High-carbohydrate, high-protein, low-fat versus low-carbohydrate, high-protein, high-fat enteral feeds for burns

  • Review
  • Intervention

Authors


Abstract

Background

Severe burn injuries increase patients' metabolic needs. Aggressive high-protein enteral feeding is used in the post-burn period to improve recovery and healing.

Objectives

To examine the evidence for improved clinical outcomes in burn patients treated with high-carbohydrate, high-protein, low-fat enteral feeds (high-carbohydrate enteral feeds) compared with those treated with low-carbohydrate, high-protein, high-fat enteral enteral feeds (high-fat enteral feeds).

Search methods

We searched the Cochrane Injuries Group Specialised Register (searched 28 Nov 2011), Cochrane Central Register of Controlled Trials (The Cochrane Library 2011, Issue 4), MEDLINE (Ovid) 1950 to Nov (Week 3) 2011, EMBASE (Ovid), ISI Web of Science: Science Citation Index Expanded (SCI-EXPANDED) (1970 to Nov 2011), ISI Web of Science: Conference Proceedings Citation Index-Science (CPCI-S) (1990 to Nov 2011), PubMed (Searched 28 Nov 2011). Online trials registers and conference proceedings were also searched to April 2010.

Selection criteria

We included all randomized controlled trials (RCTs) comparing high-carbohydrate enteral feeds to high-fat enteral feeds for treatment of patients with 10% or greater total body surface area (TBSA) burns in the immediate post-burn period, with data for at least one of the pre-specified outcomes.

Data collection and analysis

Two authors collected and analysed the following data: mortality, incidence of pneumonia and days on ventilator. Meta-analysis could only be performed for the outcomes mortality and incidence of pneumonia. A random-effects model was used for all comparisons.

Main results

Two RCTs, reporting results from 93 patients, were included in this review. Patients given a high-carbohydrate feeding formula had an odds ratio (OR) of 0.12 (95% confidence interval (CI) 0.04 to 0.39) for developing pneumonia compared to patients given a high-fat enteral formula (P value = 0.0004). Patients given a high-carbohydrate formula had an OR of 0.36 (95% CI 0.11 to 1.15) for risk of death compared to patients given a high-fat enteral formula; this difference did not reach statistical significance (P value = 0.08). Risk of bias in these studies was assessed as high and moderate.

Authors' conclusions

The available evidence suggests that use of high-carbohydrate, high-protein, low-fat enteral feeds in patients with at least 10% TBSA burns might reduce the incidence of pneumonia compared with use of a low-carbohydrate, high-protein, high-fat diet. The available evidence is inconclusive regarding the effect of either enteral feeding regimen on mortality. Note that the available evidence is limited to two small studies judged to be of moderate risk of bias. Further research is needed in this area before strong conclusions can be drawn.

Résumé scientifique

Comparaison entre une alimentation entérale riche en glucides et en protéines mais pauvre en lipides et une alimentation entérale pauvre en glucides, mais riche en protéines et en lipides chez des patients brûlés

Contexte

Des brûlures graves augmentent les besoins métaboliques des patients. Une alimentation entérale agressive et riche en protéines est administrée suite à une brûlure afin d'améliorer leur rétablissement et leur guérison.

Objectifs

Examiner les preuves de résultats cliniques améliorés chez des patients brûlés traités avec une alimentation entérale riche en glucides et en protéines, mais pauvre en lipides (alimentation entérale riche en glucides) par rapport à ceux traités avec une alimentation entérale pauvre en glucides, mais riche en protéines et en lipides (alimentation entérale riche en lipides).

Stratégie de recherche documentaire

Nous avons effectué des recherches dans le registre spécialisé du groupe Cochrane sur les blessures (recherches du 28 novembre 2011), le registre Cochrane des essais contrôlés (The Cochrane Library 2011, numéro 4), MEDLINE (Ovid) de 1950 à novembre (semaine 3) 2011, EMBASE (Ovid), ISI Web of Science : Science Citation Index Expanded (SCI-EXPANDED) (de 1970 à novembre 2011), ISI Web of Science : Conference Proceedings Citation Index-Science (CPCI-S) (de 1990 à novembre 2011), PubMed (recherches du 28 novembre 2011). Les registres d'essais en ligne et les actes de conférence ont également fait l'objet de recherches jusqu'au mois d'avril 2010.

Critères de sélection

Nous avons inclus tous les essais contrôlés randomisés (ECR) comparant une alimentation entérale riche en glucides à celle riche en lipides pour le traitement de patients dont les brûlures représentent au moins 10 % la surface corporelle totale (SCT) immédiatement après la brûlure, avec des données concernant au moins l'un des résultats spécifiés au préalable.

Recueil et analyse des données

Deux auteurs ont collecté et analysé les données suivantes : la mortalité, l'incidence de la pneumonie et les jours de ventilation. Seule une méta-analyse a pu être réalisée pour les résultats relatifs à la mortalité et l'incidence de la pneumonie. Un modèle à effets aléatoires était utilisé pour l'ensemble des comparaisons.

Résultats principaux

Deux ECR, contenant les résultats de 93 patients, étaient inclus dans cette revue. Les patients bénéficiant d'une formule alimentaire riche en glucides affichaient un rapport de cotes (RC) de 0,12 (intervalle de confiance (IC) à 95 % 0,04 à 0,39) pour le développement d'une pneumonie par rapport à ceux bénéficiant d'une formule entérale riche en lipides (valeur de P = 0,0004). Les patients bénéficiant d'une formule riche en glucides affichaient un RC de 0,36 (IC à 95 % 0,11 à 1,15) pour les risques de décès par rapport à ceux bénéficiant d'une formule entérale riche en lipides ; mais cette différence n'a pas atteint de signification statistique (valeur de P = 0,08). Les risques de biais dans ces études étaient considérés comme étant élevés à modérés.

Conclusions des auteurs

Les preuves disponibles suggèrent que l'administration d'une alimentation entérale riche en glucides et en protéines, mais pauvre en lipides chez des patients dont les brûlures représentent au moins 10 % la surface corporelle totale (SCT), peut réduire l'incidence de la pneumonie par rapport à l'administration d'une alimentation pauvre en glucides, mais riche en protéines et en lipides. Ces preuves disponibles ne sont pas concluantes quant aux effets de l'une des deux alimentations entérales sur la mortalité. Notez que ces preuves se limitent à deux études réalisées à petit échelle et dont les risques de biais sont considérés comme étant modérés. Des recherches supplémentaires doivent être effectuées dans ce domaine avant de pouvoir tirer des conclusions fiables.

Resumo

Alimentação enteral rica em carboidratos, hiperproteica e hipolipídica versus alimentação pobre em carboidratos, hiperproteica e hiperlipídica para pacientes queimados

Introdução

Queimaduras graves aumentam as necessidades metabólicas dos pacientes. Alimentação enteral agressiva, rica em proteinas, é usada no período pós-queimadura para melhorar a recuperação e a cicatrização.

Objetivos

Avaliar as evidências de melhores resultados clínicos em pacientes queimados tratados com alimentos ricos em carboidratos, hiperproteicos e hipolipídicos (alimentos enterais hipercalóricos) comparados com aqueles tratados com alimentos pobres em carboidratos, hiperproteicos e hiperlipídicos (alimentos enterais hiperlipídicos).

Métodos de busca

Nós pesquisamos a Cochrane Injuries Group Specialised Register (em 28 de novembro de2011), Cochrane Central Register of Controlled Trials (The Cochrane Library 2011, Issue 4), MEDLINE (Ovid) 1950 até novembro (terceira semana) de 2011, EMBASE (Ovid), ISI Web of Science: Science Citation Index Expanded (SCI-EXPANDED) (1970 a novembro de 2011), ISI Web of Science: Conference Proceedings Citation Index-Science (CPCI-S) (1990 a novembro de 2011), PubMed (em 28 de novembro de 2011). Plataformas de registros de ensaios clínicos e anais de conferências também foram pesquisados até abril de 2010.

Critério de seleção

Nós incluímos todos os ensaios clínicos randomizados (ECR) que compararam alimentos enterais com alto teor de carboidrato versus alimentos enterais com alto teor de gordura, para o tratamento de pacientes com 10% ou mais de área de superfície corporal queimada no período imediato pós-queimadura. Os estudos que tinham dados de pelo menos um dos resultados pré-especificados foram incluídos.

Coleta dos dados e análises

Dois autores coletaram e analisaram os seguintes dados: mortalidade, incidência de pneumonia e dias em ventilação mecânica. A metanálise só pôde ser realizada para os resultados mortalidade e incidência de pneumonia. O modelo de efeitos randômicos foi utilizado para todas as comparações.

Principais resultados

Dois ECRs, envolvendo93 pacientes, foram incluídos nesta revisão. Os pacientes que receberam uma fórmula de alimento rica em carboidratos tiveram razão de chances (odds ratio, OR) de pneumonia de 0,12 (intervalo de confiança de 95%, 95% CI, 0,04 a 0,39), comparados com pacientes que receberam fórmula enteral hiperlipidíca (p = 0,0004). Os pacientes que receberam fórmula rica em carboidratos tiveram OR de morte de 0,36 (95% CI 0,11 a 1,15), comparados com pacientes que receberam fórmula enteral hiperlipídica; essa diferença não foi significante (p = 0,08). O risco de viés desses estudos foi considerado alto e moderado.

Conclusão dos autores

As evidências disponíveis sugerem que, em pacientes com pelo menos 10% da área de superfície corporal queimada, o uso de alimento enteral rico em carboidratos, hiperproteico e hipolipídico pode reduzir a incidência de pneumonia comparado com o uso de alimentação pobre em carboidratos, hiperproteica e hiperlipídica. As evidências disponíveis são inconclusivas a respeito do efeito de qualquer regime de alimentação enteral em relação à mortalidade. Note que as evidências disponíveis são limitadas a dois estudos pequenos com risco moderado de viés. Mais pesquisas são necessárias nessa área para se chegar a conclusões robustas.

Notas de tradução

Tradução do Centro Cochrane do Brasil (José Carlos Souza)

Plain language summary

High-carbohydrate, high-protein, low-fat tube feeds versus low-carbohydrate, high-protein, high-fat tube feeds for patients with severe burns

Patients with severe burn injuries have increased metabolic needs. For this and other reasons, they are often fed a formula through a tube inserted directly into the stomach or small intestine, a process known as enteral feeding. Aggressive enteral feeding of burn patients with a high-protein diet is a core component of the treatment of severe burn injuries. However, the optimal proportions of fats and carbohydrates in feeding formulas are unknown. This review of 93 burn patients in two randomized controlled trials found that high-carbohydrate, high-protein, low-fat enteral feeds reduced incidence of pneumonia compared with low-carbohydrate, high-protein, high-fat enteral feeds in patients hospitalized with severe burn injuries. No conclusions could be made about the effect of different enteral feeding regimens on death.

The strength of this review's findings were limited by the number, size, and quality of eligible trials. Further research is needed before strong scientific conclusions and sound clinical recommendations can be made.

Résumé simplifié

Comparaison entre une alimentation par sonde riche en glucides et en protéines, mais pauvre en lipides et une alimentation par sonde pauvre en glucides, mais riche en protéines et en lipides chez des patients souffrant de graves brûlures

Les patients souffrant de graves brûlures ont des besoins métaboliques importants. C'est pour cette raison, entre autres, qu'une formule leur est administrée via une sonde insérée directement dans l'estomac ou l'intestin grêle, un processus dénommé « alimentation entérale ». Une alimentation entérale agressive de patients brûlés avec un régime alimentaire riche en protéines est un composant clé du traitement de personnes gravement brûlées. Toutefois, les proportions optimales de lipides et de glucides des formules alimentaires sont inconnues. Cette revue portant sur 93 patients brûlés dans deux essais contrôlés randomisés a révélé qu'une alimentation entérale riche en glucides et en protéines, mais pauvre en lipides, réduisait l'incidence de la pneumonie comparée à une alimentation entérale pauvre en glucides, mais riche en protéines et en lipides chez des patients hospitalisés pour des brûlures graves. Aucune conclusion n'a pu être tirée quant aux effets des différents schémas posologiques d'alimentation entérale sur la mortalité.

La portée des découvertes de cette revue était limitée par le nombre, la taille et la qualité des essais éligibles. D'autres recherches sont requises avant que des conclusions scientifiques fiables et des recommandations cliniques solides puissent être émises.

Notes de traduction

This translation refers to an older version of the review that has been updated or amended.

Traduit par: French Cochrane Centre 10th April, 2012
Traduction financée par: Ministère du Travail, de l'Emploi et de la Santé Français

Resumo para leigos

Alimentação por sonda rica em carboidratos, hiperproteica e pobre em gordura versus alimentação pobre em carboidratos, hiperproteica e rica em gordura para pacientes com queimadura grave.

Os pacientes com queimaduras graves têm suas necessidades metabólicas aumentadas. Por essa e outras razões, eles são frequentemente alimentados através de sonda inserida diretamente no estômago ou intestino delgado, um procedimento conhecido como alimentação enteral. Alimentação enteral agressiva com dieta hiperproteica é parte central do tratamento para pacientes com queimadura grave. Contudo, ainda não se sabe qual seria a proporção ideal de gorduras e carboidratos nas fórmulas para alimentação enteral. Esta revisão encontrou dois ensaios clínicos randomizados envolvendo 93 pacientes hospitalizados com queimaduras graves. A revisão chegou à conclusão de que a alimentação enteral rica em carboidratos, rica em proteínas e pobre em gordura reduz a incidência de pneumonia quando comparada com dieta enteral pobre em carboidratos, rica em proteínas e pobre em gordura. Não se chegou a nenhuma conclusão sobre o efeito dos diferentes regimes de alimentação enteral em relação à mortalidade.

A certeza quanto aos achados desta revisão foi comprometida pelo número, tamanho e qualidade dos estudos elegíveis. Portanto, mais pesquisas são necessárias para se chegar a conclusões cientificas definitivas e para que recomendações clínicas possam ser feitas.

Notas de tradução

Tradução do Centro Cochrane do Brasil (José Carlos Souza)

Background

Nutritional support is critically important in patients with extensive burn injuries. In such patients, energy and protein requirements increase to approximately double the requirements of a healthy person. The introduction of "aggressive nutrition support" in burn patients in the early 1970s dramatically improved patient survival. A review from 1971 to 1975 (Alexander 1990) of patients with burns to greater than 50% of total body surface area (TBSA) provided with adequate nutrition showed an 80% drop in mortality due to septicaemia. This decrease in septicaemia was attributed to aggressive nutritional support (Alexander 1990). It is reported that the risk of death is markedly higher in patients who lose greater than one fourth of their body weight after burn injury (Herndon 2007).

Research suggests nutritional support in burn patients will reduce the severity of the hypermetabolic response, reduce catabolism and loss of muscle mass, improve and hasten wound healing, reduce rates of infection, improve integrity and function of the gut, and reduce hospital length of stay.

Inappropriate nutritional support may lead to adverse outcomes including fatty liver disease, decreased lean muscle mass, increased body fat mass, cardiopulmonary complications, and increased infectious complications. Since these discoveries, nutritional support in burns patients has been investigated. Research has been completed on many aspects of nutritional requirements in burn patients, including timing of feeding, optimal tube placement, vitamin and mineral supplementation, fish oil supplementation, and use of immune-modulating amino acids and macronutrients. Despite the large amount of published research, optimal feeding practices have not yet been delineated. Macronutrients (carbohydrate, fat and protein) make up the base of nutritional support. If these proportions are incorrect the benefits of micronutrients may well be lost.

Description of the condition

Patients with severe burns (second or third degree burns to greater than 10% of their total body surface area) often require aggressive nutritional support. The body's reaction to severe burns is a hypermetabolic response: greatly increased energy requirements, sustained release of catabolic hormones leading to muscle breakdown, and poor glycaemic control. Enteral feeding - controlled feeding of a prepared formula directly into the stomach or small intestine via a tube - is a mainstay of treatment for patients with severe burns (Herndon 2007).

Description of the intervention

This review explored the impact of macronutrient composition of enteral feeds (tube feeds into the stomach or intestine) on clinical outcomes. Parenteral nutrition (intravenous feeding) was not investigated.

Macronutrients are "macromolecules in plant and animal structures that can be digested, absorbed and utilised by other organisms as energy sources and as a substrate for the synthesis of the carbohydrates, fats, and proteins required to maintain cell and system integrity" (Escott-Stump 2004). Macronutrients of importance in humans are fat, carbohydrate and protein. In the healthy population it is recommended that at least 55% of energy comes from carbohydrate, less than 30% of energy comes from fat, and more than 15% of energy comes from protein.

Interventions for enteral feeding in burn patients can be split into two categories:

  1. high-carbohydrate, high-protein, low-fat enteral formula (high-carbohydrate enteral formula) providing at least 55% of total energy from carbohydrates, 20-25% of total energy from protein, and less than 25% of energy from fat; or

  2. low-carbohydrate, high-protein, high-fat enteral formula (high-fat enteral formula) providing less than 55% total energy from carbohydrates, 20-25% of total energy from protein, and at least 25% of total energy from fat.

Interventions may provide varying quantities of fish oils as a fat source. At least one study has linked fish oil supplementation to decreased infectious complications, reduced length of hospital stay, reduced incidence of diarrhoea, and decreased loss of muscle mass in burn patients (Gottschlich 1990b).

Enteral feeds are usually commenced within 24 hours of admission for severe burn injuries. The feeding tube may supply the stomach (as in orogastric, nasogastric, or percutaneous endoscopic gastrostomy tubes) or the small intestine (as in nasoduodenal, nasojejunal, or percutaneous endoscopic jejunostomy tubes), depending on various clinical factors and physician preference.

How the intervention might work

The ratio of macronutrients in enteral feeds may influence clinical outcomes in burn patients in two ways:

1. Hypermetabolic response

The hypermetabolic response to significant burn injury increases energy and protein requirements to approximately twice basal requirements and leads to significant loss of lean mass and total weight loss. Furthermore, the hypermetabolic response alters the way the body metabolises and utilises macronutrients. Breakdown of muscle tissue and fat stores is increased, with the breakdown products fuelling increased hepatic gluconeogenesis. This response cannot be completely stopped with currently available treatments, but it may be possible to reduce the severity of protein breakdown.

Variations in the macronutrient composition of enteral feeds may limit the detrimental effects of the hypermetabolic response. For example, it is suggested that a high-carbohydrate energy source may increase blood insulin levels, thereby reducing the hormonal stimulus for muscle protein catabolism.

2. Gastrointestinal (GI) tolerance

Burn patients on enteral feeds often experience nausea, vomiting, diarrhoea and/or constipation, and often do not absorb dietary macronutrients as well as non-burn patients. Therefore, it is important to maximise GI tolerance in order to maximize the possibility of a patient receiving and absorbing required levels of nutrition.

It is possible that different macronutrient compositions will improve GI tolerance and thereby improve nutritional status and clinical outcomes. It is known that carbohydrates are digested more quickly than fats and proteins, and therefore it is hypothesised that a high-carbohydrate enteral feeding regimen is more quickly absorbed than a high-fat enteral feeding regimen. Decreases in absorption time may improve GI tolerance in burn patients.

A study assessing diarrhoea in burns patients showed that using a lower fat feed (less than 20% of caloric intake) reduced the frequency of diarrhoea in these patients (Gottschlich 1988). However, it remains unclear what combination of macronutrients the GI tract will tolerate best.

Why it is important to do this review

No systematic review has been completed investigating enteral feeding in burns patients and the role of fat and carbohydrates in optimising the management of a burn injury. This review will add to an overall understanding about the role of carbohydrates and fats in nutrition support in burn patients. The optimal combination of enteral macronutrients should help to optimise clinical outcomes in these patients by reducing the hypermetabolic response and muscle protein breakdown, and conversely by promoting anabolism and improved wound healing. Furthermore, optimization of enteral macronutrient proportions should improve GI tolerance, minimizing diarrhoea, constipation, nausea and vomiting, and promoting gut health generally. Assessing optimal macronutrient proportions may allow care providers to appropriately weigh patient factors (e.g. hyperglycaemia) when deciding whether or not to manipulate the macronutrient proportions in a patient's enteral formula.

This review seeks to aid in developing an evidence-based approach to the medico-nutritional management of burn patients, potentially justifying clinical recommendations and informing future research efforts.

Objectives

To examine the evidence of improved clinical outcomes in burn patients treated with high-carbohydrate enteral feeds compared with those treated with high-fat enteral feeds.

Methods

Criteria for considering studies for this review

Types of studies

We included randomized controlled trials (RCTs) comparing high-carbohydrate enteral feeding to high-fat enteral feeding in severely burned patients.

Types of participants

Patients must have burns to at least 10% of their total body surface area (TBSA) and have received enteral nutrition. Trials involving patients with any agent of burn injury (thermal, chemical, or electrical) and of any age were eligible. No criteria were set for participant characteristics (age, gender, etc.).

Types of interventions

Enteral tube feeding with a high-carbohydrate feeding regimen versus enteral tube feeding with a high-fat feeding regimen was considered. All mechanisms of enteral feeding (gastric, duodenal, and jejunal feeding) were considered. The source of macronutrients was not considered in this review.

Macronutrient composition was defined as follows:

  1. high-carbohydrate, high-protein, low-fat enteral formula (high-carbohydrate enteral formula): at least 55% of energy from carbohydrates, 20-25% of energy from protein, and less than 25% of energy from fat;

  2. low-carbohydrate, high-protein, high-fat feed (high-fat enteral formula): less than 55% of energy from carbohydrates, 20-25% of energy from protein, and at least 25% of energy from fat.

Types of outcome measures

Outcomes were divided into primary and secondary outcomes.

Primary outcomes
  • Mortality.

  • Incidence of sepsis or pneumonia.

  • Time to healing.

  • Number of days on ventilator.

Secondary outcomes
  • Glycaemic control.

  • Insulin requirements.

  • Weight change.

  • Protein turnover via tagged amino acid studies.

  • Nitrogen balance.

Search methods for identification of studies

The search for trials was not restricted by date, language or publication status.

Electronic searches

We searched the following electronic databases:

  • Cochrane Injuries Group Specialised Register (searched 28 Nov 2011),

  • Cochrane Central Register of Controlled Trials (The Cochrane Library 2011, Issue 4),

  • MEDLINE (Ovid)1950 to Nov (Week 3) 2011,

  • EMBASE (Ovid) (1980 to Nov (week 46)  2011),

  • ISI Web of Science: Science Citation Index Expanded (SCI-EXPANDED) (1970 to Nov 2011),

  • ISI Web of Science: Conference Proceedings Citation Index-Science (CPCI-S) (1990 to Nov 2011),

  • PubMed [www.ncbi.nlm.nih.gov/sites/entrez/] (Searched 28 Nov 2011).

All search strategies are reported in full in Appendix 1.

Searching other resources

We handsearched the following conference abstracts:

  • American Burn Association meetings: 1997 to 2008;

  • Dietitians Association of Australia meetings: 1990 to 2010;

  • Burns meeting abstracts: 2003 to 2010;

  • Journal of Burn Care and Research: 2003 to 2010;

  • Journal of Human Nutrition and Dietetics: 2003 to 2010;

  • Journal of the American Dietetic Association: 2003 to 2010.

We searched the reference lists of previously published reviews and other relevant material for further eligible studies.

Data collection and analysis

The Injuries Group Trials Search Co-ordinator ran the searches and collated the results using bibliographic software (EndNote X3 bld 5276), before passing them on to two authors (SA, FS) who independently screened the search results for eligibility.

Selection of studies

We obtained the full text of all relevant records and independently assessed whether each one met the pre-defined eligibility criteria. No disputes arose (kappa = 1), and no duplicate studies were found. Figure 1 illustrates the study retrieval and selection process.

Figure 1.

Study retrieval and selection process

Data extraction and management

Data were extracted independently by two authors (BM, FS). A standard form created by one author (SA) was used to extract the following:

  • study characteristics: source, authors, time period of trial, date of publication, population studied;

  • study design details: eligibility criteria, study type and allocation, method of randomization, dropouts/withdrawals, interventions, exclusions;

  • participants: age range, gender, care setting, total number, location, % TBSA burned;

  • risk of bias: sequence generation, allocation concealment, blinding, incomplete outcome data addressed, selective reporting, other biases;

  • sample size: number of patients in each arm of trial;

  • intervention details: % calories from non-enteral sources, macronutrient composition and source, other supplements, total kcal/kg/day, timing and duration of each intervention;

  • outcome measures: types of outcomes measured, time points collected, outcomes and time points reported, outcome definition, diagnostic criteria, units of measurement, missing data, summary data, effect estimate, 95% confidence interval (CI);

  • analysis: description of analysis employed;

  • miscellaneous: funding sources, authors' conclusions, references found to other relevant studies.

These data were cross-checked for accuracy. No disputes arose. Data were entered into the Cochrane Collaboration's Review Manager software, version 5 (Review Manager).

Assessment of risk of bias in included studies

Two authors independently assessed study quality (BM, FS).

The recommended approach for assessing risk of bias in studies included in systematic reviews from the Cochrane Handbook (Higgins 2011) was used. Briefly, a two-part tool was used which addressed six specific domains for assessment of bias (listed below). Each domain included one or more specific entries in a "risk of bias" table. Within each entry, the first part of the tool involved describing what was reported to have happened in the study. The second part of the tool involved assigning a judgement relating to the risk of bias for that entry. This was achieved by answering a pre-specified question about the adequacy of the study in relation to the entry, such that a judgement of "yes" indicated a low risk of bias, "no" indicated a high risk of bias, and "unclear" indicated unclear or unknown risk of bias.

The categories were as follows:

  1. Was the allocation sequence adequately generated?

  2. Was allocation adequately concealed?

  3. Was knowledge of the allocated interventions adequately concealed during the study?

  4. Were incomplete outcome data adequately addressed?

  5. Are reports of the study free of suggestion of selective outcome reporting?

  6. Was the study apparently free of any other problems that could put it at a risk of bias, for example, patient, investigator or data analysis problems?

Data synthesis

All data were entered into Review Manager software, version 5. A random-effects model was used for all comparisons.

For ordinal and nominal variables (incidence of pneumonia and mortality), both individual and pooled statistics were expressed as odds ratios (OR) with 95% confidence intervals (CI). For continuous variables (days on ventilator), mean difference with 95% CI was reported. Statistical significance was defined as a P value ≤ 0.05.

An intention-to-treat analysis could not be performed, as only one study reported data in an intention-to-treat fashion (Gottschlich 1990a), and the other study's author could not be contacted (Garrel 1995).

Results

Description of studies

We found 1121 records through the searches. Based on title and abstract screening we excluded 1116 and examined 5 studies in further detail by obtaining full text copies. We excluded three of the studies (Hart 2001; Saffle 1997; Serog 1983) because their enteral feeding regimens did not meet the study requirements. Two studies (Garrel 1995; Gottschlich 1990a) were eligible for inclusion in this systematic review. We had uniform inter-observer agreement on study inclusion eligibility (kappa = 1); no disagreements arose requiring discussion or a mediator. See Figure 2 for the results of the bias assessment tool for the two included studies.

Figure 2.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Gottschlich 1990a reported a randomized controlled trial, initially enrolling 50 patients aged three years or older with thermal injury to greater than 10% TBSA. All patients were admitted to the Cincinnati Shriners Burns Institute or Cincinnati University Hospital within five days of burn between 6 April 1986 and 20 May 1988. No patient exclusion criteria were given. Patients were followed until death or discharge from the hospital. Patients were randomized to one of three treatment groups using "a random number table stratified for age, institution and burn size". The three treatment groups were group 1, a high-fat feed group; group 2, which received a high-carbohydrate "modular tube feed" designed by the study's authors; and group 3, another high-fat feed group. All patients enrolled in the study received vitamins A and C and zinc supplementation enterally. "Parenteral nutrition was reserved for those who could not be given adequate nutritional support enterally. It was used for as short a time as possible without compromising the adequacy of nutritional support." The number of patients requiring total parenteral nutrition (TPN) and the mean number of days on TPN was not statistically significantly different between the three intervention groups.

The primary aim of this study was to "evaluate and differentiate the effects of three nutrition support regimens with one treatment group randomized to receive the new modular tube feeding." Outcomes assessed included diarrhoea (cumulative incidence as a percentage, total days with, and mean days with), constipation (number of patients experiencing, and total days with), glycaemic control (mean days requiring exogenous insulin, and mean insulin days), number of patients with bacteraemia, number of patients with wound infection, number of patients with pneumonia, mean days on ventilator, discharge weight as a percentage of pre-burn weight, and mortality. No data were available for sepsis score, time to healing, insulin requirements, protein turnover, nitrogen balance, or nutritional status serum markers.

Of the 50 patients initially enrolled in the study, 10 died before discharge. These patients' data were included in all measures other than hospital length of stay (Gottschlich correspondence), thus, this study's data were reported on an intention-to-treat basis. The study authors found no statistically significant differences in glycaemic control, days on ventilator, mortality, cumulative incidence of pneumonia or sepsis, or discharge weight as a percentage of pre-burn weight using analysis of variance between the three groups. No analysis combining groups 1 and 3 into 'high-fat' and comparing this combined group with group 2 as 'high-carbohydrate' was reported.

Garrel 1995 reported a randomized controlled trial, initially enrolling 43 adult patients with thermal injury to greater than 20% TBSA, excluding first-degree burns. All patients were admitted within 24 hours of injury between September 1990 and September 1994. Patient exclusion criteria included people aged over 65, body mass index greater than 30, pre-existing diabetes or chronic visceral insufficiency, and chronic use of either alcohol or cocaine. Patients were followed for 30 days. Patients were randomized to one of three treatment groups using a random number table. One treatment group received a high-fat enteral feeding formula, and was treated as the control group; the second group received a high-carbohydrate enteral feeding formula without omega-3 fatty acid supplementation; the third group received a high-carbohydrate enteral feeding formula with omega-3 fatty acid supplementation. All treatment groups received parenteral nutrition in addition to enteral feeding. Energy delivered through parenteral nutrition was not statistically significantly different between the three intervention groups, nor was it statistically significantly different in a combined analysis of the two high-carbohydrate groups versus the high-fat (control) group.

The primary aim of the study was to assess the effects of high-carbohydrate enteral nutritional support with and without omega-3 fatty acid supplementation on clinical outcomes, protein sparing and several serum biomarkers, and to compare both of these enteral feeding regimens with the high-fat control group. Reported end-points included mortality, sepsis score, incidence of pneumonia, exogenous insulin requirements, and hospital length of stay. No data were available for number of days on ventilator, glycaemic control, weight change, protein turnover, nitrogen balance, nutritional status or bowel habits.

Risk of bias in included studies

Assessment of bias for the RCTs was completed using the Cochrane Collaboration's 'Risk of Bias' tool (Higgins 2011). This two part tool used pre-specified questions to assess the methodological quality of the study. A judgement of 'yes' indicates low risk of bias, 'no' high risk of bias and 'unclear' if a judgement could not be made.

Random sequence generation (selection bias)

The randomization sequence was generated using a random number table in both studies.

Allocation

There was good allocation concealment in the Garrel 1995 study. Allocation concealment was not reported in the Gottschlich 1990a study.

Blinding

Both studies blinded nursing and medical staff. Garrel 1995 notified the investigator of patients' randomization groups only after allocation. Dr. Gottschlich reports via correspondence that tube feeds were delivered in cartons labelled only as "research tube feeding", while caregivers and patients were given no information on which feed they actually received (Gottschlich correspondence).

Incomplete outcome data

Both studies were inpatient studies reporting outcome data collected as part of routine care. As caregivers were blinded in both studies, authors consider there is a low risk of bias. However, Garrel 1995 excluded patients who died in the first three weeks into the study and reported "available case" data. See risk of bias table for further information.

Selective reporting

No study protocol was published for either study, and so reporting bias is unknown.

Other potential sources of bias

Of the 43 patients initially admitted to the Garrel 1995 study, six died before the end of the third week of admission. These patients were excluded from the analysis, and thus this study's data were reported on an available case basis, not an intention-to-treat basis. Thirty-seven patients were included in the reported data. Two patients who died after completion of the study were included in the reported data.

Based on these strengths and weaknesses we judge Gottschlich 1990a to have an overall moderate risk of bias and Garrel 1995 to have an overall high risk of bias.

Effects of interventions

Garrel 1995 reported available case data for 37 patients with burn injury. Mortality and cumulative incidence of pneumonia were reported as dichotomous end-points, while sepsis score, daily insulin requirements, and hospital length of stay were reported as continuous or ordinal end-points. The study authors concluded that compared to the high-fat group, the high-carbohydrate groups had a lower 30-day cumulative incidence of pneumonia, and required less exogenous insulin. Differences in mortality, sepsis score and length of care did not reach statistical significance.

Gottschlich 1990a reported intention-to-treat data for 50 patients with burn injury. Mortality, length of stay per percentage burn, cumulative incidence of pneumonia, cumulative incidence of sepsis, and cumulative incidence of constipation were reported as dichotomous end-points. Average days on ventilator, discharge weight as a percentage of pre-burn weight, mean diarrhoeal days, and mean days requiring exogenous insulin were reported as continuous end-points. This study reported statically significant reductions in wound infection (P < 0.03) and length of stay (P < 0.02) with low fat enteral feeds. It also identified trends to decreased diarrhoeal days and improved glucose tolerance although these were not statistically significant.

Common outcomes between the two studies were mortality and cumulative incidence of pneumonia. Insulin requirements were reported in both papers, but Garrel 1995 did not report standard errors or confidence intervals, and attempts to contact the study authors were unsuccessful. In meta-analysis, use of a high-carbohydrate enteral feeding regimen was associated with a mortality OR of 0.36 (95% CI 0.11 to 1.15, P = 0.08) (Analysis 1.1), and a cumulative incidence of pneumonia OR of 0.12 (95% CI 0.04 to 0.39, P = 0.004) (Analysis 1.2), compared to a high-fat enteral feeding regimen. The mortality OR did not reach statistical significance, while the cumulative incidence of pneumonia OR did reach statistical significance.

Average number of days on a ventilator was reported in Gottschlich 1990a only. Combining groups 1 and 3 from this study to form a high-fat group and comparing to group 2 as a high-carbohydrate group, use of a high-carbohydrate enteral feeding regimen was associated with a days on a ventilator mean difference (MD) of -3.30 (95% CI -5.80 to -0.80) (Analysis 1.3). This result reached statistical significance (P value = 0.01). Note that these results are not from a meta-analysis of studies, but rather an analysis combining Gottschlich 1990a's three groups into two.

No meta-analysis could be performed on the remaining outcomes data due to lack of similar outcomes reported between the two available studies.

Discussion

This systematic review of the effect of high-carbohydrate, high-protein, low-fat enteral feeding (high-carbohydrate enteral feeding) regimens versus low-carbohydrate, high-protein, high-fat enteral feeding (high-fat enteral feeding) regimens in patients with significant burn injuries includes two RCTs, one judged to be of moderate risk of bias, and the other as high. Each study had a small number of participants.

Meta-analysis of the two studies indicates the risk of pneumonia is lower in burn patients receiving a high-carbohydrate enteral feeding regimen than in those receiving a high-fat enteral feeding regimen. No conclusions could be drawn about the risk of death in patients receiving the different feeding regimens.

There may be any number of molecular and/or cellular mechanisms to explain the decreased risk of pneumonia observed in patients receiving a high-carbohydrate enteral feeding regimen compared to those receiving a high-fat enteral feeding regimen. Most immune system cells are obligate glucose consumers (Deitch 1995), and research suggests rates of glycolysis may nearly double in severely burned patients (Gore 2001; Wolfe 1979). High-carbohydrate enteral feeding regimens may supply glucose required by highly active immune system cells. Likewise, inadequate provision of dietary glucose may limit intracellular availability and subsequently impair immune system function. It is well established that chronic hyperglycaemia inhibits wound healing and increases the risk of infectious complications in the intensive care setting (Herndon 2007; Marino 2007). However, severe burn patients develop a hypermetabolic response and are monitored closely in an intensive care setting, thus they are unlikely to become hyperglycaemic even in the setting of a high-carbohydrate diet. Conversely, high dietary fat intake may increase the risk of pneumonia by increasing the production of pro-inflammatory cytokines and negatively effecting prostaglandin metabolism (Bernier 1998; Erridge 2007).

It has been suggested that high-fat formulas are beneficial for ventilated burn patients because they may decrease CO2 production (al Saady 1989) and length of time requiring mechanical ventilation (van de Berg 1994). However, combining the three study groups reported in Gottschlich 1990a into two groups - one high-carbohydrate enteral feeding group and one high-fat enteral feeding group - indicates that patients given a high-carbohydrate enteral feeding regimen spend fewer mean days mechanically ventilated than patients given a high-fat enteral feeding regimen. (Note again: this observed effect is from one relatively small study judged to have a moderate risk of bias.) If this observed effect is a true effect, then a high-carbohydrate enteral feeding regimen reduces the average time burn patients are exposed to one of the most important risk factors for developing pneumonia in the intensive care setting (Herndon 2007; Marino 2007). The question then is: Why do patients given a high-carbohydrate enteral feeding regimen spend less time on average undergoing mechanical ventilation than those given a high-fat enteral feeding regimen? Indeed, the answer may be because they are less likely to develop pneumonia and thus to require mechanical ventilation in the setting of respiratory failure.

Much remains unknown about the cellular and molecular effects of dietary carbohydrate and fat intake on the human body. Further research, particularly into the effects of dietary macronutrients on the acute phase response to injury, is essential to further our understanding of the role dietary macronutrients play in the treatment of severely burned patients, and of trauma patients generally.

Authors' conclusions

Implications for practice

Based on the currently available evidence, consisting of two relatively small studies (one with moderate risk of bias and the other with a high risk of bias), no firm conclusion can be made. There may be a possible clinical benefit to high-carbohydrate enteral feeds compared with high-fat enteral feeds in the treatment of burn injuries: a reduced risk of pneumonia. The molecular and cellular mechanisms by which this benefit may accrue remain unclear.

Based on the available evidence, one relatively small study with a moderate risk of bias, patients receiving high-carbohydrate feeds may spend less time undergoing mechanical ventilation than patients receiving a high-fat enteral feeding regimen. Again, why this may be the case is unclear.

While this review presents the currently available evidence, definitive clinical guidelines on the optimal macronutrient composition of enteral feeds in severely burned patients cannot be formulated without further research.

Implications for research

Further research is needed in order to determine the proportion of macronutrients in enteral feeding regimens that will optimize patient outcomes following severe burn injuries. Large, well-designed randomized controlled trials of burn patients would most effectively establish optimal enteral feeding practices. Animal and human research utilizing clinical outcomes as well as various biomarkers may add to our understanding of why high-carbohydrate enteral feeding regimens appear to be more beneficial to burn patients when compared to high-fat enteral feeding regimens. Continued research into the cellular and molecular effects of dietary fats and carbohydrates on the acute phase response to injury may help guide clinical research.

Acknowledgements

We would like to acknowledge the study authors whom we contacted for more information, the enteral feed manufacturers who provided information on the existence of ongoing trials, and to the staff at the Cochrane Injuries Group, specifically Karen Blackhall and Emma Sydenham for their guidance and support. Li Peng and Jane Tuckerman contributed to the protocol for this review and Taixiang Wu at the Chinese Cochrane Centre located and translated Chinese-language studies. Annette Kaufman at the National Cancer Institute and Alexandra Soroceanu at Dalhousie University assisted with searching conference abstracts.

Data and analyses

Download statistical data

Comparison 1. High-carbohydrate vs high-fat enteral feeding
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Mortality at final follow-up290Odds Ratio (M-H, Random, 95% CI)0.36 [0.11, 1.15]
2 Risk of pneumonia287Odds Ratio (M-H, Random, 95% CI)0.12 [0.04, 0.39]
3 Days on ventilator1 Mean Difference (IV, Random, 95% CI)Subtotals only
Analysis 1.1.

Comparison 1 High-carbohydrate vs high-fat enteral feeding, Outcome 1 Mortality at final follow-up.

Analysis 1.2.

Comparison 1 High-carbohydrate vs high-fat enteral feeding, Outcome 2 Risk of pneumonia.

Analysis 1.3.

Comparison 1 High-carbohydrate vs high-fat enteral feeding, Outcome 3 Days on ventilator.

Appendices

Appendix 1. Search strategy

Cochrane Injuries Group Specialised Register (searched 28 Nov 2011)
1. (feed* or fed or nutrition* or nourish*) AND (enteral or enteric or gastric or gastro* or intub* or tube*)
2. (macronutrient* or carbohydrate* or starch* or glycogen* or polysaccharide* or monosaccharide* or glucose* or protein* or amino acid* or fat* or lipid*) AND (feed* or fed or nutrition* or nourish*)
3. (enteral or enteric or gastric or gastro* or intub* or tube*) AND (feed* or nutrition*)
4. (intub* or tube*) AND (feed* or gastrointest*)
5. 1 OR 2 OR 3 OR 4
6. Burn*
7. 5 and 6

Cochrane Central Register of Controlled Trials (The Cochrane Library 2011, Issue 4)
#1 MeSH descriptor Carbohydrates explode all trees
#2 MeSH descriptor Dietary Carbohydrates explode all trees
#3 MeSH descriptor Glycogen explode all trees
#4 MeSH descriptor Polysaccharides explode all trees
#5 MeSH descriptor Monosaccharides explode all trees
#6 MeSH descriptor Glucose explode all trees
#7 MeSH descriptor Amino Acids explode all trees
#8 MeSH descriptor Amino Acids, Essential explode all trees
#9 MeSH descriptor Fats explode all trees
#10 MeSH descriptor Lipids explode all trees
#11 MeSH descriptor Starch explode all trees
#12 (#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11)
#13 ((feed* or fed or nutrition* or nourish*) near (enteral or enteric or gastric or gastro* or intub* or tube*))
#14 (#12 AND #13)
#15 MeSH descriptor Enteral Nutrition explode all trees
#16 MeSH descriptor Intubation, Gastrointestinal explode all trees
#17 (macronutrient* or carbohydrate* or starch* or glycogen* or polysaccharide* or monosaccharide* or glucose* or protein* or amino acid* or fat* or lipid*) AND ((feed* or fed or nutrition* or nourish*) near (enteral or enteric or gastric or gastro* or intub* or tube*))
#18 (enteral or enteric or gastric) near (feed* or nutrition*)
#19 (intub* or tube*) near (feed* or gastrointest*)
#20 (#14 OR #15 OR #16 OR #17 OR #18 OR #19)
#21 MeSH descriptor Burns explode all trees
#22 burn*
#23 (#21 OR #22)
#24 (#20 AND #23)

MEDLINE (Ovid)1950 to Nov (Week 3) 2011
1. exp carbohydrates/
2. exp dietary carbohydrates/
3. exp glycogen/
4. exp polysaccharides/
5. exp monosaccharides/
6. exp glucose/
7. exp amino acids/
8. exp amino acids essential/
9. exp fats/
10. exp lipids/
11. exp starch/
12. or/1-11
13. ((feed$ or fed or nutrition$ or nourish$) and (enteral or enteric or gastric or gastro$ or intub$ or tube$)).ab,ti.
14. 12 and 13
15. exp enteral nutrition/
16. exp intubation gastrointestinal/
17. (macronutrient$ or carbohydrate$ or starch$ or glycogen$ or polysaccharide$ or monosaccharide$ or glucose$ or protein$ or amino acid$ or fat$ or lipid$).ab,ti.
18. ((enteral or enteric or gastric) adj3 (feed$ or nutrition$)).ab,ti.
19. ((intub$ or tube$) adj3 (feed$ or gastrointest$)).ab,ti.
20. 14 or 15 or 16 or 17 or 18 or 19
21. exp burns/
22. burn$.ab,ti.
23. 21 or 22
24. 20 and 23
25. randomi?ed.ab,ti.
26. randomized controlled trial.pt.
27. controlled clinical trial.pt.
28. placebo.ab.
29. clinical trials as topic.sh.
30. randomly.ab.
31. trial.ti.
32. 25 or 26 or 27 or 28 or 29 or 30 or 31
33. (animals not (humans and animals)).sh.
34. 32 not 33
35. 24 and 34

EMBASE (Ovid) 1980 to week 46, Nov 2011
1. exp Carbohydrate/
2. exp carbohydrate diet/
3. exp Monosaccharide/
4. exp Glycogen/
5. exp Polysaccharide/
6. exp Glucose/
7. exp Amino Acid/
8. exp Essential Amino Acid/
9. exp Fat/
10. exp Lipid/
11. exp Starch/
12. or/1-11
13. ((feed$ or fed or nutrition$ or nourish$) and (enteral or enteric or gastric or gastro$ or intub$ or tube$)).ab,ti.
14. 12 and 13
15. exp enteral nutrition/
16. exp intubation gastrointestinal/
17. (macronutrient$ or carbohydrate$ or starch$ or glycogen$ or polysaccharide$ or monosaccharide$ or glucose$ or protein$ or amino acid$ or fat$ or lipid$).ab,ti.
18. ((enteral or enteric or gastric) adj3 (feed$ or nutrition$)).ab,ti.
19. ((intub$ or tube$) adj3 (feed$ or gastrointest$)).ab,ti.
20. 14 or 15 or 16 or 17 or 18 or 19
21. exp enteric feeding/
22. exp nose feeding/
23. exp tube feeding/
24. exp digestive tract intubation/
25. exp duodenum intubation/
26. exp esophagus intubation/
27. exp stomach intubation/
28. or/20-27
29. exp Burn/
30. burn$.ab,ti.
31. 29 or 30
32. 28 and 31
33. exp Randomized Controlled Trial/
34. exp controlled clinical trial/
35. randomi?ed.ab,ti.
36. placebo.ab.
37. *Clinical Trial/
38. randomly.ab.
39. trial.ti.
40. 33 or 34 or 35 or 36 or 37 or 38 or 39
41. exp animal/ not (exp human/ and exp animal/)
42. 40 not 41
43. 32 and 42

ISI Web of Science: Science Citation Index Expanded (SCI-EXPANDED) (1970 to Nov 2011);
ISI Web of Science: Conference Proceedings Citation Index-Science (CPCI-S) (1990 to Nov 2011)
1. (feed* or fed or nutrition* or nourish*) NEAR/3 (enteral or enteric or gastric or gastro* or intub* or tube*)
2. (enteral or enteric or gastric or gastro* or intub* or tube*) NEAR/3 (feed* or nutrition*)
3. (macronutrient* or carbohydrate* or starch* or glycogen* or polysaccharide* or monosaccharide* or glucose* or protein* or amino acid* or fat* or lipid*) NEAR/3 (feed* or fed or nutrition* or nourish*)
4. #1 or #2 or #3
5. Burn*
6. 4 AND 5
7. Topic=((singl* OR doubl* OR trebl* OR tripl*) NEAR/3 (blind* OR mask*)) OR Topic=((clinical OR control* OR placebo OR random*) NEAR/3 (trial* or group* or study or studies or placebo or controlled)) NOT Title=(Animal* or rat or rats or rodent* or mouse or mice or murine or dog or dogs or canine* or cat or cats or feline* or rabbit or rabbits or pig or pigs or porcine or swine or sheep or ovine* or guinea pig*)
8. 6 and 7

PubMed [www.ncbi.nlm.nih.gov/sites/entrez/] (Searched 28 Nov 2011)
1. "Enteral Nutrition"[Mesh] OR "Intubation, Gastrointestinal"[Mesh]
2. (enteral or enteric or gastric) and (feed* or nutrition*)
3. (macronutrient* or carbohydrate* or starch* or glycogen* or polysaccharide* or monosaccharide* or glucose or protein or amino acid* or fat or fats or lipid or lipids) AND ((feed* or fed or nutrition* or nourish*) and (enteral or enteric or gastric or gastrointestinal* or intub* or tube or tubes or tubal*))
4. 1 or 2 or 3
5. burn*
6. 4 and 5
7. ((randomized controlled trial[pt] OR controlled clinical trial[pt]) OR (randomized OR randomised OR randomly OR placebo[tiab]) OR (trial[ti]) OR ("Clinical Trials as Topic"[MeSH Major Topic])) NOT (("Animals"[Mesh]) NOT ("Humans"[Mesh] AND "Animals"[Mesh]))
8. 6 and 7

What's new

DateEventDescription
18 January 2012AmendedThe date next expected has been corrected to January 2014.

Contributions of authors

Masters B: Drafting of protocol and review, data extraction, general advice on review and clinical perspective, editing and incorporation of reviewer comments.

Aarabi S: Drafting of review, searching and selection of studies, creation of data extraction form, data analysis, data interpretation, drafting of review, editing.

Sidhwa F: Drafting of review, searching and selection of studies, contact and correspondence with authors, contact and correspondence with formula manufacturers, data extraction, data analysis, data interpretation, incorporation of reviewer comments and revision of review, recording of changes made according to reviewer comments, editing.

Wood F: Drafting of protocol, general advice on review and clinical perspective.

Declarations of interest

None known.

Differences between protocol and review

Cochrane guidance on the assessment of the risk of bias for included studies has changed since publication of the protocol on which this review was based. The instructions from the latest version of the Cochrane Handbook (Higgins 2011) were used as the basis for this review. Some grey literature sources mentioned in the protocol were excluded and others were included based on availability. Comparative data on time to healing, weight, protein turnover, nutritional status markers and bowel habits were not included in the final review as each was reported in only one of the two studies included in the review.

The protocol stated study participants would be limited to "patients with burns to 15% TBSA or greater...". This review includes patients with burns to 10% TBSA or greater. This was done so that Gottschlich 1990a could be included, and because no widely recognized definition of "severe" burn injury exists.

The protocol stated trials "allowing greater than 50% of intake from alternative sources of nutrition" and trials "that provide significant volumes of dextrose infusions in conjunction with low carbohydrate feeds" would be excluded. This exclusion criteria was modified for this review, as macronutrients provided through parenteral feeding did not differ between groups and authors were confident any feeding effects would be related to enteral feeds.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Garrel 1995

MethodsProspective randomized controlled trial, single centre
ParticipantsPatients at least 16 years of age with at least 20% total body surface area burned, excluding first-degree burns
InterventionsLow-carbohydrate, high-fat control group; high-carbohydrate, low-fat group 1; high-carbohydrate, low-fat group 2
OutcomesMortality, sepsis score, pneumonia, diarrhoea (cumulative incidence as a percentage, total days with, and mean days with), constipation (number of patients experiencing, and total days with), glycaemic control (mean days requiring exogenous insulin, and mean insulin days), number of patients with bacteraemia, number of patients with wound infection, number of patients with pneumonia, mean days on ventilator, discharge weight as a percentage of pre-burn weight
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandom number table (p482)
Allocation concealment (selection bias)Low riskInvestigator not aware of which group patients would be assigned (p482)
Blinding (performance bias and detection bias)
Patient and the outcome assessor
High riskNurses and surgeon blinded (p483); investigator aware of allocation after randomization (p482)
Incomplete outcome data (attrition bias)
All outcomes Less than 6 weeks
High riskPatients who died > 3 weeks into the study were excluded, Garrel reports "available case" data (p484)
Incomplete outcome data (attrition bias)
All outcomes More than 6 weeks
Low riskN/A. Study lasted 30 days (p483)
Selective reporting (reporting bias)Unclear riskStudy protocol not available/not identified
Other biasUnclear riskUnknown

Gottschlich 1990a

MethodsProspective randomized controlled trial, two centres
ParticipantsPatients requiring enteral feeds. No age or size of burn restriction, all patients had ≥ 10% TBSA burns
InterventionsGroup 1 low-carbohydrate, high-fat; group 2 high-carbohydrate, low-fat; group 3 low-carbohydrate, high-fat
OutcomesMortality, pneumonia, days on ventilator, sepsis, diarrhoea (cumulative incidence as a percentage, total days with, and mean days with), constipation (number of patients experiencing, and total days with), glycaemic control (mean days requiring exogenous insulin, and mean insulin days), number of patients with bacteraemia, number of patients with wound infection, number of patients with pneumonia, mean days on ventilator, discharge weight as a percentage of pre-burn weight, length of stay per percentage burn and mortality.
NotesEmail correspondence with author
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"random number table" p226 and confirmed through Gottschlich correspondence.
Allocation concealment (selection bias)Unclear riskNot stated
Blinding (performance bias and detection bias)
Patient and the outcome assessor
Low risk"double-blind study" p.225 "The tube feeding was delivered in cartons and was designated research tube feeding on the label.  The caregivers and patients had no idea which tube feeding intervention they were receiving" Gottschlich correspondence
Incomplete outcome data (attrition bias)
All outcomes Less than 6 weeks
Low riskAuthors' judgement: inpatient study, all outcomes routine clinical practise, no reported loss to follow up
Incomplete outcome data (attrition bias)
All outcomes More than 6 weeks
Low riskAuthors' judgement: short study duration, "tube feeding support ranged from 1 week to 61 days" p229, inpatient study (p 227), all outcomes routine clinical outcomes, no reported loss to follow up
Selective reporting (reporting bias)Unclear riskNo protocol published in advance Gottschlich correspondence
Other biasUnclear riskUnknown

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Hart 2001Does not compare high-carbohydrate, high-protein, low-fat regimen to low-carbohydrate, high-protein, high-fat regimen
Saffle 1997Does not compare high-carbohydrate, high-protein, low-fat regimen to low-carbohydrate, high-protein, high-fat regimen
Serog 1983Does not compare high-carbohydrate, high-protein, low-fat regimen to low-carbohydrate, high-protein, high-fat regimen