Echinacea for preventing and treating the common cold

  • Conclusions changed
  • Review
  • Intervention

Authors


Abstract

Background

Echinacea plant preparations (family Asteraceae) are widely used in Europe and North America for common colds. Most consumers and physicians are not aware that products available under the term Echinacea differ appreciably in their composition, mainly due to the use of variable plant material, extraction methods and the addition of other components.

Objectives

To assess whether there is evidence that Echinacea preparations are effective and safe compared to placebo in the prevention and treatment of the common cold.

Search methods

We searched CENTRAL 2013, Issue 5, MEDLINE (1946 to May week 5, 2013), EMBASE (1991 to June 2013), CINAHL (1981 to June 2013), AMED (1985 to February 2012), LILACS (1981 to June 2013), Web of Science (1955 to June 2013), CAMBASE (no time limits), the Centre for Complementary Medicine Research (1988 to September 2007), WHO ICTRP and clinicaltrials.gov (last searched 5 June 2013), screened references and asked experts in the field about published and unpublished studies.

Selection criteria

Randomized controlled trials (RCTs) comparing mono-preparations of Echinacea with placebo.

Data collection and analysis

At least two review authors independently assessed eligibility and trial quality and extracted data. The primary efficacy outcome was the number of individuals with at least one cold in prevention trials and the duration of colds in treatment trials. For all included trials the primary safety and acceptability outcome was the number of participants dropping out due to adverse events. We assessed trial quality using the Cochrane 'Risk of bias' tool.

Main results

Twenty-four double-blind trials with 4631 participants including a total of 33 comparisons of Echinacea preparations and placebo met the inclusion criteria. A variety of different Echinacea preparations based on different species and parts of plant were used. Evidence from seven trials was available for preparations based on the aerial parts of Echinacea purpurea.

Ten trials were considered to have a low risk of bias, six to have an unclear risk of bias and eight to have a high risk of bias. Ten trials with 13 comparisons investigated prevention and 15 trials with 20 comparisons investigated treatment of colds (one trial addressed both prevention and treatment).

Due to the strong clinical heterogeneity of the studies we refrained from pooling for the main analysis. None of the 12 prevention comparisons reporting the number of patients with at least one cold episode found a statistically significant difference. However a post hoc pooling of their results, suggests a relative risk reduction of 10% to 20%. Of the six treatment trials reporting data on the duration of colds, only two showed a significant effect of Echinacea over placebo. The number of patients dropping out or reporting adverse effects did not differ significantly between treatment and control groups in prevention and treatment trials. However, in prevention trials there was a trend towards a larger number of patients dropping out due to adverse events in the treatment groups.

Authors' conclusions

Echinacea products have not here been shown to provide benefits for treating colds, although, it is possible there is a weak benefit from some Echinacea products: the results of individual prophylaxis trials consistently show positive (if non-significant) trends, although potential effects are of questionable clinical relevance.

Résumé scientifique

Echinacea pour la prévention et le traitement du rhume banal

Contexte

Les préparations à base de la plante Echinacea (famille Asteraceae) sont largement utilisées en Europe et en Amérique du Nord pour le rhume banal. La plupart des consommateurs et des médecins ne sont pas conscients que les produits disponibles sous l'appellation Echinacea sont sensiblement différents dans leur composition, principalement en raison de l'utilisation de matières végétales et de méthodes d'extraction différentes, et de l'ajout d'autres composants.

Objectifs

Évaluer l'existence de preuves indiquant que les préparations à base d'Echinacea sont efficaces et sûres par rapport à un placebo dans la prévention et le traitement du rhume banal.

Stratégie de recherche documentaire

Nous avons effectué des recherches dans CENTRAL 2013, numéro 5, MEDLINE (de 1946 à la 5ème semaine de mai 2013), EMBASE (de 1991 à juin 2013), CINAHL (de 1981 à juin 2013), AMED (de 1985 à février 2012), LILACS (de 1981 à juin 2013), Web of Science (de 1955 à juin 2013), CAMBASE (sans limite de temps), le Centre for Complementary Medicine Research (de 1988 à septembre 2007), l'ICTRP de l'OMS et clinicaltrials.gov(dernière recherche le 5 juin 2013), passé au crible les références bibliographiques et interrogé des experts du domaine sur les études publiées et non publiées.

Critères de sélection

Essais contrôlés randomisés (ECR) comparant les monopréparations à base d' Echinacea à un placebo.

Recueil et analyse des données

Au moins deux auteurs de la revue ont indépendamment évalué l'éligibilité et la qualité des essais et extrait les données. Le principal critère de jugement de l'efficacité était le nombre d'individus présentant au moins un rhume dans les essais de prévention et la durée des rhumes dans les essais de traitement. Pour tous les essais inclus, le principal critère de jugement de la sécurité et de l'acceptabilité était le nombre de participants abandonnant en raison d'événements indésirables. Nous avons évalué la qualité des essais en utilisant l'outil Cochrane « Risque de biais ».

Résultats principaux

Vingt-quatre essais en double aveugle portant sur 4631 participants, comprenant un total de 33 comparaisons de préparations à base d'Echinacea au placebo, remplissaient les critères d'inclusion. Un éventail de différentes préparations à base d'Echinacea fabriquées à partir de différentes espèces et parties de la plante ont été utilisées. Des preuves issues de sept essais étaient disponibles pour les préparations à base des parties aériennes de l'Echinacea purpurea.

Dix essais ont été considérés comme ayant un faible risque de biais, six avaient un risque de biais incertain et huit un risque élevé de biais. Dix essais incluant 13 comparaisons examinaient la prévention et 15 essais avec 20 comparaisons examinaient le traitement des rhumes (un essai portait sur la prévention et le traitement).

En raison de la forte hétérogénéité clinique des études, nous ne les avons pas regroupées pour l'analyse principale. Aucun des 12 comparaisons de prévention rapportant le nombre de patients présentant au moins un épisode de rhume n'a trouvé une différence statistiquement significative. Cependant, un regroupement post hoc de leurs résultats suggère une réduction du risque relatif de 10 % à 20 %. Sur les six essais de traitement rapportant des données concernant la durée des rhumes, seuls deux ont montré un effet significatif de l'Echinacea par rapport au placebo. Le nombre de patients abandonnant ou signalant des effets indésirables ne différait pas significativement entre les groupes de traitement et de contrôle dans les essais de prévention et de traitement. Cependant, dans les essais de prévention, il y avait une tendance à un plus grand nombre de patients abandonnant en raison d'événements indésirables dans les groupes de traitement.

Conclusions des auteurs

Il n'a pas été démontré ici que les produits à base d'Echinacea apportent un avantage dans le traitement du rhume, bien qu'il soit possible que certains produits à base d'Echinacea aient apporté un faible bénéfice : les résultats des essais individuels portant sur la prophylaxie montrent de façon constante des tendances positives (sinon significatives), bien que les effets potentiels soient d'une pertinence clinique discutable.

アブストラクト

風邪の予防と治療におけるエキナセア

背景

エキナセアはキク科の植物製剤で欧州や北米では風邪に広く使用されている。 ほとんどの消費者や医師は、エキナセアという名称で販売されている製剤はその組成が実にさまざまであることを知らないが、これは主に、さまざまな植物原料と抽出方法を用い、さらに他の成分も添加しているためである。

目的

風邪の予防と治療に対して、エキナセア製剤がプラセボと比較して効果的かつ安全であるとするエビデンスの有無を評価すること。

検索戦略

CENTRAL 2013年5号、MEDLINE(1946年~2013年5月第5週)、EMBASE(1991年~2013年6月)、CINAHL(1981年~2013年6月)、AMED(1985年~2012年2月)、LILACS(1981年~2013年6月)、Web of Science(1955年~2013年6月)、CAMBASE(期限なし)、the Centre for Complementary Medicine Research(1988年~2007年9月)、WHO ICTRP、およびclinicaltrials.gov (最終検索:2013年6月5日)を検索し、参考文献リストを調べ、当該分野の専門家に発表済みおよび未発表の研究について尋ねた。

選択基準

エキナセア単剤とプラセボを比較したランダム化比較試験(RCT)

データ収集と分析

2名以上のレビュー著者が独立して適格性と試験の質を評価し、データを抽出した。 主要有効性アウトカムは、予防に関する試験中に1回以上風邪を引いた参加者数、および治療に関する試験中に風邪を引いていた期間とした。 調査したすべての試験で、主要安全性・許容性アウトカムは、有害事象で脱落した参加者数とした。 コクランのバイアスのリスクツールを用いて試験の質を評価した。

主な結果

4631例を対象とした24件の二重盲検試験が、エキナセア製剤とプラセボについて総計33の比較検討を実施しており、選択基準を満たすものであった。 種や部位が異なるさまざまなエキナセア製剤が使用されていた。 7件の試験では、Echinacea purpureaの地上部を用いた製剤に関するエビデンスが得られた。

バイアスのリスクについて、10件の試験は低リスク、6件はリスク不明、8件は高リスクと判断された。 10件の試験では13の比較検討で予防について調査しており、15件の試験では20の比較検討で風邪の治療について調査していた(1件の試験では予防と治療の両方について調べていた)。

研究の臨床的な異質性が大きいため、主解析のための統合は行わなかった。 1回以上風邪を引いた患者数を報告した予防に関する12の比較検討では、統計学的な有意差はなかった。 しかし、これらの結果に関する事後統合では10%~20%の相対リスク減少を示している。 風邪の期間に関するデータを報告した6件の治療に関する試験のうち、2件のみでプラセボを上回るエキナセアの有意な効果を示した。 予防および治療に関する試験では、脱落した患者数または有害作用を報告した患者数について、治療群とコントロール群に有意差はみられなかった。 しかし、予防に関する試験では、治療群でより多くの患者が有害事象により脱落する傾向がみられた。

著者の結論

エキナセア製剤について風邪の治療に有効であることは示されなかったが、エキナセア製剤の一部にはわずかな効果をもたらす可能性がある。 個々の予防投与試験の結果では、一貫して好ましい傾向を示しているが(有意差なしの場合)、効果に関する臨床的意義については疑問が残る。

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/)[2015.12.20]  《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

Echinacea for preventing and treating the common cold

Preparations of the plant Echinacea are widely used in some European countries and in North America for common colds. Echinacea preparations available on the market differ greatly as different types (species) and parts (herb, root or both) of the plant are used, different manufacturing methods (drying, alcoholic extraction or pressing out the juice from fresh plants) are used and sometimes also other herbs are added.

We reviewed 24 controlled clinical trials with 4631 participants investigating the effectiveness of several different Echinacea preparations for preventing and treating common colds or induced rhinovirus infections. Our review shows that a variety of products prepared from different Echinacea species, different plant parts and in a different form have been compared to placebo in randomized trials. Due to the significant differences in the preparations tested, it was difficult to draw strong conclusions. Five trials were rated as having a low risk of bias in all five categories of the Cochrane 'Risk of bias' tool. Five more trials were rated as low risk of bias, having an unclear risk of bias in only one category. Eight trials were rated as having a high risk of bias in at least one category and the remaining six as having an unclear risk of bias.

The majority of trials investigated whether taking Echinacea preparations after the onset of cold symptoms shortens the duration, compared with placebo. Although it seems possible that some Echinacea products are more effective than a placebo for treating colds, the overall evidence for clinically relevant treatment effects is weak. In general, trials investigating Echinacea for preventing colds did not show statistically significant reductions in illness occurrence. However, nearly all prevention trials pointed in the direction of small preventive effects. The number of patients dropping out or reporting adverse effects did not differ significantly between treatment and control groups in prevention and treatment trials. However, in prevention trials there was a trend towards a larger number of patients dropping out due to adverse events in the treatment groups.

The evidence is current to July 2013.

Laički sažetak

Ehinaceja za sprječavanje i liječenje prehlade

Preparati biljke Echinacea naveliko se koriste za prehladu u nekim europskim zemljama i u Sjevernoj Americi za liječenje prehlade. Echinacea preparati nalaze se na tržištu kao velik broj različitih pripravaka, ali ti se pripravci znatno razlikuju po tipu (vrsti), dijelovima biljke (stabljika, korijen ili oboje) i metodama proizvodnje (sušenje, alkoholna ekstrakcija ili prešanje soka iz svježih biljaka), a ponekad su pripravcima dodane i druge biljke.

Cochrane sustavni pregled analizirao je 24 kontrolirane kliničke studije s ukupno 4631 ispitanika u kojima je istražena učinkovitost nekoliko različitih pripravaka ehinaceje za sprječavanje i liječenje prehlade il infekcija rhinovirusima. Sustavni pregled literature pokazuje da je niz proizvoda koji se temelje na različitim vrstama ehinaceje, različitim dijelovima te biljke i u različitom obliku u pokusima uspoređeno s placebom. Zbog velikih razlika u pripravcima koji su korišteni u istraživanjima nije moguće donijeti čvrste zaključke. Kvaliteta studija značajno se razlikovala – samo je 5 studija ocijenjeno kao visokokvalitetno. Još pet studija ocijenjeno je niskim rizikom od pristranosti, jer su imali nejasan rizik od pristranosti samo u jednoj od pet kategorija za ocjenjivanje pristranosti. Osam studija je procijenjeno visokim rizikom od pristranosti.

Većina studija je istražila da li uzimanje Echinacea preparata nakon početka prehlade može skratiti trajanje simptoma u usporedbi s placebom. Iako se može činiti moguće da su neki proizvodi ehinaceje učinkovitiji nego placebo u liječenju prehlade, ukupni dokazi za klinički važan učinak liječenja prehlade ehinacejom su slabi. Istraživanja u kojima je proučavano može li ehinaceja smanjiti prehlade nisu pokazala statistički značajno smanjenje pojave bolesti. U studijama je bilo naznaka malog preventivnog učinka. Broj ispitanika koji je odustao od istraživanja nije se razlikovao između eksperimentalnih i pokusnih skupina u prevencijskim i terapijskim studijama. Međutim, u studijama koje su istraživale preventivni učinak ehinaceje uočen je veći broj ispitanika koji su odustajali od istraživanja zbog nuspojava u skupini koja je uzimala ehinaceju.

Dokazi se temelje na literaturi objavljenoj do srpnja 2013.

Bilješke prijevoda

Hrvatski Cochrane
Ovaj sažetak preveden je u okviru volonterskog projekta prevođenja Cochrane sažetaka. Uključite se u projekt i pomozite nam u prevođenju brojnih preostalih Cochrane sažetaka koji su još uvijek dostupni samo na engleskom jeziku. Kontakt: cochrane_croatia@mefst.hr

Résumé simplifié

Echinacea pour la prévention et le traitement du rhume banal

Les préparations à base de la plante Echinacea sont largement utilisées dans certains pays européens et en Amérique du Nord pour le rhume banal. Les préparations à base d'Echinacea disponibles sur le marché sont très différentes les unes des autres, car elles utilisent différents types (espèces) et parties (tige, racine ou les deux) de la plante, différentes méthodes de fabrication (séchage, extraction alcoolique ou expression du jus à partir de plantes fraîches) et parfois aussi l'ajout d'autres plantes.

Nous avons examiné 24 essais cliniques contrôlés portant sur 4631 participants étudiant l'efficacité de plusieurs préparations différentes à base d'Echinacea pour la prévention et le traitement du rhume banal ou d'infections à rhinovirus induites. Notre revue montre que toute une variété de produits préparés à partir de différentes espèces du genre Echinacea, de différentes parties de la plante et disponibles sous des formes différentes ont été comparés à un placebo dans des essais randomisés. En raison des différences significatives dans les préparations testées, il était difficile de tirer des conclusions solides. Cinq essais ont été considérés comme présentant un faible risque de biais dans les cinq catégories de l'outil Cochrane « Risque de biais ». Cinq autres essais ont été considérés comme étant à faible risque de biais, avec un risque de biais incertain dans seulement une catégorie. Enfin, huit essais ont été considérés à risque élevé de biais dans au moins une catégorie et les six autres comme présentant un risque de biais incertain.

La majorité des essais examinaient si la prise de préparations à base d'Echinacea après l'apparition des symptômes du rhume en réduit la durée par rapport à un placebo. Bien qu'il semble possible que certains produits à base d'Echinacea soient plus efficaces qu'un placebo pour le traitement du rhume, dans l'ensemble, les preuves sur les effets thérapeutiques cliniquement pertinents sont faibles. En général, les essais étudiant l'Echinacea dans la prévention du rhume n'ont pas montré de réduction statistiquement significative dans l'incidence de la maladie. Cependant, presque tous les essais de prévention montraient des signes de petits effets préventifs. Le nombre de patients abandonnant le traitement ou signalant des effets indésirables ne différait pas significativement entre les groupes de traitement et de contrôle dans les essais de prévention et de traitement. Cependant, dans les essais de prévention, il y avait une tendance à un plus grand nombre de patients abandonnant en raison d'événements indésirables dans les groupes de traitement.

Les preuves sont à jour en juillet 2013.

Notes de traduction

Traduit par: French Cochrane Centre 6th July, 2014
Traduction financée par: Financeurs pour le Canada : Instituts de Recherche en Santé du Canada, Ministère de la Santé et des Services Sociaux du Québec, Fonds de recherche du Québec-Santé et Institut National d'Excellence en Santé et en Services Sociaux; pour la France : Ministère en charge de la Santé

Резюме на простом языке

Эхинацея для профилактики и лечения простуды

Препараты из растения Эхинацея широко используются при простуде в некоторых европейских странах и в Северной Америке. Эхинацеи препараты, имеющиеся на рынке, существенно различаются как по использованию её разных типов (видов), частей растения (трава, корни или и то, и другое), так и по использованию различных методов производства (высушивание, алкогольная экстракция или выжимка сока из свежих растений), а иногда и добавляются другие травы.

Мы проанализировали 24 контролируемых клинических испытания с 4631 участниками по изучению эффективности различных Эхинацеи для профилактики и лечения простудных заболеваний или заболеваний, индуцированных риновирусными инфекциями. Наш обзор показывает, что целый ряд продуктов, полученных из различных видов Эхинацеи, различных частей растения и разные лекарственные формы были сравнены с плацебо в рандомизированных испытаниях. Из-за существенных различий в протестированных препаратах было трудно сделать обоснованные выводы. Пять испытаний были оценены как имеющие низкий риск смещения по всем пяти категориям Кокрейновского инструмента «Риск смещения». Еще пять испытаний были оценены как испытания с низким риском смещения, имея неяcный риск смещения только в одной категории. Восемь испытаний были оценены как имеющие высокий риск смещения, по меньшей мере, в одной категории, а остальные шесть –как имеющие неяcный риск смещения.

Большинство испытаний изучали, может ли прием препаратов Эхинацеи после появления симптомов простуды, сокращать их продолжительность, по сравнению с плацебо. Хотя кажется, что возможно, что некоторые продукты Эхинацеи более эффективны, чем плацебо, для лечения простуды, общие доказательства клинически значимых эффектов лечения слабы. В целом, испытания, изучавшие Эхинацею для профилактики простудных заболеваний, не показали статистически значимого уменьшения наступления болезни. Однако, почти все испытания по профилактике указывали на небольшой профилактический эффект. Число пациентов, которые выбывали из испытаний, или сообщали о неблагоприятных эффектах, значимо [статистически] не различалось между группами лечения и контроля в испытаниях по профилактике и лечению. Однако, в испытаниях по профилактике наблюдалась тенденция к большему числу пациентов, выбывавших из-за неблагоприятных событий в группах лечения.

Доказательства актуальны по июль 2013 года.

Заметки по переводу

Перевод: Александрова Эльвира Григорьевна. Редактирование: Гамирова Римма Габдульбаровна, Зиганшина Лилия Евгеньевна. Координация проекта по переводу на русский язык: Казанский федеральный университет. По вопросам, связанным с этим переводом, пожалуйста, свяжитесь с нами по адресу: lezign@gmail.com

Laienverständliche Zusammenfassung

Echinacea zur Vorbeugung und Behandlung von Erkältungskrankheiten

Präparate aus der Pflanze Echinacea (Sonnenhut) werden in einigen Ländern Europas und in Nordamerika häufig bei Erkältungskrankheiten eingesetzt. Die im Handel erhältlichen Präparate aus Echinacea unterscheiden sich stark hinsichtlich der verwendeten Pflanzenarten und -bestandteile (Staude, Wurzel oder beidem) und der Verarbeitungsmethode (Trocknung, alkoholische Auszug oder Auspressen des Saftes der frischen Pflanze). In manchen Fällen werden auch weitere Kräuter hinzugefügt.

In diesem Review wurden 24 kontrollierte klinische Studien mit 4631 Teilnehmern eingeschlossen, die die Wirksamkeit verschiedener Echinacea-Präparate zur Vorbeugung und Behandlung von Erkältungen untersuchten. Der Review zeigt, dass in den randomisierten Studien eine Vielzahl von Echinacea-Produkten, die aus verschiedenen Pflanzengattungen und -bestandteilen für unterschiedliche Darreichungsformen hergestellt wurden, mit Plazebo verglichen wurden. Aufgrund der erheblichen Unterschiede der getesteten Präparate war es schwierig, eindeutige Schlussfolgerungen zu ziehen. Das Bias-Risiko wurde in fünf Studien, anhand des Cochrane 'Risiko für Bias Tool' in allen fünf Kategorien als niedrig bewertet. Fünf weitere Studien wurden mit niedrigem Bias-Risiko eingestuft, jedoch war es in einer der Kategorien unklar. Acht Studien hatten ein hohes Bias-Risiko in mindestens einer Kategorie und die übrigen sechs Studien hatten ein unklares Bias-Risiko.

Die meisten Studien untersuchten, ob die Einnahme von Echinacea Präparaten nach Einsetzen der Erkältung verglichen mit Plazebo die Dauer der Krankheit verringern kann. Obwohl es möglich zu sein scheint, dass manche Echinacea Präparate zur Behandlung von Erkältungen wirksamer sind als Plazebo, ist insgesamt die Evidenz für eine klinisch relevante Wirkung schwach. Allgemein fanden Studien, die Echinacea für die Vorbeugung von Erkältungskrankheiten untersuchten, keine statistisch signifikante Verringerung von Krankheitsfällen. Jedoch wiesen fast alle Präventionssstudien auf eine kleine vorbeugende Wirkung von Echinacea hin. Die Zahl von Patienten in den Behandlungs- und Kontrollgruppen, unterschieden sich weder in Studien zur Vorbeugung noch solchen zur Behandlung signifikant. Jedoch war in den Präventionsstudien die Zahl von Studienabbrechern aufgrund von unerwünschten Ereignissen in den Behandlungsgruppen höher .

Die Evidenz ist auf dem Stand von Juli 2013.

Anmerkungen zur Übersetzung

E. von Elm, Koordination durch Cochrane Schweiz.

平易な要約

風邪の予防と治療におけるエキナセア

エキナセアの植物製剤は欧州の一部の国々や北米で風邪に広く使用されている。 市販されているエキナセア製剤は実にさまざまである。なぜなら、異なるタイプ(種)や部位(草の葉や根または両方)が使われており、さまざまな製造方法(乾燥、アルコール抽出、または新鮮な植物から汁を搾り出す)で作られ、さらに他のハーブを添加する場合もあるためである。

風邪または誘発性ライノウイルス感染症の予防および治療において、4631例を対象に複数の異なるエキナセア製剤の有効性を調べた24件の比較臨床試験について評価した。 本レビューでは、種、部位、剤形が異なるさまざまなエキナセア 製剤が、ランダム化試験でプラセボと比較されてきたことを明らかにしている。 試験に使われた製剤がかなり異なるため、確実な結論を出すのは困難であった。 Cochraneのバイアスのリスクツールにおける5つのカテゴリーのすべてで、5件の試験が 低リスクと評価された。 さらに5件の試験が低リスクと評価され、1つのカテゴリーでリスク不明とされた。 8件の試験が1つ以上のカテゴリーで高リスクと評価され、残りの6件がリスク不明とされた。

大部分の試験では、風邪の症状が現れた後にエキナセア製剤を服用し、プラセボと比較して罹患期間が短縮されるかについて調べていた。 風邪の治療において、一部のエキナセア製剤はプラセボより有効な可能性があるが、臨床的に意義があるとするエビデンスは全般的に弱い。 概して、風邪の予防についてエキナセアを調べた試験では、風邪の発症が統計学的に有意に減少するという結果を示さなかった。 しかし、ほぼすべての予防に関する試験で、わずかな予防効果の傾向がみられた。 予防および治療に関する試験では、脱落した患者数または有害作用を報告した患者数について、治療群とコントロール群に有意差はみられなかった。 しかし、予防に関する試験では、治療群でより多くの患者が有害事象により脱落する傾向がみられた。

本エビデンスは2013年7月現在のものである。

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/)[2015.12.20]  《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

Description of the condition

Common cold is the most frequent disease in humans. A large US-American survey showed that over 70% of the population annually was suffering from at least one viral respiratory tract infection. The authors concluded that the economic burden in the USA was almost USD 40 billion annually (Fendrick 2003). Viral agents causing common colds are mostly picornaviruses (rhinoviruses and enteroviruses), coronaviruses, adenoviruses, parainfluenza viruses and respiratory syncytial viruses (Denny 1995; Monto 1987). The incidence of the common cold has a peak in the winter months. There are several different hypotheses and explanations for this. One is that cooling of the nasal airway decreases the effectiveness of local respiratory defenses such as mucociliary clearance and leucocyte phagocytosis (Eccles 2002).

Description of the intervention

Extracts of the plant Echinacea (of the family Asteraceae) are widely used by consumers and practitioners in some European countries and in the US for preventing and treating upper respiratory tract infections (Barrett 2003). In the US mainstream market, Echinacea preparations are among the second top-selling herbal products (Blumenthal 2005).

Assessment of the effectiveness of Echinacea preparations is complicated by the limited comparability of the available preparations for the following reasons.

  1. Three different species are in medical use: Echinacea purpurea (E. purpurea),Echinacea pallida (E. pallida) and Echinacea angustifolia (E. angustifolia).

  2. Different parts of the plant are used (root, herb, flower or whole plant).

  3. Different methods of extraction are used.

  4. In some preparations other plant extracts or homeopathic components are added.

The evidence available from clinical trials on its effectiveness has been considered inconsistent in several reviews (Barrett 1999; Caruso 2005; Linde 2006; Melchart 1994; Melchart 1999). Two meta-analyses pooling trials using different heterogeneous Echinacea preparations for the treatment of induced rhinovirus infections (Schoop 2006b) or the common cold (Shah 2007) found more positive results for the effect of Echinacea. These results have to be interpreted with caution, as the great heterogeneity of tested Echinacea preparations makes comparison and pooling of data methodologically questionable.

How the intervention might work

The exact mechanisms of action for the immunomodulating effects of Echinacea preparations are unclear. Four classes of compounds are known to contribute to the immunomodulatory activity of Echinacea extracts: alkamides, glycoproteins, polysaccharides and caffeic acid derivatives (CADs). Phenolic compounds include caffeic, cichoric, caftaric and chlorogenic acid, as well as cynarin and echinacoside and are found in differing concentrations in the roots of both E. angustifolia and E. purpurea but also in the aerial parts of E. purpurea. Alkamides (alkylamides; fatty acid amides) are characteristic constituents of E. angustifolia roots, but are also found in roots and aerial parts of E. purpurea. Flavonoids, essential oils, polyacetylenes, ketones and pyrrolizidine alkaloids have also been isolated from Echinacea species. It is important to note that the pharmacologic effects associated with the constituents of Echinacea may result from independent or synergistic interactions with single or multiple constituents.

E. purpurea extracts rich in glycoproteins, polysaccharides and CADs have long been reported to demonstrate immunoactivity. Research in mice more than two decades ago demonstrated activation of macrophages and natural killer cells (Bauer 1989). Since then, numerous studies have supported these findings and have reported a variety of additional effects on adaptive and innate immune mechanisms (Chavez 2007; Gurbuz 2010; Hall 2007; Ramasahayam 2011; Ritchie 2011; Sadigh-Eteghad 2011; Yamada 2011; Zhai 2007).  

In contrast, early research on alkamide-rich extracts of E. angustifolia and E. purpurea suggested anti-inflammatory activity (Müller-Jakic 1994). Since then, however, studies have reported immunoactivity attributable to alkamides (Lalone 2009; Matthias 2008) and have suggested that influences on inflammatory pathways are complex, with both pro- and anti-inflammatory effects reported (Birt 2008; Qiang 2013; Yu 2013). Echinacea alkamides have been shown to be absorbed into the blood (Matthias 2005; Woelkart 2005b; Woelkart 2006; Woelkart 2008) and appear to exert a variety of effects through the endocannabinoid system (Chicca 2009; Woelkart 2005a; Woelkart 2007). Research on the effects of gene expression and signaling pathways is well underway (Altamirano-Dimas 2007; Gertsch 2004; Uluisik 2012).

In addition to research on immune and inflammatory pathways, indications of antiviral activity have been reported (Bodinet 2002; Ghaemi 2009; Sharma 2006). Finally, recent research suggests potential anti-anxiety properties (Haller 2013), potentially due to neuro-synaptic modulation in the hippocampus (Hajos 2012).

In summary, while it is clear that variousEchinacea extracts and constituents have demonstrated pharmacological activities in a variety of biological assays, there is as yet no evidence-based conceptual framework to explain howEchinacea might effectively prevent or treat acute respiratory infections.  

Why it is important to do this review

The common cold has a high prevalence and although it is a self limiting condition effective treatment options which lessen the severity and duration of symptoms would be of major importance. Echinacea products are widely used but their effectiveness is uncertain. We completed a first version of this review in 1998 (Melchart 1999), updated it in 2006 (Linde 2006) and again in 2008 (Linde 2008). The last literature search was conducted in 2007 and did not detect new publications on the issue. Now, six years later, several new trials have been published and evidence may have changed. Therefore, a major update of this review was necessary.

Objectives

To assess whether there is evidence that Echinacea preparations are effective and safe compared to placebo in the prevention and treatment of the common cold.

Methods

Criteria for considering studies for this review

Types of studies

Randomized controlled trials (RCTs).

Types of participants

We included studies if participants were:

  1. individuals with non-specific viral upper respiratory tract infections (URTIs) with a clinical diagnoses of common cold, influenza-like syndrome or viral URTI (it was not possible to apply a standard definition of common cold across all trials);

  2. volunteers without acute URTIs but treated for preventative purposes (prevention studies);

  3. volunteers without acute URTIs but challenged with rhinovirus treated for preventative or therapeutic purposes (or both).

We did not include studies of individuals suffering from other URTIs with a defined etiology (for example influenza) or a more specific symptomatology (for example acute sinusitis, angina tonsillaris).

Types of interventions

We included trials of oral Echinacea mono-preparations versus placebo. We excluded trials on combinations of Echinacea and other herbs and trials comparing Echinacea with no treatment or another treatment than placebo.

Types of outcome measures

Selected trials had to include clinical outcome measures related to occurrence (prevention studies) and severity or duration of infections (prevention and treatment studies). We excluded trials focusing solely on physiological parameters (such as phagocytosis activity). We did not include or exclude studies based on their primary outcome measure if at least one clinically relevant outcome measure listed above was reported.

Primary outcomes
  • The primary efficacy outcome measure for prevention trials was the number of participants experiencing at least one cold episode.

  • The primary efficacy outcome measure for treatment trials was duration in days.

  • The primary outcome for safety and acceptability for both prevention and treatment trials was the number of participants dropping out due to side effects or adverse events.

Secondary outcomes
  • Secondary efficacy outcome measures for prevention trials were the number of participants experiencing more than one cold episode; cold duration in days; severity scores.

  • Secondary efficacy outcome measures for treatment trials were total severity and duration measures (e.g. area under the curve); severity of symptoms at days two to four and at days 5 to 10; in trials with very early onset of treatment also the number of participants who developed the 'full picture of a cold'.

  • Secondary safety and acceptability outcome measures for both prevention and treatment trials were the total number of drop-outs and the number of participants reporting side effects or adverse events.

Search methods for identification of studies

Electronic searches

For this 2013 update we searched for new studies published since the last publication of our review and also searched for older trials. This was done as search methods have evolved over time and the inclusion criteria of our reviews have changed considerably.

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) 2013, Issue 5, part of The Cochrane Library, www.thecochranelibrary.com (accessed 5 June 2013) which contains the Acute Respiratory Infections Group's Specialized Register, MEDLINE (1946 to May week 4, 2013), EMBASE (1991 to June 2013), CINAHL (1981 to June 2013), AMED (1985 to February 2012), LILACS (1981 to June 2013) and Web of Science (1955 to June 2013).

We used the following search strategy to search CENTRAL and MEDLINE. We combined the MEDLINE search strategy with the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity-maximizing version (2008 revision); Ovid format (Lefebvre 2011).

MEDLINE (Ovid)

1 Echinacea/
2 echinac*.tw.
3 coneflower*.tw.
4 ("E. purpurea" or "E. angustifolia" or "E. pallida").tw.
5 1 or 2 or 3 or 4

We adapted the search strategy to search EMBASE (Appendix 1), CINAHL (Appendix 2), AMED (Appendix 3), LILACS (Appendix 4) and Web of Science (Appendix 5). Searches for the first review (published in 1999) and the 2007 update are described in Appendix 6.

Searching other resources

We searched WHO ICTRP and clinicaltrials.gov (latest search 8 October 2012), the Centre for Complementary Medicine Research (1988 to September 2007) and CAMBASE (latest search 5 June 2013). We screened bibliographies of identified trials and review articles for further potentially relevant publications. We contacted experts in the field and asked about further published and unpublished studies.

Data collection and analysis

Selection of studies

One review author (MKV) screened the titles and abstracts, where available, of all identified references and eliminated non-human studies and trials without a control group. We obtained and checked further copies of all other references for eligibility. At least two review authors independently checked all potentially relevant publications or reports identified by the screening process for fulfillment of the selection criteria. We resolved disagreements by discussion. We assessed eligibility of trials in which one of the review authors was involved by review authors not involved in the trial.

Data extraction and management

At least two authors independently extracted descriptive information on patients, interventions, outcomes, results, drop-outs and side effects using a standard data extraction form. Details on extraction of outcomes used for analyses are described below. Trials in which review authors were involved were extracted and assessed by review authors not involved in the trial. We contacted trial authors or manufacturers and sponsors and asked them to provide lacking or additional data if the information in the available publications or reports was incomplete. A pharmacist with specific expertise on Echinacea (KAW) extracted information on the Echinacea preparations.

Assessment of risk of bias in included studies

At least two authors independently assessed the methodological quality of the included trials using the Cochrane 'Risk of bias' tool (Higgins 2011b). We assessed the generation of the random sequence, allocation concealment, blinding, incomplete outcome data and selective reporting. For an overall assessment we considered as low risk of bias only trials in which at least four of the five items were rated as low risk and none high risk. Any trial with one or more items rated as high risk was considered high risk in the overall assessment. The remaining trials were considered as unclear risk.

Measures of treatment effect

For dichotomous efficacy outcomes we calculated risk ratios (RR) and for safety/adverse event outcomes we calculated odds ratios (OR). For continuous efficacy outcomes we calculated mean differences (MD) if the same scale of measure was used (e.g. number of days) and standardized MD if measurement tools or scales varied. For all effect estimates we calculated 95% confidence intervals (CI).

Unit of analysis issues

In all trials individual patients were randomized. However, in one trial (Taylor 2003) investigating early self treatment more than one cold could be treated by participants (on average participants treated about 1.5 cold episodes) and the results for duration and severity presented in the publication were based on the number of cold episodes. For effect size calculation we used the number of cold episodes, because using the number of patients would only lead to a small change of the weight of this trial.

Dealing with missing data

In case of missing outcome data we tried to obtain additional information from study authors. If, in case of continuous outcomes, means were presented but standard deviations were missing we calculated standard deviations from standard errors, P values or confidence intervals as described in the Cochrane Handbook for Systematic Reviews of Interventions Section 7.7.3 (Higgins 2011a). In case of missing dichotomous data we assumed that no event occurred.

Data synthesis

We expected a priori that meaningful quantitative meta-analyses of the studies would not be possible for the following reasons: 1) a variety of Echinacea preparations are used in trials (phytochemical comparability is unclear); 2) the approach of studies differs (some investigate prevention, some self treatment, some treatment, some treatment and/or prevention of experimentally induced colds); 3) outcome measures differ in the trials; and 4) the presentation of results in available reports often includes insufficient detail to allow effect size estimation. However, we aimed to calculate effect size estimates for relevant outcome measures in single trials whenever possible. If an outcome was probably measured but data for effect size calculation were not reported we documented this.

For our main analysis we had predefined criteria to consider pooling (fixed-effect model) of data from different trials: 1) treatment given for the same purpose (prevention or treatment); 2) use of the same or a very similar (regarding plant species, part and extraction mode) preparation and in similar dosage; and 3) at least two trials that met the criteria 1 and 2. Because of these criteria we ended up with multiple subgroups and most subgroups consisted of only one trial (with a maximum of two trials when criteria were met for pooling) we decided to run additional exploratory random-effects meta-analyses including all available trials regardless of the type of Echinacea product tested. For these meta-analyses studies with more than one Echinacea group were entered only once (pooling data from the Echinacea groups) to avoid duplicate use of placebo data. The meta-analyses serve to provide a crude overview of the overall direction and magnitude of the available study results and to investigate consistency and heterogeneity of the findings. We considered pooled effect sizes as clinically interpretable – at least with caution – when a) at least two-thirds of trials measuring an outcome actually could be included in the meta-analysis; b) at least five trials could be pooled; c) the I² statistic was ≤ 40%; and d) the P value of the Chi² test for heterogeneity was ≥ 0.25. All other pooled effect sizes were not considered clinically interpretable and only used to check whether results differed between studies.

Duration of colds was analyzed and reported in highly variable manner in the primary studies. While only two presented MD with some measure of variability (the measure we would have preferred for meta-analysis), some provided median duration and P values from log rank text, Cox regression or a Wilcoxon rank sum test. According to our protocol we included only the two trials reporting mean duration. To provide at least a crude summary of study findings in a post hoc secondary analysis we used an overall estimate of effect for each study rather than summary data for each intervention group for an inverse variance analysis. For this exploratory analysis we interpreted medians as means for calculating the MD and calculated standard errors as if P values were derived from a t-test. We did not pool findings from individual studies due these liberal assumptions and the heterogeneity of the study findings but included the resulting forest plot only for giving a graphical impression of the overall evidence.

In general we used the number of patients randomized when calculating effect estimates for dichotomous outcome and the number of patients analyzed for continuous measures. However, for some study approaches this was considered inappropriate. For example, in five trials of self treatment, participants were randomized to receive an Echinacea product or placebo medication to take at home but told to take their medication only in case a cold occurred. In these cases we used the number of patients in whom a cold actually occurred for analyses.

In the case of pooling we examined heterogeneity between trials by calculating a Chi² test, the I² statistic and the Tau² statistic. An I² statistic value of 0% to 40% was not considered to be important heterogeneity; 40% to 60% was considered moderate heterogeneity; 60% to 90% was considered substantial and an I2 statistic value greater than 90% indicated considerable heterogeneity (Higgins 2003). We generated funnel plots for meta-analyses including at least four studies. We carried out all calculations in RevMan 2012, version 5.2.

Results

Description of studies

Results of the search

The database searches identified a total of 556 hits (Figure 1). Three additional records were identified through other sources. We identified a total of 82 full-text articles describing trials which tested preparations of Echinacea in humans (alone or in combination with other plant extracts).

Figure 1.

study flow diagram.

Included studies

Twenty-four studies (in 29 publications) met the inclusion criteria (Characteristics of included studies table). Of these, 15 had been included in the previous version of the review (Barrett 2002; Brinkeborn 1999; Bräunig 1992; Dorn 1997; Galea 1996; Goel 2004; Goel 2005; Grimm 1999; Hoheisel 1997; Kim 2002a; Lindenmuth 2000; Melchart 1998; Schulten 2001; Taylor 2003; Yale 2004), two of the now included studies on induced colds had initially been excluded in the previous version of the review (Sperber 2004; Turner 2000) and seven studies have been newly included (Barrett 2010; Hall 2007; Jawad 2012; O'Neill 2008; Tiralongo 2012; Turner 2005; Zhang 2003). One study (Taylor 2003) was in children, the other 23 in adults. Three trials (Brinkeborn 1999; Bräunig 1992; Melchart 1998; Turner 2005) had more than one experimental group receiving an Echinacea product (different dosage in one and different extracts in three) so there were a total of 33 comparisons of an Echinacea preparation with placebo. One study had both a placebo and a no treatment control group (Kim 2002a) and one study had a no treatment and an unblinded Echinacea group in addition to the blinded Echinacea and placebo groups (Barrett 2010).

Twelve trials originated from the USA, five from Germany, three from Canada, two from Sweden, one from the United Kingdom and one from Australia. Two trials from the USA and one from Canada were only available as unpublished manuscripts (Galea 1996; Kim 2002a; Zhang 2003).

Excluded studies

Forty-nine studies (in 53 publications) did not meet the inclusion criteria (Characteristics of excluded studies table; Figure 1). In 19 studies Echinacea in combination with other remedies was used. Eleven studies examined other conditions than common cold. Six studies examined other conditions than common cold using Echinacea in combination with other remedies. Ten studies were not randomized controlled trials, five of which also examined other conditions than common cold. Four studies only examined laboratory measurement results, without clinical outcomes and three of them examined other conditions than common cold.

Among the 49 excluded studies there was one which was included in the previous version of this review (Spasov 2004). It was an unblinded study without placebo control.

Study approaches

As described in the methods section, we separated prevention trials (treatment of healthy volunteers without cold symptoms to avoid occurrence of colds or reduce severity and duration of occurring cold) and treatment trials (treatment of individuals with colds or of early cold symptoms).

Ten studies were prevention trials (with a total of 13 Echinacea groups). In four of these trials 431 healthy volunteers (Hall 2007) or persons being challenged by inoculation with rhinovirus (Sperber 2004; Turner 2000; Turner 2005) were treated over a shorter period (two to four weeks). Six trials including 1391 participants (Grimm 1999; Jawad 2012; Melchart 1998; O'Neill 2008; Tiralongo 2012; Zhang 2003) treated healthy volunteers over a longer period (6 to 16 weeks) for preventative purposes.

Fifteen studies (with 20 comparisons) were categorized as treatment trials, but among these studies two different approaches were used: five placebo-controlled trials with a total of seven Echinacea groups (Brinkeborn 1999; Galea 1996; Goel 2004; Goel 2005; Taylor 2003) investigated self treatment. Healthy volunteers were randomized and instructed to start treatment only if they caught a cold. These trials randomized a total of 1910 participants (range 150 to 559). However, 846 did not start treatment as they did not catch a cold during the study period; therefore only 1064 (62 to 436) actually started treatment. In 10 trials with a total of 13 Echinacea groups (Barrett 2002; Barrett 2010; Bräunig 1992; Dorn 1997; Hoheisel 1997; Kim 2002a; Lindenmuth 2000; Schulten 2001; Turner 2005; Yale 2004) individuals with cold symptoms were randomized and treated. These 10 trials included a total of 1538 participants (range 57 to 359 in each). Typically, they tried to start treatment as early as possible. Three trials (Hoheisel 1997; Schulten 2001; Turner 2005) explicitly not only investigated duration and severity of symptoms but also tried to prevent the development of a 'full cold' by treatment of first symptoms. Two of these trials were performed in industrial plants where employees could access treatment very fast (Hoheisel 1997; Schulten 2001). Development of a "full cold" was also tested by one trial examining experimental rhinovirus colds (Turner 2005).

Echinacea preparations tested

In the 33 experimental groups of the 24 included trials, widely different Echinacea preparations were used (Table 1). A large proportion of the preparations used in the trials were pressed juices (stabilized with alcohol), alcohol tinctures or tablets made from dried extracts. In six trials preparations from the pressed juice of the aerial parts of E. purpurea were used (Grimm 1999; Hoheisel 1997; Schulten 2001; Sperber 2004 Taylor 2003; Yale 2004). In five of these six trials (Grimm 1999; Hoheisel 1997; Schulten 2001; Sperber 2004; Taylor 2003) the same product was used. Preparations based on E. purpurea root alone were used in two trials, with three experimental groups (Bräunig 1992; Zhang 2003). An identical product based on a mixture of E. purpurea root (5%) and herb (95%) was used in two trials with three experimental groups (Brinkeborn 1999; Jawad 2012). One of these trials (Brinkeborn 1999) tested two different concentrations of the same product. Two trials by the same study group (Goel 2004; Goel 2005) investigated the effectiveness of an extract of 'various' parts of E. purpurea. The particular aspect of this preparation was that it was standardized for its content of three bioactive components (alkamides, cichoric acid and polysaccharides).

Table 1. Details of the preparations used in the included trials
  1. E. ang: Echinacea angustifolia
    E. pallida: Echinacea pallida
    E. purp: Echinacea purpurea
    v/v: volume/volume

ReferencePreparationManufacturerSpecies, partExtractionContent detailsGalenic formDosageTreatment periodRemarks
Barrett 2002Not reportedShaklee Tecninca, Pleasanton, CaliforniaUnrefined E. ang. root (50%), E. purp. herb (25%) and root (25%)Not applicable (dried Echinacea)0.20% to 0.26% echinacoside, 0.77% to 0.84% dichroic acid, 0.82% alkamides, 0.03% chlorogenic acid, 0.33% cafeolytartaric acidCapsules6 x 4 capsules (6 g) during first 24 hours, then 3 x 4 capsules (3 g/day)10 days

Barrett

2010

Not reportedMediHerb, Warwick, Queensland, Australia E. purp. root, E.ang. rootNot reported2.1 mg of alkamides per tabletTablets4 x 2 tablet during first 24 hours, then 4 x 1 tablet per day for the next 4 days; that means 10.2 g of dried Echinacea root during first 24 hours, 5.1 g during each of the next days5 days
Brinkeborn 1999, Group 1EchinaforceBioforce, Roggwil, Switzerland E. purp. herb (95%) and root (5%)Alcoholic aqueous extract6.78 mg crude extractTablets2 x 3 (40.68 mg/day)Max. 7 days
Brinkeborn 1999, Group 2No brand nameBioforce, Roggwil, Switzerland E. purp. herb (95%) and root (5%)Alcoholic aqueous extract48.27 mg crude extractTablets2 x 3 (289.62 mg/day)Max. 7 days
Brinkeborn 1999, Group 3No brand nameBioforce, Roggwil, Switzerland E. purp. rootAlcoholic aqueous extract29.60 mg crude extractTablets2 x 3 (177.6 mg/day)Max. 7 days
Bräunig 1992, group 1Not reportedNot reported E. purp. root55% v/v ethanolic extract, DER 1:5Not reportedTincture180 drops (900 mg/day)Probably 8 to 10 days
Bräunig 1992, group 2Not reportedNot reported E. purp. root55% v/v ethanolic extract, DER 1:5Not reportedTincture90 drops (450 mg/day)Probably 8 to 10 days
Dorn 1997Not reportedNot reported E. pallida rootNot reportedNot reportedTincture90 drops (900 mg/day)8 to 10 days
Galea (unpublished)Not reportedLocal pharmacist E. ang. (part not specified)Not reportedPowder standardized at 4% content of echinacosideCapsules3 x 1 (750 mg/day)10 days
Goel 2004EchinilinNatural Factors Nutritional Products, Inc., Vancouver, BC, Canada E. purp. (various parts)Aqueous and alcoholic extract combined to a 40% ethanolic formulationStandardized for 0.25 mg/ml alkamides, 2.5 mg/ml cichoric acid, 25 mg/ml polysaccharidesLiquid10 x 4 ml day 1, then 6 days 4 x 4 ml7 daysExtract standardized on the basis of 3 known active components
Goel 2005EchinilinNatural Factors Nutritional Products, Inc., Vancouver, BC, Canada E. purp. (various parts)Aqueous and alcoholic extract combined to a 40% ethanolic formulationStandardized for 0.25 mg/ml alkamides, 2.5 mg/ml cichoric acid, 25 mg/ml polysaccharidesLiquid8 x 5 ml day 1, then 6 days 3 x 5 ml7 daysExtract standardized on the basis of 3 known active components
Grimm 1999EchinacinMadaus AG, Cologne, Germany E. purp. aerial partsFresh expressed juice of whole flowering plants harvested without roots, containing 22% alcoholNot reportedLiquid2 x 4 ml day8 weeks
Hall 2007 Echinacea StandardizedNature´s Way, Springville, UT (USA) E. purp. (part not specified)Not reportedNot reportedCapsules8 capsules/day (3 x 2 with each meal and 2 at bedtime)4 weeks
Hoheisel 1997Echinagard (Echinacin)Madaus AG, Cologne, Germany E. purp. aerial partsFresh expressed juice of whole flowering plants harvested without roots, stabilized with 20% ethanolNot reportedLiquid20 drops every 2 hours day 1, then 3 x 20 drops/dayMax. 10 days
Jawad 2012Echinaforce dropsA. Vogel, Bioforce, Switzerland E. purp. (95% herb, 5% roots)Alcohol (57%) extract from freshly harvested E. purp.5 mg/100g of dodecatetraenoic acid isobutylamideLiquid3 x 0.9 ml/day (2400 mg of extract/day); in case of cold: 5 x 0.9ml/day (4000 mg of extract/day)4 monthsEach single dose was diluted in water and retained in mouth for 10 seconds
Kim (unpublished)Not reportedNature's Way products, Inc. R/O America's Natural healthcare Company, Springville, Utah E. purp. herb (80%), E. ang. roots (20%)No detailed information; final tincture with 25% to 35% alcoholNot reportedLiquid101 ml (1000 mg dry plant) per dayAt least 5 days

Lindenmuth

2000

Echinacea Plus herbal teaDry extract ingredient by Emil Flachsmann AG, Zurich, Switzerland E. purp. and E. ang. aerial parts; E. purp. roots E. purp. root water soluble dry extract DER 6:11.275 mg per tea bag servingTea bag5 to 6 cups day 1, titration to one cup on day 55 days
Melchart 1998 Group 1No brand namePlantapharmazie, Göttingen, Germany E. ang. root30% ethanolic extract, DER 1:11Extract contained 1007.9 µg/ml glycoproteins/polysaccharides and echinacosideTincture2 x 50 drops12 weeks (intake on 5 days per week)
Melchart 1998 Group 2No brand namePlantapharmazie, Göttingen, Germany E. purp. root30% ethanolic extract, DER 1:11Extract contained 1026.2 µg/ml glycoproteins/polysaccharides and cichoric acidTincture2 x 50 drops12 weeks (intake on 5 days per week)
O'Neil 2008No brand nameNatures Resource, Mission Hill, California E. purp. (part not specified)Not reportedNot reportedCapsules3 x 2 capsules/day (1 capsule containing 300 mg E. purp.)8 weeks
Schulten 2001EchinacinMadaus AG, Cologne, Germany E. purp. aerial partsFresh expressed juice of whole flowering plants harvested without roots, stabilized with 20% ethanol, DER 1.7-2.5:1Not reportedLiquid2 x 5 ml/day10 days
Sperber 2004EchinaGuardMadaus AG, Cologne, Germany E. purp. aerial partsPressed juice in 22% alcohol baseNot reportedLiquid3 x 2.5 ml/day14 days
Taylor 2003EchinacinMadaus AG, Cologne, Germany E. purp. aerial partsPressed juice, combined with syrupNot reportedLiquid (dried pressed juice dissolved in syrup)2 x 3.75 ml/day (children 2 to 5 years), 2 x 5 ml/day (6 to 11 years)Max. 10 days
Tiralongo 2012MediHerbIntegria Healthcare Pty Ltd., Australia E. purp. root, E. ang. rootExtract, details not reportedTablets standardized for a content of 4.4 mg alkylamides with 112.5 mg E. purp. 6:1 extract and 150 mg E. ang. 4:1 extract (detailed alkamide composition is summarized in a table with e.g. a content of 1.504 mg/tablet dodecatetraenoic acid isobutyl amides)TabletsPriming dose 2 x 1 tablet/day, flying dose 2 x 2 tablets/day, overseas dose 2 x 1 tablet/day, after-travel dose 2 x 1 tablet/day, sick dose 2 x 3 tablets/day35 days if 1 week of travel (14 days primary dose, 7 days overseas, 14 days after-travel dose) or 63 days if 5 weeks of travel (11 days with priming dose, 10 days flying dose, 25 days overseas dose, 10 days flying dose and 7 days after travel dose)
Turner 2005 Group 1 and 4No brand nameNot reported E. ang. rootExtraction with supercritical carbon dioxideNo polysaccharides, 73.8% alkamides, no echinacosidesLiquid3 x 1.5 ml/day (3 x equivalent of 300 mg Echinacea root)13 days (virus challenge on day 8)
Turner 2005 Group 2 and 5No brand nameNot reported E. ang. rootExtraction with 60% ethanol48.9% polysaccharides, 2.3% alkamides, no echinacosides, 0.16 mg/ml cynarine)Liquid3 x 1.5 ml/day (3 x equivalent of 300 mg Echinacea root)13 days (virus challenge on day 8)
Turner 2005 Group 3 and 6No brand nameNot reported E. ang. rootExtraction with 20% ethanol42.1% polysaccharides, 0.1% alkamides, no echinacosidesLiquid3 x 1.5 ml/day (3 x equivalent of 300 mg Echinacea root)13 days (virus challenge on day 8)
Turner 2000No brand nameNot reported E. purp. and E. ang.4% phenolic extract0.16% cichoric acid, almost no echinacosides or alkamidesCapsules with powder3 x 300 mg/day19 days (virus challenge on day 14)
Yale 2004EchinaFreshEnzymatic Therapy, Green Bay, Wisconsin E. purpurea aerial partsPressed juiceStandardized for a content of 2.4% soluble beta-1,2-D-fructofuranosidesCapsules with freeze dried juice3 x 1 capsule/day7 days, if symptoms not resolved max. 14 days
Zhang 2003No brand nameNot reported E. purp. rootRoot powder4.4 mg cichoric acidCapsules with 294 mg powder1 capsule/day8 weeks

The 14 remaining Echinacea preparations were only used in one trial each: tinctures or extracts prepared from E. pallida root (Dorn 1997), E. angustifolia root (Melchart 1998) or E. angustifolia, without information on the plant parts used (Galea 1996); a dried plant preparation based on 50% E. angustifolia root, 25% E. purpurea root and 25% E. purpurea herb (Barrett 2002); a preparation based on E. purpurea root and E. angustofolia root, without information on extraction details (Barrett 2010); a tincture from 80% E. purpurea herb and 20% E. angustifolia root (Kim 2002a); a preparation based on an extract of E. purpurea root and E. angustofolia root (extraction details not reported) (Tiralongo 2012); a 4% phenolic extract of E. purpurea and E. angustofolia (Turner 2000); three preparations based on E. angustofolia root using three different extraction methods (20% alcohol; 60% alcohol and CO2) (Turner 2005); two kinds of E. purpurea capsules (parts and extraction details not reported (Hall 2007; O'Neill 2008); and a tea preparation based on dry extracts from the aerial parts of E. purpurea and E. angustifolia (Lindenmuth 2000).

Outcome measurement

All 10 prevention trials investigated the occurrence of cold. Other outcomes, like the number of people with more than one cold episode, duration and severity scores were only measured in some of the prevention trials. Among the self treatment and treatment trials, methods for outcome measurements and the results actually presented varied greatly (there were differences regarding instruments used, timing of measurements, type of analysis and descriptive statistics).

Risk of bias in included studies

'Risk of bias' judgements are given in the Characteristics of included studies table and in Figure 2 and Figure 3. We rated five trials (Barrett 2002; Barrett 2010; Brinkeborn 1999; Goel 2004; Taylor 2003) as having a low risk of bias in all five categories of the Cochrane 'Risk of bias' tool (Higgins 2011b). We rated five more trials as low risk of bias, having an unclear risk of bias in only one category. We rated eight trials as having a high risk of bias (Bräunig 1992; Dorn 1997; Galea 1996; Hoheisel 1997; Jawad 2012; Lindenmuth 2000; O'Neill 2008; Zhang 2003) and six as having an unclear risk of bias.

Figure 2.

'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3.

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Random sequence generation was performed appropriately in at least 17 studies. Additional information by authors and sponsors was taken into account. In one study (Lindenmuth 2000), allocation to groups appeared to follow an alternating sequence rather than true randomization. However, the allocation process was handled by an independent and blinded secretary after inclusion of participants in the trial. This study was already included in the previous version of the review. An adequate method of concealment was used in at least 14 studies (taking into account additional information received from authors and sponsors). In 10 studies sufficiently detailed information on allocation concealment was not reported.

Blinding

All 24 trials were described as blinded. In 16 trials we considered the risk of performance and detection bias as low. One older trial was not adequately blinded (Bräunig 1992). In this three-armed trial, one Echinacea group received 180 drops daily while the other Echinacea group and the placebo group received 90 drops. One trial was performed among employees of the manufacturer (Schulten 2001) who may have recognized the taste of their Echinacea product; the success of blinding was not tested. In one trial capsules filled with vegetable oil were used as placebo and may have been distinguishable from Echinacea capsules by taste (Galea 1996). In two more trials (Hoheisel 1997; Lindenmuth 2000) Echinacea preparation and placebo could possibly have been distinguishable by taste. Thirteen trials reported a test for the success of blinding. In 11 trials blinding seemed to have been successful while in two there was evidence of some unblinding (Melchart 1998) or major unblinding (Zhang 2003).

Incomplete outcome data

The risk of attrition bias was considered low in 14 trials, having reported less than 20% attrition and performed an intention-to-treat analysis or reported generally less than 5% attrition. The risk of attrition bias was unclear in three and high in seven trials.

Selective reporting

We assessed 17 trials as having a low risk of reporting bias. In four trials important relevant outcomes are not reported/examined (Bräunig 1992; Hall 2007; Hoheisel 1997; Jawad 2012). In two trials the reports are not systematically biased, but outcomes have been reported insufficiently to allow effect size calculation (Galea 1996; Schulten 2001).

Effects of interventions

Primary outcomes for prevention trials and for treatment or self treatment trials

Primary efficacy outcome measure for prevention trials: number of participants experiencing at least one cold episode

Nine of the 10 prevention trials reported the number of patients experiencing at least one cold. None of the 12 comparisons of Echinacea preparations and placebo in these nine trials (Grimm 1999; Hall 2007; Melchart 1998; O'Neill 2008; Sperber 2004; Tiralongo 2012; Turner 2000; Turner 2005; Zhang 2003) demonstrated statistically significant results in comparison to placebo (Figure 4; Analysis 1.1).

Figure 4.

Forest plot of comparison: 1 Echinacea versus placebo to prevent common cold, outcome: 1.1 Number of participants with at least 1 cold episode.

Only two trials investigated the same Echinacea product; the pooled risk ratio (RR) also did not show significant effects over placebo (RR 0.82, 95% confidence interval (CI) 0.67 to 1.02; P = 0.07). However, in our exploratory meta-analysis pooling all trials (1167 patients totally), regardless of the product used, prophylactic treatment with Echinacea products was associated with a reduced risk of experiencing a cold (RR 0.83, 95% CI 0.75 to 0.92; P < 0.001). Study findings were highly consistent across studies with an I² statistic of 0%, a Tau² of 0.00 and a P value of 0.98 in the Chi² test for heterogeneity. The funnel plot showed some asymmetry (Eggers test p = 0.03) but point estimates in the single trials were similar and including only the four most precise trials in meta-analysis reduced the pooled estimate only marginally (RR 0.85, 95% CI 0.75 to 0.96). We could not include the largest prevention study (Jawad 2012) in the quantitative analyses as it did not report the number of patients with at least one cold but only the total number of cold episodes. There were 149 cold episodes in the Echinacea group versus 188 in the placebo group; this finding seems very compatible with the reduced RR suggested by our meta-analysis.

The observed risk ratio of 0.83 in our meta-analysis corresponds to an absolute risk reduction of 10% (95% CI 5% to 16%) and a number needed to treat of 10 (95% CI 6 to 20). As more participants in the control groups experienced a cold the absolute risk reduction in the four most precise trials was 11% (94% CI 4% to 19%) and the number needed to treat 9 (95% CI 5 to 25) in spite of the slightly larger risk ratio of 0.85.

Primary efficacy outcome measure for treatment or self treatment trials: duration in days

Only one trial of a mixture of 50% E. angustifolia root, 25% E. purpurea root and 25% E. purpurea herb (Barrett 2002) and one trial of a mixture of E. purpurea root and E. angustifolia root (Barrett 2010) reported the mean duration of colds (Figure 5; Analysis 2.1). Neither trial found a significant difference compared to placebo. A total of six trials could be included in our post hoc secondary inverse variance analysis also using other data on duration (Analysis 2.2). Study findings were heterogeneous (I² statistic = 77%, Tau² = 0.88, P = 0.0002 in Chi² test) with two trials (Lindenmuth 2000; Schulten 2001) finding a significantly shorter duration in the Echinacea group.  

Figure 5.

Forest plot of comparison: 2 Echinacea versus placebo to treat patients with common cold, outcome: 2.1 Duration: mean difference.

Primary safety and acceptability outcome for prevention trials: number of participants dropping out due to side effects or adverse events

Our main outcome measure for the safety and acceptability analysis, the number of patients dropping out due to adverse effects, was reported in eight comparisons in seven studies (Grimm 1999; Jawad 2012; Melchart 1998; O'Neill 2008; Sperber 2004; Tiralongo 2012; Turner 2000; see Figure 6; Analysis 1.5). There were no significant differences in the single trials but the confidence intervals are wide as the number of patients dropping out was generally low. If study findings were pooled regardless of the Echinacea product used (heterogeneity indicators I² statistic = 0%; Tau² = 0.00; P = 0.87 in Chi² test) 2.4% in Echinacea groups dropped out from the studies due to side effects compared to 0.8% from the placebo groups (odds ratio (OR) 2.17, 95% CI 0.85 to 5.53; P = 0.10).

Figure 6.

Forest plot of comparison: 1 Echinacea versus placebo to prevent common cold, outcome: 1.5 Number of patients dropping out due to adverse effects.

Primary safety and acceptability outcome for treatment or self treatment trials: number of participants dropping out due to side effects or adverse events

For 11 trials (14 comparisons) the number of patients dropping out due to side or adverse effects could be extracted. Only three of 1088 patients who received an Echinacea product and none of the 930 patients who received placebo dropped out for these reasons (Analysis 2.7).

Secondary outcomes for prevention trials and for treatment or self treatment trials

Secondary efficacy outcome measures for prevention trials: number of participants experiencing more than one cold episode; cold duration in days; severity scores

Three trials (with four comparisons) reported the number of patients with more than one cold episode (Analysis 1.2). None of the trials found significant differences but confidence intervals were very wide. Four trials (with five comparisons) reported the duration of cold episode (Analysis 1.3). Only one small trial (Hall 2007) found a very large (5.2 days on average) statistically significant effect over placebo. Point estimates in the other comparisons vary between 1.2 days shorter to 0.6 days longer duration in the Echinacea groups. Five trials (with seven comparisons) presented sufficient data for calculations on effect sizes for severity of cold episodes (Analysis 1.4). Again, none of the single trials found a significant effect over placebo. However, as study findings were similar across trials (I² statistic = 0% and Tau² = 0.00 and P = 0.86 in Chi² test) we report the pooled standardized mean difference (SMD) for this outcome which is -0.24 (95% CI -0.07 to -0.40; P = 0.005), indicating a small effect over placebo.

Secondary efficacy outcome measures for treatment or self treatment trials: total severity and duration measures; severity of symptoms at days two to four and at days 5 to 10; in trials with very early onset of treatment also the number of participants who developed the 'full picture of a cold'

In nine trials (nine comparisons) measures integrating both severity and duration were presented in sufficient detail, or were provided by the authors, to calculate effect sizes (Analysis 2.2). The two trials on E. purpurea herb pressed juice preparations also reporting duration again reported conflicting results (Schulten 2001; Taylor 2003). The pooled SMD was not significant (SMD -0.26, 95% CI -0.75 to 0.23; P = 0.30; I² statistic = 76%). This applies also to two trials testing a standardized extract of E. purpurea (SMD -0.18, 95% CI -0.57 to 0.21; P = 0.20; I² = 40%). Trials of other extracts did not find any significant differences. While heterogeneity seems limited (I² statistic = 17%; Tau² 0.00; P = 0.29 in Chi² test) our SMD from meta-analysis of all available studies has to be interpreted with great caution (SMD -0.09, 95% CI -0.20 to 0.02; P = 0.10). Data on severity scores after two to four days (Analysis 2.4) and five to 10 days of treatment (Analysis 2.5) were reported in seven trials (eight comparisons) and eight trials (11 comparisons), respectively. Significant differences were found in two comparisons after two to four days and four comparisons after five to 10 days. As study findings were heterogeneous (I² statistic = 76% and 90%, respectively) we do not report a pooled effect estimate of all pooled data. For the standardized E. purpurea extract tested in two trials pooled SMDs were non-significant (after two to four days: SMD -0.20, 95% CI -0.88 to 0.48; P = 0.56 and after five to 10 days SMD -0.31, 95% CI -0.75 to -1.00; P = 0.18). Only three trials reported the number of patients developing a "full" cold after the early treatment of prodromes (self treatment) (Analysis 2.6). For the two studies using the same Echinacea product (E. purpurea herb pressed juice) the pooled RR was non-significant (RR 0.79, 95% CI 0.54 to 1.14; P = 0.21).

Secondary safety and acceptability outcomes for prevention trials: total number of drop-outs and the number of participants reporting side effects or adverse events

For 12 comparisons in nine trials the number of patients dropping out (Analysis 1.6) has been reported. Two studies reported that significantly more patients were dropping out in the Echinacea group than in the placebo group (Jawad 2012 and the comparison using E. angustofolia root extracted with CO2 in Turner 2005). The other trials found no significant differences in the number of drop-outs. As study results seem broadly consistent (I² statistic = 8%, Tau² = 0.02; P = 0.37 in Chi² test) we also report pooled results. The percentage of participants in the Echinacea groups terminating studies early was 12.7% compared to 9.0% in the placebo groups (OR 1.37, 95% CI 0.98 to 1.91; P = 0.06). For nine comparisons in eight trials the number of persons reporting adverse effects (Analysis 1.7) has been reported. Results were non-significant in all trials except for one trial which found significantly fewer persons reporting adverse effects in the placebo group (Zhang 2003). In most of the trials there was a trend towards fewer adverse effects in the placebo groups. As some heterogeneity cannot be ruled out (I² statistic = 25%, Tau² = 0.10; P = 0.23 in Chi² test) our pooled findings are hard to interpret. 11.8% versus 8.6% of patients reported side or adverse effects (RR 1.49, 95% CI 0.95 to 2.35; P = 0.09).

Secondary safety and acceptability outcomes for treatment or self treatment trials: total number of drop-outs and the number of participants reporting side effects or adverse events

Numbers of participants dropping out were similar in Echinacea and placebo groups in the trials presenting these data (Analysis 2.8 to Analysis 2.9), except for E. angustofolia root extracted with 60% ethanol which led to more drop-outs in the Echinacea group (Turner 2005). The number of patients reporting adverse effects did not differ significantly between treatment and control groups in the single trials. Heterogeneity was low (I² statistic = 0%, Tau² = 0.00, P = 0.84 Chi² test). Meta-analysis showed a significant difference in the number of patients reporting side effects in the placebo groups (32.6%) and in the treatment groups (34.1%) (OR 1.28, 95% CI 1.02 to 1.60, P = 0.03). One trial in children using a preparation made from pressed juice of E. purpurea herb found an increased frequency of rash in the experimental group (Taylor 2003).

Discussion

Summary of main results

Our review shows that a variety of products prepared from different Echinacea species, different plant parts and in different forms have been compared to placebo in randomized trials. These preparations contain quite different amounts of bioactive components and hence are not biochemically comparable. Furthermore, trial approaches and methods for cold assessment were highly variable. Taken together, results from prevention trials suggest that a number of Echinacea products slightly reduce the risk of getting a cold in healthy individuals. If this conclusion is true, the lack of significance in individual trials could be due to a lack of statistical power (too few patients included in single studies). Although it seems possible that some Echinacea products also have effects over placebo for treating colds, the overall evidence for clinically relevant treatment effects over placebo is weak.

Overall completeness and applicability of evidence

For our review we could identify and include three unpublished trials which did not find significant effects over placebo (Galea 1996; Kim 2002a; Zhang 2003). We are also aware of at least one further unpublished, negative trial from the USA. During the search for an earlier review on Echinacea (Melchart 1994) one of the authors was informed by an expert in the field through personal communication that there were several negative trials from Germany (however, possibly partly of combinations). It seems possible that there are additional unpublished trials, though we think that it is unlikely that the conclusions of our review would change substantially; the evidence regarding treatment effects is weak and cannot justify recommendations to take Echinacea or a specific product. All single prevention trials included in our review yielded non-significant results. Particularly in smaller trials, P values were far from statistically significant.

Quality of the evidence

The great heterogeneity of preparations tested makes conclusions difficult. Several of the newer trials tested products which were standardized on the content of a bioactive ingredient. However, the available research indicates that the clinical effects of (some) Echinacea preparations are likely to be due to several components which may have synergistic effects. Two of the tested products were standardized for known bioactive components, namely alkamides, cichoric acid and polysaccharides (Goel 2004; Goel 2005; Tiralongo 2012). Components (or one component) of the study medication have been analyzed and documented for several trials (Barrett 2002; Barrett 2010; Galea 1996; Jawad 2012; Melchart 1998; Turner 2000; Turner 2005; Yale 2004; Zhang 2003). Testing preparations that have been standardized to specific components seems like a desirable way to move forward. The quality of the included trials was heterogeneous as we considered 38% of the trials to have a high risk of bias while we considered 42% of the trials to have a low risk of bias.

In 2005, a further systematic review on the effectiveness of Echinacea for the treatment (not prevention) of colds was published (Caruso 2005). The authors concluded that the possible therapeutic effectiveness of Echinacea had not been established. A major criticism was that most studies (apart from two negative trials) lacked a proof of blinding. While we agree that successful blinding is crucial for the validity of a trial, we find it problematic to overemphasize this criterion when assessing the available evidence. First, the vast majority (93%) of placebo-controlled RCTs provide no evidence of blinding success and, of those that do, the majority report less than satisfactory results (Fergusson 2004). Second, participant guesses at the end of a trial are also influenced by the perceived outcome and are not necessarily evidence of bias. Nevertheless, we agree with the authors of this review that the available evidence is far from convincing and that a lack of blinding can be a relevant problem in trials of Echinacea products.

Potential biases in the review process

Study selection and data extraction were performed by at least two review authors independently. Studies in which one of the authors were involved were handled by another review author. We checked study findings entered for effect size calculation against the original publications. However, it was a major challenge for the authors of this review to summarize the results of the included studies in a manner that is both concise and reflects the heterogeneity adequately. In our main analysis we did not pool studies unless they clearly investigated comparable Echinacea products. Yet, deviating from our protocol, we included some pooled estimates from meta-analysis across different Echinacea preparations in the text. We believe that this decision is justified as a) it allows us to check whether study findings are consistent across studies and products and b) it provides a crude idea of the possible size of potential effects. However, we urge that these results have to be interpreted with caution and should not be interpreted as 'average' effects of Echinacea products.

Agreements and disagreements with other studies or reviews

A meta-analysis (Schoop 2006b) of the three trials on induced rhinovirus infections included in our review (Sperber 2004; Turner 2000; Turner 2005) found that the likelihood of participants experiencing a clinical cold was significantly lower in the Echinacea groups. The results were pooled, although different Echinacea preparations were examined in the trials. These findings could indicate that the trials were too small to detect a small effect of the tested preparations. This conclusion is consistent with our results.

In 2007 another meta-analysis of RCTs investigating the effectiveness of Echinacea products for preventing and treating common colds was published (Shah 2007). This review drew more favorable conclusions, especially on the effect of Echinacea on the cold duration, than we do. Shah 2007 used different inclusion criteria and also included trials investigating combinations of Echinacea. These authors heavily relied on meta-analysis, pooling findings from studies investigating very different Echinacea preparations and from treatment and prevention trials in one analysis. If all these trials are interpreted as investigating the same treatment for the same purpose, then the evidence can be considered as more positive and the conclusions reasonable. For our main analysis we refrained from pooling studies testing different Echinacea preparations.

Other complementary and alternative medicine (CAM) interventions for prevention and treatment have been investigated in systematic reviews. The evidence that vitamin C supplementation or probiotics used for prevention of the common cold, and zinc used for treatment of the common cold, are effective (Hao 2011; Hemilä 2013; Singh 2013a) is stronger than the evidence for Echinacea. Evidence for the effects of other CAM interventions is similarly limited (Pelargonium sidoides, Timmer 2009) or even weaker (saline nasal irrigation, Kassel 2010; increased fluid intake Guppy 2011; heated humidified air, Singh 2013b; garlic, Lissiman 2012; and Chinese medical herbs, Zhang 2010).

Authors' conclusions

Implications for practice

The most important recommendation for consumers and clinicians is to be aware that the available Echinacea products differ greatly. The overwhelming majority of these products have not been tested in clinical trials. It has been shown that labeling of products marketed in health food stores can be incorrect (Gilroy 2003). Our exploratory meta-analyses suggest that at least some Echinacea preparations may reduce the relative risk of catching a cold by 10% to 20%. A risk reduction of 15% would mean that if 500 out of 1000 persons receiving a placebo would catch a cold this figure would be 425 of 1000 persons with an Echinacea product. This is clearly a small effect of unclear clinical relevance. Furthermore, we cannot say which Echinacea products have an effect of this size, or a greater or lesser effect. While there are some hints that both alcoholic extracts and pressed juices that are based primarily on the aerial parts of E. purpurea have beneficial effects on cold symptoms in adults, the evidence for clinically relevant treatment effects is weak. There are still many remaining doubts due to the fact that not all trials using such preparations show even a trend towards an effect.

As randomized controlled trials include limited numbers of participants and often exclude persons with relevant co-morbidity, a review of such trials can only contribute limited knowledge on safety issues. The number of patients dropping out or reporting adverse effects did not differ significantly between treatment and control groups in prevention and treatment trials. However, in prevention trials there was a trend towards a larger number of patients dropping out due to side effects or reporting side effects in the treatment groups.The most relevant potential adverse effects of Echinacea preparations are probably allergic reactions (Huntley 2005; Mullins 2002). One trial suggested an absolute 5% increase in rash in children (Taylor 2003). Parenteral application of Echinacea preparation should be discouraged, as there is no evidence of either safety or effectiveness.

Implications for research

In principle, further research is clearly desirable given the widespread use of Echinacea products. However, given the multiplicity and diversity of products on the market applying the knowledge gained from such studies will remain a challenge to persons without in-depth knowledge of herbal preparations. The use of chemically well-defined preparations is recommended to improve comparability of results from different studies. It would be desirable if experts in research on common colds could develop recommendations for a core set of outcome measures to be used and reported in randomized clinical trials. Trials investigating the prevention of colds need large sample sizes as the potential effects of Echinacea products are likely to be small.

Acknowledgements

We would like to thank Sarah Thorning (Trials Search Co-ordinator of the Cochrane ARI Group) for her help with the literature search and Liz Dooley (Managing Editor of the Cochrane ARI Group) for her patience when waiting for this update to be completed. Furthermore we would like to thank Dieter Melchart for his contribution to the previous versions of the review. The authors would also like to thank the following people for commenting on drafts of this review: Janet Wale, Ann Fonfa, Ajima Olaghere, Steven Yale, Zaina AlBalawi, Sree Nair, Mark Jones and Meenu Singh. We are indebted to the authors of primary studies who provided additional information, often to a considerable extent.

Bruce Barrett is supported by a K24 Mid-Career Investigator Award from the National Center for Complementary and Alternative Medicine at the U.S. National Institutes of Health.
David Kiefer is a post-doctoral research fellow, supported by a T32 National Research Service Award from the National Center for Complementary and Alternative Medicine at the U.S. National Institutes of Health.

Data and analyses

Download statistical data

Comparison 1. Echinacea versus placebo to prevent common cold
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Number of participants with at least 1 cold episode9 Risk Ratio (IV, Fixed, 95% CI)Totals not selected
1.1 E. purpurea herb1 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.2 E. purpurea herb pressed juice2 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.3 E. purpurea dried plant extract1 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.4 E. purpurea root extract1 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.5 E. purpurea root alcoholic extract1 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.6 E. purpurea root and E. angustofolia root1 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.7 4% phenolic extract of E. purpurea and E. angustofolia1 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.8 E. angustifolia root alcoholic extract1 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.9 E. angustofolia root extract with CO21 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.10 E. angustofolia root extract with 60% ethanol1 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.11 E. angustofolia root extract with 20% ethanol1 Risk Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
2 Number of participants with more than 1 cold3 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 E. purpurea herb pressed juice1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.2 E. purpurea alcoholic extract with 95% herb and 5% root1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.3 E. purpurea root alcoholic extract1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.4 E. angustifolia root alcoholic extract1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
3 Duration: mean difference4 Mean Difference (IV, Fixed, 95% CI)Totals not selected
3.1 E. purpurea herb1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.2 E. purpurea dried plant extract1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.3 E. purpurea root extract1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.4 E. purpurea root alcoholic extract1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.5 E. angustofolia root alcoholic extract1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Total severity score5 Std. Mean Difference (IV, Fixed, 95% CI)Totals not selected
4.1 E. purpurea herb pressed juice1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.2 E. purpurea root extract1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.3 E. purpurea root and E. angustofolia root1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.4 4% phenolic extract of E. purpurea and E. angustofolia1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.5 E. angustofolia root extract with CO21 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.6 E. angustofolia root extract with 60% ethanol1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.7 E. angustofolia root extract with 20% ethanol1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5 Number of patients dropping out due to adverse effects7 Odds Ratio (IV, Fixed, 95% CI)Totals not selected
5.1 E. purpurea herb pressed juice2 Odds Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.2 E. purpurea dried plant extract1 Odds Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.3 E. purpurea alcoholic extract with 95% herb and 5% root1 Odds Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.4 E. purpurea root alcoholic extract1 Odds Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.5 E. purpurea root and E. angustofolia root1 Odds Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.6 4% phenolic extract of E. purpurea and E. angustofolia1 Odds Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.7 E. angustifolia root alcoholic extract1 Odds Ratio (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
6 Number of patients dropping out9 Odds Ratio (M-H, Fixed, 95% CI)Totals not selected
6.1 E. purpurea herb pressed juice2 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.2 E. purpurea dried plant extract1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.3 E. purpurea alcoholic extract with 95% herb and 5% root1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.4 E. purpurea root extract1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.5 E. purpurea root alcoholic extract1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.6 E. purpurea root and E. angustofolia root1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.7 4% phenolic extract of E. purpurea and E. angustofolia1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.8 E. angustifolia root alcoholic extract1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.9 E. angustofolia root extract with CO21 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.10 E. angustofolia root extract with 60% ethanol1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.11 E. angustofolia root extract with 20% ethanol1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7 Number of patients reporting adverse effects8 Odds Ratio (M-H, Fixed, 95% CI)Totals not selected
7.1 E. purpurea herb pressed juice2 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.2 E. purpurea dried plant extract1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.3 E. purpurea alcoholic extract with 95% herb and 5% root1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.4 E. purpurea root extract1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.5 E. purpurea root alcoholic extract1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.6 E. purpurea root and E. angustofolia root1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.7 4% phenolic extract of E. purpurea and E. angustofolia1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.8 E. angustifolia root alcoholic extract1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
Analysis 1.1.

Comparison 1 Echinacea versus placebo to prevent common cold, Outcome 1 Number of participants with at least 1 cold episode.

Analysis 1.2.

Comparison 1 Echinacea versus placebo to prevent common cold, Outcome 2 Number of participants with more than 1 cold.

Analysis 1.3.

Comparison 1 Echinacea versus placebo to prevent common cold, Outcome 3 Duration: mean difference.

Analysis 1.4.

Comparison 1 Echinacea versus placebo to prevent common cold, Outcome 4 Total severity score.

Analysis 1.5.

Comparison 1 Echinacea versus placebo to prevent common cold, Outcome 5 Number of patients dropping out due to adverse effects.

Analysis 1.6.

Comparison 1 Echinacea versus placebo to prevent common cold, Outcome 6 Number of patients dropping out.

Analysis 1.7.

Comparison 1 Echinacea versus placebo to prevent common cold, Outcome 7 Number of patients reporting adverse effects.

Comparison 2. Echinacea versus placebo to treat patients with common cold
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Duration: mean difference2 Mean Difference (IV, Fixed, 95% CI)Totals not selected
1.1 Mixture of E. angustifolia root (50%), E. purpurea root (25%) and E. purpurea herb (25%)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.2 E. purpurea root and E. angustofolia root1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
2 Duration (inverse variance)6 Mean Difference (Random, 95% CI)Totals not selected
2.1 E. purpurea herb freeze-dried pressed juice1 Mean Difference (Random, 95% CI)0.0 [0.0, 0.0]
2.2 E. purpurea herb pressed juice2 Mean Difference (Random, 95% CI)0.0 [0.0, 0.0]
2.3 Tea E. purpurea and angustifolia1 Mean Difference (Random, 95% CI)0.0 [0.0, 0.0]
2.4 Mixture of E. angustifolia root (50%), E. purpurea root (25%) and E. purpurea herb (25%)1 Mean Difference (Random, 95% CI)0.0 [0.0, 0.0]
2.5 E. purpurea root and E. angustofolia root1 Mean Difference (Random, 95% CI)0.0 [0.0, 0.0]
3 Total severity and duration measures (for example, area under the curve)9 Std. Mean Difference (IV, Fixed, 95% CI)Totals not selected
3.1 E. purpurea pressed juice2 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.2 E. purpurea standardized2 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.3 Mixture of 80% E. purpurea herb and 20% E. angustifolia roots1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.4 Mixture of E. angustifolia root (50%), E. purpurea root (25%) and E. purpurea herb (25%)1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.5 E. purpurea root and E. angustofolia root1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.6 E. angustifolia1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.7 E. angustofolia root extract with CO21 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.8 E. angustofolia root extract with 60% ethanol1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.9 E. angustofolia root extract with 20% ethanol1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Sum score after 2 to 4 days7 Std. Mean Difference (IV, Fixed, 95% CI)Totals not selected
4.1 E. purpurea herb freeze-dried pressed juice1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.2 E. purpurea herb pressed juice1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.3 E. purpurea standardized2 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.4 E. purpurea root 450 mg1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.5 E. purpurea root 900 mg1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.6 Mixture of E. angustifolia root (50%), E. purpurea root (25%) and E. purpurea herb (25%)1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.7 E. purpurea root and E. angustofolia root1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5 Sum scores after 5 to 10 days8 Std. Mean Difference (IV, Fixed, 95% CI)Totals not selected
5.1 E. purpurea herb freeze-dried pressed juice1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.2 E. purpurea herb pressed juice1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.3 E. purpurea standardized2 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.4 95% E. purpurea herb and 5% E. purpurea root (39 mg crude extract/day)1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.5 95% E. purpurea herb and 5% E. purpurea root (290 mg crude extract/day)1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.6 E. purpurea root (178 mg crude extract/day)1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.7 E. purpurea root 450 mg1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.8 E. purpurea root 900 mg1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.9 Mixture of E. angustifolia root (50%), E. purpurea root (25%) and E. purpurea herb (25%)1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.10 E. purpurea root and E. angustofolia root1 Std. Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
6 Early treatment of prodromi: patients developing a 'full' cold3 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
6.1 E. purpurea herb pressed juice2 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.2 E. angustofolia root extract with CO21 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.3 E. angustofolia root extract with 60% ethanol1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.4 E. angustofolia root extract with 20% ethanol1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7 Number of patients dropping out due to adverse effects11 Odds Ratio (M-H, Fixed, 95% CI)Totals not selected
7.1 E. purpurea herb pressed juice3 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.2 E. purpurea standardized2 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.3 95% E. purpurea herb and 5% E. purpurea root (39 mg crude extract/day)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.4 95% E. purpurea herb and 5% E. purpurea root (290 mg crude extract/day)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.5 E. purpurea root (178 mg crude extract/day)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.6 E. purpurea root 450 mg1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.7 E. purpurea root 900 mg1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.8 Tea E. purpurea and angustifolia1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.9 Mixture of E. angustifolia root (50%), E. purpurea root (25%) and E. purpurea herb (25%)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.10 E. purpurea root and E. angustofolia root1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7.11 E. pallida root1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8 Number of patients dropping out13 Odds Ratio (M-H, Fixed, 95% CI)Totals not selected
8.1 E. purpurea herb freeze-dried pressed juice1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.2 E. purpurea herb pressed juice3 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.3 E. purpurea standardized2 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.4 95% E. purpurea herb and 5% E. purpurea root (39 mg crude extract/day)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.5 95% E. purpurea herb and 5% E. purpurea root (290 mg crude extract/day)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.6 E. purpurea root (178 mg crude extract/day)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.7 E. purpurea root 450 mg1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.8 E. purpurea root 900 mg1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.9 Tea E. purpurea and angustifolia1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.10 Mixture of E. angustifolia root (50%), E. purpurea root (25%) and E. purpurea herb (25%)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.11 E. purpurea root and E. angustofolia root1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.12 E. angustofolia root extract with CO21 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.13 E. angustofolia root extract with 60% ethanol1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.14 E. angustofolia root extract with 20% ethanol1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
8.15 E. pallida root1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
9 Number of patients reporting adverse effects8 Odds Ratio (M-H, Fixed, 95% CI)Totals not selected
9.1 E. purpurea herb pressed juice3 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
9.2 E. purpurea standardized1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
9.3 95% E. purpurea herb and 5% E. purpurea root (39 mg crude extract/day)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
9.4 95% E. purpurea herb and 5% E. purpurea root (290 mg crude extract/day)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
9.5 E. purpurea root (178 mg crude extract/day)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
9.6 Tea E. purpurea and angustifolia1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
9.7 Mixture of E. angustifolia root (50%), E. purpurea root (25%) and E. purpurea herb (25%)1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
9.8 E. purpurea root and E. angustofolia root1 Odds Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
Analysis 2.1.

Comparison 2 Echinacea versus placebo to treat patients with common cold, Outcome 1 Duration: mean difference.

Analysis 2.2.

Comparison 2 Echinacea versus placebo to treat patients with common cold, Outcome 2 Duration (inverse variance).

Analysis 2.3.

Comparison 2 Echinacea versus placebo to treat patients with common cold, Outcome 3 Total severity and duration measures (for example, area under the curve).

Analysis 2.4.

Comparison 2 Echinacea versus placebo to treat patients with common cold, Outcome 4 Sum score after 2 to 4 days.

Analysis 2.5.

Comparison 2 Echinacea versus placebo to treat patients with common cold, Outcome 5 Sum scores after 5 to 10 days.

Analysis 2.6.

Comparison 2 Echinacea versus placebo to treat patients with common cold, Outcome 6 Early treatment of prodromi: patients developing a 'full' cold.

Analysis 2.7.

Comparison 2 Echinacea versus placebo to treat patients with common cold, Outcome 7 Number of patients dropping out due to adverse effects.

Analysis 2.8.

Comparison 2 Echinacea versus placebo to treat patients with common cold, Outcome 8 Number of patients dropping out.

Analysis 2.9.

Comparison 2 Echinacea versus placebo to treat patients with common cold, Outcome 9 Number of patients reporting adverse effects.

Appendices

Appendix 1. Embase.com search strategy

#10 #1 AND #9
#9 #4 NOT #8
#8 #5 NOT #7
#7 #5 AND #6
#6 'human'/de
#5 'animal'/de OR 'nonhuman'/de OR 'animal experiment'/de
#4 #2 OR #3
#3 random*:ab,ti OR placebo*:ab,ti OR crossover*:ab,ti OR 'cross-over':ab,ti OR (doubl* NEXT/1 blind*):ab,ti OR allocat*:ab,ti OR trial:ti
#2 'randomized controlled trial'/exp OR 'single blind procedure'/exp OR 'double blind procedure'/exp OR 'crossover procedure'/exp
#1 1922
#1.6 #1.1 OR #1.2 OR #1.3 OR #1.4 OR #1.5
#1.5 coneflower*:ab,ti
#1.4 'e. purpurea':ab,ti OR 'e. pallida':ab,ti OR 'e.angustifolia':ab,ti
#1.3 'echinacea purpurea'/de OR 'echinacea extract'/de OR 'echinacea purpurea extract'/de OR 'echinacea pallida extract'/de OR 'echinacea angustiflora extract'/de
#1.2 echinac*:ab,ti
#1.1 'echinacea'/exp

Appendix 2. CINAHL (EBSCO) search strategy

S15 S5 and S14
S14 S6 or S7 or S8 or S9 or S10 or S11 or S12 or S13
S13 TI placebo* OR AB placebo* S
S12 TI clinic* W1 trial* OR AB clinic* W1 trial*
S11 (MH "Quantitative Studies")
S10 (MH "Placebos")
S9 TI random* OR AB random*
S8 TI ((singl* or doubl* or tripl* or trebl*) W1 (blind* or mask*)) OR AB ((singl* or doubl* or tripl* or trebl*) W1 (blind* or mask*))
S7 PT clinical trial
S6 (MH "Clinical Trials+")
S5 S1 or S2 or S3 or S4
S4 TI coneflower* OR AB coneflower*
S3 TI ("E. purpurea" or "E. pallida" or "E. angustifolia") OR AB ("E. purpurea" or "E. pallida" or "E. angustifolia")
S2 TI echinac* OR AB echinac*
S1 (MH "Echinacea")

Appendix 3. AMED (Ovid) search strategy

1 exp echinacea/
2 echinac*.tw.
3 1 or 2
4 randomized controlled trials/
5 exp clinical trials/
6 random allocation/
7 double blind method/
8 (clin* adj25 trial*).tw.
9 ((singl* or doubl* or trebl* or tripl*) adj25 (blind* or mask*)).tw.
10 placebos/
11 placebo*.tw.
12 random*.tw.
13 or/4-12
14 3 and 13

Appendix 4. LILACS (BIREME) search strategy

> Search > MH:"Echinacea angustifolia" OR MH:"Echinacea purpurea" OR MH:echinacea OR MH:HP4.018.251.105 OR MH:HP4.018.251.116 OR MH:B01.650.940.800.575.100.100.310 OR echinac$ OR "E. angustifolia" OR "E. pallida" OR "E. purpurea"

Appendix 5. Web of Science (Thomson Reuters) search strategy

# 3 209

#2 AND #1

Databases=SCI-EXPANDED, CPCI-S Timespan=All Years

Lemmatization=On  

# 2 1,292,302

Topic=(random* or placebo* or ((singl* or doubl*) NEAR/1 blind*) or allocat* or crossover* or "cross over") OR Title=(trial)

Databases=SCI-EXPANDED, CPCI-S Timespan=All Years

Lemmatization=On  

# 1 1,605

Topic=(echinac* OR "E. purpurea" OR "E. angustifolia" OR "E. pallida" OR coneflower*)

Databases=SCI-EXPANDED, CPCI-S Timespan=All Years

Lemmatization=On  

Appendix 6. Previous searches

For the first publication of this review in The Cochrane Library, 1999, Issue 1 (Melchart 1999) the following sources were searched:

  1. MEDLINE (1966 to 1998): all hits for Echinac* screened;

  2. EMBASE (1991 to 1998): all hits for Echinac* screened;

  3. the Cochrane Acute Respiratory Infections Group Specialized Register: all hits for Echinac* screened;

  4. the database of the Cochrane Field Complementary Medicine: all hits for Echinac* screened;

  5. the database Phytodok (Munich, specialized on Phytomedicine) screening all clinical studies for Echinac*;

  6. bibliographies of identified articles;

  7. existing reviews;

  8. manufacturers and researchers in the field (who were contacted to identify published and unpublished trials);

  9. proceedings of phytomedicine congresses (International Congresses on Phytomedicine and Congresses of the German Society of Phytotherapy) (screened).

For the 2007 update we searched the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2007, Issue 3); PubMed (1997 to September 2007); EMBASE (1998 to April 2007); AMED (to August 2005) and the Centre for Complementary Medicine Research (in Munich) (1988 to September 2007). We also contacted experts and screened references of reviews.

We searched PubMed and CENTRAL using the following terms combined with the highly sensitive search strategy devised by Dickersin (Dickersin 1994).

1 exp ECHINACEA/
2 Echinacea
3 or/1-2

We searched EMBASE and AMED using adapted terms. We searched the database of the Centre of Complementary Medicine Research in Munich for controlled trials of Echinacea.

We screened bibliographies of identified trials and review articles for further potentially relevant publications. We contacted experts in the field and asked about further published and unpublished studies.

Feedback

Duration of Echinacea dosage

Summary

I wish to comment on the Cochrane review 'Echinacea for preventing and treating the common cold'. It is claimed in the Hot Topic of the Month (Relief from coughs and colds), August 2001, p. 5, para. 6.1, that "the German drug regulatory authority recommends that it be used for no longer than eight weeks at a time". I have asked the Consumer Network about the evidence for this and been told that it is not available. Nevertheless, I think that if it is indeed a recommendation of the German drug regulatory authority, it should be mentioned in both the review and the abstract.

I certify that I have no affiliations with or involvement in any organisation or entity with a direct financial interest in the subject matter of my criticisms.

Reply

We appreciate this comment. It is correct that the German drug regulatory authority recommends that Echinacea preparations should not be taken for longer than eight weeks at a time. While there is no evidence that longer intake can be harmful, such a precaution seems justified in the absence of data on long-term use. We included a statement on this issue in the review conclusion section 'Implications for practice'.

Klaus Linde

Contributors

David Potter
Comment posted 02/06/2005

What's new

DateEventDescription
2 July 2014AmendedA mistake in the Abstract regarding treatment trials reporting data on the duration of colds has been corrected.

History

Protocol first published: Issue 4, 1997
Review first published: Issue 1, 1999

DateEventDescription
5 June 2013New citation required and conclusions have changed

Change to authorship byline: new first author Marlies Karsch-Völk. Dieter Melchart was not involved in this update.

Inclusion criteria changed: now only randomized controlled trials and also studies on induced rhinovirus infections included.

Outcome measures changed: duration of cold is now a primary outcome of treatment trials.

Conclusions changed: possibly slight effect of Echinacea in the prevention of colds. Evidence for treatment effectiveness is weak.

5 June 2013New search has been performed

Searches updated. Seven new studies were included (Barrett 2010; Hall 2007; Jawad 2012; O'Neill 2008; Tiralongo 2012; Turner 2005; Zhang 2003). Two formerly excluded studies are now included (Sperber 2004; Turner 2000).

One formerly included study is now excluded (Spasov 2004) and 12 new trials were excluded (Di Pierro 2012; Hauke 2002; Heinen-Kammerer 2005; Isbaniah 2011; Minetti 2011; Narimanian 2005; Naser 2005; Saunders 2007; Schapowal 2009; Schoop 2006a; Wahl 2008; Yakoot 2011).

15 March 2010New search has been performedSearches conducted
6 August 2009AmendedContact details updated.
8 May 2009AmendedContact details updated.
16 January 2008AmendedConverted to new review format.
5 October 2007New search has been performedConclusions remain unchanged.
12 September 2005New search has been performedSearches conducted.
1 June 2005Feedback has been incorporatedFeedback and reply added.
16 November 1998New search has been performedReview first published.

Contributions of authors

Co-ordination of the review update: MKV
Planning of updates: MKV, KL, BB, DK
Searches in addition to searches done by the Cochrane ARI Group: MKV
Study selection, data extraction and quality assessment: MKV, KL, BB, DK
Extraction of pharmaceutical data/pharmaceutical expertise: KAW, RB
Statistical analyses: MKV, KL
Drafting of the manuscript: MKV, KL
Interpretation of results, critical feedback on draft versions: BB, DK, KAW, RB 

Declarations of interest

Klaus Linde was involved in one, Karin Ardjomand-Woelkart in one, Bruce Barrett in two and Rudolf Bauer in four of the studies included in this review. Marlies Karsch-Völk and David Kiefer have no conflict of interest. Authors did not carry out data extraction or quality assessment of studies they were involved in.

Sources of support

Internal sources

  • Centre for Complementary Medicine Research, Technische Universität München, Germany.

  • Institute of Pharmaceutical Sciences, Department of Pharmacognosy, Karl-Franzens-University, Graz, Austria.

  • Department of Family Medicine, University of Wisconsin, USA.

External sources

  • National Center for Complementary and Alternative Medicine at the U.S. National Institutes of Health (RO1 AT001428), USA.

  • Robert Wood Johnson Foundation Generalist Physician Scholars Program, USA.

Differences between protocol and review

Change to authorship byline: New first author is Marlies Karsch-Völk. Dieter Melchart was not involved in this update.
Inclusion criteria changed: Now only randomized controlled trials and also studies on induced rhinovirus infections have been included.
Outcome measures changed: Duration of cold is now a primary outcome of treatment trials.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Barrett 2002

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 73 randomized to Echinacea, 75 to placebo; 69 analyzed in the Echinacea group and 73 in the placebo group
Setting: University Family Medicine Dept., USA

Participants: university students

Demographics: mean age 21 years, 69% female

Main selection criteria: at least 2 of 15 cold symptoms for less than 36 hours

Interventions

Echinacea: capsules containing 50% E. angustifolia root (123 mg per capsule), 25% E. purpurea root (62 mg), 25% E. purpurea herb (62 mg) and thyme and peppermint to disguise taste and flavor
Placebo: capsules containing 333 mg alfalfa
Dosage and treatment duration: 6 x 4 capsules in the first 24 hours, then 3 x 4 capsules up to 10 days

Concurrent medication: "Patients using antibiotics, antihistamines or decongestants were excluded." No further information on the actual intake of concurrent medication is reported

Outcomes

Primary: severity score, duration, severity of single symptoms (daily reporting)

Secondary: not defined

Notes

Funding source: U.S. Department of Health and Human Services and the National Institutes of health; Shaklee Tecnica

Conflict of interest: none disclosed
Additional information provided by author

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"...random number-generator (MS Excel) and a balanced blocks-of-four design."
Allocation concealment (selection bias)Low riskSequentially labeled bottles, allocation concealed until data had been collected, entered and cleaned
Blinding (performance bias and detection bias)
All outcomes
Low risk

Exit interview: blinding successful

Placebo and Echinacea indistinguishable

Incomplete outcome data (attrition bias)
All outcomes
Low risk< 5% drop-outs, available case analysis considered appropriate
Selective reporting (reporting bias)Low risk

Relevant outcomes presented in the manuscript

Additional information provided from author

Barrett 2010

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 184 randomized to the blinded Echinacea group, 179 to the blinded placebo group, 183 analyzed in the blinded Echinacea group and 176 in the blinded placebo group

Setting: 2 centers in Wisconsin

Participants: recruited via newspaper advertisements, posters community talks, targeted mailings, e-mails and word of mouth

Demographics: mean age 34 years, 65% female

Main selection criteria: cold symptoms since up to 36 hours

Interventions

Echinacea: tablets containing 675 mg of E. purpurea root and 600 mg of E. angustifolia root, each standardized to 2.1 mg of alkamides. Manufacturer MediHerb (Australia)

Placebo: "Tablet excipients included calcium acid phosphate, cellulose, silica, sodium starch glycolate, hypromellose and magnesium stearate. Placebo and Echinacea tablets contained the same proportion of inert ingredients and were covered with identical digestible coatings."

Dosage and treatment duration: 2 tablets doses 4 times within the first 24 hours of enrolment, then 1 tablet 4 times daily for 4 days; that means 10.2 g of dried Echinacea root during first 24 hours, 5.1 g during each of the next days

Concurrent medication: "Patients receiving antibiotics, antivirals, nasal steroids, decongestants, antihistamines, combination of cold formulas, Echinacea, zinc or vitamin C were excluded." No further information on the actual intake of concurrent medication is provided

Outcomes

Primary: area under the curve for global severity, with duration and severity assessed twice daily by self report, duration, severity score (Wisconsin Upper Respiratory Symptom Survey, short version; WURSS-21)

Secondary: self report on psycho-social questionnaires and biomarker of immune response and inflammation

Notes

Funding source: National Center for Complementary and Alternative Medicine at the National Institute of Health, Robert Wood Johnson Foundation Generalist Physician Faculty Scholars Program; MediHerb (Queensland , Australia) provided the placebo and Echinacea tablets and conducted the phytochemical assays; financial support facilitated by Deans Robert Golden and Paul DeLuca of the University of Wisconsin Medical School

Conflict of interest: no relevant conflict of interest reported

The trial had 2 additional arms (open Echinacea and no treatment) to investigate placebo and expectation effects in colds. These 2 arms were not included in this review

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"We used SAS (SAS Institute, Cary, North Carolina) to generate a single block of 804 unique identification numbers so that each of 12 cells (3 clinician groups by 4 pill groups) was represented equally."
Allocation concealment (selection bias)Low risk"Using these codes, the UW Hospitals Pharmaceutical Research Center Investigational Drug Service prepared consecutively numbered, sealed envelopes to direct allocation. An envelope-within envelope strategy was used, so that group assignment would be revealed as soon as the participant gave consent and the research assistant opened the larger outer envelope.
Allocation concealment for the 2 blinded pill groups was accomplished by using identical coated tablets and plastic pill bottles. For the 2 thirds of the sample who would see a clinician, a second, smaller envelope that directed allocation to a standard or enhanced visit group was opened by the study clinician before entering the examination room.
The randomized allocation key was not shared with investigators until after all data were collected, entered and cleaned and analysis strategies were determined."
Blinding (performance bias and detection bias)
All outcomes
Low risk"Blinding was tested at the exit interview by asking participants which group they thought they had been assigned to." Group differences were not statistically significant. "Placebo and echinacea tablets were covered with identical digestible coatings." (There were also an open-label group and a no treatment group examined in the study, which are not taken into account for this review)
Incomplete outcome data (attrition bias)
All outcomes
Low risk"... 719 were enrolled and randomly assigned. Retention was high. 2 participants were lost to follow-up and 4 withdrew before primary outcome data could be gathered;..."
Selective reporting (reporting bias)Low riskAll predefined main outcome measures and secondary outcomes are well reported

Brinkeborn 1999

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 3 Echinacea groups and 1 placebo group, 559 randomized, 246 got a cold and started treatment, 2 dropped out, 181 in per protocol analysis (41/49/44/46), 246 in intent-to-treat analysis (55/64/63/64)
Setting: infectious disease center in Sweden

Demographics: 74% female, mean age 41 years

Main selection criteria: healthy volunteers, prone to common cold

Interventions

Echinacea 1: Echinaforce tablets (6.78 mg E. purpurea crude extract based on 95% herb and 5% roots)
Echinacea 2: concentrate preparation (48.27 mg of the same extract)
Echinacea 3: 29.6 mg crude extract based on root only

Placebo: not described
Dosage and treatment duration: 3 x 2 tablets daily up to a maximum of 7 days

Concurrent medication: patients were excluded if they were taking "other medications which may affect the immune system like immunostimulants and antibiotics or may influence the symptoms like nose-drops or anticoughs". No further information on the actual intake of concurrent medication is reported

Outcomes

Primary: relative reduction of the complaint index (= sum score) based on 12 symptoms according to doctor's record

Secondary: relative reduction of the complaint index according to the patient's diary, assessment of efficacy and tolerance by the investigator and the patients, the frequency and severity of adverse events

Notes

Funding source: Bioforce AG, Switzerland

Conflict of interest: not stated

Additional information provided by sponsor

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"computer-generated randomization list in blocks of four"
Allocation concealment (selection bias)Low risk"consecutively numbered drug bottles"
Blinding (performance bias and detection bias)
All outcomes
Low risk"The trial doses of the four treatments were presented in identical vials with identical labels and could almost not be distinguished from one another by their smell or taste."
Incomplete outcome data (attrition bias)
All outcomes
Low riskOnly 2 drop-outs and ITT-analysis available (although methods for replacing data not clear)
Selective reporting (reporting bias)Low riskPredefined relevant outcomes adequately reported

Bräunig 1992

Methods

Approach: treatment trial

Design: randomized; placebo-controlled. Comparison of the lower dosage of Echinacea and placebo probably double-blind, group receiving the higher dosage probably not blinded (it seems that all patients received 90 drops placebo)

Participants

Numbers: 60 persons were randomized to treatment group 1, 60 to treatment group 2 and 60 to placebo group, only 1 drop-out in the placebo group
Setting: 1 general practice, Germany

Demographics: not reported

Main selection criteria: influenza-like URTI

Interventions

Echinacea 1: Echinacea purpurea root extract 90 drops (450 mg) daily for 8 to 10 days
Echinacea 2: Echinacea purpurea root extract 180 drops (900 mg) daily for 8 to 10 days
Placebo: mixture of ethanol and water (50 vol. %) with caramel color

Dosage and treatment duration: Echinacea (see above), for placebo not mentioned explicitly

Concurrent medication: patients were excluded, if they were using antihistamines, antibiotics or other relevant medication influencing the clinical picture. No further information on the actual intake of concurrent medication is provided

Outcomes

Duration of disease was described as a main outcome measure but results were not reported

2 scores on 5 medical findings (assessed by the physician after 3 to 4 and 8 to 10 days) and 8 symptoms (assessed by the patient)

Notes

Funding source: Fink GmbH Herrenberg, Firma Salus-Haus Bruckmühl

Conflict of interest: not stated

Study not truly double-blind
Additional information provided by author

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk"randomized", not described
Allocation concealment (selection bias)Unclear riskNot described
Blinding (performance bias and detection bias)
All outcomes
Unclear risk

For treatment group 1: low

For treatment group 2: high

Double-blinding stated but 1 group received more medication (180 drops instead of 90 drops in the first Echinacea group and in the placebo group) and was therefore not truly double-blind

Incomplete outcome data (attrition bias)
All outcomes
Low risk1 drop-out in the placebo group is mentioned in the publication, but it is not stated if the results of this participant were included in the analysis. (Author confirmed in personal communication that there were no drop-outs and withdrawals)
Selective reporting (reporting bias)High riskDuration of disease was described as a main outcome measure but results were not reported

Dorn 1997

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 160 randomized to Echinacea group and analyzed, 160 randomized to placebo group and analyzed
Setting: 1 general practice, Germany

Demographics: 48% female, age not reported

Man inclusion criteria: upper respiratory tract infection (viral origin was suspected in 114 patients (70 Echinacea versus 44 placebo), bacterial in 46 (10 Echinacea versus 36 placebo))

Interventions

Echinacea:Echinacea pallida root extract
Placebo: "coloured aqueous alcoholic solution"

Dosage and treatment duration: 90 drops (900 mg Echinacea) daily for 8 to 10 days

Concurrent medication: patients were excluded if they "were being treated with other drugs that might interact with a herbal preparation". No detailed information on the actual intake of concurrent medication is provided

OutcomesLength of illness, symptoms, clinical findings (assessed after 3 to 4 and 8 to 10 days), lymphocytes and granulocytes, effect of the frequency of URTI during the previous 3 years on the treatment
Notes

Funding source: not stated, but probably funding by Pascoe GmbH, Germany

Conflict of interest: not stated
Additional information provided by author

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated randomization list (provided by author)
Allocation concealment (selection bias)Low risk"neutral packaging with appropriate code numbers" (author confirmed "consecutively numbered")
Blinding (performance bias and detection bias)
All outcomes
Low risk"The study was double blind: ...received a coloured aqueous alcoholic solution that mimicked and was indistinguishable..."
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskReport suggests that there were no drop-outs or withdrawals
Selective reporting (reporting bias)High riskInsufficiently reported trial; relevant group differences regarding suspected cause (more bacterial infections in the placebo group than in Echinacea group: 36 versus 10)

Galea 1996

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 395 randomized, 235 returned symptom checklist, 192 had a cold, 2 incomplete checklists due to drop-out, 106 participants in the Echinacea group and 84 in the placebo group completed the trial
Setting: Family Medicine Department, Canada

Demographics: no information

Main selection criteria: healthy students

Interventions

Echinacea: capsules with 250 mg E. angustifolia (standardized on 4% content of echinacoside)
Placebo: placebo capsules (filled with vegetable oil)
Dosage and treatment duration: 3 x 1 capsule for 10 days

Concurrent medication: not reported

OutcomesNumber of symptoms (4 major and 4 minor; single and summed) over 10-day period
Notes

Funding source: the authors thank C.E. Jamieson and Co. Ltd. and R. P. Scherer Canada Inc (Windsor, Ontario) for supplying the Echinacea, placebo and remuneration for the trial participants. No direct financial support stated

Conflict of interest: not stated

Simple, unpublished trial; only short report available

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot mentioned
Allocation concealment (selection bias)Unclear riskInsufficient information
Blinding (performance bias and detection bias)
All outcomes
Unclear riskPlacebo capsules were filled with vegetable oil. They may have been distinguishable from Echinacea capsules by taste
Incomplete outcome data (attrition bias)
All outcomes
High risk> 20% attrition for unclear reasons, not partitioned between treatment and placebo groups
Selective reporting (reporting bias)Unclear riskReporting does not seem systematically biased but several outcomes are reported insufficiently to allow effect size calculation

Goel 2004

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 282 randomized, 128 contracted a cold and were included in the analysis (ITT: 59 Echinacea, 69 placebo; PP: 54 Echinacea, 57 placebo)
Demographics: 54% in Echinacea and 75% in placebo group female, mean age 32 years

Setting: University of Alberta, Canada

Main selection criteria: healthy adults with at least 2 colds last year

Interventions

Echinacea: Echinacea purpurea extract containing 0.25 mg/ml alkamides, 2.5 mg/ml cichoric acid, 25.5 mg/ml polysaccharides
Placebo: "Placebo made to look, taste and smell like the echinacea extract"
Dosage and treatment duration: 10 x 4 ml the first day, then 4 x 4 ml for 6 days

Concurrent medication: "Volunteers were excluded if they...were on immunosuppressive drugs such as corticosteroids or cyclosporine...The subjects were instructed not to take any other medication during the treatment." Subjects using "concomitant relief medications on a regular basis during the treatment period" were excluded from the PP population

OutcomesTotal daily symptom score, 13 symptoms (reported daily by patients and on day 3 and 8 by nurse)
Notes

Funding source: treatment provided by Factors R&D Technologies, Canada

Conflict of interest: not stated

5 (Echinacea) versus 12 (placebo) drop-outs/excluded patients
Additional information provided by author

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"The randomization list was generated using a computer program."
Allocation concealment (selection bias)Low riskNumbered drugs (information from author)
Blinding (performance bias and detection bias)
All outcomes
Low risk"Placebo made to look, taste and smell like the echinacea extract." Blinding tested after completion of the study and was maintained adequately during the treatment period
Incomplete outcome data (attrition bias)
All outcomes
Low riskTreatment group with 9% attrition, placebo group with 17% attrition, ITT analysis done. Plausible method for replacing missing data described
Selective reporting (reporting bias)Low riskThe published report includes all expected outcomes, including those that were pre-specified

Goel 2005

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 150 randomized, 62 contracted a cold (26 Echinacea, 36 placebo), 56 were analyzed (25 Echinacea, 31 placebo)
Setting: University of Alberta, Canada

Demographics: 59% female, mean age 40 years

Main selection criteria: volunteers between 18 and 65 years with 2 or more colds previous year

Interventions

Echinacea: Echinacea purpurea extract containing 0.25 mg/ml alkamides, 2.5 mg/ml cichoric acid, 25.5 mg/ml polysaccharides
Placebo: "The placebo contained similar ingredients without the echinacea." (40% ethanol)
Dosage and treatment duration: 8 x 5 ml the first day, then 3 x 5 ml for 6 days

Concurrent medication: "Volunteers were excluded if they...were on immunosuppressive drugs such as corticosteroids or cyclosporine...Participants reporting the concomitant use of other relief medications on a regular basis during their cold were excluded from the study."

OutcomesDiary with 8 cold symptoms, multiple physiological measures
Notes

Funding source: Echinacea and placebo "were provided by Factors R & D Technologies, Burnaby, BC, Canada." No direct financial support stated

Conflict of interest: not stated

Subsequent trial after the Goel 2004 trial focusing on physiological outcomes and reporting clinical outcomes only briefly. Author (T. K. Basu) provided additional information. Patients in the Echinacea group had higher symptom scores at baseline. Analyses of % change from day 1 showed significant differences in favor of the Echinacea group, however, absolute values (provided for us by the author) at defined time points were similar in both groups

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"The randomization codes were generated using an Excel 97 computer program."
Allocation concealment (selection bias)Low riskNumbered bottles
Blinding (performance bias and detection bias)
All outcomes
Low riskDouble-blind ("indistinguishable as to appearance, color, or flavor")
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskAlmost 10% of the participants excluded from analysis, no ITT analysis performed
Selective reporting (reporting bias)Low riskPublication includes all expected outcomes, including those that were pre-specified

Grimm 1999

Methods

Approach: prevention trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 109 randomized, 108 analyzed (54 Echinacea, 54 placebo)
Setting: 1 general practice, Germany

Demographics: 62% female, mean age 40 years

Main selection criteria: volunteers with more than 3 URTIs in the previous 6 months of winter

Interventions

Echinacea: pressed juice of Echinacea purpurea herb in 22% alcohol

Placebo: alcohol/water solution with artificial color
Dosage and treatment duration: 2 x 4 ml daily for 8 weeks

Concurrent medication: "Use of immunostimulating drugs within 4 weeks before study entry" was an exclusion criterion. No further information on the actual intake of concurrent medication reported

Outcomes

Primary: incidence, number, severity and duration of infections; time to first infection, CD4/CD8-ratio

Secondary: side effects

Notes

Funding source: "The study was sponsored by Madaus AG, Cologne, Germany."

Conflict of interest: not stated
No additional information sought

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"computer-generated randomization list"
Allocation concealment (selection bias)Unclear riskProbably numbered drug containers, not clearly described
Blinding (performance bias and detection bias)
All outcomes
Low risk"According to the manufacturer, verum and placebo were indistinguishable as to appearance, color, flavor." Adequacy of blinding assessed at the end of follow-up: "27 (50%) of 54 patients in the echinacea group stated that they would like to continue to take the allocated medicine compared to 23 (43%) of 54 patients in the placebo group (P=0.54)"
Incomplete outcome data (attrition bias)
All outcomes
Low riskDrop-outs: 1 of 109 (before starting treatment)
Selective reporting (reporting bias)Low riskExtractable data provided

Hall 2007

MethodsApproach: prevention trial
Design: randomized; placebo-controlled, double-blind
Participants

Numbers: 32 analyzed (Echinacea: 18; placebo: 14); number randomized not reported

Setting: not reported, probably Dept. of Kinesiology, Elmhurst College (USA)

Demographics: mean age 26 years, gender not reported

Main selection criteria: active, non-smoking adults aged 19 to 46 years

Interventions

Echinacea: E. purpurea herb 1.2 g in capsules

Placebo: gelatin capsules containing sugar mixture

Dosage and treatment duration: 8 capsules/day for 4 weeks

Concurrent medication: "Only those subjects that were...not taking any medications and / or dietary supplements...were included in the study." No further information on the actual intake of concurrent medication is provided

OutcomesNumber and duration of URTI additionally to salival IgA, secretion rate of IgA and salival flow rate
Notes

Funding source: study medication was provided by Nature's Way, Springville (USA), no financial support reported

Conflict of interest: not stated

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"...randomly assigned to a group based on a computer generated table of random numbers..."
Allocation concealment (selection bias)Low risk"The placebo and echinacea capsules were placed into identical containers... and were randomly assigned and coded, by a third party,to assure allocation concealment"
Blinding (performance bias and detection bias)
All outcomes
Low riskPlacebo: "same weight, shape and color as the echinacea capsules", identical containers
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNumber randomized and drop-outs not mentioned
Selective reporting (reporting bias)Unclear riskGender not reported, side effects not examined/reported, relevant results without numerical standard deviation or not shown numerically, unclear if other outcomes have been examined

Hoheisel 1997

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 120 randomized and analyzed (60 Echinacea, 60 placebo)
Setting: company physician of a large furniture-making factory

Demographics: 10% female, mean age 36.5 years

Main inclusion criteria: patients presenting with first symptoms of an URTI and having a history of recurrent URTI (> 3 episodes in the previous 12 months) (employees of the furniture-making factory)

Interventions

Echinacea: pressed juice of Echinacea purpurea herb
Placebo: "identical in color and ethanol concentration"
Dosage and treatment duration: on day 1 every 2 hours 20 drops, then up to maximally 10 days 3 x 20 drops daily

Concurrent medication: "Nine patients in the Echinagard group and 4 in the placebo group reported taking concomitant medication, mainly analgesics and anti-ulcer agents." These patients were at least included in the ITT-analysis

Outcomes

Primary: number of patients who developed a 'full' common cold and days until improvement

Secondary: symptom diary, global assessments

Notes

Funding source: not stated
Conflict of interest: not stated
No diary data presented, subjective patient definition what was considered a 'full' cold (could be a major problem if patients should have been unblinded)

No additional information received

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Patients were randomized to treatment groups (using the programme Random V5.0)..."
Allocation concealment (selection bias)Unclear riskNot described
Blinding (performance bias and detection bias)
All outcomes
Unclear risk"Placebo identical in colour and ethanol concentration; identical bottles". Taste not mentioned, no test of blinding, no description of development/testing of liquid placebo
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll 120 patients randomized completed the study and were analyzed
Selective reporting (reporting bias)High risk"Recorded subjective symptoms daily in diary card" not shown in tables. Subsample record only of those who report "real cold" may introduce bias

Jawad 2012

Methods

Approach: prevention trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: Echinacea: 379 randomized, 325 analyzed; placebo: 376 randomized, 348 analyzed

Setting: Common Cold Center in Cardiff University (United Kingdom)

Demographics: Echinacea: 68.7% female, mean age 23.6. Placebo: 62.7% female, mean age 23.2

Main selection criteria: healthy adults with 2 or more colds per year

Interventions

Echinacea: alcoholic extraction from freshly harvested E. purpurea with 95% herb and 5% roots

Placebo: "Placebo drops were similar in shape, color, consistency, odor, flavor and they contained the same amount of alcohol."

Dosage and treatment duration: 3 x 0.9 ml per day for illness prevention (2400 mg of extract per day), during acute stages of cold dose was increased to 5 x 0.9 ml per day (4000 mg of extract per day). Treatment duration was 4 months

Concurrent medication: "The exclusion criteria were...currently taking antimicrobial or antiviral medication...corticosteroid treated asthma, medicinal treated atopy or allergy...In the Echinacea and placebo groups, 58 and 88 episodes, respectively, were treated with aspirin, paracetamol, or ibuprofen. Thus, significantly more (+52%) cold episodes in the placebo group were additionally treated with pain medication..."

OutcomesAdverse effects, multiple physiological measures, prevalence of colds, symptom scores
Notes

Funding source: not stated in the publication. Sponsoring by A. Vogel Bioforce AG, Switzerland stated in additional information by the authors

Conflict of interest: no conflicts of interest of first and last author, conflicts of interest of the other authors not stated

Some information relevant for this review is missing in the publication. Additional information was requested from the authors and has partly been provided

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"The randomization code was prepared in block-sizes of 6 with "RANCODE Professional 3.6" program."
Allocation concealment (selection bias)Low risk"Each participant received treatment based on his/her identification number, which was allocated according to the time point of inclusion."
Blinding (performance bias and detection bias)
All outcomes
Low risk"...double blind...", "Placebo drops similar in shape, colour, consistency, odor, flavor and they contained the same amount of alcohol.", "Primary and secondary packaging was identical..." Blinding was tested in 79 persons and was found to be adequate
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskPopulation not clearly described. Drop-outs relatively few, but distribution disproportionate: 54/379 versus 28/376
Selective reporting (reporting bias)High riskMost results adequately reported but N of patients with at least 1 infection not reported. No P value for accumulated number of cold episodes reported. Severity and duration mentioned, but not reported

Kim 2002a

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: Echinacea: 27; placebo: 30 (no treatment: 25)
Setting: Cornell University, Ithaca, NY, USA

Demographics: 55% female, mean age 20 years

Main outcome criteria: healthy students

Interventions

Echinacea: extract from 80% E. purpurea herb, 20% E. angustifolia root
Control 1: placebo: parsley juice and orange extract.
Control 2: no treatment
Dosage and treatment duration: 101 ml containing 1000 mg (dry plant) per day for at least 5 days

Concurrent medication: "Screening criteria included...those who were taking medication for seasonal allergies and those who had taken Echinacea." Further information on the actual intake of concurrent medication is not provided

OutcomesDuration, retrospective assessment of 11 cold symptoms
Notes

Funding source: not stated

Conflict of interest: not stated

Unpublished 3-armed study

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomization not described
Allocation concealment (selection bias)Low riskIdentically labeled bottles
Blinding (performance bias and detection bias)
All outcomes
Low riskPost cold survey to test success of blinding: blinding effective
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskLoss to follow-up and missing data not described
Selective reporting (reporting bias)Low riskStudy focuses on single cold symptoms; no evidence of selective reporting

Lindenmuth 2000

Methods

Approach: treatment trial

Design: placebo-controlled, double-blind; quasi-randomized but adequate concealment likely

Participants

Numbers: 95 allocated, 95 analyzed (Echinacea: 48; placebo: 47)
Setting: nursing and rehabilitation center in York, PN, USA

Demographics: 85% women, mean age 40 years

Main inclusion criteria: employees of a nursing and rehabilitation center with earliest symptoms of a cold

Interventions

Echinacea: tea preparation from aerial parts of E. purpurea and E. angustifolia and E. purpurea root. The Echinacea tea contained small amounts of 2 flavoring components (lemon grass leaf and spearmint leaf)
Placebo: placebo tea (cinnamon bark, ginger rhizome, peppermint leaf, sweet fennel seed, rose hip, papaya leaf, alfalfa leaf)
Dosage and treatment duration: 5 to 6 cups on day 1, titration to 1 cup on day 5

Concurrent medication: "Subjects who were excluded included...those who...were already taking antibiotics." No further information on the actual intake of concurrent medication is reported

OutcomesGlobal assessment of effectiveness and duration
Notes

Funding source: donation of treatment and placebo teas from Traditional Medicinals Inc., no direct financial funding stated

Conflict of interest: not stated
Additional information sought but no answer received

Simple study with insufficient outcome measurement. Not truly randomized but adequately concealed alternate allocation. Duration of symptoms has been coded reciprocally ("1" for "more then 5 days" and "5" for "immediately", "4" for "2 days" etc.). Thus the numeric results in the publication dissemble an effect favoring placebo. Therefore, the group results were exchanged between Echinacea and placebo group for the analysis with regard to the fact that the text of the publication describes results favoring Echinacea

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)High riskNot truly randomized: "...allocation was accomplished by utilization of alternation for assignment..."
Allocation concealment (selection bias)Unclear risk"The random assignment was accomplished by specific trained secretary personnel not associated with the study and having no prior knowledge of the groups or which of the 2boxes of tea bags contained the packets ofEchinacea Plus tea bags and that contained the packets of placebo tea bags. These personnel used a numbered system..."
Blinding (performance bias and detection bias)
All outcomes
Unclear risk"....each subject was given 21 tea bags of the same appearance..." "...this tea... does not have any obvious or easily recognizable flavor characteristics that would make it easily distinguishable from the Echinacea Plus tea by an untrained palate."
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll included patients remained in the study and were analyzed
Selective reporting (reporting bias)Low riskResults for all measured outcomes reported

Melchart 1998

Methods

Approach: prevention trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 302 were randomized; 289 participants were analyzed by intention-to-treat, 214 complied fully with the protocol. (Echinacea 1: 103/99/85; Echinacea 2: 103/100/84; placebo: 96/90/75)
Setting: 5 centers (4 military centers, 1 industrial plant, Munich, Germany)

Demographics: 29% female, mean age 29.5 years

Main inclusion criteria: healthy volunteers

Interventions

Echinacea 1: Echinacea angustifolia root extract (plant extract ratio 1:11 in 30% alcohol)
Echinacea 2: Echinacea purpurea root extract (plant extract ratio 1:11 in 30% alcohol)
Placebo: colored ethanolic solution
Dosage and treatment duration: 2 x 50 drops daily from Monday to Friday for 12 weeks

Concurrent medication: exclusion criteria were: "systemic intake of corticosteroids, antibiotics, or immunostimulants in the previous 2 weeks." Further information on the actual intake of concurrent medication is not provided

OutcomesPredefined main outcome measure: time until occurrence of first URTI
Secondary outcomes: proportions with at least 1 URTI; number, severity and duration of episodes, global assessment. Predefined subgroup analysis on patients with more than 3 infections in the previous 12 months
Notes

Funding source: Bavarian Parliament, Plantapharmazie (Göttingen, Germany)

Conflict of interest: not stated
Treatment preparations characterized by the content of glycoproteins/polysaccharides (additional information from manufacturer)

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"stratified for participants with up to 3 and more than 3 colds in the last year; computer-generated randomization list, block size 15..."
Allocation concealment (selection bias)Low risk"concealment by consecutively numbered medication"
Blinding (performance bias and detection bias)
All outcomes
Unclear riskMore volunteers in treatment groups correctly guessed that they had received a true treatment (P < 0.001; some unblinding likely)
Incomplete outcome data (attrition bias)
All outcomes
Low riskDrop-outs: 58/302, intention-to-treat analysis performed
Selective reporting (reporting bias)Low riskAll relevant outcomes reported

O'Neill 2008

MethodsApproach: prevention trial
Participants

Numbers: 90 randomized, 58 analyzed (Echinacea: 45 randomized, 28 analyzed; placebo: 45 randomized, 30 analyzed)

Setting: University Medical Center and Family Health Center at the University of California San Francisco-Fresno (USA)

Demographics: mean age 40 years, sex not reported

Main selection criteria: healthy adults recruited from hospital personnel

Interventions

Echinacea: E. purpurea dried plant extract in capsules

Placebo: parsley capsules

Dosage and treatment duration: 3 capsules 2 times daily for 8 weeks (300 mg per capsule)

Concurrent medication: "Persons with...immunosuppressive therapy...were excluded...Individuals currently using Echinacea were not considered for the study, whereas those using other upper respiratory tract infection preventive measures were allowed to continue their use...Participants with symptoms were asked about...any medications used to treat symptoms (e.g., aspirin, acetaminophen, vitamins and cold formulas)"

OutcomesDays with symptoms, days missed from work, medications used to treat symptoms, adverse effects
Notes

Funding source: grant from Health Resources and Services Administration Border Health Education and Training Center, Medication used was donated by Natures Resource, Mission Hills, California (USA)

Conflict of interest: none

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"...list generated using the random-number generator in a spreadsheet program (Microsoft Excel....)"
Allocation concealment (selection bias)Unclear riskThe method of concealment is not described in sufficient detail to allow a definite judgement
Blinding (performance bias and detection bias)
All outcomes
Low risk"...indistinguishable in size, color and smell...Participants, the main investigator and all persons involved in the study remained blinded to the identity of each group until data analyses were completed."
Incomplete outcome data (attrition bias)
All outcomes
High riskLarge number of drop-outs (32/90), no ITT analysis
Selective reporting (reporting bias)Low riskAll mentioned outcomes reported

Schulten 2001

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 80 randomized and analyzed in ITT-analysis (ITT: Echinacea: 41; placebo: 39; per protocol: Echinacea: 37; placebo: 33)
Setting: industrial plant in Germany

Demographics: 49% female, mean age 39 years

Main inclusion criteria: patients (employees of the manufacturer) with subjective sensation of a cold and at least 1 of 8 symptoms

Interventions

Echinacea: E. purpurea herb (pressed juice from fresh flowering plants)
Placebo: not described
Dosage and treatment duration: 2 x 5 ml for 10 days

Concurrent medication: "Therapy with immunosuppressants in the week prior to the trial and therapy with immunostimulants (herbal immunostimulants, cytokines, thymus fractions), zinc or antibiotics during 2 weeks before commencement of the trial were not allowed." No further information on the actual intake of concurrent medication is provided

Outcomes

Primary: number of days with illness, number of patients with a "full" cold, area under the curve for the daily symptom score (modified Jackson score) and single symptoms, symptom score

Secondary: subjective efficacy assessment by patients

Notes

Funding source: Madaus AG, Cologne, Germany

Conflict of interest: not stated, but study was performed by and with Madaus employees
Additional information sought but not received

Rigorous and well-reported study, however, performed in employees of the manufacturer without testing the success of blinding. Adaptive design with early stopping after 80 patients as a significant difference was detected

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"computer generated randomization list"
Allocation concealment (selection bias)Low riskConsecutive randomization, identical packaging
Blinding (performance bias and detection bias)
All outcomes
Unclear risk"indistinguishable in terms of appearance, taste, smell, colour and packaging", but we are uncertain: Can Madaus employees recognize the taste of their Echinacea product?
Incomplete outcome data (attrition bias)
All outcomes
Low risk3/80, ITT analysis performed
Selective reporting (reporting bias)Unclear riskMedian duration versus mean (not reported); were URIs defined a priori?

Sperber 2004

Methods

Approach: prevention trial (experimental rhinovirus colds)

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 48 randomized, 46 analyzed (Echinacea: 24; placebo: 22)

Setting: not clearly described, probably Department of Internal Medicine, Hackensack University Center, Hackensack (USA)

Demographics: mean age 33 years, Echinacea: 50% female, placebo: 58% female

Main selection criteria: healthy adults, aged 18 to 65 years, antibody titers of =< 1:2 to Rhinovirus type 39

Interventions

Echinacea: pressed juice of E. purpurea in 22% alcohol

Placebo: "matching placebo", not described

Dosage and treatment duration: 2.5 ml 3 times daily for 14 days (intranasal virus inoculation with RV-39 after 7 days)

Concurrent medication: "Individuals who had received medication known to affect rhinorrhea, cough, or nasal congestion within 7 days (4 weeks for cromolyn sodium and long-acting antihistamines) before study initiation were excluded." Further information on actual intake of concurrent medication is not reported

Outcomes

Primary: occurrence and severity of symptoms

Secondary: increase in RV-39 neutralizing antibody titer and/or recovery of rhinovirus on viral culture

Notes

Funding source: Madaus AG, Cologne, Germany

Conflict of interest: not stated

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomization only stated
Allocation concealment (selection bias)Unclear riskNot described
Blinding (performance bias and detection bias)
All outcomes
Low riskDouble-blinding stated. "The active medication and placebo were identical in appearance, taste and smell and were packaged in identical 100ml bottles."
Incomplete outcome data (attrition bias)
All outcomes
Low riskOnly 2/48 excluded (< 5%)
Selective reporting (reporting bias)Low riskMeasured outcomes well reported

Taylor 2003

MethodsApproach: treatment trial
Design: randomized; placebo-controlled, double-blind
Participants

Numbers: 524 randomized, probably 436 with infection, 407 analyzed (Echinacea: 200; placebo: 207)
Setting: 7 private practices and 1 inner-city clinic, USA

Demographics: 50% female, mean age 5.5 years
Main inclusion criteria: healthy children between 2 and 11 years

Interventions

Echinacea: E. purpurea herb harvested at flowering, in syrup
Placebo: syrup
Dosage and treatment duration: 7.5 ml/day in children 2 to 5 years and 10 ml/day in those 6 to 11 years up to a maximum of 10 days

Concurrent medication: "...and those receiving chronic medications of any kind or herbal, mineral, or specific vitamin supplements were excluded...parents were asked to not give their child any medication other than the study medication and acetaminophen (if desired) unless prescribed by a physician. However, if another medication was administered, the parent was requested to record the name." Use of concomitant medication was reported (secondary outcome)

Outcomes

Primary: duration, severity, adverse events

Secondary: peak severity, number of days with fever, parental global assessment of severity, concomitant medication

Notes

Funding source: grant from the National Center for Complementary and Alternative Medicine. Study medication and placebo provided by Madaus AG, Cologne, Germany
Conflict of interest: not stated
Additional information on results in the first cold episode was provided from authors. Additional publication: Weber 2005. In the "Data and analyses" section of this review the outcome "the number of adverse effects per URI" reported in this trial is used as an equivalent to the outcome "number of patients reporting adverse effects per number of participants"

Rigorous and well-reported study; complex analysis as several infections per child were monitored

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Randomization was performed using a computer-generated randomization list and was stratified by site and in blocks of 10."
Allocation concealment (selection bias)Low risk"...unique study number corresponding to the numbers on the bottles of study medication"
Blinding (performance bias and detection bias)
All outcomes
Low risk"Placebo identical in appearance and similar in taste and smell..." Blinding tested: only 35.1%/22.7% correct guesses
Incomplete outcome data (attrition bias)
All outcomes
Low riskDrop-outs: 36/436. Missing outcome data balanced in numbers across intervention groups, with similar reasons for missing data across groups
Selective reporting (reporting bias)Low riskAll mentioned outcomes have been reported

Tiralongo 2012

Methods

Approach: prevention trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 175 randomized, 170 analyzed (ITT; Echinacea: 85; placebo: 85)

Setting: not clearly described, probably School of Pharmacy, Griffith Health Institute, Griffith University, Gold Coast Australia

Demographics: mean age 43 years, 67% female

Main inclusion criteria: healthy adults, traveling on intercontinental flights, 18 to 65 years

Interventions

Echinacea: tablets containing 112.5 mg E. purpurea root 6:1 extract (equivalent to 675 mg dry root) and 150 mg E. angustofolia root 4:1 extract (equivalent to 600 mg dry root)

Placebo: not clearly described, tablets covered identically

Dosage and treatment duration: days -14 to -3: 1 tablet. twice a day; days -2 to +7: 2 tablets twice a day; +8 to +32 1 tablet twice a day; +33 to +42 2 tablets twice a day; +43 to +49 1 tablet twice a day. For shorter travels the dose between day 8 and 32 was shortened, reflecting the time the patient was spending abroad

Concurrent medication: "Volunteers were excluded if they...were on regular treatment with Echinacea, antibiotics, corticosteroids, antihistamines and immunosuppressants." Further details on the actual intake of concurrent medication are not provided

Outcomes

Primary: symptom score, URI symptom-related quality of life

Secondary: occurrence and duration of jet lag, headache, sleep pattern, herpes simplex scores

Notes

Funding source: Integria Healthcare Pty. Ltd.

Conflict of interest: not stated

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"The random allocation sequence provided by the sponsor was computer generated using a randomization plan from http://www.randomization.com/ with randomization in blocks of 10."
Allocation concealment (selection bias)Low risk"...allocation was concealed by providing each participant with a number...Labelling only identified the patient number."
Blinding (performance bias and detection bias)
All outcomes
Low risk

"Placebo tablets were manufactured to match the Echinacea tablets in size, excipients and colour. Both sets of tablets were coated with a brown colour and hypromellose to make them indistinguishable. Tablets were packed in identical amber glass bottles with identical labelling."

Blinding tested and successful

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk27 patients lost to follow-up (16% drop-outs, no real ITT)
Selective reporting (reporting bias)Low riskAll outcomes reported

Turner 2000

Methods

Approach: prevention trial (experimental rhinovirus colds)

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 117 randomized, 92 challenged with rhinovirus type 23, infection occurred in 22 of 50 in Echinacea group and in 24 of 42 in placebo group

Setting: not clearly stated, probably Department of Pediatrics and Medicine, Medical University of South Carolina (USA)

Demographics: not reported

Main selection criteria: healthy young adults with a titer of neutralizing antibody of ≤ 1:4 to rhinovirus type 23

Interventions

Echinacea: 4% phenolic extract of a mixture of E. purpurea and E. angustofolia formulated as powder, containing 0.16% cichoric acid with almost no echinacosides or alkamides

Placebo: not described

Dosage and treatment duration: 1 capsule containing 900 mg Echinacea of placebo once a day, for 14 days prior to virus challenge and 5 days after virus challenge

Concurrent medication: according to additional information provided by author exclusion criteria were "use of any anti-inflammatory (steroids or NSAIDs) or couch/cold preparation in the 2 weeks prior to the study, use of astemizole in the month prior to the study..." No detailed information on the actual intake of concurrent medication is reported

OutcomesNumber of rhinovirus infections, incidence of clinical colds, severity of symptoms
Notes

Funding source: grant from the Procter & Gamble Company, Cincinnati, Ohio (USA)

Conflict of interest: not stated, but the second author is a Procter & Gamble employee

Publication is a short report. Some additional information (study protocol) was provided by author

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskUnpublished study protocol provided by author states that allocation is randomized but does not report further details
Allocation concealment (selection bias)Unclear riskSee above
Blinding (performance bias and detection bias)
All outcomes
Unclear riskBlinding not mentioned; assessment of blinding with some (non-significant) differences in correct guesses - therefore blinding of participants done; blinding of providers unclear
Incomplete outcome data (attrition bias)
All outcomes
Unclear risk24 drop-outs before virus challenge
Selective reporting (reporting bias)Low riskRelevant outcomes seem to be reported

Turner 2005

Methods

Approach: prevention and treatment trial (experimental rhinovirus colds)

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: the trial consisted of a prophylaxis phase of 7 days, then a challenge with rhinovirus inoculation, followed by a 5-day treatment phase. Participants were allocated to 7 study groups. 3 groups received 1 of 3 different Echinacea preparations in the prophylaxis phase and in the treatment phase. 3 groups received placebo in the prophylaxis phase and 1 of the 3 different Echinacea preparations in the treatment phase. The control group received placebo in both the prophylaxis and the treatment phase. 437 were randomized and 399 analyzed. E1/E1: 52; E2/E2: 52; E3/E3: 45; placebo/E1: 48; placebo/E2: 51; placebo/E3: 48; placebo/placebo: 103 (Echinacea = E)

Setting: University of Virginia School of Medicine, Charlottesville (USA)

Demographics: mean age 21 years, 60% female

Main selection criteria: healthy young adults, serum-neutralizing antibody titer =< 1:4 to rhinovirus type 39

Interventions

Echinacea: E. angustofolia root extract with CO2, with 60% ethanol or with 20% ethanol

Placebo: mixture of alcoholic beverages, denatonium benzoate and tap water

Dosage and treatment duration: 1.5 ml (equivalent of 300 mg Echinacea root) tincture 3 times daily for 13 days (prophylaxis groups) or placebo for 7 days and Echinacea for 5 days (treatment groups), the control group received placebo for 13 days

Concurrent medication: not reported

Outcomes

Primary: symptom score

Secondary: testing of infection with rhinovirus type 39, nasal secretion weight, assessment of inflammation in nasal lavage specimens

Notes

Funding source: grant from the National Center for Complementary and Alternative Medicine (NCCAM) of the National Institutes of Health

Conflict of interest: first author has served as a consultant for Wyeth Consumer Healthcare, Schering-Plough Research Institute, Nordic Phytopharma A7S, the Dial Corporation and Procter & Gamble and having received grant support from Biopolymer Engineering, the Dial Corporation and Procter & Gamble. The last author has received grant support from the U.S. Department of Defense

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk"randomly assigned in blocks" no further description of the sequence generation
Allocation concealment (selection bias)Unclear riskNot clearly described
Blinding (performance bias and detection bias)
All outcomes
Low riskBlinding tested and successful
Incomplete outcome data (attrition bias)
All outcomes
Low riskMissing outcome data balanced in numbers across intervention groups, with similar reasons for missing data across groups
Selective reporting (reporting bias)Low riskRelevant outcomes reported

Yale 2004

Methods

Aproach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 128 (Echinacea: 63, placebo: 65)
Setting: clinic network, USA

Demographics: 86% female, mean age 38 years

Main inclusion criteria: adult patients presenting with acute sneezing and nasal discharge for 6 to 24 hours

Interventions

Echinacea: Echinacea fresh capsule (freeze-dried pressed juice from E. purpurea herb standardized on a content of 2.4% soluble beta-1,2-D-fructofuranosides)
Placebo: capsules with lactose
Dosage and treatment duration: 3 x 100 mg up to a maximum of 14 days

Concurrent medication: "Exclusion from the study occurred if thee patient...received antibiotics, antihistamines, decongestants, nasal sprays, or corticosteroids in the 48 hours before enrolment,...used corticosteroids during the 8 weeks before enrolment...Patients were encouraged not to take any other cold remedies, including antihistamines, decongestants, cough suppressants, corticosteroids, or isotonic sodium chloride solution during the study. Acetaminophen was allowed, if needed." Actual intake is reported

OutcomesSeverity score, single symptoms (daily reporting), duration, global assessment, adverse effects
NotesFunding source: Marshfield Clinic Research Foundation, Wisconsin (USA), Enzymatic Therapy provided study materials
Conflict of interest: not stated
Some additional information received from author
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"...randomization list sent from the Biostatistics and Bioinformatics core to the Investigational Drug Pharmacy..." (information from author)
Allocation concealment (selection bias)Unclear riskNo exact description but all persons were blinded except the clinical pharmacist
Blinding (performance bias and detection bias)
All outcomes
Low risk"...identical-appearing lactose placebo capsule...", "treatments and group assignments remained blinded to patients and investigators, with the exception of the clinical pharmacist." Blinding tested and successful
Incomplete outcome data (attrition bias)
All outcomes
Low riskAccording to information provided by author only 1 non-completer
Selective reporting (reporting bias)Low riskAdditional results provided by author

Zhang 2003

  1. a

    Abbreviations:

    URTI = upper respiratory tract infection
    URI = upper respiratory infection
    ITT = intention-to-treat
    PP = per protocol

Methods

Approach: prevention trial

Design: randomized; placebo-controlled, double-blind

Participants

Numbers: 120 randomized, 111 analyzed (Echinacea: 54; placebo: 57)

Setting: University of Wolverhampton (USA)

Demographics: 57% female, median age 25

Main inclusion criteria: healthy participants, 18 to 65 years old

Interventions

Echinacea: E. purpurea root extract (capsules containing 294 mg E. purpurea root extract, 4.4 mg cichoric acid and 500 mg of non-specified herb powder)

Placebo: capsules containing 500 mg herb powder

Dosage and treatment duration: 1 capsule daily for 8 weeks

Concurrent medication: "The general exclusion criteria were...use of antibiotics / corticosteroids within 2 weeks or immunostimulating drugs within 4 weeks of study entry..." No detailed information on the actual intake of concurrent medication is provided

OutcomesSymptom and severity score, total episodes of URI, missed doses, experience of side effects
Notes

Funding source: not stated

Conflict of interest: not stated

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"...stratified randomization procedure...", "In order of their recruitment subjects in each pre-treatment URI score group were grouped into blocks of 4 and 2 subjects in each block of 4 were allocated randomly to Echinacea and the other 2 to placebo by using separate sets of randomized numbers provided for each of the 2 pre-treatment URI categories. The randomized number sequences were generated by a statistician..."
Allocation concealment (selection bias)Low risk"The Echinacea and placebo dose-pots were then labeled with the numbers that had been allocated to individual subjects..."
Blinding (performance bias and detection bias)
All outcomes
High risk

" Echinacea and the identical placebo doses were enclosed in cellulose capsules...This enabled concealment of colour differences.", "Subjects were advised not to open the capsules since the taste/smell/colour of the contents might lead to identification of the doses."

By testing of blinding 82% in the placebo group correctly identified their group

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskDrop-outs: 9/120, no ITT
Selective reporting (reporting bias)Low riskAll outcomes reported

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Baetgen 1984Not randomized
Not common cold (pertussis)
Baetgen 1988Not randomized
Not common cold (acute bronchitis)
Bendel 1989Not common cold (reduction of immunosuppressive side effects of combined radio-chemotherapy in patients with breast cancer), combination of Echinacea with other herbal extracts
Bendel 1990Not common cold (reduction of immunosuppressive side effects of radiotherapy in patients with breast cancer), combination of Echinacea with other herbal extracts
Blumröder 1985Not common cold (angina lacunaris), combination of Echinacea with other herbal extracts
Coeugniet 1986Not randomized
Not common cold (recurrent vaginal candida infection)
Cohen 2004Randomized trial of a combination
Cubasch 1992Only measurement of laboratory and psychological parameters
Di Pierro 2012Randomized trial about enhancement of immune response to influenza vaccines, not common cold
Dorn 1989Randomized trial of a combination of Echinacea with other herbal extracts
Ehringer 1968Not common cold (venous insufficiency)
Engel 1988Not randomized
Forth 1981Randomized trial of a combination of Echinacea with other herbal extracts
Freitag 1984Not common cold (pertussis)
Freyer 1974Trial of a combination of Echinacea with other herbal extracts, with alternate allocation
Hauke 2002Trial of Esberitox® as supportive therapy to antibiotics, not common cold (acute exacerbation of chronic bronchitis)
Heinen-Kammerer 2005Open study, not a randomized controlled trial
Helbig 1961Trial of a combination of Echinacea with other herbal extracts, with alternate allocation
Henneicke 1999Randomized trial of a combination of Echinacea with other herbal extracts
Hill 1993Not common cold (topical treatment of insect bites)
Hill 1995Not common cold (topical treatment of insect bites)
Hill 1996Not common cold (topical treatment of insect bites)
Isbaniah 2011Trial of a combination of Echinacea with other agents, not common cold (chronic obstructive pulmonary disease)
Kim 2002bMeasurement of physiological outcomes in healthy volunteers
Kleinschmidt 1965Trial of a combination of Echinacea with other herbal extracts with alternate allocation
Melchart 1995aMeasurement of immunological parameters in healthy volunteers
Melchart 1995bMeasurement of immunological parameters in healthy volunteers
Melchart 1995cMeasurement of immunological parameters in healthy volunteers
Melchart 1995dMeasurement of immunological parameters in healthy volunteers
Melchart 1995eMeasurement of immunological parameters in healthy volunteers
Minetti 2011Trial of a combination of Echinacea with other herbal extracts
Narimanian 2005Randomized trial of a combination of Echinacea with other herbal extracts
Naser 2005Randomized trial of a combination of Echinacea with other herbal extracts
Pohl 1970Not randomized
Not common cold (reduction of radiotherapy-induced leukopenia)
Quadripur 1976Not common cold (skin infections)
Reitz 1990Randomized trial of a combination of Echinacea with other herbal extracts
Sartor 1972Not common cold (reduction of radiotherapy-induced leukopenia)
Saunders 2007Open-label study, not a randomized controlled trial
Scaglione 1995Randomized trial of a combination of Echinacea with other herbal extracts
Schapowal 2009Randomized trial of a combination of Echinacea with other herbal extracts
Schmidt 1990Randomized trial of a combination of Echinacea with other herbal extracts
Schoop 2006aOpen study, not a randomized controlled trial
Schwarz 2002Measurement of immunological parameters in healthy volunteers
Spasov 2004Unblinded study without placebo control
Stolze 1986Not randomized
Not common cold (variety of respiratory tract infections treated with antibiotics)
Thom 1997Randomized trial of a combination
Timmermanns 1990Not common cold (urinary dysfunction)
Vonau 2001Randomized trial of Echinacea purpurea for recurrent genital herpes
Vorberg 1984Randomized trial of a combination of Echinacea with other herbal extracts
Vorberg 1989Randomized trial of a combination of Echinacea with other herbal extracts
Wahl 2008Randomized trial of Echinacea in combination with osteopathic treatment in children with recurrent otitis media
Yakoot 2011Randomized trial of a combination of Echinacea with other herbal extracts
Zimmer 1985Not common cold (acute sinusitis)

Characteristics of studies awaiting assessment [ordered by study ID]

Rahmati 2012

Methods

Approach: treatment trial

Design: randomized; placebo-controlled, double-blind

Participants100 children aged 5 to 11 years with upper respiratory tract infection
InterventionsEchinacea root extract or placebo
OutcomesDuration and severity of symptoms
NotesStudy in Persian language identified in the update search

Ancillary