) herbal medicines versus no treatment;
(2) herbal medicines versus placebo;
(3) herbal medicines versus conventional medicine; and
(4) herbal medicines plus conventional medicine versus conventional medicine alone.

Types of outcome measures

Primary outcomes

The primary outcome was an improvement of 30% or more in a validated clinical global symptom score, such as a visual analogue scale (VAS) score (Revill 1976), or total symptom score (Ametov 2003), at least four weeks after randomisation. Where this outcome was not available, the primary outcome was global symptom improvement by whatever criteria the trial authors used. The outcome was measured after at least four weeks of treatment but preferably after 12 weeks of treatment.

Secondary outcomes
  1. Change in or absolute values of quality of life measured with a validated scale such as the Short-Form 36 Health Survey (SF-36) (Ware 1992), after 12 weeks of treatment or more.

  2. Change in or absolute values of motor or sensory nerve conduction velocity measured by validated methods after 12 weeks of treatment or more.

  3. Adverse events monitored and measured for the duration of treatment with CHM and divided into all adverse events, adverse events which led to cessation of treatment, and serious adverse events which were life-threatening, fatal, or required or prolonged hospitalisation.

Search methods for identification of studies

Electronic searches

On 14 May 2012, we searched the Cochrane Neuromuscular Disease Group Specialized Register, CENTRAL (2012, Issue 4, in The Cochrane Library), MEDLINE (January 1966 to May 2012), EMBASE (January 1980 to May 2012), and AMED (January 1985 to May 2012). In October 2012, we searched the Chinese BioMedical Database (CBM) (1979 to October 2012), Chinese National Knowledge Infrastructure Database (CNKI) (1979 to October 2012), and VIP Chinese Science and Technique Journals Database (1989 to October 2012). We searched for unpublished literature in the Chinese Conference Papers Database and the Chinese Dissertation Database (from inception to October 2012).

The detailed search strategies are in the appendices: CENTRAL (Appendix 1), MEDLINE (Appendix 2), EMBASE (Appendix 3), AMED (Appendix 4), CBM (Appendix 5), CNKI (Appendix 6), VIP (Appendix 7), Chinese Conference Papers Database (Appendix 8), and Chinese Dissertation Database (Appendix 9).

Searching other resources

We checked the reference lists of identified RCTs and review articles in order to find randomised trials not identified by the electronic searches or handsearches.

Data collection and analysis

Selection of studies

Two review authors (XXL, GYY) independently selected the trials according to the prespecified selection criteria. The review authors resolved any disagreement by discussion or by consultation with a third party (JPL).

Data extraction and management

Two review authors (XXL, GYY) independently extracted data concerning details of study population, intervention, and outcomes using a self developed data extraction form. The data extraction form included the following items.

  1. General information: published or unpublished, title, authors, source, contact address, country where study performed, language of publication, year of publication, sponsoring, and study setting.

  2. Trial characteristics: design, randomisation (and method), allocation concealment (and method), blinding (participants, people administering treatment, and outcome assessors), and any check for the effectiveness of blinding.

  3. Participants: diagnostic criteria, inclusion or exclusion criteria, total number and number in comparison groups, sex, age, baseline characteristics, similarity of groups at baseline (including any co-morbidity), assessment of compliance, withdrawals or losses to follow-up (reasons or description), and subgroups.

  4. Intervention(s): single herb or compound herbs, dose, timing, mode of administration; comparison intervention(s) (dose, route, timing), and co-medication(s) (dose, route, and timing).

  5. Outcomes: outcomes specified above, duration of treatment, any other clinical relevant outcomes assessed, adverse events, and length of follow-up for outcomes.

We resolved differences in data extraction by consensus. When necessary, we sought information from the investigators of the primary studies. One author entered data into the Cochrane software Review Manager 5 (RevMan 5) and another checked the data to reduce the possibility of data entry errors.

Assessment of risk of bias in included studies

We undertook the 'Risk of bias' assessments of individual trials according to guidance in the Cochrane Handbook for Systematic Reviews of Interventions (Chapter 8.5) (Higgins 2011). This involved the following domains: sequence generation, allocation concealment, blinding of participants, personnel and outcome assessors, incomplete outcome data, selective outcome reporting, and other sources of bias. We completed a 'Risk of bias' table according to Cochrane guidelines. We made judgements on each of these criteria relating to the risk of bias: low, high or unclear (indicating unclear or unknown risk of bias).

Measures of treatment effect

We expressed dichotomous data as risk ratios (RRs) with 95% confidence intervals (CIs), and continuous data as mean differences (MDs) with 95% CIs.

Assessment of heterogeneity

We tested heterogeneity using the I2 statistic with significance set at ≥ 50%, and the Chi2 statistic with significance set at P < 0.10. If we had identified significant heterogeneity, the random-effects model would have been used. We had planned to assess possible sources of heterogeneity by subgroup and sensitivity analyses where data were available.

Assessment of reporting biases

We investigated potential biases using the funnel plot or other corrective analytical methods according to Egger 1997, where data were available.

Data synthesis

We performed quantitative meta-analysis within comparisons where more than two included studies compared the same herb versus the same control intervention. Analyses were performed by intention to treat, where possible. For dichotomous outcomes, we planned to include participants with incomplete or missing data in a sensitivity analysis by counting them as treatment failures to explore the possible effect of loss to follow-up on the findings ('worst-case' scenario). The analyses would have been carried out in RevMan.

We planned to tabulate the following comparisons, where data were available:
(1) herbal medicines versus no treatment;
(2) herbal medicines versus placebo;
(3) herbal medicines versus conventional medicine;
(4) herbal medicines plus conventional medicine versus conventional medicine alone.

Subgroup analysis and investigation of heterogeneity

If we had identified a sufficient number of RCTs, we would have performed subgroup analyses according to: clinical course (type 1 or type 2 diabetes mellitus), formulation of herbs (extract, single herb, or mixture of herbs), and treatment duration (short and long term).

Furthermore, if we had identified a sufficient number of RCTs, we had planned to perform sensitivity analyses to explore the influence of risk of bias on effect estimates. The risk of bias components included adequacy of generation of allocation sequence, concealment of allocation, double blinding, and the use of intention to treat analysis (high or low risk).

Results

Description of studies

Results of the search

Our primary searches identified 1647 references from the above databases. After reading titles and abstracts, we excluded 1594 of these articles because of obvious ineligibility. We retrieved a total of 53 references published in Chinese or English for further assessment. Of these, we excluded 14 references because they did not meet our inclusion criteria. The reasons for exclusion are listed under Characteristics of excluded studies.

In the updating stage, we identified a total of 1017 new references, including 293 in English (207 were translated abstracts, original language in Chinese), and 724 in Chinese. We removed 247 by deduplication and excluded 755 because they did not meet the inclusion criteria for the review. We retrieved a total of 15 references for further assessment. Of these, we excluded five references for the reasons listed under Characteristics of excluded studies.

Included studies

See: Characteristics of included studies

This review includes a total of 49 randomised controlled trials (RCTs). The trials reported random allocation of participants with diabetic peripheral neuropathy (DPN) to Chinese herbal medicines or control interventions (placebo in one trial, no treatment in one trial, and conventional medicine in 23 trials); or to Chinese herbal medicines plus conventional medicine versus conventional medicine alone (26 trials). Two trials (Jiao 2007; Zhao 2008) were three-arm trials: Jiao 2007 tested Chinese herbal medicine (CHM) versus CHM plus conventional medicine versus conventional medicine, and Zhao 2008 tested CHM versus conventional medicine versus placebo. The remaining trials reported two parallel arms in their studies. All the trials claimed a positive effect favouring CHM. The 49 randomised trials have been listed under Characteristics of included studies. They were all conducted and published in China.

Participants

A total of 3639 participants with DPN were randomised in the 49 trials. The average size of the trial was 74 participants, ranging from 36 to 134 per trial. All trials were hospital based. Seventeen trials enrolled participants with DPN and type 2 diabetes, two trials enrolled participants with DPN and either type 1 or 2 diabetes, and 30 trials did not report the type of diabetes. Twenty-seven trials included both inpatients and outpatients, seven trials included inpatients, four trials included outpatients, and the other 11 trials did not report whether they included inpatients or outpatients. All patients were Chinese adults. One trial did not provide data on age (Niu 2008).

Diagnosis

The diagnostic criteria were based on the American Diabetes Association (ADA) criteria in nine trials (Liu 2001; Li 2001; Liu 2004; He 2005; Zhou 2006; Zhang 2007; Lin 2008a; Sun 2009; Liu 2011), International Diabetes Federation (IDF) criteria in two trials (Hu 2005; Li 2010), Chinese medical textbooks in three trials (Cao 2011; Wu 2010; Zhou 2010a), and World Health Organization (WHO) criteria in the other 35 trials.

Interventions

There were large variations in the formulation, dosage, administration, and duration of treatment of both the Chinese herbal medicines tested, and control interventions (Table 1). In total, the included trials tested 38 different herbal medicines, including four single herbs (including extracts from a single herb), eight Chinese traditional patent medicines, and 26 self composed Chinese herbal compound prescriptions. Six trials investigated modified Huang Qi Gui Zhi Wu Wu Tang (Li 2005; Chen 2006; Liu 2006; Zhou 2006; Shu 2007; Liu 2011); however, the specific herbal composition of the formulae were a little different between the six trials (Table 1). Four trials investigated Tongxinluo capsule (herbal compound) (Li 2001; Liu 2004; Jiao 2007; Li 2008b). Two trials tested modified Buyang Huanwu Tang (Liu 2008; Sun 2009), with few differences in the specific compositions of herbs (Table 1). Three different trials tested formulations of Xuesaitong (Xuesaitong capsule in He 2005, Xuesaitong injection in Xin 2003, and Xuesaitong tablet in Hao 2010). The formulations of herbal medicines were different and included capsules, oral liquid, decoction, and injection. The duration of treatment varied from one month to three months. No trial reported the quality standard of the herbal preparations. The conventional medicines included mecobalamin, vitamin B, and cobamamide. Hypoglycemic therapy was a co-intervention in all the 49 trials, including an oral hypoglycaemic drug, insulin, and exercise.

Table 1. Herbal medicines and adverse effects in the included trials
Study IDName of HerbsFormulationCompositionsAdverse events
Cao 2010Xiaoke Tongluo TangDecoctionRadix Angelicae Pubescentis, Radix Gentianae Macrophyllae, Herba Epimedii, Radix Paeoniae Alba, Cortex Eucommiae, Radix Rehmanniae Recens, Radix Achyranthis Bidentatae, Rhizoma Ligustici Chuanxiong, Herba Taxilli, Caulis Spatholobi, Herba Lycopodii, Rhizoma Corydalis, and ScorpioNot reported
Cao 2011Zhuyu Huatan TangDecoctionRadix Paeoniae Alba, Radix Angelicae Sinensis, Radix Salviae Miltiorrhizae, Hirudo, Lumbricus, Caulis Spatholobi, Ramulus Cinnamomi, Rhizoma Pinelliae Preparata, Pericarpium Citri Reticulatae, Poria, Arisaema cum Bile, Semen Sinapis, Radix et Rhizoma GlycyrrhizaeNot reported
Chen 2006Modified Huangqi Guizhi Wuwu TangDecoctionRadix Astragali seu Hedysari, Radix Salviae Miltiorrhizae, Ramulus Cinnamomi, Radix Paeoniae Rubra, Radix Paeoniae Alba, Radix Rehmanniae Recens, Radix Angelicae Sinensis, Radix Glycyrrhizae, Fructus Jujubae, Rhizoma Zingiberis RecensNot reported
Chen 2009Tongbi TangDecoctionRadix Astragali seu Hedysari, Radix Angelicae Sinensis, Radix Rehmanniae Preparata, Fructus Lycii, Rhizoma Dioscoreae, Rhizoma Ligustici Chuanxiong, Radix Salviae Miltiorrhizae, Semen Persicae, Scorpio, Lumbricus, Ramulus Cinnamomi, Herba Asari, Radix Achyranthis BidentataeNot reported
Cheng 2005Modified Sishen decoctionDecoctionRadix Astragali seu Hedysari, Herba Dendrobii, Radix Polygalae, Flos Lonicerae, Caulis Spatholobi, Rhizoma Ligustici Chuanxiong, and Radix Paeoniae AlbaNot reported
Ding 2010Tang Mai YinDecoctionRadix Astragali seu Hedysari, Caulis Spatholobi, Rhizoma Atractylodis, Radix Salviae Miltiorrhizae, Radix Achyranthis Bidentatae, Radix Angelicae Sinensis, Fructus Liquidambaris, Semen Vaccariae, Flos Carthami, Rhizoma Anemarrhenae, Scorpio. Herba Asari and Scolopendra were added for participants with serious pain; Radix Aconiti Lateralis Preparata and Rhizoma Zingiberis were added for participants feeling cold in the extremities; Radix Astragali seu Hedysari was added for participants with weaknessNot reported
Gao 2008Compound Qiying granuleGranuleRadix Astragali seu Hedysari, Rhizoma Polygonati, Radix Salviae Miltiorrhizae, Periostracum Cicadae, Cicer arietinum Linn Fufang, etcNot reported
Hao 2010Xuesaitong tabletTabletTotal saponins of panax notoginsengNot reported
He 2005Xuesaitong capsuleCapsuleTotal saponins of panax notoginsengNot reported
Hu 2005Jianpi Yiqi Tongluo TangDecoctionRadix Pseudostellariae, Rhizoma Atractylodis Macrocephalae, Poria, Radix Astragali seu Hedysari, Radix Achyranthis Bidentatae, Rhizoma Dioscoreae, Herba Taxilli, Cortex Eucommiae, Rhizoma Ligustici Chuanxiong, and Radix GlycyrrhizaeNot reported
Jiao 2007Tongxinluo capsuleCapsuleRadix Ginseng, Hirudo, Scorpio, Eupolyphaga Seu Steleophaga, Scolopendra, Periostracum Cicadae, Radix Paeoniae Rubra, Borneolum Syntheticum, etcNot reported
Li 2001Tongxinluo capsuleCapsuleRadix Ginseng, Hirudo, Scorpio, Eupolyphaga Seu Steleophaga, Scolopendra, Periostracum Cicadae, Radix Paeoniae Rubra, Borneolum Syntheticum, etcAbdominal discomfort and pain in 4 cases (not reported in which group)
Li 2005Modified Huangqi Guizhi Wuwu TangDecoctionRadix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Olibanum, Myrrha, Flos Carthami, Caulis Spatholobi, Rhizoma Zingiberis Recens, and Fructus JujubaeNo adverse event was identified
Li 2008bTongxinluo capsuleCapsuleRadix Ginseng, Hirudo, Scorpio, Eupolyphaga Seu Steleophaga, Scolopendra, Periostracum Cicadae, Radix Paeoniae Rubra, Radix Paeoniae Rubra, Borneolum Syntheticum, Lignum Santali Albi, etcNot reported
Li 2009Yiqi Bushen Huoxue TangDecoctionRadix Astragali seu Hedysari, Radix Codonopsis, Radix Rehmanniae Recens, Radix Rehmanniae Preparata, Semen Cuscutae, Fructus Corni, Fructus Lycii, Poria, Radix Angelicae Sinensis, Radix Paeoniae Rubra, Radix Paeoniae Alba, Rhizoma Ligustici Chuanxiong, Radix Clematidis, Caulis Spatholobi, and Caulis TrachelospermiNot reported
Li 2010Dispelling dampness and dredging collaterals TCM decoctionDecoctionRhizoma Atractylodis, Radix Astragali seu Hedysari, Poria, Rhizoma Arisaematis, Radix Saposhnikoviae, Ramulus Cinnamomi, Caulis Spatholobi, Lumbricus, Radix Clematidis, Herba Trachelospermi jasminoides, and Pericarpium Citri ReticulataeNot reported
Lin 2008Yiqi Huoxue Tongluo compound prescriptionDecoctionRadix Rehmanniae Recens, Radix Scrophulariae, Radix Astragali seu Hedysari, Radix Notoginseng, Radix Angelicae Sinensis, Radix Paeoniae Rubra, Radix et Rhizoma Rhei, Exocarpium Citri Rubrum, Rhizoma Arisaematis Cum Bile, Lumbricus, Scorpio, and Radix PuerariaeNot reported
Lin 2008aXuefu Zhuyu capsuleCapsuleSemen Persicae, Flos Carthami, Radix Paeoniae Rubra, Rhizoma Ligustici Chuanxiong, Fructus Aurantii, Radix Bupleuri, Radix Platycodonis, Radix Angelicae Sinensis, Rehmannia glutinosa Libosch, Radix Achyranthis Bidentatae, and Radix GlycyrrhizaeNo adverse event was identified
Liu 2001Aloe AloeNo adverse event was identified
Liu 2004Tongxinluo capsulecapsuleRadix Ginseng, Hirudo, Scorpio, Eupolyphaga Seu Steleophaga, Scolopendra, Periostracum Cicadae, Radix Paeoniae Rubra, Radix Paeoniae Rubra, etcNo adverse event was identified
Liu 2006Modified Huangqi Guizhi Wuwu TangDecoctionRadix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Radix Puerariae, Herba Dendrobii, Radix Salviae Miltiorrhizae, Caulis Spatholobi, Scorpio, Rhizoma Zingiberis Recens, Fructus JujubaeNo adverse event was identified
Liu 2008Modified Buyang Huanwu TangDecoctionRadix Astragali seu Hedysari, Semen Persicae, Flos Carthami, Radix Angelicae Sinensis, Radix Paeoniae Rubra, Radix Rehmanniae Preparata, Rhizoma Ligustici Chuanxiong, Lumbricus, Scorpio, Radix Puerariae, Radix GlycyrrhizaeNo adverse event was identified
Liu 2011Modified Huangqi Guizhi Wuwu TangDecoctionRadix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Rhizoma Zingiberis Recens, Fructus Jujubae, Radix et Rhizoma Glycyrrhizae, Radix Angelicae Sinensis, Radix Rehmanniae Recens, Rhizoma Ligustici Chuanxiong, Radix Paeoniae Rubra, Rhizoma Anemarrhenae, Flos Carthami, Semen PersicaeNot reported
Luo 2008Unnamed Chinese herbal prescriptionDecoctionRadix Aconiti Preparata, Radix Aconiti Kusnezoffii Preparata, Herba Asari, Radix Angelicae Dahuricae, Lumbricus, Caulis Clematis Armanoii, Radix Astragali seu Hedysari, Semen Persicae, Flos Carthami, Radix Angelicae Sinensis, Rhizoma Ligustici Chuanxiong, Caulis SpatholobiNo adverse event was identified
Luo 2009Yiyiren TangDecoctionSemen Coicis, Rhizoma Atractylodis, Rhizoma et Radix Notopterygii, Radix Angelicae Pubescentis, Radix Saposhnikoviae, Herba Ephedrae, Ramulus Cinnamomi, Radix Aconiti Preparata, Radix Angelicae Sinensis, Rhizoma Ligustici Chuanxiong, Radix Glycyrrhizae, Rhizoma Zingiberis RecensNot reported
Ma 2008Fufang Danshen injectionInjectionRadix Salviae Miltiorrhizae, Lignum Dalbergiae OdoriferaeNo adverse event was identified
Niu 2008Unnamed Chinese herbal prescriptionDecoctionRadix Astragali seu Hedysari, Radix Rehmanniae Recens, Rhizoma Atractylodis, Radix Scrophulariae, Radix Puerariae, Radix Salviae Miltiorrhizae, Radix Achyranthis Bidentatae, Ramulus Mori, Ramulus Cinnamomi, Scorpio, Olibanum, MyrrhaNot reported
Peng 2009Yiqi Tongluo TangDecoctionRadix Astragali seu Hedysari, Lumbricus, Rhizoma Dioscoreae, Radix Ophiopogonis, Rhizoma Sparganii, Rhizoma Curcumae, Olibanum, MyrrhaNot reported
Shen 2009Tangmaining capsuleCapsuleRhizoma Acori Tatarinowii, Radix Puerariae, Radix Achyranthis Bidentatae, Radix Salviae Miltiorrhizae, Radix Dipsaci, Herba Leonuri, etcNo adverse event was identified
Shu 2007Modified Huangqi Guizhi Wuwu TangDecoctionRadix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Radix Paeoniae Rubra, Radix Angelicae Sinensis, Radix Salviae Miltiorrhizae, Caulis Spatholobi, Scorpio, Fructus JujubaeNot reported
Sun 2008Buyang Huanwu TangDecoctionRadix Angelicae Sinensis, Radix Paeoniae Rubra, Radix Astragali seu Hedysari, Rhizoma Ligustici Chuanxiong, Lumbricus, Semen Persicae, Flos CarthamiNot reported
Sun 2009Modified Buyang Huanwu TangDecoctionRadix Astragali seu Hedysari, Radix Angelicae Sinensis, Rhizoma Ligustici Chuanxiong, Flos Carthami, Semen Persicae, Radix Paeoniae Rubra, Lumbricus, Radix Rehmanniae Preparata, Radix Cyathulae, Radix DipsaciNot reported
Sun 2010Furongtong capsuleCapsuleHirudo, Lumbricus, Scorpio, Radix Puerariae, Radix Scrophulariae, yam, Radix Astragali seu Hedysari, Radix Achyranthis Bidentatae, Radix GlycyrrhizaeNot reported
Sun 2010bModified Duhuo Jisheng decoction plus Wujia Canger Jiangtang decoctionDecoctionHerba Taxilli, Radix Angelicae Pubescentis, Radix Rehmanniae Recens, Radix Angelicae Sinensis, Radix Paeoniae Rubra, Ramulus Mori, Caulis Spatholobi, Rhizoma Sparganii Stoloniferi, Rhizoma Sparganii, Radix Gentianae Macrophyllae, Cortex Acanthopanacis Gracilistyli Radicis, Poria, Rhizoma Atractylodis Macrocephalae, Alismatis Rhizoma, Rhizoma Polygonati, Radix Glycyrrhizae, Radix Notoginseng powder. For heat in lung and stomach, adding gypsum and Rhizoma Anemarrhenae; for qi deficiency, adding Radix Astragali seu Hedysari and Radix Dioscoreae Oppositae; for yin deficiency, adding Radix Polygoni Multiflori, Rhizoma Polygonati Odorati, Herba Dendrobii and Fructus Armeniaca mume; for yang deficiency in spleen and kidney, adding Cortex Eucommiae, Radix Himalayan Teasel and Ramulus Cinnamomi Cassiae; for dampness, adding Herba Artemisiae Yinchenhao, Semen Plantaginis, and Radix Sophorae Flavescentis; for blood stasis, adding Radix Salviae Miltiorrhizae and HirudoNot reported
Wang 2008Guizhi Gegen TangDecoctionRamulus Cinnamomi, Radix Paeoniae Alba, Radix Puerariae, Radix Angelicae Sinensis, Scorpio, Rhizoma Zingiberis Recens, Fructus Jujubae, Radix GlycyrrhizaeNo adverse event was identified
Wu 2003ErigeronInjectionErigeron extractNo adverse event was identified
Wu 2006Ciwujia injectionInjectionRadix Acanthopanacis Senticosi extractNot reported
Wu 2010Yangyin Huoxue TangDecoctionRadix Rehmanniae Recens, Radix Rehmanniae Preparata, Radix Asparagi, Succus Ophiopogonis, Rhizoma Anemarrhenae, Rhizoma Polygonati, Herba Taxilli, Radix Dipsaci, Radix et Rhizoma Salviae Miltiorrhizae, Flos Carthami, Semen Persicae, Radix Angelicae Sinensis, Lumbricus, HirudoNo adverse event was identified
Xie 2008Taoren Honghua JianDecoctionRadix Salviae Miltiorrhizae, Rhizoma Ligustici Chuanxiong, Semen Persicae, Flos Carthami, prepared Rhizoma Cyperi, Rhizoma Corydalis, Radix Paeoniae Rubra, Pericarpium Citri Reticulatae Viride, Radix Angelicae Sinensis, Radix Rehmanniae RecensNot reported
Xin 2003Xuesaitong injectionInjectionTotal saponins of panax notoginsengNo adverse event was identified
Yan 2006Modified Danggui Sini TangDecoctionRadix Angelicae Sinensis, Ramulus Cinnamomi, Medulla Tetrapanacis, Radix Salviae Miltiorrhizae, Radix Paeoniae Alba, Herba Asari, Ramulus Euonymi, Caulis Spatholobi, Scorpio, Radix Glycyrrhizae, Fructus JujubaeNo adverse event was identified
Yan 2008Ginkgo biloba extractInjectionGinkgo biloba extractNot reported
Yang 2011Qingxiao Tangbi prescriptionDecoctionRadix Astragali seu Hedysari, Radix Codonopsis, Radix Ophiopogonis, Rhizoma Anemarrhenae, Radix Paeoniae Alba, Semen Persicae, Flos Carthami, Radix Salviae Miltiorrhizae, Radix Mongolian SnakegourdNo adverse event was identified
Yao 2010Tonifying kidney and activating blood formulaDecoctionRadix Rehmanniae Recens, Rhizoma Anemarrhenae, Cortex Eucommiae, Herba Taxilli, Radix Cyathulae, Radix Himalayan Teasel, Fructus Psoraleae, Radix Salviae Miltiorrhizae, Flos Carthami, Semen Persicae, Radix Angelicae Sinensis, Radix Ligustici Chuanxiong, Lumbricus, Fructus LiquidambarisNo adverse event was identified
Yin 2011Tongluo Tangtai DecoctionDecoctionRadix Astragali seu Hedysari, Rhizoma Dioscoreae, Radix Scrophulariae, Radix Ophiopogonis, Radix et Rhizoma Salviae Miltiorrhizae, Rhizoma Chuanxiong, Caulis Spatholobi, Radix Puerariae, Gypsum Fibrosum, Rhizoma Anemarrhenae, Rhizoma Atractylodis, Herba Agastachis, Semen Sinapis, Hirudo, Rhizoma CorydalisNo adverse event was identified
Zhang 2007Yiqi Huoxue Wenyang prescriptionDecoctionRadix Ginseng, Radix Astragali seu Hedysari, Semen Persicae, Flos Carthami, Radix Paeoniae Rubra, Radix Angelicae Sinensis, Olibanum, Myrrha, Lumbricus, Ramulus Cinnamomi, prepared Radix Aconiti Lateralis Preparata, Herba Ephedrae, Herba Asari,and Radix Achyranthis BidentataeNot reported
Zhao 2008Tangluoan capsuleCapsuleRadix Astragali seu Hedysari, Radix Rehmanniae Recens, Herba Eupatorii, Rhizoma Coptidis, Scorpio, Rhizoma Ligustici Chuanxiong, Fructus Lycii, etcNo adverse event was identified
Zhou 2006Modified Huangqi Guizhi Wuwu TangDecoctionRadix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Rhizoma Zingiberis Recens, Fructus Jujubae, Semen Sinapis Albae, Hirudo, Bombyx Batryticatus, Radix Puerariae, Radix et Caulis Acanthopanacis SenticosiOne mild skin eruption was identified in the control group
Zhou 2010aYiqi Huayu Tongbi TangDecoctionRadix Astragali seu Hedysari, Caulis Spatholobi, Radix Puerariae, Radix Angelicae Sinensis, Rhizoma Ligustici Chuanxiong, Hirudo, Eupolyphaga Seu Steleophaga, Radix Pseudostellariae,  Rhizoma Polygonati, Radix Clematidis, Herba Lycopodii, Ramulus Mori, Ramulus Cinnamomi, Radix Cyathulae, Fructus ChaenomelisNot reported
Outcomes

No trial reported the incidence of complications, quality of life outcomes, or health economics. The reported outcomes included global symptom improvement (including improvement in numbness or pain), change in motor or sensory nerve conduction velocity, and adverse effects. Investigators collected data on changes in motor sensory nerve conduction velocity in 11 trials (Liu 2001; Sun 2008; Zhao 2008; Ding 2010; Sun 2010; Hao 2010; Wu 2010; Yang 2011; Yao 2010; Yin 2011; Zhou 2010a). Two trials reported that there were adverse events, but the information was incomplete. Eighteen trials found no adverse events, and the other 29 trials did not mention whether they monitored adverse events. All the reported outcomes were measured at the end of treatment. No trial reported follow-up after the end of the intervention period.

Excluded studies

See: Characteristics of excluded studies.

Risk of bias in included studies

We assessed none of the included trials as having a low risk of bias. The trials provided very limited information about design and methodology. Although all trials claimed to have generated the allocation sequence adequately, insufficient information was provided to judge whether or not randomisation was conducted properly. No trial stated how allocation concealment or blinding was performed. One trial reported the number of dropouts (Yin 2011); however, the investigators did not use intention-to-treat analysis, just cancelled the data of participants who dropped out. Selective reporting was difficult to assess in the selected RCTs as we were not able to obtain access to the trial protocols. No trial had a pretrial estimation of sample size. We have received no response to our requests for relevant information from the investigators. Figure 1 shows the review authors' judgements about each 'Risk of bias' item for each included study. The review authors considered all studies to be of low or very low quality and to be at a high risk of bias.

Figure 1.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study. Red: high risk of bias; green: low risk of bias; yellow: unclear risk of bias.

Effects of interventions

Single herbs (including extracts from a single herb)

There were four different single herbs (including extracts from a single herb) tested for their effectiveness on DPN.

Aloe

One trial tested aloe in 40 participants with DPN, compared with no treatment (Liu 2001). Aloe did not show a favourable effect on global symptom improvement compared with no treatment (RR 1.67, 95% CI 0.96 to 2.88). However, aloe had a significantly better effect on peroneal motor nerve conduction velocity (MD 8.60, 95% CI 7.18 to 10.02), peroneal sensory nerve conduction velocity (MD 8.60, 95% CI 7.58 to 9.62), median motor nerve conduction velocity (MD 7.40, 95% CI 5.09 to 9.71), and median sensory nerve conduction velocity (MD 10.00, 95% CI 8.08 to 11.92) (all in m/s).

Ciwujia injection

One trial tested Ciwujia injection in 88 participants with DPN (Wu 2006). The combination of Ciwujia injection with vitamin B1 and vitamin B12 showed a better effect on global symptom improvement compared with vitamin B1 and vitamin B12 (RR 1.78, 95% CI 1.34 to 2.36) (Figure 2).

Figure 2.

Forest plot of comparison: 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, outcome: 4.1 Global symptom improvement.

Erigeron injection

One trial tested Erigeron infusion in 79 participants with DPN (Wu 2003). Erigeron infusion had a better effect on global symptom improvement compared with injected vitamin B1 (RR 1.61, 95% CI 1.15 to 2.25) (Figure 3). There was no control for the administration method.

Figure 3.

Forest plot of comparison: 3 Chinese herbal medicine vs conventional medicine, outcome: 3.1 Global symptom improvement.

Ginkgo biloba extract injection

One trial tested Ginkgo biloba extract in 70 participants with DPN (Yan 2008). The combination of Ginkgo biloba extract with vitamin B1 and vitamin B12 did not differ significantly from vitamin B1 and vitamin B12 regarding global symptom improvement (RR 1.25, 95% CI 0.96 to 1.62) (Figure 2).

Chinese traditional patent medicine

Eight Chinese traditional patent medicines were tested in 12 trials (four trials tested Tongxinluo capsule in DPN participants and two trials tested two formulations of Xuesaitong).

Compound Qiying granule

One trial tested compound Qiying granule in 62 participants with DPN (Gao 2008). There was no significant difference between compound Qiying granule and cobamamide on global symptom improvement (RR 1.33, 95% CI 0.98 to 1.82) (Figure 3).

Dispelling dampness and dredging collaterals TCM decoction

One trial tested Dispelling dampness and dredging collaterals TCM decoction in 82 participants with DPN (Liu 2011). Dispelling dampness and dredging collaterals TCM decoction had a better effect on global symptom improvement compared with mecobalamin alone (RR 2.27, 95% CI 1.40 to 3.68) (Figure 3).

Fufang Danshen injection

One trial tested Fufang Danshen injection in 100 participants with DPN (Ma 2008). The combination of Fufang Danshen injection with mecobalamin had a better effect on global symptom improvement compared with mecobalamin alone (RR 1.53, 95% CI 1.21 to 1.95) (Figure 2).

Furong Tongmai capsule

One trial tested Furong Tongmai capsule in 86 participants with DPN (Sun 2010). Furong Tongmai capsule had a better effect on global symptom improvement compared with mecobalamin alone (RR 1.46, 95% CI 1.12 to 1.90) (Figure 3), motor nerve conduction velocity in the common peroneal nerve (MD 4.60 m/s, 95% CI 1.94 to 7.26), and sensory nerve conduction velocity in the common peroneal nerve (MD 5.30 m/s, 95% CI 1.66 to 8.94).

Tangluoan capsule

One three-arm trial tested Tangluoan capsule in 87 participants with DPN (Zhao 2008). Tangluoan capsule had a better effect compared with placebo (57 participants) on global symptom improvement (RR 2.31, 95% CI 1.24 to 4.28) (Figure 4), motor nerve conduction velocity in the common peroneal nerve (MD 3.84 m/s, 95% CI 1.38 to 6.30), sensory nerve conduction velocity in the common peroneal nerve (MD 5.07 m/s, 95% CI 2.57 to 7.57), and a significantly better effect on sensory nerve conduction velocity in the median nerve (MD 5.67 m/s, 95% CI 3.02 to 8.32) while it did not have a significantly better effect on motor nerve conduction velocity in the median nerve (MD 1.64 m/s, 95% CI -1.70 to 4.98).

Figure 4.

Forest plot of comparison: 2 Chinese herbal medicine vs placebo, outcome: 2.1 Global symptom improvement.

When compared with mecobalamin (61 participants), Tangluoan capsule did not have a significantly better effect on global symptom improvement (RR 1.18, 95% CI 0.82 to 1.71) (Figure 3), motor nerve conduction velocity in the common peroneal nerve (MD 1.92 m/s, 95% CI -0.64 to 4.48), motor nerve conduction velocity in the median nerve (MD -0.76 m/s, 95% CI -4.05 to 2.53), and sensory nerve conduction velocity in the median nerve (MD 1.76 m/s, 95% CI -1.02 to 4.54) while it had a better effect on sensory nerve conduction velocity in the common peroneal nerve (MD 2.57 m/s, 95% CI 0.08 to 5.06).

Tangmaining capsule

One trial tested Tangmaining capsule in 100 participants with DPN (Shen 2009). There was no significant difference between Tangmaining capsule and cobamamide in global symptom improvement (RR 1.02, 95% CI 0.93 to 1.12) (Figure 3).

Tongxinluo capsule

Four trials tested Tongxinluo capsule in 261 participants with DPN (Li 2001; Liu 2004; Jiao 2007; Li 2008b). The trial conducted by Jiao 2007 was a three-arm study in which Tongxinluo capsule plus mecobalamin, Tongxinluo capsule alone, and mecobalamin were tested for their effects on global symptom improvement. Meta-analysis showed that there was no significant difference in global symptom improvement between the combination of Tongxinluo capsule with mecobalamin and mecobalamin alone (overall effect RR 1.24, 95% CI 0.95 to 1.62) (Figure 2) in Jiao 2007 and Li 2008b (105 participants), and between Tongxinluo capsule and mecobalamin (overall effect RR 1.25, 95% CI 0.98 to 1.58) (Figure 3) in Li 2001 and Jiao 2007 (121 participants). Tongxinluo capsule plus vitamin B1 and vitamin B12 showed a significantly better effect on global symptom improvement compared with vitamin B1 and vitamin B12 (RR 1.87, 95% CI 1.22 to 2.86) (Figure 2) in Liu 2004 in 65 participants.

Tonifying kidney and activating blood formula

One trial tested tonifying kidney and activating blood formula in 100 participants with DPN (Yao 2010). There was no significant difference between tonifying kidney and activating blood formula and mecobalamin on global symptom improvement (RR 1.35, 95% CI 0.98 to 1.86) (Figure 3); however, tonifying kidney and activating blood formula showed better effect on motor nerve conduction velocity in the common peroneal nerve (MD 1.00, 95% CI 0.64 to 1.36), sensory nerve conduction velocity in the common peroneal nerve (MD 8.83, 95% CI 8.08 to 9.58), motor nerve conduction velocity in the median nerve (MD 3.53, 95% CI 2.38 to 4.68), and sensory nerve conduction velocity in the median nerve (MD 2.44, 95% CI 1.43 to 3.45) (all in m/s).

Xuefu Zhuyu capsule

One trial tested Xuefu Zhuyu capsule in 60 participants with DPN (Lin 2008a). The combination of Xuefu Zhuyu capsule with mecobalamin did not differ significantly from mecobalamin alone regarding global symptom improvement (RR 1.23, 95% CI 0.96 to 1.57) (Figure 2).

Xuesaitong capsule/injection/tablet

Three trials tested three different formulations of Xuesaitong: a Xuesaitong capsule (He 2005), a Xuesaitong injection (Xin 2003) and a Xuesaitong tablet (Hao 2010), in 238 participants with DPN. The main components of Xuesaitong were Panax notoginsenosides and flavonoids. The combination of Xuesaitong capsule/tablet and mecobalamin showed better global symptom improvement than mecobalamin (RR 1.27, 95% CI 1.02 to 1.58) (Figure 2) in He 2005 and Hao 2010 (total of 177 participants). Xuesaitong tablet plus mecobalamin showed better effect on motor nerve conduction velocity (m/s) in the median nerve (MD 6.47, 95% CI 5.14 to 7.80), sensory nerve conduction velocity in the median nerve (MD 1.04, 95% CI 0.03 to 2.05), and sensory nerve conduction velocity in the common peroneal nerve (MD 8.50, 95% CI 7.93 to 9.07) than mecobalamin, but the effect of Xuesaitong tablet plus mecobalamin on common peroneal motor nerve conduction velocity did not differ significantly from that of mecobalamin alone (MD -0.55, 95% CI -1.97 to 0.87) (Hao 2010, 112 participants). Xuesaitong injection plus vitamin B1 and vitamin B12 showed a significantly better effect on global symptom improvement compared with vitamin B1 and vitamin B12 (RR 1.86, 95% CI 1.19 to 2.90) (Figure 2) in Xin 2003 (61 participants).

Self composed Chinese herbal compound prescription

Seventeen different self composed Chinese herbal compound prescriptions were tested in DPN patients, in 23 trials (modified Buyang Huanwu Tang in two trials and modified Huangqi Guizhi Wuwu Tang in five trials).

Buyang Huanwu Tang

One trial tested Buyang Huanwu Tang in 84 participants with DPN (Sun 2008). The combination of Buyang Huanwu Tang with mecobalamin had a better effect compared with mecobalamin alone on global symptom improvement (RR 1.50, 95% CI 1.08 to 2.08) (Figure 2). It also improved motor nerve conduction velocity in common peroneal nerve (MD 5.95 m/s, 95% CI 3.75 to 8.15) and sensory nerve conduction velocity in common peroneal nerve (MD 5.68 m/s, 95% CI 3.40 to 7.96).

Guizhi Gegen Tang

One trial tested Guizhi Gegen Tang in 60 participants with DPN (Wang 2008). The combination of Guizhi Gegen Tang with vitamin B1 and vitamin B6 did not differ significantly from vitamin B1 and vitamin B6 regarding global symptom improvement (RR 1.29, 95% CI 0.99 to 1.67) (Figure 2).

Jianpi Yiqi Tongluo Tang

One trial tested Jianpi Yiqi Tongluo Tang in 48 participants with DPN (Hu 2005). Jianpi Yiqi Tongluo Tang showed a better effect on global symptom improvement compared with cobamamide (RR 2.22, 95% CI 1.29 to 3.84) (Figure 3).

Modified Buyang Huanwu Tang

Two trials tested modified Buyang Huanwu Tang in 224 participants with DPN (Liu 2008; Sun 2009); however, the specific compositions of herbs in the two trials were a little different. In Liu 2008 (110 participants), the combination of modified Buyang Huanwu Tang with vitamin B1 and vitamin B6 had a better effect on global symptom improvement compared with vitamin B1 and vitamin B6 (RR 1.28, 95% CI 1.06 to 1.55) (Figure 2). In the trial by Sun 2009 (134 participants), modified Buyang Huanwu Tang had a better effect on global symptom improvement compared with vitamin B1 and vitamin B12 (RR 1.25, 95% CI 1.02 to 1.53) (Figure 3).

Modified Danggui Sini Tang

One trial tested modified Danggui Sini Tang in 70 participants with DPN (Yan 2006). The combination of modified Danggui Sini Tang with vitamin B1 and vitamin B12 had a better effect on global symptom improvement compared with vitamin B1 and vitamin B12 (RR 1.33, 95% CI 1.04 to 1.71) (Figure 2).

Modified Duhuo Jisheng decoction plus Wujia Canger Jiangtang decoction

One trial tested modified Duhuo Jisheng decoction plus Wujia Canger Jiangtang decoction in 40 participants with DPN (Sun 2010b). The modified Duhuo Jisheng decoction plus Wujia Canger Jiangtang decoction had a better effect on global symptom improvement compared with mecobalamin and vitamin B1 (RR 1.58, 95% CI 1.09 to 2.30) (Figure 3).

Modified Huangqi Guizhi Wuwu Tang

Six trials tested Modified Huangqi Guizhi Wuwu Tang in 434 participants with DPN (Li 2005; Chen 2006; Liu 2006; Zhou 2006; Shu 2007; Liu 2011); however, the specific compositions of these herbal formulae were a little different in each of the six trials. Meta-analysis of three trials (186 participants) showed that the combination of modified Huangqi Guizhi Wuwu Tang and mecobalamin had a better effect on global symptom improvement than mecobalamin alone (overall effect RR 1.29, 95% CI 1.11 to 1.50) (Figure 2) (Chen 2006; Zhou 2006; Liu 2011); however, the combination of modified Huangqi Guizhi Wuwu Tang with vitamin B1 and vitamin B6 did not have a better effect on global symptom improvement than vitamin B1 and vitamin B6 alone (RR 1.13, 95% CI 0.93 to 1.36) (Figure 2) in Liu 2006 (110 participants). Modified Huangqi Guizhi Wuwu Tang showed a better effect on global symptom improvement compared with mecobalamin (RR 1.29, 95% CI 1.02 to 1.64) in the trial of Shu 2007 (68 participants) and compared with vitamin B1 plus vitamin B6 (RR 1.33, 95% CI 1.01 to 1.76) (Figure 3) in the trial by Li 2005 (70 participants).

Modified Sishen Jian

One trial tested modified Sishen Jian in 70 participants with DPN (Cheng 2005). Modified Sishen Jian had a better effect on global symptom improvement compared with vitamin B1 and vitamin B12 (RR 1.43, 95% CI 1.11 to 1.85) (Figure 3).

Qingxiao Tangbi prescription

One trial tested Qingxiao Tangbi prescription in 60 participants with DPN (Yang 2011). The combination of Qingxiao Tangbi prescription and mecobalamin showed better effect than mecobalamin alone on global symptom improvement (RR 1.63, 95% CI 1.13 to 2.34) (Figure 2), motor nerve conduction velocity in the left common peroneal nerve (MD 1.29, 95% CI 0.46 to 2.12), motor nerve conduction velocity in the right common peroneal nerve (MD 2.17, 95% CI 1.50 to 2.84), sensory nerve conduction velocity in the left common peroneal nerve (MD 1.49, 95% CI 1.12 to 1.86), and sensory nerve conduction velocity in the right common peroneal nerve (MD 2.37, 95% CI 1.92 to 2.82) (all m/s).

Tang Mai Yin

One trial tested Tang Mai Yin in 62 participants with DPN (Ding 2010). Tang Mai Yin had a better effect on global symptom improvement compared with mecobalamin (RR 1.64, 95% CI 1.15 to 2.35) (Figure 3). For nerve conduction velocity, Tang Mai Yin did not have a significantly better effect on motor nerve conduction velocity in the common peroneal nerve (MD -0.60, 95% CI -2.65 to 1.45), sensory nerve conduction velocity in the common peroneal nerve (MD 0.70, 95% CI -1.38 to 2.78), or motor nerve conduction velocity in the tibial nerve (MD -0.30, 95% CI -2.99 to 2.39) while it had a significantly better effect on sensory nerve conduction velocity in the tibial nerve (MD 3.10, 95% CI 0.80 to 5.40) (all m/s).

Taoren Honghua Jian

One trial tested Taoren Honghua Jian in 80 participants with DPN (Xie 2008). The combination of Taoren Honghua Jian with mecobalamin had a better effect on global symptom improvement compared with mecobalamin alone (RR 1.36, 95% CI 1.04 to 1.79) (Figure 2).

Tongbi Tang

One trial tested Tongbi Tang in 86 participants with DPN (Chen 2009). Tongbi Tanghad a better effect on global symptom improvement compared with mecobalamin and vitamin B1 (RR 1.31, 95% CI 1.04 to 1.66) (Figure 3).

Tongluo Tangtai Decoction

One trial tested Tongluo Tangtai Decoction in 40 participants with DPN (Yin 2011). The Tongluo Tangtai Decoction showed better effect on improvement in pain (RR 1.60, 95% CI 1.03 to 2.50), and sensory nerve conduction velocity in the common peroneal nerve (MD 8.22 m/s, 95% CI 4.78 to 11.66) than mecobalamin. There were no significant difference between Tongluo Tangtai Decoction and mecobalamin on global symptom improvement (RR 1.27, 95% CI 0.96 to 1.66) (Figure 3), improvement in numbness (RR 1.36, 95% CI 1.00 to 1.84), sensory nerve conduction velocity in median nerve (MD 5.20 m/s, 95% CI -0.01 to 10.41) and sensory nerve conduction velocity in ulnar nerve (MD 4.13 m/s, 95% CI -0.78 to 9.04).

Unnamed Chinese herbal prescription (Niu 2008)

One trial tested a self composed Chinese herbal prescription with a unique recipe in 84 participants with DPN (Niu 2008). The combination of self composed Chinese herbal prescription with mecobalamin did not differ significantly from mecobalamin alone regarding global symptom improvement (RR 1.16, 95% CI 0.94 to 1.42) (Figure 2).

Unnamed Chinese herbal prescription (Luo 2008)

One trial tested a self composed Chinese herbal prescription in 90 participants with DPN (Luo 2008). CHM showed a better effect on improvement in subjective numbness compared with mecobalamin (RR 1.26, 95% CI 1.06 to 1.50) but no significant improvement in pain (RR 1.45, 95% CI 0.96 to 2.18).

Xiaoke Tongluo Tang

One trial tested Xiaoke Tongluo Tang in 60 participants with DPN (Cao 2010). Xiaoke Tongluo Tang did not show a better effect on global symptom improvement compared with vitamin B1 and vitamin B6 (RR 1.39, 95% CI 1.00 to 1.94) (Figure 3).

Yangyin Huoxue Tang

One trial tested Yangyin Huoxue Tang in 72 participants with DPN (Wu 2010). The combination of Yangyin Huoxue Tang and mecobalamin showed better effect than mecobalamin alone on global symptom improvement (RR 1.34, 95% CI 1.02 to 1.77) (Figure 2), sensory nerve conduction velocity in the common peroneal nerve (MD 9.08 m/s, 95% CI 8.26 to 9.90), motor nerve conduction velocity in the median nerve (MD 3.19 m/s, 95% CI 1.93 to 4.45), and sensory nerve conduction velocity in the median nerve (MD 3.14 m/s, 95% CI 1.92 to 4.36), but not on motor nerve conduction velocity in common peroneal nerve (MD 0.05 m/s, 95% CI -0.49 to 0.59).

Yiqi Bushen Huoxue Tang

One trial tested Yiqi Bushen Huoxue Tang in 36 participants with DPN (Li 2009). The combination of Yiqi Bushen Huoxue Tang with indometacin and vitamin B did not differ significantly from indometacin and vitamin B in global symptom improvement (RR 1.42, 95% CI 1.00 to 2.00) (Figure 2).

Yiqi Huoxue Tongluo compound prescription

One trial tested a self prescribed Yiqi Huoxue Tongluo compound prescription in 79 participants with DPN (Lin 2008). The combination of self prescribed Yiqi Huoxue Tongluo compound prescription with mecobalamin had a better effect on global symptom improvement compared with mecobalamin alone (RR 1.32, 95% CI 1.04 to 1.67) (Figure 2).

Yiqi Huoxue Wenyang prescription

One trial tested Yiqi Huoxue Wenyang prescription in 60 participants with DPN (Zhang 2007). The Yiqi Huoxue Wenyang prescription showed a better effect on global symptom improvement compared with mecobalamin alone (RR 1.53, 95% CI 1.09 to 2.16) (Figure 3).

Yiqi Huayu Tongbi Tang

One trial tested Yiqi Huayu Tongbi Tang in 98 participants with DPN (Zhou 2010a). The comparison treatment was mecobalamin, the Yiqi Huayu Tongbi Tang showed better effect on global symptom improvement (RR 1.31, 95% CI 1.08 to 1.59) (Figure 3). Some participants (35/52 in intervention group and 33/46 in control group) underwent nerve conduction velocity examination and results showed that Yiqi Huayu Tongbi Tang had a better effect on motor nerve conduction velocity in the common peroneal nerve (MD 3.30, 95% CI 1.55 to 5.05), sensory nerve conduction velocity in the common peroneal nerve (MD 4.30, 95% CI 2.01 to 6.59), motor nerve conduction velocity in the median nerve (MD 2.60, 95% CI 0.78 to 4.42), and sensory nerve conduction velocity in the median nerve (MD 3.00, 95% CI 0.74 to 5.26) (all m/s).

Yiqi Tongluo Tang

One trial tested Yiqi Tongluo Tang in 90 participants with DPN (Peng 2009). The combination of Yiqi Tongluo Tangwith mecobalamin had a better effect on global symptom improvement compared with mecobalamin alone (RR 1.46, 95% CI 1.15 to 1.87) (Figure 2).

Yiyiren Tang

One trial tested Yiyiren Tang in 74 participants with DPN (Luo 2009). The combination of Yiyiren Tang with mecobalamin had a better effect on global symptom improvement compared with mecobalamin alone (RR 1.26, 95% CI 1.03 to 1.55) (Figure 2).

Zhuyu Huatan Tang

One trial tested Zhuyu Huatan Tang in 80 participants with DPN (Cao 2011). The Zhuyu Huatan Tang showed a better effect on global symptom improvement compared with vitamin B1 and vitamin B12 combined (RR 1.30, 95% CI 1.01 to 1.66) (Figure 3).

Discussion

Summary of main results

In our review, all the trials claimed that CHM had a positive effect though some of the trials turned out to have negative findings when analysed by standard statistical techniques using risk ratios (RRs) or mean differences (MDs). Based on this systematic review, there is no evidence to support the objective effectiveness and safety of Chinese herbal medicines for DPN. No well-designed, randomised placebo controlled trial with objective outcome measures has been conducted.

Quality of the evidence

All the randomised trials included in this review were of very low quality in terms of design, reporting, and methodology. They provided only limited descriptions of study design, randomisation, allocation concealment, and baseline data. Although all trials stated the randomisation procedure they used, they provided insufficient information to judge whether randomisation was conducted properly. No trial stated how the investigators concealed allocation. In addition, 16 trials (Chen 2009; Li 2001; Li 2005; Li 2008b; Li 2009; Lin 2008a; Liu 2004; Liu 2006; Liu 2008; Liu 2011; Sun 2010b; Wang 2008; Wu 2003; Wu 2006; Zhang 2007; Zhou 2010a) had only one author, which makes it impossible to perform a randomised controlled trial (RCT) properly in terms of the randomisation procedure and allocation concealment. No trial performed blinding. One trial reported the number of dropouts (Yin 2011); however, the investigators did not use an intention-to-treat analysis, but cancelled the data of participants who dropped out, which exposed the results to bias. None of the trials mentioned intention-to-treat analysis or had a pre-trial estimation of sample size. Methodologically poorly designed trials show larger differences between experimental and control groups than those conducted rigorously (Schulz 1995; Moher 1998; Kjaergard 2001). The small improvements in qualitative outcomes in these trials should therefore be regarded with caution.

The included trials were heterogeneous in the type of Chinese herbal medicine (CHM) (few of the herbal medicines were tested more than twice), controls used, and outcomes. No multicentre, large-scale RCTs were identified. It should be noted that there was a lack of knowledge among the trial authors of the meaning of placebo control in the trials. Among the 49 included trials, only one (Zhao 2008) used a formal placebo control. However, in this trial the author stated clearly that he did not use 'blinding' and thus the placebo effect could not be excluded. The positive effect of Tangluoan capsule should be regarded as inconclusive. The other 48 trials used either no control or another active control. The majority of trials in this review used conventional medicine as the control treatment. Other than those that compared CHM with no treatment or placebo control, the trials compared CHM to B vitamins, namely, vitamin B12, B1, or mecobalamin. The effectiveness of vitamin B12 on diabetic peripheral neuropathy has been shown by RCTs and a ystematic review (Sun 2005; Li 2008). However, there has not been sufficient evidence so far to support the efficacy of mecobalamin. We could not conclude that the CHM interventions worked even if the trials had positive results. In addition, when trials compare two medicines, it is essential that both groups receive the comparator in the same formulation as the active treatment so that they appear the same. If not, then the trial is not blinded. In most of the RCTs in our review, the formulations of control medicine and CHM were quite different. The lack of blinding risks exaggerating the effects of CHM. As a result, any interpretation of the positive findings of treatment with CHM should be made with caution.

Trials of Chinese herbal injections should also highlight the potential effect of a positive placebo effect or adverse effects of an injection. The included 49 trials used a total of five Chinese herbal injections, that is Fufang Danshen (Ma 2008), Erigeron (Wu 2003), Ciwujia (Wu 2006), Xuesaitong (Xin 2003), and Ginkgo biloba extract (Yan 2008). One trial (Wu 2003) compared an Erigeron infusion with vitamin B1 injection, and the other four trials (Xin 2003; Wu 2006; Ma 2008; Yan 2008) compared Chinese herbal injection plus non-injected conventional medicine (vitamin B or mecobalamin) with conventional medicine. It is known that an injection alone has a strong potential placebo effect; however, these five trials used no adequate placebo control to offset the placebo effect. Two trials highlight the strong placebo effect of injection (Xin 2003; He 2005). These two trials tested Xuesaitong in different formulations, one a Xuesaitong injection and the other a Xuesaitong capsule (Figure 2). When compared with conventional medicines, the effect was nonsignificant if the substance was taken by capsule and significant when injected. Therefore, the overall effect of Chinese herbal injection could not rule out a placebo effect of the injection. The positive effect should be interpreted conservatively. In other cases, participants took oral CHM daily versus a daily control injection, with global symptom improvement as an outcome measure. The effect of daily intramuscular injections may have an adverse and negative effect in this instance.

There were large variations in the formulation, dosage, administration, duration of treatment, and control interventions in the included trials for the Chinese herbal medicines tested. In total, the trials tested 38 different herbal medicines and few more than twice. Even for a herbal intervention of the same name, there were still differences in the specific composition of the Chinese herbal medicine and the treatment regimen including the dosage and duration of treatment. Therefore, it was difficult to undertake subgroup analyses to explore factors that may have an impact on the effects of a treatment regimen. There is still a lack of information about quality control for the development of the herbal preparations or for the manufacture of herbal products. Future trials should provide information about standardisation, including composition, quality control, a detailed regimen, and duration of treatment.

The outcome measures to show effectiveness in this review were not appropriately objective for the assessment of unbiased outcomes. The use of composite qualitative and subjective outcomes such as 'markedly effective', 'effective', and 'ineffective' in most of trials limited the generalisability of the findings. The three classifications for 'global symptom improvement' as an outcome measure are not internationally recognised but global symptom improvement is commonly used in Chinese trials. In light of the lack of blinding and other significant bias any positive reported outcomes should be interpreted with caution. In the inclusion criteria, we extracted data on change in nerve conduction velocity only if the treatment duration lasted more than 12 weeks, because we believe that less than 12 weeks of treatment is unlikely to affect nerve conduction velocity. There were only five trials with a treatment period that exceeded 12 weeks and within these, conduction was measured using various nerves. No trial reported the incidence of complications, quality of life, or health economics. Most trials focused on short-term or intermediate-term rather than long-term outcomes. Data from individual participants were not available.

There was inadequate reporting of adverse events in the included trials. Most of the trials (29/49) did not mention whether they monitored for adverse effects. In 18 trials, no adverse events were identified. Only two trials reported adverse events, one occurred in the control group and in the other trial it was unclear in which group the adverse events occurred. It is not possible to draw conclusions about the safety of herbal medicines from this review due to the limited and inadequate recording and reporting of adverse events. In China, there is a general perception that it is safe to use herbal medicines for various conditions. The low level of reporting on adverse events may reflect this. However, there are increasing reports of liver toxicity and other adverse events associated with Chinese herbal medicines (Ishizaki 1996; Melchart 1999; Gottieb 2000). For this reason, clinical trials need to monitor and report the safety of herbal medicines.

Potential biases in the review process

We conducted funnel plots to investigate the potential publication bias, see Figure 5 and Figure 6. Both plots demonstrated asymmetrical funnel plots. Funnel plots are a visual aid to identify publication bias or systematic heterogeneity. We have included all CHMs in these funnel plots recognising the substantial heterogeneity of the intervention, trial size and design. However, as shown in Figure 5 and Figure 6, there were no studies of any size which found a negative or nil effect, which classically indicates publication bias. Although we undertook extensive searches for unpublished material, we found no unpublished 'negative' studies, as also found in previous studies (Vickers 1998). Asymmetry may also be due to the poor methodological quality of the included trials. The less precise the study (higher standard error), the more exaggerated the intervention effect estimate produced, either as a result of small sample size, or poor quality studies.

Figure 5.

Funnel plot of comparison: 3 Chinese herbal medicine vs conventional medicine, outcome: 3.1 Global symptom improvement.

Figure 6.

Funnel plot of comparison: 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, outcome: 4.1 Global symptom improvement.

Authors' conclusions

Implications for practice

Based on this systematic review, there is no evidence to support the objective effectiveness and safety of Chinese herbal medicines for diabetic peripheral neuropathy. No well-designed, randomised, placebo controlled trial with objective outcome measures has been conducted.

Implications for research

More studies of high methodological quality, performed worldwide, with large numbers of participants and good reporting are required to provide evidence for any benefit of Chinese herbal medicines. Trial reports should provide information about species of herb, geographical origin of herbs, season for collecting, and quality of the preparations (Gagnier 2006), where this is thought to affect efficacy. Investigators should also pay attention to the definition of outcome measures and the incidence of adverse reactions. Trialists should address the following methodological issues: (i) methods used to generate the allocation sequence and allocation concealment; (ii) double blinding with the use of adequate placebo; (iii) clear descriptions of withdrawals and dropouts during the trial, and use of intention-to-treat analysis; and (iv) reporting on trials according to the CONSORT Statement (www.consort-statement.org/) (Moher 2010; Schulz 2010).

Future trials might initially concentrate on Chinese herbal medicines where some evidence for changes in objective outcomes has been presented, namely aloe vera, Furong Tongmai capsule, Buyang Huanwu Tang, and possibly the Tangluoan capsule.

Acknowledgements

This work was supported by a grant from the National Basic Research Program of China ("973" Program, no 2006CB504602).

The authors thank the Cochrane Neuromuscular Disease Group for helping with the development of the protocol and the literature search of non-Chinese databases (Cochrane Neuromuscular Disease Group Specialized Register, CENTRAL, MEDLINE, EMBASE, and AMED) and we thank all the volunteers on this review for their great work in the selection of the tremendous literature.

The editorial base of the Cochrane Neuromuscular Disease Group is supported by the MRC Centre for Neuromuscular Diseases.

Data and analyses

Download statistical data

Comparison 1. Chinese herbal medicine vs no treatment
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Global symptom improvement1 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Aloe vs no treatment140Risk Ratio (M-H, Fixed, 95% CI)1.67 [0.96, 2.88]
2 Change of motor nerve conduction velocity in common peroneal nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
2.1 Aloe vs no treatment140Mean Difference (IV, Fixed, 95% CI)8.60 [7.18, 10.02]
3 Change of sensory nerve conduction velocity in common peroneal nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
3.1 Aloe vs no treatment140Mean Difference (IV, Fixed, 95% CI)8.60 [7.58, 9.62]
4 Change of motor nerve conduction velocity in median nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
4.1 Aloe vs no treatment140Mean Difference (IV, Fixed, 95% CI)7.40 [5.09, 9.71]
5 Change of sensory nerve conduction velocity in median nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
5.1 Aloe vs no treatment140Mean Difference (IV, Fixed, 95% CI)10.0 [8.08, 11.92]
Analysis 1.1.

Comparison 1 Chinese herbal medicine vs no treatment, Outcome 1 Global symptom improvement.

Analysis 1.2.

Comparison 1 Chinese herbal medicine vs no treatment, Outcome 2 Change of motor nerve conduction velocity in common peroneal nerve (m/s).

Analysis 1.3.

Comparison 1 Chinese herbal medicine vs no treatment, Outcome 3 Change of sensory nerve conduction velocity in common peroneal nerve (m/s).

Analysis 1.4.

Comparison 1 Chinese herbal medicine vs no treatment, Outcome 4 Change of motor nerve conduction velocity in median nerve (m/s).

Analysis 1.5.

Comparison 1 Chinese herbal medicine vs no treatment, Outcome 5 Change of sensory nerve conduction velocity in median nerve (m/s).

Comparison 2. Chinese herbal medicine vs placebo
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Global symptom improvement1 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Tangluoan capsule vs placebo157Risk Ratio (M-H, Fixed, 95% CI)2.31 [1.24, 4.28]
2 Change of motor nerve conduction velocity in common peroneal nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
2.1 Tangluoan capsule vs placebo157Mean Difference (IV, Fixed, 95% CI)3.84 [1.38, 6.30]
3 Change of sensory nerve conduction velocity in common peroneal nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
3.1 Tangluoan capsule vs placebo157Mean Difference (IV, Fixed, 95% CI)5.07 [2.57, 7.57]
4 Change of motor nerve conduction velocity in median nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
4.1 Tangluoan capsule vs placebo157Mean Difference (IV, Fixed, 95% CI)1.64 [-1.70, 4.98]
5 Change of sensory nerve conduction velocity in median nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
5.1 Tangluoan capsule vs placebo157Mean Difference (IV, Fixed, 95% CI)5.67 [3.02, 8.32]
Analysis 2.1.

Comparison 2 Chinese herbal medicine vs placebo, Outcome 1 Global symptom improvement.

Analysis 2.2.

Comparison 2 Chinese herbal medicine vs placebo, Outcome 2 Change of motor nerve conduction velocity in common peroneal nerve (m/s).

Analysis 2.3.

Comparison 2 Chinese herbal medicine vs placebo, Outcome 3 Change of sensory nerve conduction velocity in common peroneal nerve (m/s).

Analysis 2.4.

Comparison 2 Chinese herbal medicine vs placebo, Outcome 4 Change of motor nerve conduction velocity in median nerve (m/s).

Analysis 2.5.

Comparison 2 Chinese herbal medicine vs placebo, Outcome 5 Change of sensory nerve conduction velocity in median nerve (m/s).

Comparison 3. Chinese herbal medicine vs conventional medicine
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Global symptom improvement22 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Erigeron injection vs vit B1179Risk Ratio (M-H, Fixed, 95% CI)1.61 [1.15, 2.25]
1.2 Compound Qiying granule vs cobamamide162Risk Ratio (M-H, Fixed, 95% CI)1.33 [0.98, 1.82]
1.3 Modified Huangqi Guizhi Wuwu Tang vs mecobalamin168Risk Ratio (M-H, Fixed, 95% CI)1.29 [1.02, 1.64]
1.4 Modified Huangqi Guizhi Wuwu Tang vs vit B1+ vit B6170Risk Ratio (M-H, Fixed, 95% CI)1.33 [1.01, 1.76]
1.5 Jianpi Yiqi Tong Luo Tang vs cobamamide148Risk Ratio (M-H, Fixed, 95% CI)2.22 [1.29, 3.84]
1.6 Tongbi Tang vs mecobalamin186Risk Ratio (M-H, Fixed, 95% CI)1.31 [1.04, 1.66]
1.7 Modified Sishen Jian decoction vs vit B1+ vit B12170Risk Ratio (M-H, Fixed, 95% CI)1.43 [1.11, 1.85]
1.8 Tangluoan capsule vs mecobalamin161Risk Ratio (M-H, Fixed, 95% CI)1.18 [0.82, 1.71]
1.9 Tangmaining capsule vs mecobalamin1100Risk Ratio (M-H, Fixed, 95% CI)1.02 [0.93, 1.12]
1.10 Furong Tongmai capsule vs mecobalamin186Risk Ratio (M-H, Fixed, 95% CI)1.46 [1.12, 1.90]
1.11 Tongxinluo capsule vs mecobalamin2121Risk Ratio (M-H, Fixed, 95% CI)1.25 [0.98, 1.58]
1.12 Tang Mai Yin vs mecobalamin162Risk Ratio (M-H, Fixed, 95% CI)1.64 [1.15, 2.35]
1.13 Modified Buyang Huanwu Tang vs vit B1+ vit B 121134Risk Ratio (M-H, Fixed, 95% CI)1.25 [1.02, 1.53]
1.14 Yiqi Huoxue Wenyang prescription vs mecobalamin160Risk Ratio (M-H, Fixed, 95% CI)1.53 [1.09, 2.16]
1.15 Xiaoke Tongluo Tang vs vit B1 + vit B6160Risk Ratio (M-H, Fixed, 95% CI)1.39 [1.00, 1.94]
1.16 Dispelling dampness and dredging collaterals TCM decoction vs mecobalamin182Risk Ratio (M-H, Fixed, 95% CI)2.27 [1.40, 3.68]
1.17 Modified Duhuo Jisheng decoction + Wujia Canger Jiangtang decoction vs mecobalamin + vit B1140Risk Ratio (M-H, Fixed, 95% CI)1.58 [1.09, 2.30]
1.18 Zhuyu Huatan Tang vs vit B1 + vit B12180Risk Ratio (M-H, Fixed, 95% CI)1.30 [1.01, 1.66]
1.19 Tonifying kidney and activating blood formula vs mecobalamin1100Risk Ratio (M-H, Fixed, 95% CI)1.35 [0.98, 1.86]
1.20 Tongluo Tangtai Decoction vs mecobalamin140Risk Ratio (M-H, Fixed, 95% CI)1.27 [0.96, 1.66]
1.21 Yiqi Huayu Tongbi Tang vs mecobalamin198Risk Ratio (M-H, Fixed, 95% CI)1.31 [1.08, 1.59]
2 Improvement in numbness2 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
2.1 Unnamed Chinese herbal prescription (Luo 2008) vs mecobalamin183Risk Ratio (M-H, Fixed, 95% CI)1.26 [1.06, 1.50]
2.2 Tongluo Tangtai Decoction vs mecobalamin140Risk Ratio (M-H, Fixed, 95% CI)1.36 [1.00, 1.84]
3 Improvement in pain2 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
3.1 Unnamed Chinese herbal prescription (Luo 2008) versus mecobalamin174Risk Ratio (M-H, Fixed, 95% CI)1.45 [0.96, 2.18]
3.2 Tongluo Tangtai decoction vs mecobalamin136Risk Ratio (M-H, Fixed, 95% CI)1.6 [1.03, 2.50]
4 Change of motor nerve conduction velocity in common peroneal nerve (m/s)5 Mean Difference (IV, Fixed, 95% CI)Subtotals only
4.1 Tangluoan capsule vs mecobalamin161Mean Difference (IV, Fixed, 95% CI)1.92 [-0.64, 4.48]
4.2 Furong Tongmai capsule vs mecobalamin186Mean Difference (IV, Fixed, 95% CI)4.60 [1.94, 7.26]
4.3 Tang Mai Yin vs mecobalamin162Mean Difference (IV, Fixed, 95% CI)-0.60 [-2.65, 1.45]
4.4 Yiqi Huayu Tongbi Tang versus mecobalamin168Mean Difference (IV, Fixed, 95% CI)3.30 [1.55, 5.05]
4.5 Tonifying kidney and activating blood formula vs mecobalamin1100Mean Difference (IV, Fixed, 95% CI)1.0 [0.64, 1.36]
5 Change of sensory nerve conduction velocity in common peroneal nerve (m/s)6 Mean Difference (IV, Fixed, 95% CI)Subtotals only
5.1 Tangluoan capsule vs mecobalamin161Mean Difference (IV, Fixed, 95% CI)2.57 [0.08, 5.06]
5.2 Furong Tongmai capsule vs mecobalamin186Mean Difference (IV, Fixed, 95% CI)5.30 [1.66, 8.94]
5.3 Tang Mai Yin vs mecobalamin162Mean Difference (IV, Fixed, 95% CI)0.70 [-1.38, 2.78]
5.4 Tonifying kidney and activating blood formula vs mecobalamin1100Mean Difference (IV, Fixed, 95% CI)8.83 [8.08, 9.58]
5.5 Tongluo Tangtai decoction vs mecobalamin140Mean Difference (IV, Fixed, 95% CI)8.22 [4.78, 11.66]
5.6 Yiqi Huayu Tongbi Tang vs mecobalamin168Mean Difference (IV, Fixed, 95% CI)4.30 [2.01, 6.59]
6 Change of motor nerve conduction velocity in median nerve (m/s)3 Mean Difference (IV, Fixed, 95% CI)Subtotals only
6.1 Tangluoan capsule vs mecobalamin161Mean Difference (IV, Fixed, 95% CI)-0.76 [-4.05, 2.53]
6.2 Tonifying kidney and activating blood formula vs mecobalamin1100Mean Difference (IV, Fixed, 95% CI)3.53 [2.38, 4.68]
6.3 Yiqi Huayu Tongbi Tang vs mecobalamin168Mean Difference (IV, Fixed, 95% CI)2.60 [0.78, 4.42]
7 Change of sensory nerve conduction velocity in median nerve (m/s)4 Mean Difference (IV, Fixed, 95% CI)Subtotals only
7.1 Tangluoan capsule vs mecobalamin161Mean Difference (IV, Fixed, 95% CI)1.76 [-1.02, 4.54]
7.2 Tonifying kidney and activating blood formula versus mecobalamin1100Mean Difference (IV, Fixed, 95% CI)2.44 [1.43, 3.45]
7.3 Tongluo Tangtai decoction vs mecobalamin140Mean Difference (IV, Fixed, 95% CI)5.20 [-0.01, 10.41]
7.4 Yiqi Huayu Tongbi Tang vs mecobalamin168Mean Difference (IV, Fixed, 95% CI)3.0 [0.74, 5.26]
8 Change of motor nerve conduction velocity in tibial nerve1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
8.1 Tang Mai Yin vs mecobalamin162Mean Difference (IV, Fixed, 95% CI)-0.30 [-2.99, 2.39]
9 Change of sensory nerve conduction velocity in tibial nerve1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
9.1 Tang Mai Yin vs mecobalamin162Mean Difference (IV, Fixed, 95% CI)3.10 [0.80, 5.40]
10 Change of sensory nerve conduction velocity in ulnar nerve1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
10.1 Tongluo Tangtai decoction vs mecobalamin140Mean Difference (IV, Fixed, 95% CI)4.13 [-0.78, 9.04]
Analysis 3.1.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 1 Global symptom improvement.

Analysis 3.2.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 2 Improvement in numbness.

Analysis 3.3.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 3 Improvement in pain.

Analysis 3.4.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 4 Change of motor nerve conduction velocity in common peroneal nerve (m/s).

Analysis 3.5.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 5 Change of sensory nerve conduction velocity in common peroneal nerve (m/s).

Analysis 3.6.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 6 Change of motor nerve conduction velocity in median nerve (m/s).

Analysis 3.7.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 7 Change of sensory nerve conduction velocity in median nerve (m/s).

Analysis 3.8.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 8 Change of motor nerve conduction velocity in tibial nerve.

Analysis 3.9.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 9 Change of sensory nerve conduction velocity in tibial nerve.

Analysis 3.10.

Comparison 3 Chinese herbal medicine vs conventional medicine, Outcome 10 Change of sensory nerve conduction velocity in ulnar nerve.

Comparison 4. Chinese herbal medicine plus conventional medicine vs conventional medicine
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Global symptom improvement26 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Ciwujia injection + vit B1+ vit B12 vs vit B1+ vit B12188Risk Ratio (M-H, Fixed, 95% CI)1.78 [1.34, 2.36]
1.2 Fufang Danshen injection + mecobalamin vs mecobalamin1100Risk Ratio (M-H, Fixed, 95% CI)1.53 [1.21, 1.95]
1.3 Ginkgo biloba extract injection + vit B1+ vit B12 vs vit B1+ vit B12170Risk Ratio (M-H, Fixed, 95% CI)1.25 [0.96, 1.62]
1.4 Xuesaitong injection + vit B1+ vit B12 vs vit B1+ vit B12161Risk Ratio (M-H, Fixed, 95% CI)1.86 [1.19, 2.90]
1.5 Xuesaitong capsule/tablet + mecobalamin vs mecobalamin2177Risk Ratio (M-H, Fixed, 95% CI)1.27 [1.02, 1.58]
1.6 Buyang Huanwu decoction + mecobalamin vs mecobalamin184Risk Ratio (M-H, Fixed, 95% CI)1.5 [1.08, 2.08]
1.7 Modified Buyang Huanwu Tang + vit B1+ vit B6 vs vit B1+ vit B61110Risk Ratio (M-H, Fixed, 95% CI)1.28 [1.06, 1.55]
1.8 Modified Danggui Sini Tang + vit B1+ vit B6+ vit B12 vs vit B1+ vit B6+ vit B12170Risk Ratio (M-H, Fixed, 95% CI)1.33 [1.04, 1.71]
1.9 Guizhi Gegen Tang + vit B1+ vit B6 vs vit B1+ vit B6160Risk Ratio (M-H, Fixed, 95% CI)1.29 [0.99, 1.67]
1.10 Yiqi Bushen Huoxue Tang + indometacin + vit B1+ vit B12 vs indometacin + vit B1+ vit B12136Risk Ratio (M-H, Fixed, 95% CI)1.42 [1.00, 2.00]
1.11 Huangqi Guizhi Wuwu Tang + mecobalamin vs mecobalamin3186Risk Ratio (M-H, Fixed, 95% CI)1.29 [1.11, 1.50]
1.12 Modified Huangqi Guizhi Wuwu Tang + vit B1+ vit B6 vs vit B1+ vit B61110Risk Ratio (M-H, Fixed, 95% CI)1.13 [0.93, 1.36]
1.13 Taoren Honghua Jian + mecobalamin vs mecobalamin180Risk Ratio (M-H, Fixed, 95% CI)1.36 [1.04, 1.79]
1.14 Tongxinluo capsule + mecobalamin vs mecobalamin2105Risk Ratio (M-H, Fixed, 95% CI)1.24 [0.95, 1.62]
1.15 Tongxinluo capsule + vit B1+ vit B12 vs vit B1+ vit B12165Risk Ratio (M-H, Fixed, 95% CI)1.87 [1.22, 2.86]
1.16 Xuefu Zhuyu capsule + mecobalamin vs mecobalamin160Risk Ratio (M-H, Fixed, 95% CI)1.23 [0.96, 1.57]
1.17 Yiyiren Tang + mecobalamin vs mecobalamin174Risk Ratio (M-H, Fixed, 95% CI)1.26 [1.03, 1.55]
1.18 Yiqi Huoxue Tongluo compound prescription + mecobalamin vs mecobalamin179Risk Ratio (M-H, Fixed, 95% CI)1.32 [1.04, 1.67]
1.19 Unnamed Chinese herbal prescription (Niu 2008) + mecobalamin vs mecobalamin184Risk Ratio (M-H, Fixed, 95% CI)1.16 [0.94, 1.42]
1.20 Yiqi Tongluo Tang + mecobalamin vs mecobalamin190Risk Ratio (M-H, Fixed, 95% CI)1.46 [1.15, 1.87]
1.21 Yangyin Huoxue Tang + mecobalamin vs mecobalamin172Risk Ratio (M-H, Fixed, 95% CI)1.34 [1.02, 1.77]
1.22 Qingxiao Tangbi prescription + mecobalamin vs mecobalamin160Risk Ratio (M-H, Fixed, 95% CI)1.63 [1.13, 2.34]
2 Change of motor nerve conduction velocity in common peroneal nerve (m/s)3 Mean Difference (IV, Fixed, 95% CI)Subtotals only
2.1 Buyang Huanwu decoction + mecobalamin vs mecobalamin184Mean Difference (IV, Fixed, 95% CI)5.95 [3.75, 8.15]
2.2 Xuesaitong tablet + mecobalamin vs mecobalamin1112Mean Difference (IV, Fixed, 95% CI)-0.55 [-1.97, 0.87]
2.3 Yangyin Huoxue Tang + mecobalamin vs mecobalamin172Mean Difference (IV, Fixed, 95% CI)0.05 [-0.49, 0.59]
3 Change of sensory nerve conduction velocity in common peroneal nerve (m/s)3 Mean Difference (IV, Fixed, 95% CI)Subtotals only
3.1 Buyang Huanwu decoction + mecobalamin vs mecobalamin184Mean Difference (IV, Fixed, 95% CI)5.68 [3.40, 7.96]
3.2 Xuesaitong tablet + mecobalamin vs mecobalamin1112Mean Difference (IV, Fixed, 95% CI)8.5 [7.93, 9.07]
3.3 Yangyin Huoxue Tang + mecobalamin vs mecobalamin172Mean Difference (IV, Fixed, 95% CI)9.08 [8.26, 9.90]
4 Change of motor nerve conduction velocity in median nerve (m/s)2 Mean Difference (IV, Fixed, 95% CI)Subtotals only
4.1 Xuesaitong tablet + mecobalamin vs mecobalamin1112Mean Difference (IV, Fixed, 95% CI)6.47 [5.14, 7.80]
4.2 Yangyin Huoxue Tang + mecobalamin vs mecobalamin172Mean Difference (IV, Fixed, 95% CI)3.19 [1.93, 4.45]
5 Change of sensory nerve conduction velocity in median nerve (m/s)2 Mean Difference (IV, Fixed, 95% CI)Subtotals only
5.1 Xuesaitong tablet + mecobalamin vs mecobalamin1112Mean Difference (IV, Fixed, 95% CI)1.04 [0.03, 2.05]
5.2 Yangyin Huoxue Tang + mecobalamin vs mecobalamin172Mean Difference (IV, Fixed, 95% CI)3.14 [1.92, 4.36]
6 Change of motor nerve conduction velocity in left common peroneal nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
6.1 Qingxiao Tangbi prescription + mecobalamin vs mecobalamin160Mean Difference (IV, Fixed, 95% CI)1.29 [0.46, 2.12]
7 Change of motor nerve conduction velocity in right common peroneal nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
7.1 Qingxiao Tangbi prescription + mecobalamin vs mecobalamin160Mean Difference (IV, Fixed, 95% CI)2.17 [1.50, 2.84]
8 Change of sensory nerve conduction velocity in left common peroneal nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
8.1 Qingxiao Tangbi prescription + mecobalamin vs mecobalamin160Mean Difference (IV, Fixed, 95% CI)1.49 [1.12, 1.86]
9 Change of sensory nerve conduction velocity in right common peroneal nerve (m/s)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
9.1 Qingxiao Tangbi prescription + mecobalamin vs mecobalamin160Mean Difference (IV, Fixed, 95% CI)2.37 [1.92, 2.82]
Analysis 4.1.

Comparison 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, Outcome 1 Global symptom improvement.

Analysis 4.2.

Comparison 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, Outcome 2 Change of motor nerve conduction velocity in common peroneal nerve (m/s).

Analysis 4.3.

Comparison 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, Outcome 3 Change of sensory nerve conduction velocity in common peroneal nerve (m/s).

Analysis 4.4.

Comparison 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, Outcome 4 Change of motor nerve conduction velocity in median nerve (m/s).

Analysis 4.5.

Comparison 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, Outcome 5 Change of sensory nerve conduction velocity in median nerve (m/s).

Analysis 4.6.

Comparison 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, Outcome 6 Change of motor nerve conduction velocity in left common peroneal nerve (m/s).

Analysis 4.7.

Comparison 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, Outcome 7 Change of motor nerve conduction velocity in right common peroneal nerve (m/s).

Analysis 4.8.

Comparison 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, Outcome 8 Change of sensory nerve conduction velocity in left common peroneal nerve (m/s).

Analysis 4.9.

Comparison 4 Chinese herbal medicine plus conventional medicine vs conventional medicine, Outcome 9 Change of sensory nerve conduction velocity in right common peroneal nerve (m/s).

Appendices

Appendix 1. CENTRAL search strategy

#1(diabetes or diabetic)
#2MeSH descriptor Peripheral Nervous System Diseases explode all trees
#3(neuropath* or polyneuropath*)
#4(#2 OR #3)
#5(#1 AND #4)
#6MeSH descriptor Drugs, Chinese Herbal, this term only
#7MeSH descriptor Medicine, Chinese Traditional explode all trees
#8MeSH descriptor Medicine, East Asian Traditional, this term only
#9(chinese near (traditional or medicine*))
#10herbs or herbal or herb
#11plant or plants
#12MeSH descriptor Phytotherapy, this term only
#13(traditional near medicine*)
#14(#6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13)
#15(#5 AND #14)

Appendix 2. MEDLINE (OvidSP) search strategy

Database: Ovid MEDLINE(R) <1946 to May Week 1 2012>
Search Strategy:
--------------------------------------------------------------------------------
1 randomized controlled trial.pt. (326354)
2 controlled clinical trial.pt. (84043)
3 randomized.ab. (230515)
4 placebo.ab. (130935)
5 drug therapy.fs. (1528972)
6 randomly.ab. (166436)
7 trial.ab. (238406)
8 groups.ab. (1094464)
9 or/1-8 (2838403)
10 exp animals/ not humans.sh. (3712810)
11 9 not 10 (2410011)
12 exp diabetes mellitus/ (279973)
13 diabet$.mp. (385890)
14 12 or 13 (386989)
15 neuropath$.mp. (84520)
16 exp peripheral nervous system diseases/ (113998)
17 polyneuropath$.mp. (11599)
18 or/15-17 (165943)
19 14 and 18 (18006)
20 exp diabetic neuropathies/ (15444)
21 diabetic neuropath$.mp. (12388)
22 diabetic polyneuropath$.mp. (689)
23 or/20-22 (16669)
24 19 or 23 (21518)
25 drugs, chinese herbal/ or plants medicinal/ or medicine, herbal/ (69430)
26 Medicine, Chinese Traditional/ (9227)
27 medicine, oriental tradition/ (3215)
28 phytotherapy/ (24688)
29 (chinese adj7 (traditional or medicine$)).tw. (12574)
30 (herb or herbs or herbal).mp. (40024)
31 (plant or plants).mp. (417249)
32 (traditional adj8 medicine$).tw. (13090)
33 or/25-32 (453342)
34 24 and 33 (217)
35 11 and 34 (106)

Appendix 3. EMBASE (OvidSP) search strategy

Database: Embase <1980 to 2012 Week 19>
Search Strategy:
--------------------------------------------------------------------------------
1 crossover-procedure/ (33755)
2 double-blind procedure/ (108636)
3 randomized controlled trial/ (321249)
4 single-blind procedure/ (15834)
5 (random$ or factorial$ or crossover$ or cross over$ or cross-over$ or placebo$ or (doubl$ adj blind$) or (singl$ adj blind$) or assign$ or allocat$ or volunteer$).tw. (1115305)
6 or/1-5 (1192574)
7 human/ (13371590)
8 6 and 7 (885619)
9 nonhuman/ or human/ (16488235)
10 6 not 9 (201000)
11 8 or 10 (1086619)
12 exp diabetes mellitus/ (483819)
13 diabet$.mp. (571905)
14 12 or 13 (574676)
15 neuropath$.mp. (177434)
16 exp peripheral nervous system diseases/ (43955)
17 polyneuropath$.mp. (18043)
18 or/15-17 (186479)
19 14 and 18 (29988)
20 exp diabetic neuropathies/ (15847)
21 diabetic neuropath$.mp. (17365)
22 diabetic polyneuropath$.mp. (1123)
23 or/20-22 (17512)
24 19 or 23 (29988)
25 herbaceous agent/ (29160)
26 chinese medicine/ or herbal medicine/ or traditional medicine/ (44018)
27 medicine, oriental tradition/ (2497)
28 phytotherapy/ (14043)
29 (chinese adj7 (traditional or medicine$)).tw. (19573)
30 (herb or herbs or herbal).tw. (34544)
31 (plant or plants).tw. (277831)
32 (traditional adj8 medicine$).tw. (21087)
33 or/25-32 (364504)
34 11 and 24 and 33 (46)
35 34 and 201000:201219.(em). (9)
36 crossover-procedure.sh. (33755)
37 double-blind procedure.sh. (108636)
38 single-blind procedure.sh. (15834)
39 randomized controlled trial.sh. (321249)
40 (random$ or crossover$ or cross over$ or placebo$ or (doubl$ adj blind$) or allocat$).tw,ot. (863399)
41 trial.ti. (129634)
42 or/36-41 (989109)
43 (animal/ or nonhuman/ or animal experiment/) and human/ (1176120)
44 animal/ or nonanimal/ or animal experiment/ (3263880)
45 44 not 43 (2705699)
46 42 not 45 (906403)
47 limit 46 to embase (701645)
48 exp diabetes mellitus/ (483819)
49 diabet$.mp. (571905)
50 48 or 49 (574676)
51 neuropath$.mp. (177434)
52 exp peripheral nervous system diseases/ (43955)
53 polyneuropath$.mp. (18043)
54 or/51-53 (186479)
55 50 and 54 (29988)
56 exp diabetic neuropathies/ (15847)
57 diabetic neuropath$.mp. (17365)
58 diabetic polyneuropath$.mp. (1123)
59 or/56-58 (17512)
60 55 or 59 (29988)
61 herbaceous agent/ (29160)
62 chinese medicine/ or herbal medicine/ or traditional medicine/ (44018)
63 medicine, oriental tradition/ (2497)
64 phytotherapy/ (14043)
65 (chinese adj7 (traditional or medicine$)).tw. (19573)
66 (herb or herbs or herbal).tw. (34544)
67 (plant or plants).tw. (277831)
68 (traditional adj8 medicine$).tw. (21087)
69 or/61-68 (364504)
70 47 and 60 and 69 (26)

Appendix 4. AMED (OvidSP) search strategy

Database: AMED (Allied and Complementary Medicine) <1985 to May 2012>
Search Strategy:
--------------------------------------------------------------------------------
1 exp diabetes mellitus/ (1975)
2 diabet$.mp. (3721)
3 1 or 2 (3721)
4 neuropath$.mp. (1411)
5 exp peripheral nervous system disease/ (2624)
6 polyneuropath$.mp. (105)
7 or/4-6 (3413)
8 3 and 7 (858)
9 exp diabetic neuropathies/ (606)
10 diabetic neuropath$.mp. (263)
11 diabetic polyneuropath$.mp. (17)
12 or/9-11 (679)
13 8 or 12 (858)
14 drugs chinese herbal/ or plants medicinal/ or herbal drugs/ (19381)
15 traditional chinese medicine/ or traditional medicine/ (1363)
16 herbalism/ (2272)
17 phytotherapy/ (1801)
18 (chinese adj7 (traditional or medicine$)).tw. (5553)
19 (herb or herbs or herbal).tw. (8874)
20 (plants or plant).tw. (21612)
21 (traditional adj8 medicine$).tw. (7447)
22 or/14-21 (31035)
23 13 and 22 (30)
24 Randomized controlled trials/ (1526)
25 Random allocation/ (302)
26 Double blind method/ (435)
27 Single Blind Method.mp. (29)
28 exp Clinical Trials/ (3187)
29 (clin$ adj25 trial$).tw. (5410)
30 ((singl$ or doubl$ or treb$ or trip$) adj25 (blind$ or mask$ or dummy)).tw. (2223)
31 placebos/ (518)
32 placebo$.tw. (2493)
33 random$.tw. (12690)
34 research design/ (1671)
35 Prospective Studies/ (452)
36 cross?over studies.mp. (6)
37 meta analysis/ (108)
38 (meta?analys$ or systematic review$).tw. (1828)
39 control$.tw. (27400)
40 (multicenter or multicentre).tw. (720)
41 ((study or studies or design$) adj25 (factorial or prospective or intervention or crossover or cross-over or quasi-experiment$)).tw. (9674)
42 or/24-41 (42237)
43 23 and 42 (12)

Appendix 5. CBM search strategy

#1 diabetic neuropathy
#2 diabetic polyneuropathy
#3 diabetic peripheral neuropathy
#4 #1?#3/or
#5 traditional Chinese herbal medicine
#6 Chinese herbal medicine
#7 herbal medicine
#8 traditional Chinese medicine
#9 traditional medicine
#10 Chinese medicine
#11 #5?#10/or
#12 random
#13 #4 and #11 and #12
#14 exp animals/ not humans.sh.
#15#13not #14

All of the search terms were translated to Chinese terms when we conducted the searches in CBM database.

Appendix 6. CNKI search strategy

1. exp Randomized Controlled trials / all subheadings
2. Random*
3. 1 or 2
4. exp traditional Chinese medicine/ all subheadings
5. exp integrated Chinese and western medicine/ all subheadings

6. exp herbal medicine / all subheadings
7. or/ 4-6
8. diabetic neuropathy
9. diabetic polyneuropathy 10. diabetic peripheral neuropathy

11. or/8-10
12. Human
13. 3 and 7 and 11 and 12

All of the search terms were translated to Chinese terms when we conducted the searches in CNKI database.

Appendix 7. VIP search strategy

Title/Topic/Keyword=( "diabetic peripheral neuropathy" or "diabetic neuropathy" or "diabetic polyneuropathy") AND Title/Topic/Keyword=("Chinese medicine" or "Chinese traditional medicine" or "Chinese herb*" or "oriental medicine" or "oriental traditional medicine" or "medicinal plant*" or herb*)

Timespan= 1979 to June 2010

All of the search terms were translated to Chinese terms when we conducted the searches in VIP database.

Appendix 8. Chinese Conference Papers Database search strategy

Title/Topic/Keyword=( "diabetic peripheral neuropathy" or "diabetic neuropathy" or "diabetic polyneuropathy") AND Title/Topic/Keyword=("Chinese medicine" or "Chinese traditional medicine" or "Chinese herb*" or "oriental medicine" or "oriental traditional medicine" or "medicinal plant*" or herb*)

Timespan= 1979 to June 2010

All of the search terms were translated to Chinese terms when we conducted the searches.

Appendix 9. Chinese Dissertation Database search strategy

Title/Topic/Keyword=( "diabetic peripheral neuropathy" or "diabetic neuropathy" or "diabetic polyneuropathy") AND Title/Topic/Keyword=("Chinese medicine" or "Chinese traditional medicine" or "Chinese herb*" or "oriental medicine" or "oriental traditional medicine" or "medicinal plant*" or herb*)

Timespan= 1979 to June 2010

All of the search terms were translated to Chinese terms when we conducted the searches.

What's new

DateEventDescription
10 January 2013New citation required but conclusions have not changedTen new trials incorporated but no overall change to conclusions. Two new co-authors, Xinxue Li and Guoyan Yang
14 May 2012New search has been performedSearches updated

Contributions of authors

Chen W developed and revised the review.

Zhang Y assisted with the literature searching, study selection, and data extraction.

Li XX and Yang GY assisted with the literature searching, study selection, and data extraction in the updated review.

Liu JP conceived and revised the review.

Declarations of interest

We declare that we have no conflicts of interest.

Sources of support

Internal sources

  • Centre for Evidence-Based Chinese Medicine, Beijing University of Chinese Medicine, Not specified.

External sources

  • National Basic Research Program of China ('973' Program, No. 2006CB504602), China.

  • International Cooperation Project (grant number 2009DFA31460), China.

Differences between protocol and review

Chinese herbal medicines are widely used for treating DPN in China and the number of published, relevant articles is huge. The primary research identified nearly 500 randomised trials on this topic. Previous studies have shown that RCTs without details of the randomisation sequence generation were actually not true RCTs (Tang 1999). Therefore, we restricted our inclusion criteria to only include RCTs describing a correct randomisation procedure.

We assessed 'Risk of bias' according to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008). The naming of review author judgements was revised in 2011 (Higgins 2011).

The original protocol was authored by Wei Chen, Zanhua Li and Jian Ping Liu. For the full review Yin Zhang replaced Zanhua Li. For this update there were two new authors, Xinxue Li and Guoyan Yang.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Cao 2010

Methods

RCT

No estimation of sample size

Participants

60 participants were randomised into intervention group (n = 30, M/F 16/14, mean age 50.53 ± 5.33 years), and control group (n = 30, M/F 17/13, mean age 59.33 ± 6.17 years)
Type of DM: not reported

Study location: China
Study setting: inpatient
Diagnostic criteria: World Health Organization criteria 1999

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Xiaoke Tongluo Tang. Xiaoke Tongluo Tang contains Radix Angelicae Pubescentis, Radix Gentianae Macrophyllae, Herba Epimedii, Radix Paeoniae Alba, Cortex Eucommiae, Radix Rehmanniae Recens, Radix Achyranthis Bidentatae, Rhizoma Ligustici Chuanxiong, Herba Taxilli, Caulis Spatholobi, Herba Lycopodii, Rhizoma Corydalis, and Scorpio

The treatments lasted for 6 weeks

(2) Control: vitamin B1 10 mg and vitamin B6 10 mg taken orally 3 times per day for 6 weeks

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not available
Other biasLow riskNo evidence of other bias

Cao 2011

MethodsGeneration of allocation sequence: random number table. No estimation of sample size
Participants

80 participants were randomised into intervention group (n = 40, M/F 28/12, mean age 58.93 ± 10.12 years), and control group (n = 40, M/F 26/14, mean age 59.21 ± 9.25 years).

Type of DM: type 2 DM

Study location: China

Study setting: inpatients and outpatients

Diagnostic criteria: Chinese criteria according to Practice of Internal Medicine and Clinical endocrinology

Inclusion criteria: not specified

Exclusion criteria: DPN caused by other disease

Interventions

(1)  Intervention: Zhuyu Huatan Tang. The Zhuyu Huatan Tang contains Radix Paeoniae Alba, Radix Angelicae Sinensis, Radix Salviae Miltiorrhizae, Hirudo, Lumbricus, Caulis Spatholobi, Ramulus Cinnamomi, Rhizoma Pinelliae Preparata, Pericarpium Citri Reticulatae, Poria, Arisaema cum Bile, Semen Sinapis, and Radix et Rhizoma Glycyrrhizae

(2) Control: vitamin B1 100 mg was injected once daily, vitamin B12 0.5 mg was injected once daily.

The treatments lasted for 4 weeks.

Participantss in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise and education.

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in tibial nerve (m/s)

Change of sensory nerve conduction velocity in tibial nerve (m/s)

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information provided to judge whether or not randomisation was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Chen 2006

Methods

RCT

No estimation of sample size

Participants

62 participants were randomised into intervention group (n = 34, M/F 16/18, mean age 58.6 years, range 34 to 78), and control group (n = 28, M/F 13/15, mean age 54.6 years, range 36 to 75).
Type of DM: not reported

Study location: China
Study setting: inpatient
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: modified Huangqi Guizhi Wuwu Tang and mecobalamin tablets

The basic Huangqi Guizhi Wuwu Tang formula contains Radix Astragali seu Hedysari, Radix Salviae Miltiorrhizae, Ramulus Cinnamomi, Radix Paeoniae Rubra, Radix Paeoniae Alba, Radix Rehmanniae Recens, Radix Angelicae Sinensis, Radix Glycyrrhizae, Fructus Jujubae, and Rhizoma Zingiberis Recens. The basic formula was modified according to the specific symptoms of the participants. Mecobalamin tablets of 500 ug were taken orally 3 times per day. All the treatments lasted for 6 weeks

(2) Control: mecobalamin tablets for 6 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not randomisation was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Chen 2009

Methods

RCT

No estimation of sample size

Participants

86 participants were randomised into intervention group (n = 43, M/F 23/20, mean age 56.4 ± 8 years), and control group (n = 43, M/F 24/195, mean age 57.5 ± 7.6 years)
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Tongbi Tang. The basic Tongbi Tang formula contains Radix Astragali seu Hedysari, Radix Angelicae Sinensis, Radix Rehmanniae Preparata, Fructus Lycii, Rhizoma Dioscoreae, Rhizoma Ligustici Chuanxiong, Radix Salviae Miltiorrhizae, Semen Persicae, Scorpio, Lumbricus, Ramulus Cinnamomi, Herba Asari, and Radix Achyranthis Bidentatae

The treatments lasted for 30 days

(2) Control: mecobalamin 500 ug given by intramuscular injection every 2 days. Vitamin B1 tablet 20 mg taken orally 3 times per day for 30 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report. This trial has only one author, which makes proper randomisation and allocation concealment unlikely
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Cheng 2005

Methods

RCT

No estimation of sample size

Participants

70 participants were randomised into intervention group (n = 35, M/F 19/16, mean age 56.3 years, range 34 to 67), and control group (n = 35, M/F 21/14, mean age 53.6 years, range 36 to 71)
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: modified Sishen decoction. The Sishen decoction contains Radix Astragali seu Hedysari, Herba Dendrobii, Radix Polygalae, Flos Lonicerae, Caulis Spatholobi, Rhizoma Ligustici Chuanxiong, and Radix Paeoniae Alba.

The treatments lasted for 1 month

(2) Control: vitamin B1100 mg and vitamin B12 500 ug injected once per day for 1 month

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Ding 2010

Methods

RCT

No estimation of sample size

Participants

62 participants were randomised into intervention group (n = 32, M/F 18/14, mean age 56.12 ± 3.87 years), and control group (n = 30, M/F 16/14, mean age 58.32 ± 2.78 years).
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization criteria

Inclusion criteria: not specified
Exclusion criteria: having serious liver, heart of kidney complications; having another serious primary disease; having had diabetic ketoacidosis or serious infection within 1 month; peripheral neuropathy not due to diabetes

Interventions

(1) Intervention: Tang Mai Yin decoction. The basic Tang Mai Yin contains Radix Astragali seu Hedysari, Caulis Spatholobi, Rhizoma Atractylodis, Radix Salviae Miltiorrhizae, Radix Achyranthis Bidentatae, Radix Angelicae Sinensis, Fructus Liquidambaris, Semen Vaccariae, Flos Carthami, Rhizoma Anemarrhenae, and Scorpio. Herba Asari and Scolopendra were added for participants with serious pain; Radix Aconiti Lateralis Preparata and Rhizoma Zingiberis were added for participants feeling cold in the extremities; Radix Astragali seu Hedysari was added for participants with weakness. The treatments lasted for 3 months

(2) Control: mecobalamin 500 ug taken orally 3 times per day for 3 months

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in tibial nerve (m/s)

Change of sensory nerve conduction velocity in tibial nerve (m/s)

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Gao 2008

Methods

RCT

No estimation of sample size

Participants

62 participants were randomised into intervention group (n = 32, M/F 15/17, mean age 64.93 ± 8.42 years), and control group (n = 30, M/F 12/18, mean age 63.30 ± 9.44 years).
Type of DM: Type 2 DM

Study location: China
Study setting: not reported
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: FBG < 13.9 mmol/L
Exclusion criteria: pregnant or lactating women; liver or kidney dysfunction; cancer; myocardial infarction less than 1 year previously; heart failure, stroke, serious mental illness or neurosis; chronic alcoholism; having severe diabetic complications; ankle hyperreflexia; peripheral neuropathy not due to diabetes

Interventions

(1) Intervention: compound Qiying granule. Compound Qiying granule contains Radix Astragali seu Hedysari, Rhizoma Polygonati, Radix Salviae Miltiorrhizae, Periostracum Cicadae, and Cicer arietinum Linn Fufang. Qiying granule 7.5 g taken orally 3 times a day.

The treatments lasted for 4 weeks

(2) Cobamamide tablets of 0.5 mg taken orally 3 times per day for 4 weeks

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskA random number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Hao 2010

Methods

RCT

No estimation of sample size

Participants

112 participants were randomised into intervention group (n = 60, M/F 38/22, mean age 51.6 ± 3.7 years), and control group (n = 52, M/F 30/22, mean age 50.4 ± 7.1 years).

Type of DM: not reported 

Study location: China

Study setting: outpatients

Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: meeting DM diagnosis criteria; having DPN symptoms; peripheral nerve damage

Exclusion criteria: with other diabetic complications; DPN due to other diseases.

Interventions

(1) Intervention: Xuesaitong tablet and mecobalamin. The main components of Xuesaitong capsule were total saponins of panax notoginseng (the exact amount was not reported). Xuesaitong 50 mg, taken 3 times a day and 0.5 mg mecobalamin orally 3 times per day. Treatment lasted for 3 months

(2) Control: 0.5 mg mecobalamin orally 3 times per day. Treatment lasted for 3 months

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in tibial nerve (m/s)

Change of sensory nerve conduction velocity in tibial nerve (m/s)

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

He 2005

MethodsRCT. No estimation of sample size
Participants

65 participants were randomised into intervention group (n = 33, M/F 15/18, mean age 56.0 ± 7.9 years), and control group (n = 32, M/F 17/15, mean age 55.0 ± 8.0 years)
Type of DM: not reported

Study location: China
Study setting: not reported
Diagnostic criteria: American Diabetes Association 1997 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Xuesaitong capsule and mecobalamin. The main components of Xuesaitong capsule were total saponins of panax notoginseng (the exact amount was not reported). 120 mg Xuesaitong capsule taken orally twice a day. 500 ug mecobalamin taken orally 3 times per day

All the treatments were for 2 months

(2) Control: mecobalamin. Same regimens as above for 2 months

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Hu 2005

MethodsRCT. No estimation of sample size
Participants

48 participants were randomised into intervention group (n = 24, M/F 14/10, mean age 53.2 ± 11.2 years), and control group (n = 24, M/F 16/8, mean age 56.7 ± 10.3 years)
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: International Diabetes Federation (IDF) criteria 1997

Inclusion criteria: not specified
Exclusion criteria: pregnant or lactating women; being allergic to the drugs; heart, liver or kidney dysfunction; having serious mental illness

Interventions

(1) Intervention: Jianpi Yiqi Tongluo Tang. The Jianpi Yiqi Tongluo Tang contains Radix Pseudostellariae, Rhizoma Atractylodis Macrocephalae, Poria, Radix Astragali seu Hedysari, Radix Achyranthis Bidentatae, Rhizoma Dioscoreae, Herba Taxilli, Cortex Eucommiae, Rhizoma Ligustici Chuanxiong, and Radix Glycyrrhizae

The treatments were for 4 weeks

(2) Control: cobamamide injection. Cobamamide 1 mg injected once per day for 4 weeks

Participants in the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Jiao 2007

MethodsRCT. No estimation of sample size
Participants

90 participants were randomised into intervention group 1 (Tongxinluo capsule and mecobalamin) (n = 30, M/F 12/18, mean age 58.1 years, range 40 to 70), intervention group 2 (Tongxinluo capsule) (n = 30, M/F 15/15, mean age 57.3 years, range 44 to 67), and control group (mecobalamin) (n = 30, M/F 13/17, mean age 56.9 years, range 41 to 70)
Type of DM: Type 2 DM

Study location: China
Study setting: not reported
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: stable blood glucose control for 1 month before the trial
Exclusion criteria: having severe diabetic complications; combined with serious heart, liver, kidney, brain, or blood system disease; having mental illness; Achilles tendon hyperreflexia; peripheral neuropathy not related to diabetes

Interventions

(1) Intervention 1: Tongxinluo capsule and mecobalamin. Tongxinluo capsule contains Radix Ginseng, Hirudo, Scorpio, Eupolyphaga Seu Steleophaga, Scolopendra, Periostracum Cicadae, Radix Paeoniae Rubra,Borneolum Syntheticum, and etc. 0.5 mg mecobalamin taken orally 3 times per day

All the treatments were for 8 weeks

(2) Intervention 2: Tongxinluo capsule. Same regimen as above for 8 weeks
(3) Control: mecobalamin. Same regimen as above for 8 weeks

Participants in all the 3 groups accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Li 2001

MethodsRCT. No estimation of sample size
Participants

61 participants were randomised into intervention group (n = 31, no report on number of M/F, and mean age), and control group (n = 30, no report on number of M/F, and mean age)
Type of DM: Type 2 DM

Study location: China
Study setting: not reported
Diagnostic criteria: criteria of American Diabetes Association 1997

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Tongxinluo capsule. Tongxinluo capsule contains Radix Ginseng, Hirudo, Scorpio, Eupolyphaga Seu Steleophaga, Scolopendra, Periostracum Cicadae, Radix Paeoniae Rubra,Borneolum Syntheticum, and etc.

The treatments lasted for 60 days

(2) Control: 500 ug mecobalamin taken orally 3 times per day for 60 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report. This trial has only one author, proper randomisation procedure and allocation concealment unlikely
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasUnclear riskThe comparability of baseline characteristics was not reported

Li 2005

MethodsRCT. No estimation of sample size
Participants

70 participants were randomised into intervention group (n = 36, M/F 20/16, mean age 56.24 ± 5.18 years), and control group (n = 34, M/F 21/13, mean age 55.38 ± 6.16 years)
Type of DM: Type 2 DM

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1997 criteria

Inclusion criteria: not specified
Exclusion criteria: pregnant or lactating women; combined with serious heart, liver, kidney, brain, or blood system disease

Interventions

(1) Intervention: modified Huangqi Guizhi Wuwu Tang. The modified Huangqi Guizhi Wuwu Tang formula contains Radix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Olibanum, Myrrha, Flos Carthami, Caulis Spatholobi, Rhizoma Zingiberis Recens, and Fructus Jujubae

The treatments were for 30 days

(2) Control: vitamin B1 100 mg and vitamin B6 100 mg injected once per day for 30 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Li 2008b

MethodsRCT. No estimation of sample size
Participants

45 participants were randomised into intervention group (n = 24, M/F 17/7, mean age 55.0 ± 3.6 years), and control group (n = 21, M/F 15/6, mean age 54.0 ± 4.8 years)
Type of DM: type 2 DM

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization criteria

Inclusion criteria: not specified
Exclusion criteria: less than 20 or more than 70 years of age; pregnant women; heart, liver, or kidney dysfunction; or combined with other serious diseases

Interventions

(1) Intervention: Tongxinluo capsule and mecobalamin. Tongxinluo capsule contains Radix Ginseng, Hirudo, Scorpio, Eupolyphaga Seu Steleophaga, Scolopendra, Periostracum Cicadae, Radix Paeoniae Rubra, Radix Paeoniae Rubra, Borneolum Syntheticum, Lignum Santali Albi, etc. 500 ug mecobalamin taken orally 3 times per day

All the treatments were for 60 days

(2) Control: mecobalamin. Same regimens as above for 60 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Li 2009

MethodsRCT. No estimation of sample size
Participants

36 participants were randomised into intervention group (n = 18, M/F 11/7, mean age 45.6 ± 12.3 years), and control group (n = 18, M/F 10/8, mean age 44.8 ± 11.9 years)
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: less than 25 years of age or more than 70 years of age; pregnant or lactating women; having acute complications of diabetes; serious liver and kidney damage; and peripheral neuropathy not related to diabetes

Interventions

(1) Intervention: Yiqi Bushen Huoxue Tang, indometacin enteric-coated tablets, and vitamin B1, B12. Yiqi Bushen Huoxue Tang contains Radix Astragali seu Hedysari, Radix Codonopsis, Radix Rehmanniae Recens, Radix Rehmanniae Preparata, Semen Cuscutae, Fructus Corni, Fructus Lycii, Poria, Radix Angelicae Sinensis, Radix Paeoniae Rubra, Radix Paeoniae Alba, Rhizoma Ligustici Chuanxiong, Radix Clematidis, Caulis Spatholobi, and Caulis Trachelospermi. 25 mg indometacin enteric-coated tablets taken orally 3 times per day. 100 mg vitamin B1 and 0.5 mg vitamin B12 injected once per day

All the treatments were for 4 weeks

(2) Control: indometacin enteric-coated tablets and vitamin B1, B12. Same regimens as above for 4 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to the generate allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Li 2010

MethodsRCT. No estimation of sample size
Participants

82 participants were randomised into intervention group (n = 42, M/F  24/18, mean age 54.2 ± 11.2 years), and control group (n = 40, M/F  19/21 , mean age 55.7 ± 10.3 years).

Type of DM: not reported

Study location: China

Study setting: inpatients and outpatients

Diagnostic criteria: International diabetes association in 2005

Inclusion criteria: meeting DM/DPN diagnosis criteria; having sensory/motor neuropathy symptoms; pollex vibration threshold value over 10V; meeting phlegm dampness and blood stasis syndrome 

Exclusion criteria: Pregnant or lactating women; being allergic to the drugs; heart, liver or kidney dysfunction; having serious mental illness; having acute complications; complicated non-diabetes neuropathy; those who are nonadherent to treatment

Interventions

(1) Intervention: Dispelling dampness and dredging collaterals TCM decoction, which contains Rhizoma Atractylodis, Radix Astragali seu Hedysari, Poria, Rhizoma Arisaematis, Radix Saposhnikoviae, Ramulus Cinnamomi, Caulis Spatholobi, Lumbricus, Radix Clematidis, Herba Trachelospermi jasminoides, and Pericarpium Citri Reticulatae. The decoction was taken once a day, in divided doses, 400 ml each.

(2) Control: mecobalamin 0.5 mg taken orally 3 times per day.

Both the treatments were for 6 weeks.

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Palmesthesia threshold

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Lin 2008

MethodsRCT. No estimation of sample size
Participants

79 participants were randomised into intervention group (n = 41, M/F 25/16, mean age 51.9 ± 5.7 years), and control group (n = 38, M/F 23/15, mean age 52.1 ± 5.9 years)
Type of DM: not reported

Study location: China
Study setting:inpatients
Diagnostic criteria: World Health Organization criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Yiqi Huoxue Tongluo compound prescription and mecobalamin. Yiqi Huoxue Tongluo compound prescription contains Radix Rehmanniae Recens, Radix Scrophulariae, Radix Astragali seu Hedysari, Radix Notoginseng, Radix Angelicae Sinensis, Radix Paeoniae Rubra, Radix et Rhizoma Rhei, Exocarpium Citri Rubrum, Rhizoma Arisaematis Cum Bile, Lumbricus, Scorpio, and Radix Puerariae. The basic formula was modified according to the specific symptoms of the participants. 500 ug mecobalamin taken orally once per day

All the treatments were for 4 weeks

(2) Control: mecobalamin. Same regimens as above for 4 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Lin 2008a

MethodsRCT. No estimation of sample size
Participants

60 participants were randomised into intervention group (n = 30, M/F 18/12, mean age 51.2 ± 7.3 years), and control group (n = 30, M/F 17/13, mean age 52.01 ± 6.9 years)
Type of DM: nor reported

Study location: China
Study setting: not reported
Diagnostic criteria: criteria of American Diabetes Association 1997

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Xuefu Zhuyu capsule and mecobalamin. Xuefu Zhuyu capsule contains Semen Persicae, Flos Carthami, Radix Paeoniae Rubra, Rhizoma Ligustici Chuanxiong, Fructus Aurantii, Radix Bupleuri, Radix Platycodonis, Radix Angelicae Sinensis, Rehmannia glutinosa Libosch, Radix Achyranthis Bidentatae, and Radix Glycyrrhizae. 500 ug mecobalamin taken orally 3 times per day

All the treatments were for 2 months

(2) Control: mecobalamin. Same regimens as above for 2 months

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Liu 2001

MethodsRCT. No estimation of sample size
Participants

40 participants were randomised into intervention group (n = 20, M/F was not reported, mean age 50.0 ± 17.1 years), and control group (n = 42, M/F was not reported, mean age 50.7 ± 19.8 years)
Type of DM: Type 1 and type 2 DM

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: criteria of American Diabetes Association 1997

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Aloe. 30 g aloe taken orally everyday for 3 months

(2) Control: no intervention

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Change of motor nerve conduction velocity in median nerve (m/s)

Change of sensory nerve conduction velocity in median nerve (m/s)

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasUnclear riskThe comparability of baseline characteristics was not reported

Liu 2004

MethodsRCT. No estimation of sample size
Participants

65 participants were randomised into intervention group (n = 33, M/F 19/14, mean age 55.9 ± 6.3 years), and control group (n = 32, M/F 17/15, mean age 55.1 ± 7.5 years).
Type of DM: not reported

Study location: China
Study setting: not reported
Diagnostic criteria: criteria of American Diabetes Association 1997

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Tongxinluo capsule and vitamin B Tongxinluo capsule contains Radix Ginseng, Hirudo, Scorpio, Eupolyphaga Seu Steleophaga, Scolopendra, Periostracum Cicadae, Radix Paeoniae Rubra, Radix Paeoniae Rubra, etc. Vitamin B1100 mg and vitamin B12 0.5 mg was injected once a day

All the treatments were for 4 weeks

(2) Control: vitamin B. Same regimens as above for 4 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Liu 2006

MethodsRCT. No estimation of sample size
Participants

110 participants were randomised into intervention group (n = 58, M/F 38/20, mean age 55.62 ± 3.68 years), and control group (n = 52, M/F 31/21, mean age 55.49 ± 5.27 years)
Type of DM: Type 2 DM

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization criteria 1997

Inclusion criteria: not specified
Exclusion criteria: combined with serious heart, liver, kidney, brain, or blood system disease; having mental illness; pregnant or lactating women; being allergic to the drugs

Interventions

(1) Intervention: modified Huangqi Guizhi Wuwu Tang and vitamin B. The basic Buyang Huanwu Tang formula contains Radix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Radix Puerariae, Herba Dendrobii, Radix Salviae Miltiorrhizae, Caulis Spatholobi, Scorpio, Rhizoma Zingiberis Recens, and Fructus Jujubae. Vitamin B1100 mg and vitamin B6 100 mg was injected once per day. All the treatments were for 30 days.

(2) Control: vitamin B. Same regimens as above for 30 days.

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Liu 2008

MethodsRCT. No estimation of sample size
Participants

110 participants were randomised into intervention group (n = 58, M/F 31/27, mean age 55.63 ± 5.95 years), and control group (n = 52, M/F 29/23, mean age 56.08 ± 4.96 years)
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: stable blood glucose control for a month before the trial; 18 to 65 years of age
Exclusion criteria: pregnant or lactating women; being allergic to the drugs; having serious liver, kidney or hematopoietic system disease

Interventions

(1) Intervention: modified Buyang Huanwu Tang and vitamin B. Buyang Huanwu Tang formula contains Radix Astragali seu Hedysari, Semen Persicae, Flos Carthami, Radix Angelicae Sinensis, Radix Paeoniae Rubra, Radix Rehmanniae Preparata, Rhizoma Ligustici Chuanxiong, Lumbricus, Scorpio, Radix Puerariae, and Radix Glycyrrhizae. Vitamin B1 100 mg and vitamin B6 100 mg was injected once per day

All the treatments were for 30 days

(2) Control: vitamin B. Same regimens as above for 30 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Liu 2011

MethodsRCT. No estimation of sample size
Participants

80 participants were randomised into intervention group (n = 40, M/F 21/19 , mean age 56 ± 13 years), and control group (n = 40, M/F 23/17 , mean age 56 ±11 years).

Type of DM: not reported

Study location: China

Study setting: inpatients and outpatients

Diagnostic criteria: American Diabetes Association and Neuropathy Association 1988 San Antonio Meeting

Inclusion criteria: not specified

Exclusion criteria: pregnant women; having acute diabetic complications; having diabatic gangrene, combined with serious heart, liver, kidney dysfunction; having cardiovascular diseases, cancer, or mental illness; peripheral neuropathy not due to diabetes; bad compliance; using other medicine during the observation period

Interventions

(1)  Intervention: modified Huangqi Guizhi Wuwu Tang. The modified Huangqi Guizhi Wuwu Tang contains Radix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Rhizoma Zingiberis Recens, Fructus Jujubae, Radix et Rhizoma Glycyrrhizae, Radix Angelicae Sinensis, Radix Rehmanniae Recens, Rhizoma Ligustici Chuanxiong, Radix Paeoniae Rubra, Rhizoma Anemarrhenae, Flos Carthami, and Semen Persicae. 0.5 mg mecobalamin taken orally 3 times per day.

The treatments lasted for 8 weeks.

(2) Control: 0.5 mg mecobalamin taken orally 3 times per day.

The treatments lasted for 8 weeks.

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise and education.

Outcomes

Global symptom improvement

Imporvement of pain

Improvement of numbness

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Luo 2008

MethodsRCT. No estimation of sample size
Participants

90 participants were randomised into intervention group (n = 48, M/F 29/19, mean age 46.4 ± 8.5 years), and control group (n = 42, M/F 25/17, mean age 45.7 ± 6.5 years)
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Unnamed Chinese herbal prescription. Unnamed Chinese herbal prescription contains Radix Aconiti Preparata, Radix Aconiti Kusnezoffii Preparata, Herba Asari, Radix Angelicae Dahuricae, Lumbricus, Caulis Clematis Armanoii, Radix Astragali seu Hedysari, Semen Persicae, Flos Carthami, Radix Angelicae Sinensis, Rhizoma Ligustici Chuanxiong, and Caulis Spatholobi, modified according to the specific symptoms of the participants

The treatments were for 4 weeks

(2) Control: 1.0 mg mecobalamin given by intravenous infusion once per day for 4 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Improvement in numbness

Improvement in pain

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Luo 2009

MethodsRCT. No estimation of sample size
Participants

74 participants were randomised into intervention group (n = 38, M/F 23/15, mean age 45.6 ± 8.2 years), and control group (n = 36, M/F 22/14, mean age 46.6 ± 7.5 years)
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria in 1999

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Yiyiren Tang and mecobalamin. The basic Yiyiren Tang contains Semen Coicis, Rhizoma Atractylodis, Rhizoma et Radix Notopterygii, Radix Angelicae Pubescentis, Radix Saposhnikoviae, Herba Ephedrae, Ramulus Cinnamomi, Radix Aconiti Preparata, Radix Angelicae Sinensis, Rhizoma Ligustici Chuanxiong, Radix Glycyrrhizae, and Rhizoma Zingiberis Recens. The basic formula modified according to the specific symptoms of the participants. 1.0 mg mecobalamin given by intravenous infusion once per day

All the treatments were for 4 weeks

(2) Control: mecobalamin. Same regimens as above for 4 weeks.

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Ma 2008

MethodsRCT. No estimation of sample size
Participants

100 participants were randomised into intervention group (n = 50, M/F 26/24, mean age 51.2 ± 9.5 years), and control group (n = 50, M/F 27/23, mean age 52.3 ± 10.5 years)
Type of DM: Type 2 DM

Study location: China
Study setting: not reported
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Fufang Danshen injection and mecobalamin. The main components of Fufang Danshen injection were Radix Salviae Miltiorrhizae and Lignum Dalbergiae Odoriferae. Mecobalamin 0.5 mg taken orally 3 times per day. All the treatments were for 6 weeks

(2) Control: mecobalamin. Same regimens as above for 6 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Niu 2008

MethodsRCT. No estimation of sample size
Participants

84 participants were randomised to intervention group (n = 42, sex and mean age were not reported), or control group (n = 42, sex and mean age were not reported)
Type of DM: type 2 DM

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Unnamed Chinese herbal prescription and mecobalamin. Unnamed Chinese herbal prescription contains Radix Astragali seu Hedysari, Radix Rehmanniae Recens, Rhizoma Atractylodis, Radix Scrophulariae, Radix Puerariae, Radix Salviae Miltiorrhizae, Radix Achyranthis Bidentatae, Ramulus Mori, Ramulus Cinnamomi, Scorpio, Olibanum, and Myrrha. 500 ug mecobalamin injected once per day

All the treatments were for 30 days

(2) Control: mecobalamin. Same regimens as above for 30 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Peng 2009

MethodsRCT. No estimation of sample size
Participants

90 participants were randomised to intervention group (n = 45, M/F 25/20, mean age 58 ± 5.2 years), or control group (n = 45, M/F 26/19 and mean age 57 ± 5 years)
Type of DM: not reported

Study location: China
Study setting: inpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Yiqi Tongluo Tang and mecobalamin. Yiqi Tongluo Tang contains Radix Astragali seu Hedysari, Lumbricus, Rhizoma Dioscoreae, Radix Ophiopogonis, Rhizoma Sparganii, Rhizoma Curcumae, Olibanum, and Myrrha. The basic formula was modified according to the specific symptoms of the participants. 500 ug mecobalamin taken orally 3 times per day

All the treatments were for 8 weeks

(2) Control: mecobalamin. Same regimens as above for 8 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Shen 2009

MethodsRCT. No estimation of sample size
Participants

100 participants were randomised into intervention group (n = 50, M/F 21/29, mean age 60 ± 4.2 years), and control group (n = 50, M/F 27/23, mean age 61 ± 5.3 years)
Type of DM: Type 2 DM

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Tangmaining capsule. Tangmaining capsule contains Rhizoma Acori Tatarinowii, Radix Puerariae, Radix Achyranthis Bidentatae, Radix Salviae Miltiorrhizae, Radix Dipsaci, Herba Leonuri, and etc.

The treatments were for 8 weeks

(2) Control: 500 ug mecobalamin taken orally 3 times per day for 8 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Shu 2007

MethodsRCT. No estimation of sample size
Participants

68 participants were randomised into intervention group (n = 34, M/F 16/18, mean age 49.43 ± 7.24 years), and control group (n = 34, M/F 19/15, mean age 51.7 ± 9.27 years)
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: modified Huanqi Guizhi Wuwu Tang. Modified Huanqi Guizhi Wuwu Tang formula contains Radix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Radix Paeoniae Rubra, Radix Angelicae Sinensis, Radix Salviae Miltiorrhizae, Caulis Spatholobi, Scorpio, and Fructus Jujubae. The formula was modified according to the specific symptoms of the participants

The treatments were for 60 days

(2) Control: mecobalamin 500 ug was taken orally 3 times per day for 60 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Sun 2008

MethodsRCT. No estimation of sample size
Participants

84 participants were randomised into intervention group (n = 42, M/F 23/19, age range 44 to 73 years), and control group (n = 42, M/F 22/20, age range 45 to 72 years)
Type of DM: type 2 DM

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Buyang Huanwu Tang and mecobalamin. Buyang Huanwu Tang contains Radix Angelicae Sinensis, Radix Paeoniae Rubra, Radix Astragali seu Hedysari, Rhizoma Ligustici Chuanxiong, Lumbricus, Semen Persicae, and Flos Carthami. The basic formula was modified according to the specific symptoms of the participants. Mecobalamin was injected 0.5 ml once per day

All the treatments were for 3 months

(2) Control: mecobalamin. Same regimens as above for 3 months

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Sun 2009

MethodsRCT. No estimation of sample size
Participants

134 participants were randomised into intervention group (n = 78, M/F 49/29, mean age 53.8 ± 7.8 years), and control group (n = 56, M/F 34/22, mean age 55.2 ± 6.6 years)
Type of DM: not reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: American Diabetes Association criteria in 1997

Inclusion criteria: not specified
Exclusion criteria: combined with serious heart, liver, kidney dysfunction; having other serious diseases such as cancer, stroke, rheumatism, rheumatoid arthritis, or mental illness; taking other drugs that would influence the curative effect

Interventions

(1) Intervention: modified Buyang Huanwu Tang and vitamin B. Modified Buyang Huanwu Tang contains Radix Astragali seu Hedysari, Radix Angelicae Sinensis, Rhizoma Ligustici Chuanxiong, Flos Carthami, Semen Persicae, Radix Paeoniae Rubra, Lumbricus, Radix Rehmanniae Preparata, Radix Cyathulae, and Radix Dipsaci. Vitamin B1 100 mg and vitamin B12 500 ug were injected once per day

All the treatments were for 8 weeks

(2) Control: vitamin B. Same regimens as above for 8 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Sun 2010

MethodsRCT. No estimation of sample size
Participants

86 participants were randomised into intervention group (n = 43, M/F 23/20, mean age 59.30 ± 8.73 years), and control group (n = 43, M/F 21/22, mean age 57.91 ± 8.67 years)
Type of DM: type 2 DM

Study location: China
Study setting: outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: pregnant women; having acute diabetic complications; having diabatic gangrene, combined with serious heart, liver, kidney dysfunction; having cardiovascular diseases, cancer, or mental illness; peripheral neuropathy not due to diabetes

Interventions

(1) Intervention: Furong Tongmai capsule. The Furong Tongmai capsule contains Hirudo, Lumbricus, Scorpio, Radix Puerariae, Radix Scrophulariae, yam, Radix Astragali seu Hedysari, Radix Achyranthis Bidentatae, and Radix Glycyrrhizae

The treatments lasted for 12 weeks

(2) Control: mecobalamin 0.5 mg taken orally 3 times per day for 12 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Sun 2010b

MethodsRCT. No estimation of sample size
Participants

40 participants were randomised into intervention group (n =20, M/F 9 /11, mean age 60.2 ± 7.3 years), and control group (n =  20, M/F 10/10, mean age 58.9 ± 7.7 years).

Type of DM: Type 2 DM

Study location: China

Study setting: inpatients and outpatients

Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: meeting DM/DPN diagnosis criteria; having sensory abnormally (pains, numbness, fever et al), signs of reaction bluntness; MCV<45m/s, SCV<40m/s; TCM syndrome meeting liver-kidney deficiency patterns, qi-blood stasis patterns; age < 70 year; given inform consent

Exclusion criteria: those with peripheral neuropathy due to other diseases; pregnant or lactating women; having systems damages, with psychopathy, chronic infection or diabetic ketoacidosis.

Interventions

(1) Intervention: modified Duhuo Jisheng decoction and Wujia Canger Jiangtang decoction, which contains Herba Taxilli, Radix Angelicae Pubescentis, Radix Rehmanniae Recens, Radix Angelicae Sinensis, Radix Paeoniae Rubra, Ramulus Mori, Caulis Spatholobi, Rhizoma Sparganii Stoloniferi, Rhizoma Sparganii, Radix Gentianae Macrophyllae, Cortex Acanthopanacis Gracilistyli Radicis, Poria, Rhizoma Atractylodis Macrocephalae, Alismatis Rhizoma, Rhizoma Polygonati, Radix Glycyrrhizae, and Radix Notoginseng powder. For heat in lung and stomach, adding gypsum and Rhizoma Anemarrhenae; for qi deficiency, adding Radix Astragali seu Hedysari and Radix Dioscoreae Oppositae; for yin deficiency, adding Radix Polygoni Multiflori, Rhizoma Polygonati Odorati, Herba Dendrobii, and Fructus Armeniaca mume; for yang deficiency in spleen and kidney, adding Cortex Eucommiae, Radix Himalayan Teasel and Ramulus Cinnamomi Cassiae; for dampness, adding Herba Artemisiae (Yinchenhao), Semen Plantaginis, and Radix Sophorae Flavescentis; for blood stasis, adding Radix Salviae Miltiorrhizae and Hirudo. One prescription daily.

(2) Control: mecobalamin 0.5 mg taken orally 3 times per day, vitamin B1 10 mg, 3 times per day.

The treatments course not reported

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Wang 2008

MethodsRCT. No estimation of sample size
Participants

60 participants were randomised into intervention group (n = 30, M/F 17/13, mean age 56.12 ± 4.55 years), and control group (n = 30, M/F 19/11, mean age 59.98 ± 3.12 years)
Type of DM: Type 2 DM

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: stable blood glucose control for 1 month before the trial
Exclusion criteria: combined with serious heart, liver, kidney, brain, or blood system disease; having mental illness; pregnant or lactating women; being allergic to the drugs

Interventions

(1) Intervention: Guizhi Gegen Tang and vitamin B. Buyang Huanwu Tang contains Ramulus Cinnamomi, Radix Paeoniae Alba,Radix Puerariae, Radix Angelicae Sinensis, Scorpio, Rhizoma Zingiberis Recens, Fructus Jujubae, and Radix Glycyrrhizae. Vitamin B1 100 mg and vitamin B6 100 mg was injected once per day

All the treatments were for 30 days

(2) Control: vitamin B. Same regimens as above for 30 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Wu 2003

MethodsRCT. No estimation of sample size
Participants

79 participants were randomised into intervention group (n = 40, M/F not reported, age range 32 to 78 years), and control group (n = 39, M/F not reported, age range 35 to 71 years).
Type of DM: Type 1 and type 2 DM

Study location: China
Study setting: not reported
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Erigeron injection. The main component of Erigeron injection was Erigeron extract. Erigeron injection 40 ml given intravenously by infusion once per day for 4 weeks

(2) Control: vitamin B1 100 mg injected once per day. All the treatments were for 4 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasHigh riskThe comparability of baseline characteristics was not reported

Wu 2006

MethodsRCT. No estimation of sample size
Participants

88 participants were randomised into intervention group (n = 48, M/F 25/23, age range 41 to 69 years), and control group (n = 40, M/F 23/17, age range 40 to 70 years)
Type of DM: type 2 DM.

Study location: China
Study setting: not reported.
Diagnostic criteria: World Health Organization 1997 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Ciwujia injection and vitamin B. The main component of Ciwujia injection was Radix Acanthopanacis Senticosi extract. Ciwujia injection was given by intravenous infusion once per day. Vitamin B1100 mg and vitamin B12 500 ug were injected once daily.

All the treatments were for 30 days

(2) Control: vitamin B. Same regimens as above for 30 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Wu 2010

MethodsRCT. No estimation of sample size
Participants

72 participants were randomised into intervention group (n = 38, M/F 21/17 , mean age 51.6 ± 3.9 years), and control group (n = 34, M/F 20/14, mean age 50.0 ± 4.1 years).

Type of DM: not reported

Study location: China

Study setting: outpatients

Diagnostic criteria: Chinese criteria: Modern Diabetology

Inclusion criteria: DNP, and willing to participate in the trial

Exclusion criteria: having serious diabetic complications ; impaired mobility of the legs due to cancer, osteoarthrosis or rheumatoid arthritis, cerebrovascular disease in acute stage and sequela period; hepatic and kidney dysfunction; peripheral neuropathy not due to diabetes

Interventions

(1)  Intervention: Self prescribed Yangyin Huoxue Tang. The Self prescribed Yangyin Huoxue Tang contains Radix Rehmanniae Recens, Radix Rehmanniae Preparata, Radix Asparagi, Succus Ophiopogonis, Rhizoma Anemarrhenae, Rhizoma Polygonati, Herba Taxilli, Radix Dipsaci, Radix et Rhizoma Salviae Miltiorrhizae, Flos Carthami, Semen Persicae, Radix Angelicae Sinensis, Lumbricus, and Hirudo.

0.5 mg mecobalamin taken orally 3 times per day.

The treatments lasted for 12 weeks.

(2) Control: 0.5 mg mecobalamin taken orally 3 times per day.

The treatments lasted for 12 weeks.

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise and education.

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in tibial nerve (m/s)

Change of sensory nerve conduction velocity in tibial nerve (m/s)

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias [?]

Xie 2008

MethodsRCT. No estimation of sample size
Participants

80 participants were randomised into intervention group (n = 40, M/F 24/16, mean age 59.6 ± 5.7 years), and control group (n = 40, M/F 26/14, mean age 58.9 ± 5.3 years)
Type of DM: not reported

Study location: China
Study setting: inpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Taoren Honghua Jian and mecobalamin. Taoren Honghua Jian contains Radix Salviae Miltiorrhizae, Rhizoma Ligustici Chuanxiong, Semen Persicae, Flos Carthami, prepared Rhizoma Cyperi, Rhizoma Corydalis, Radix Paeoniae Rubra, Pericarpium Citri Reticulatae Viride, Radix Angelicae Sinensis, and Radix Rehmanniae Recens, and was modified according to the specific symptoms of the participants. 500 ug mecobalamin taken orally 3 times per day

All the treatments were for 8 weeks

(2) Control: mecobalamin. Same regimen as above for 8 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Xin 2003

MethodsRCT. No estimation of sample size
Participants

61 participants were randomised into intervention group (n = 31, M/F 17/14, mean age 54.6 ± 9.2 years), and control group (n = 30, M/F 14/16, mean age 53.2 ± 9.9 years)
Type of DM: not reported

Study location: China
Study setting: inpatients
Diagnostic criteria: World Health Organization criteria

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Xuesaitong injection and vitamin B. 2 ml Xuesaitong injection, which contains 200 mg Panax notoginsenosides, was given by intravenous infusion once per day. Vitamin B1 and vitamin B12 were injected, while the exact amount was not specified

All the treatments were for 30 days

(2) Control: vitamin B. Same regimen as above for 30 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Yan 2006

MethodsRCT. No estimation of sample size
Participants

70 participants were randomised into intervention group (n = 35, M/F 19/16, mean age 56.32 ± 3.18 years), and control group (n = 35, M/F 20/15, mean age 56.79 ± 3.42 years)
Type of DM: Type 2 DM

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: stable blood glucose control for 1 month before the trial
Exclusion criteria: combined with serious heart, liver, kidney, or blood system disease; pregnant or lactating women; being allergic to drugs

Interventions

(1) Intervention: modified Danggui Sini Tang and vitamin B. Modified Danggui Sini Tang formula contains Radix Angelicae Sinensis, Ramulus Cinnamomi, Medulla Tetrapanacis, Radix Salviae Miltiorrhizae, Radix Paeoniae Alba, Herba Asari, Ramulus Euonymi, Caulis Spatholobi, Scorpio, Radix Glycyrrhizae, and Fructus Jujubae. Vitamin B12 500 ug injected twice a week. Vitamin B1 100 mg and vitamin B6 100 mg injected once per day

All the treatments were for 30 days

(2) Control: vitamin B. Same regimen as above for 30 days

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Yan 2008

MethodsRCT. No estimation of sample size
Participants

70 participants were randomised into intervention group (n = 40, M/F 34/6, mean age 52.6 years, range 35 to 75), and control group (n = 30, M/F 26/4, mean age 54.5 years, range 35 to 71)
Type of DM: not reported

Study location: China
Study setting: not reported
Diagnostic criteria: World Health Organization criteria (year not reported)

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Ginkgo biloba extract and vitamin B. Ginkgo biloba extract injection 20 ml was given intravenously once per day. Vitamin B1 100 mg and vitamin B12 500 ug was injected once a day

All the treatments were for 4 weeks

(2) Control: vitamin B. Same regimens as above for 4 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskDrawing lots was used to generate allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear due to the inaccessibility to the trial protocol
Other biasLow riskNo evidence of other bias

Yang 2011

MethodsRCT. No estimation of sample size
Participants

60 participants were randomised into intervention group (n = 30, M/F 17/13  , mean age 58.30 ± 8.33 years), and control group (n = 30 , M/F 16/14, mean age 58.35 ± 8.72 years).

Type of DM: not reported

Study location: China

Study setting: inpatients and outpatients

Diagnostic criteria: Shanghai National Diabetes Chronic Complications Meeting in 1993, and Practical Neurologies

Inclusion criteria: with Qi-yin deficiency syndrome, or static blood blocking collaterals syndrome according to TCM syndrome differentiation criteria; 45-70 years of age; stable blood glucose control  by receiving diabetes comprehensive treatment; signed Inform Consent

Exclusion criteria: Pregnant or lactating women; being allergic to the drugs; heart, liver or kidney dysfunction; having serious mental illness; having acute complications; complicated non-diabetes neuropathy; with diabetes foot I grade above.

Interventions

(1) Intervention: Qingxiao Tangbi prescription. The prescription contains Radix Astragali seu Hedysari, Radix Codonopsis, Radix Ophiopogonis, Rhizoma Anemarrhenae, Radix Paeoniae Alba, Semen Persicae, Flos Carthami, Radix Salviae Miltiorrhizae, Radix Mongolian Snakegourd. For qi deficiency, adding Radix Codonopsis; for blood heat due to yin-deficiency, adding Radix Rehmanniae Recens; for blood stasis, adding Flos Carthami and Semen Persicae;for pain, adding Rhizoma Curcumae Longae; for disease in upper limbs, adding Ramulus Mori; for disease in lower limbs, adding Radix Cyathulae and Cydonia sinensis Thouin.

And 0.5 mg mecobalamin was taken orally 3 times per day for 8 weeks

(2) Control: 0.5 mg mecobalamin was taken orally 3 times per day for 8 weeks

The treatments lasted for 8 weeks

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Using Toronto clinical neuropathy scoring system (TCSS):

  • change of motor nerve conduction velocity in tibial nerve (m/s);

  • change of sensory nerve conduction velocity in tibial nerve (m/s);

  • change of motor nerve conduction velocity in common peroneal nerve (m/s);

  • change of sensory nerve conduction velocity in common peroneal nerve (m/s).

Adverse events

Notes

Sources of funding not stated

Dissertation for Master's degree

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskComputer was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Yao 2010

MethodsRCT. No estimation of sample size
Participants

100 participants were randomised into intervention group (n = 50, M/F  22/28, mean age 49 ± 9 years), and control group (n = 50, M/F 27/23, mean age 50 ±11 years).

Type of DM: not reported

Study location: China

Study setting: inpatients and outpatients

Diagnostic criteria: World Health Organization criteria from 1980 to 1985, and International diabetes association

Inclusion criteria: not specified

Exclusion criteria: age > 70 years; having other diabetic complications; having cancer, heart failure, cerebrovascular acute diseases, kidney or liver failure; peripheral neuropathy due to other diseases

Interventions

(1) Intervention: self prescribed tonifying kidney and activating blood formula, which contains Radix Rehmanniae Recens, Rhizoma Anemarrhenae, Cortex Eucommiae, Herba Taxilli, Radix Cyathulae, Radix Himalayan Teasel, Fructus Psoraleae, Radix Salviae Miltiorrhizae, Flos Carthami, Semen Persicae, Radix Angelicae Sinensis, Radix Ligustici Chuanxiong, Lumbricus, and Fructus Liquidambaris. Taken in divided dose, each 100 ml. Treatment lasted for 6 weeks.

(2) Control: 0.5 mg mecobalamin taken orally 3 times per day for 6 weeks

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Change of motor nerve conduction velocity in tibial nerve (m/s)

Change of sensory nerve conduction velocity in tibial nerve (m/s)

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Yin 2011

MethodsRCT. No estimation of sample size
Participants

43 participants were randomised into intervention group (n = 22) and control group (n = 21), but only 40 participants completed the trial and were analysed. (n = 20, M/F 13/7, mean age 58.4 ± 10.67  years in intervention group, and n = 20, M/F 9/11, mean age 61.7 ± 4.92 years in the control group).

Type of DM: not reported

Study location: China

Study setting: outpatients

Diagnostic criteria: World Health Organization criteria in 1999, Chinese criteria: Diagnosis and Effectiveness Standard of Traditional Chinese Medicine(1994), Standard of Syndrome Differentiation of TCM, Guiding Principle for New Drug Clinical Research(2002), Guide to Prevention and Treatment of Diabetes in China(2008)

Inclusion criteria: not specified

Exclusion criteria: not specified

Interventions

(1) Intervention: Tongluo Tangtai Decoction. The Tongluo Tangtai Decoction contains Radix Astragali seu Hedysari, Rhizoma Dioscoreae, Radix Scrophulariae, Radix Ophiopogonis, Radix et Rhizoma Salviae Miltiorrhizae, Rhizoma Chuanxiong, Caulis Spatholobi, Radix Puerariae, Gypsum Fibrosum, Rhizoma Anemarrhenae, Rhizoma Atractylodis, Herba Agastachis, Semen Sinapis, Hirudo, and Rhizoma Corydalis.

The treatments lasted for 8 weeks.

(2) Control: 0.5 mg mecobalamin was taken orally 3 times per day.

The treatments lasted for 8 weeks.

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise and education.

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in tibial nerve (m/s)

Change of sensory nerve conduction velocity in tibial nerve (m/s)

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High risk2/22 missing from the Intervention group, one due to home moving, and reason of the other was not reported. 1/21 withdrawing from the control group due to respiratory tract infection and did not complete the trial. Data of the 40 participants who complete the trial was analysed. Intention-to-treat analysis was not conducted
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasHigh riskNo evidence of other bias

Zhang 2007

MethodsRCT. No estimation of sample size
Participants

60 participants were randomised into intervention group (n = 30, M/F 13/17, mean age 56.81 ± 9.36 years), and control group (n = 30, M/F 16/14, mean age 55.64 ± 9.73 years)
Type of DM: nor reported

Study location: China
Study setting: inpatients and outpatients
Diagnostic criteria: American Diabetes Association criteria in 1997

Inclusion criteria: not specified
Exclusion criteria: not specified

Interventions

(1) Intervention: Yiqi Huoxue Wenyang prescription. Yiqi Huoxue Wenyang prescription contains Radix Ginseng, Radix Astragali seu Hedysari, Semen Persicae, Flos Carthami, Radix Paeoniae Rubra, Radix Angelicae Sinensis, Olibanum, Myrrha, Lumbricus, Ramulus Cinnamomi, prepared Radix Aconiti Lateralis Preparata, Herba Ephedrae, Herba Asari, and Radix Achyranthis Bidentatae. The treatment was for 8 weeks

(2) Control: 50 mg mecobalamin taken orally 3 times per day for 8 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment.
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Zhao 2008

MethodsRCT. No estimation of sample size
Participants

87 participants were randomised into intervention group (n = 31, M/F 15/16, mean age 58.3 ± 5.4 years), control group (n = 30, M/F 16/14, mean age 57.3 ± 4.0 years), and placebo group (n = 26, M/F 14/12, mean age 56.8 ± 4.1 years)
Type of DM: Type 2 DM

Study location: China
Study setting: not reported
Diagnostic criteria: World Health Organization 1999 criteria

Inclusion criteria: 20 to 65 years of age; stable blood glucose control (fasting blood glucose 5.0 to 7.0 mmol/L, 2-hr postprandial plasma glucose 7.0 to 10.0mmol/L) for 1 month before the trial; suspended other drugs for DPN.
Exclusion criteria: pregnant or lactating women; combined with serious heart, liver, kidney or blood system disease; peripheral neuropathy not due to diabetes

Interventions

(1) Intervention: Tangluoan capsule. Tangluoan capsule contains Radix Astragali seu Hedysari, Radix Rehmanniae Recens, Herba Eupatorii, Rhizoma Coptidis, Scorpio, Rhizoma Ligustici Chuanxiong, Fructus Lycii, etc

for 3 months

(2) Control: mecobalamin 500 ug taken orally 3 times per day for 3 months

(3) Placebo: placebo which resembled the Tangluoan capsule taken orally 3 times per day for 3 months

Participants in all the 3 groups accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Change of motor nerve conduction velocity in median nerve (m/s)

Change of sensory nerve conduction velocity in median nerve (m/s)

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Zhou 2006

MethodsRCT. No estimation of sample size
Participants

44 participants were randomised into intervention group (n = 22, M/F 6/18, mean age 45.2 years, range 42 to 66), and control group (n = 22, M/F 10/12, mean age 44.9 years, range 40 to 70)
Type of DM: not reported

Study location: China
Study setting: outpatients
Diagnostic criteria: criteria of American Diabetes Association in 1997

Inclusion criteria: 42 to 66 years of age; did not use nerve-nourishing drugs for 1 month before the trial
Exclusion criteria: unstable blood glucose control; peripheral neuropathy not due to diabetes; combined with serious liver, kidney, brain disease

Interventions

(1) Intervention: modified Huangqi Guizhi Wuwu Tang and mecobalamin. The basic Huangqi Guizhi Wuwu decoction formula contains Radix Astragali seu Hedysari, Ramulus Cinnamomi, Radix Paeoniae Alba, Rhizoma Zingiberis Recens, Fructus Jujubae, Semen Sinapis Albae, Hirudo, Bombyx Batryticatus, Radix Puerariae, and Radix et Caulis Acanthopanacis Senticosi. Mecobalamin 0.5 mg taken orally 3 times per day

All the treatments were for 8 weeks

(2) Control: mecobalamin. Same regimen as above for 8 weeks

Participants in both the intervention group and control group accepted hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise

Outcomes

Global symptom improvement

Adverse events

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Zhou 2010a

  1. a

    DM: diabetes mellitus

MethodsRCT. No estimation of sample size
Participants

98 participants were randomised into intervention group (n = 52, M/F 33/19 , mean age 41.5 ± 68  years), and control group (n = 46, M/F 30/16 , mean age 42.5 ± 67 years).

Type of DM: not reported

Study location: China

Study setting: inpatients and outpatients

Diagnostic criteria: Chinese criteria of Western medicine and Chinese medicine: Integrative Medicine for the Diagnosis and Treatment of Diabetes and the Complications, Integrative Medicine for Diabetes

Inclusion criteria: not specified

Exclusion criteria: DPN caused by other disease

Interventions

(1) Intervention: Yiqi Huayu Tongbi Tang. The Yiqi Huayu Tongbi Tang contains Radix Astragali seu Hedysari, Caulis Spatholobi, Radix Puerariae, Radix Angelicae Sinensis, Rhizoma Ligustici Chuanxiong, Hirudo, Eupolyphaga Seu Steleophaga, Radix Pseudostellariae, Rhizoma Polygonati, Radix Clematidis, Herba Lycopodii, Ramulus Mori, Ramulus Cinnamomi, Radix Cyathulae, and Fructus Chaenomelis. The prescription was modified according to the symptoms of participants.

The treatments lasted for 4 weeks.

(2) Control: mecobalamin 0.5 mg taken orally 3 times per day, or injected 0.5 mg once daily, 3 times per week.

The treatments lasted for 4 weeks.

Participants in both the intervention group and control group received hypoglycaemic therapy, which included oral hypoglycaemic drug, insulin, exercise and education.

Outcomes

Global symptom improvement

Change of motor nerve conduction velocity in tibial nerve (m/s)

Change of sensory nerve conduction velocity in tibial nerve (m/s)

Change of motor nerve conduction velocity in common peroneal nerve (m/s)

Change of sensory nerve conduction velocity in common peroneal nerve (m/s)

Adverse events: not reported

NotesSources of funding not stated
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandom number table was used to generate the allocation sequence, but insufficient information was provided to judge whether or not it was conducted properly
Allocation concealment (selection bias)High riskNot mentioned in the report.This trial has only one author, which is impossible for a RCT to be done properly in terms of randomisation procedure and the allocation concealment
Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants or investigators
Incomplete outcome data (attrition bias)
All outcomes
High riskNot mentioned in the report
Selective reporting (reporting bias)Unclear riskUnclear as the trial protocol was not accessible
Other biasLow riskNo evidence of other bias

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Chrubasik 2002A commentary, not a RCT
He 2005aData were not available for analysis due to inadequate reporting
Incandela 2001Not diabetic peripheral neuropathy
Jamal 1990The intervention was not Chinese herbal medicine
Lane 2006Not a RCT
Leonard 2004The intervention was not Chinese herbal medicine
Li 1999The randomisation procedure was not reported
Li 2009aThe randomisation procedure described was not correct
Ren 2000The randomisation procedure was not reported
Sakamoto 1987Not diabetic peripheral neuropathy
Shu 2011Thought to be a duplicate of another study by the same author owing to a high degree of similarity of the content
Tawata 1994Not a RCT
Tong 2006Not a RCT
Wang 2012Intervention in the control was not specified
Yang 2011aWithout definite diagnostic criteria.
Zhang 2011Unclear description of the compared medicine
Zhou 2007Two interventions were used in the intervention group
Zhou 2010Thought to be a duplicate of Liu 2008 due to high degree of similarity of the content
Zhou 2010bThe randomisation procedure was not correct

Ancillary