Antibiotics for treating infected burn wounds

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To summarize the evidence for the effectiveness and harm of antibiotics for treating people with infected burn wounds.

Background

Description of the condition

Burns can be caused by many sources including chemicals, electricity, heat, sunlight or nuclear radiation. According to the World Health Organization (WHO), burns are the fourth most common type of trauma worldwide, following traffic accidents, falls, and interpersonal violence (WHO 2004). Worldwide, over 300,000 people die from fire‐related burn injuries each year, most of these occur in low‐ and middle‐income countries (WHO 2008). In the USA, approximately 500,000 people seek medical treatment for burn injuries annually and 40,000 of these need to be hospitalised.

Infection is common in burns because the injury causes the skin to lose its natural barrier to microbes; this allows pathogens to have a direct entry route to the wound. In addition, burn wounds usually produce high levels of exudate, which creates a suitably moist, nutrient‐rich environment for bacterial growth (Cakir 2004; Greenfield 1997) and may result in infection (McManus 1994; Mathangi 2006). Infection of burn wounds can delay healing and encourage scarring. Furthermore, without appropriate treatment, an infection may develop into something more serious, such as multi‐organ dysfunction syndrome or invasive infections, such as bacteraemia, or sepsis (Ekrami 2007; Fitzwater 2003). Data from the French Burns Centres indicates that 19% of inpatients with burns develop an infection (Ravat 2011). Despite substantial advancements in burn care and antibiotic management, infection remains the leading cause of death in people with a burn wound (Keen 2010). Bacteria and fungi are the most common pathogens found in burn wounds; these can originate from the patient’s own skin, gut and respiratory microflora, as well as through contact with the hospital environment or healthcare workers (D'Avignon 2008; Gosain 2005; Weber 2004).

Unfortunately, it is difficult to diagnose whether a burn wound is infected or not because the burn injury itself can appear inflamed, which is a symptom of infection. Secondly, the presence of micro‐organisms in the wound does not necessarily mean that it is infected. Different types of microbes can colonise burn wounds harmlessly, and it is difficult to differentiate skin infection from colonisation. Sometimes, colonisation can shift to infection: this progression depends on three factors, namely, host immune function, the level of bacterial inoculum and bacterial virulence (Brusselaers 2010; Ravat 2011).

Description of the intervention

In recent years, wound excision (surgical removal) and topical (local) antimicrobial treatments have been the most widely accepted interventions for moderate to severe burns (Church 2006). Whilst surgery can remove eschar (dead tissue), which reduces the inflammation and minimises scarring, early surgical debridement (removal of dead tissue) can only go so far towards decreasing infectious complications, and the rate of wound infection remains high in people with burn wounds (Appelgren 2002). Sliver compounds that deliver local treatments of polyclonal immunoglobulin and antibiotics have been used to treat infected burn wounds and have shown promise in clinical practice.

Antibiotics can be used to treat underlying infections, thus reducing morbidity and preventing mortality (Magnotti 2005; Nagoba 2010). According to the guidelines of the French Society for Burn Injuries (SFETB), antibiotics should be administered once an infection has been diagnosed. Topical antimicrobial therapy is often applied to prevent or treat infections, or used as an adjunct (in addition) to surgical treatment or systemic (whole body treatment) antibiotics. Treatments should be matched to the species of microbes causing the infection, and should not encourage drug resistance. Oral antibiotics have not been used widely in these situations because evidence shows that they are ineffective (Church 2006).

Use of antibiotics in people with infected burn wounds is more complicated than in other diseases. This is because the pharmacokinetic parameters of antibiotics (absorption, distribution, metabolism and excretion) are altered in the burned person, and significant variations exist between individuals (Conil 2006; Kiser 2006; Mohr 2008). The main change is insufficient tissue concentration of antibiotics, which leads directly to failure of therapy and drug resistance; subsequently, drug resistance reduces the efficacy of usual doses of antibiotics (Craig 2001). Consequently, the standard recommended regimen may not be suitable for burn patients. In addition, the effect of antibiotics such as ciprofloxacin or penicillins resistant to the enzyme penicillinase has waned for Staphylococcus aureus, which is the most frequent pathogen identified in wound infection sites, and there is uncertainty around the efficacy of new products that have become available in recent years (D'Avignon 2008; Guggenheim 2009). Therefore, in burn patients, adaption, administration and treatment duration of antibiotics are more difficult to establish, and practitioners should take all factors causing morbidity and mortality into account when considering their use.

Although SFETB has published the recommendations made by a group of experts, there are no strong data to support them. For example, the time point at which antibiotic treatment for infected burn wounds should be discontinued remains unclear, as it is difficult to balance the desired antibacterial effects of antibiotics against the potential development of microbial resistance if treatment with them continues for too long.

Why it is important to do this review

There is uncertainty about the use of antibiotics for treating infected burn wounds, especially regarding the most effective antibiotic and treatment regimen. A summary of high quality evidence is needed. This will hopefully lead to a general consensus on treatment that is beneficial for people with burns.

Objectives

To summarize the evidence for the effectiveness and harm of antibiotics for treating people with infected burn wounds.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials (RCTs), including parallel and cluster designs of studies reporting an objective measure of burn wound infection (i.e. clinical examination and culture data). However, infections diagnosed only by clinical examination will also be eligible. Quasi‐RCTs will be excluded.

We will exclude the following: RCTs in which the presence or absence of a specific antibiotic intervention was not the only systematic difference between treatment groups; evaluations of antibiotics used as preparations for skin grafting of burn wounds; evaluations of physical therapies with purported antimicrobial effects and evaluations of antibiotics used for the surgical treatment of burns (i.e. where antibiotics were used to treat surgical site infection).

Types of participants

People with infected burn wounds, regardless of age, race and gender. We will include any severity of burn injury (determined by percentage of total body surface area burned (TBSA) or objective assessments) and any type of burn. We will accept the authors' definitions of 'infected burn wound' used in each original trial, which may be conducted in any setting (inpatient, outpatient, nursing home plus any others). Each diagnostic criterion will be described in the 'Characteristics of included studies' table.

Types of interventions

Interventions of interest are antibiotics (topical or systemic) prescribed for infected burn wounds. Control regimens may include placebo, an alternative antibiotic, any other therapy (e.g. surgery), standard care, or no treatment. Both intervention and control regimens may consist of antibiotics administered singly or in combination. Interventions may be delivered in any setting (inpatient, outpatient, nursing home or any other location).

Types of outcome measures

Primary outcomes

• Time to complete healing (analysed using appropriate statistics e.g. survival analysis)

• Proportion of wounds healed within a specified time period (during the follow‐up)

• Adverse events (as reported by the patient or trial investigator).

Secondary outcomes

• Infection‐related mortality: i.e. mortality due to infection of burn wounds, sepsis, or another infective complication.

• Proportion of graft loss in a specified time period.

• Development of resistant organisms in isolates including Clostridium difficile.

• Development of any persistent, or worsening signs or symptoms of infection associated with the isolation of a new pathogen.

• Proportion of wounds requiring re‐grafting.

• Adverse events, including those related to antibiotics during hospitalisation: gastrointestinal symptoms, reduced appetite, abdominal bloating, urticaria (rash) and itching.

Search methods for identification of studies

Electronic searches

We will search the following electronic databases to identify reports of relevant randomised clinical trials:

• The Cochrane Wounds Specialised Register;

• The Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library) (latest issue);

• Ovid MEDLINE (1946 to present);

• Ovid MEDLINE (In‐Process & Other Non‐Indexed Citations) (latest issue);

• Ovid EMBASE (1974 to present);

• EBSCO CINAHL Plus (1937 to present);

• Wan Fang Database (via http://www.wanfangdata.com/) (1980 to present);

• Chinese Biomedical Literature Database (CBM) (1978 to present);

• China National Knowledge Infrastructure (CNKI) (1974 to present).

The provisional search strategy for The Cochrane Central Register of Controlled Trials (CENTRAL) can be found in Appendix 1.

We will adapt this strategy to search Ovid MEDLINE, Ovid EMBASE and EBSCO CINAHL. The Ovid MEDLINE search will be combined with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity‐ and precision‐maximizing version (2008 revision) (Lefebvre 2011). We will combine the EMBASE search with the Ovid EMBASE filter developed by the UK Cochrane Centre (Lefebvre 2011). We will combine the CINAHL search with the trial filters developed by the Scottish Intercollegiate Guidelines Network (SIGN 2013). We will not restrict studies with respect to language, date of publication or study setting.

We will also search the following clinical trials registries:

• ClinicalTrial.gov (https://clinicaltrials.gov/);

• WHO Chinese Clinical Trial Register ( ChiCTR) (http://www.who.int/ictrp/network/chictr2/en/);

• ISRCTN registry (http://www.controlled‐trials.com/);

• WHO International Clinical Trials Registry Platform (ICTRP) (http://apps.who.int/trialsearch/Default.aspx).

Searching other resources

We will scan the reference lists of retrieved articles and reviews to identify further potential relevant trials. In addition, we plan to identify any other systematic reviews or meta‐analyses on this topic and to handsearch their references lists. We will contact pharmaceutical companies and researchers active in this field for unpublished data. We will impose no language or publication restrictions.

Data collection and analysis

Selection of studies

Independently, two review authors (MY, JL) will scan the initial search results, assessing them by title, and abstract where available, and excluding studies that are clearly irrelevant. We will obtain the full‐text of potentially eligible studies and again independently assess them against the inclusion criteria. We will resolve any disagreements during this process by discussion.

Data extraction and management

We will used a standardised extraction form suggested by the Cochrane Wounds Group to collect data from the included studies; two review authors (TX, JRY) will extract data, working independently. The form will comply with the recommendations in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

The review authors will extract the following data: care setting (including country), study population, source of funding, design of the trial, baseline data about the wound (e.g. depth and size of burn wound, burn type), unit of analysis if more than one burn per patient, type and duration of treatment, outcomes and length of follow‐up. We will extract all the necessary data from published reports and will contact the original trial author if data is missing, if necessary. We will resolve disagreements by discussion or consultation with a third review author. One review author (JRY) will check and enter data into Revman 5.3 for further statistical analysis (RevMan 2014).

Assessment of risk of bias in included studies

Independently, two review authors (TX, JL) will assess the quality of included studies and their risk of bias. The criteria for assessment of risk of bias will include: random sequence generation; allocation concealment; blinding of participants, personnel and outcomes; incomplete outcome data; selective outcome reporting; and other sources of bias according to methods recommended in theCochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Disagreements between TX and JL about the quality of trials will be resolved by consensus, or the arbitration of a third author, if required.

Each criterion will be accessed as being at: low risk of bias; high risk of bias; or unclear risk of bias (due either to a lack of information or uncertainty over the potential for bias). We will complete a 'Risk of bias' table for each eligible study and present assessment of risk of bias using a 'Risk of bias' summary figure.

Measures of treatment effect

The measures of treatment effects will be defined as follows :

Dichotomous outcomes

We will present dichotomous outcomes for individual trials as risk ratios (RR) with 95% confidence intervals (CI). We will discuss the findings of each study and, if possible, pool feasible data. If the results are statistically significant, we will calculate the absolute risk reduction (ARR) or absolute risk increase (ARI) and the number needed to treat for an additional beneficial outcome (NNTB), or number need to treat for an additional harmful outcome (NNTH).

Continuous outcomes

We will present continuous outcomes with mean differences (MD) and 95% CIs. MD will be used preferentially because it is easier for clinicians and readers to interpret. We will use standardised mean difference (SMD) when different scales are used to measure the same outcome between trials.

Time‐to‐event outcomes

We will present time‐to‐event data as hazard ratios (HR) with 95% CI, if HR unavailable, log‐rank and Cox model will be used to estimated HR indirectly,

Ordinal outcomes

We will measure the types of adverse events and complications as ordinal data, which will be reported qualitatively.

Unit of analysis issues

In wound trials, it is not uncommon for more than one wound per patient to be included. Our approach will be: i) If the number of burns is similar to the number of participants, we will use the participant as the unit of analysis; ii) if multiple burns on one participant have been randomised to different treatments, we will report the data descriptively and comment on the unit of analysis issue in the risk of bias assessment; iii) if participants with more than one burn are randomised individually but results are analysed by wound, the situation is similar to a cluster‐randomised trial. In this case, we will not adjust for clustering but report data descriptively and comment on the unit of analysis issue in the risk of bias assessment.

Cluster RCTs

If cluster RCTs are identified, these studies will be re‐analysed by calculating the effective sample sizes, where possible, according to the methods in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), and, if necessary, an estimate of the intra‐cluster coefficient (ICC) will be calculated by using external estimates derived from similar studies.

Dealing with missing data

If trial reports are missing data, the randomisation process is compromised, so we will attempt to contact the original trial author for more information. If the data are not available, we will record this explicitly in the data extraction form and the narrative of the review.

For the outcome 'time to healing', survival analysis will take account of missing data as long as missing data and their reasons are similar across groups. If 'proportion of wounds healed' data is missing, we will assume that if participants were not included in the analysis, their burn did not heal (i.e. they will be included in the denominator but not the numerator).

No continuous outcomes are included. However, time to healing data may be reported incorrectly as mean and standard deviation. This approach has a potential to introduce bias, because times are known only for participants in whom the burn healed. If continuous data are reported incorrectly, results will be excluded from any meta‐analysis and the issue will be discussed in the risk of bias assessment.

Assessment of heterogeneity

Firstly, we will assess the clinical and methodological heterogeneity present across interventions by noting any differences in the population, intervention and outcome measurements between studies, control groups and outcomes, Secondly, statistical heterogeneity will be assessed by the Chi² test and by calculating the I² statistic. For the Chi² test, P values of less than 0.10 will indicate significant statistical heterogeneity; an I² value of 0% indicates no statistical heterogeneity; larger I² values indicate increasing heterogeneity (Higgins 2011).

We will judge the importance of the observed value of I² according to the magnitude and direction of effects and the strength of evidence for heterogeneity. If I² does equal 50% or more, the relevant data will be described individually instead of pooling. When the I² is less than 50%, outcomes will be combined in meta‐analyses where possible.

Assessment of reporting biases

If at least 10 studies are included for the analysis of the primary outcome, we will conduct funnel plots to identify any potential reporting biases (Sterne 2011).

Data synthesis

The details of included studies, i.e. the baseline characteristics (population, severity of burns), the comparison between intervention and comparator and the time point of outcome measurements will be included in the narrative of the review.

Clinical and methodological heterogeneity will be considered and data pooling will be undertaken if the included studies appear appropriately similar in terms of burn type and severity, intervention type, duration of treatment and outcome assessment. The random‐effects model will be used when the clinical or/and statistical heterogeneity is detected. The fixed effects model will be only used when clinical heterogeneity is defined as minimal by the review authors and statistical heterogeneity is estimated as non‐statistically significant for the Chi² value and 0% for the I² assessment (Kontopantelis 2013). If clinical heterogeneity is acceptable or of interest, meta‐analysis will be conducted even if statistical heterogeneity is high. However, we will attempt to report the source of heterogeneity or consider using meta‐regression for that purpose, if possible (Thompson 1999; Thompson 2002).

We will present data using forest plots if possible. For dichotomous outcomes, a risk ratio (RR) with 95% CIs will be used for the summary estimate. For continuous outcomes, a pooled mean difference (MD) with 95% CIs will be used where outcomes are measured the same way across studies. We plan to pool standardised mean difference (SMD) estimates where the same outcome is measured using different methods. For time‐to‐event data, hazard rations (HRs) (using the generic inverse variance method in Review Manager 5.3 (RevMan 2014))and 95% CIs will be used for plotting or pooling. Where time‐to‐healing is presented as a continuous measure but we cannot judge whether all wounds were healed, we will report the outcomes narratively.

Subgroup analysis and investigation of heterogeneity

Subgroup analyses can identify potential clinical heterogeneity in characteristics of participants, comorbidity or treatments (different types, duration etc). If possible, we will conduct subgroup analyses separately for the following groups:

• age: children (aged from 0 to 18 years) versus adults (over 18 years of age);

• severity of burn wounds: burns involving less than 20% TBSA versus burns involving more than 20% TBSA;

• types of drug delivery: topical versus systemic administration;

• different types of infectious organism, e.g. C difficile vsS aureus ;

• different doses of antibiotics;

• different endpoints of outcomes.

Sensitivity analysis

If sufficient studies are available, the following sensitivity analyses will be performed:

• exclusion of cluster studies, separately, from the analysis to assess the impact of study design on the results;

• exclusion of studies judged to be at high risk of bias from the analysis to assess the impact of methodological quality of studies on the results;

• assessment of the impact of imputation of missing data in the analyses versus no imputation.

'Summary of findings' tables

We will present the main results of the review in 'Summary of findings' tables. These tables present key information concerning the quality of the evidence, the magnitude of the effects of the interventions examined, and the sum of the available data for the main outcomes (Schunemann 2011a). The 'Summary of findings' tables also include an overall grading of the evidence related to each of the main outcomes using the GRADE (Grades of Recommendation, Assessment, Development and Evaluation) approach. The GRADE approach defines the quality of a body of evidence as the extent to which one can be confident that an estimate of effect or association is close to the true quantity of specific interest. The quality of a body of evidence involves consideration of within‐trial risk of bias (methodological quality), directness of evidence, heterogeneity, precision of effect estimates and risk of publication bias (Schunemann 2011b).

We plan to present the following outcomes in the 'Summary of findings' tables:

• wound healing (time to complete ulcer healing, proportion of ulcers completely healing during the trial period);

• mortality;

• adverse events.

Appendices

Appendix 1. The Cochrane Central Register of Controlled Trials (CENTRAL) provisional search strategy

#1 MeSH descriptor Burns explode all trees #2 (burn or burns or burned or scald*):ti,ab,kw #3 (thermal NEXT injur*):ti,ab,kw #4 (#1 OR #2 OR #3) #5 MeSH descriptor Infection explode all trees #6 (infect* or swell* or swollen or erythema* or odour or odor or hypertherm* or coloni* or contamin* or inflamm* or purulent or exudat* or devital*):ti,ab,kw #7 (positive NEAR/5 culture*):ti,ab,kw #8 (pain* NEAR/5 wound*):ti,ab,kw #9 (dirty NEAR/5 wound*):ti,ab,kw #10 (#5 OR #6 OR #7 OR #8 OR #9) #11 MeSH descriptor Anti‐Bacterial Agents explode all trees #12 (antibiotic* or antimicrobial* or amoxicillin or amoxycillin or ampicillin or bacitracin or cephalothin or cefazolin or cefotaxime or cefoperazone or ceftazidime or ceftriaxone or cefuroxime or erythromycin or cloxacillin or furazolidone or gentamicin or gramicidin or imipenem or "mafenide acetate" or mupirocin or natamycin or neomycin or nitrofurazone or oxacillin or penicillin or piperacillin or polymyxin or rifam* or "silver nitrate" or "silver sulfadiazine" or "sulfacetamide sodium" or tobramycin or amphotericin or tazocin or teicoplanin or tetracycline or sulfamethoxazole or vancomycin or chloramphenicol or ciprofloxacin or clarithromycin or clindamycin or cloxacillin or colistin or colymycin):ti,ab,kw #13 (#11 OR #12) #14 (#4 AND #10 AND #13)

What's new

Date	Event	Description
3 July 2017	Amended	This protocol has been withdrawn as authors were unable to progress the review.

Contributions of authors

Jing Lu: conceived the review question; developed the protocol; produced the first draft of the protocol; contributed to writing or editing the protocol; advised on the protocol; and approved the final version of the protocol prior to submission.

Ming Yang: conceived the review question; developed the protocol; produced the first draft of the protocol; contributed to writing or editing the protocol; and advised on the protocol.

Mei Zhan: conceived the review question; produced the first draft of the protocol; contributed to writing or editing the protocol; and advised on the protocol.

Xuewen Xu: conceived the review question; coordinated the protocol development; produced the first draft of the protocol; made an intellectual contribution to the protocol; and is a guarantor of the protocol.

Jirong Yue: conceived the review question; coordinated the protocol development; produced the first draft of the protocol; and made an intellectual contribution to the protocol.

Ting Xu: conceived the review question; produced the first draft of the protocol; made an intellectual contribution to the protocol; and is a guarantor of the protocol.

Contributions of the editorial base:

Joan Webster, Editor edited the protocol; advised on methodology, interpretation and protocol content. Approved the final protocol prior to submission. Sally Bell‐Syer, Gill Rizzello: Managing Editors: coordinated the editorial process. Advised on interpretation and content. Edited the protocol. Ruth Foxlee, Trial Search Co‐ordinator, drafted the search strategy.

Sources of support

Internal sources

• Chinese Cochrane Center, China.

External sources

• This project was supported by the National Institute for Health Research, via Cochrane Infrastructure funding to Cochrane Wounds. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, NHS or the Department of Health, UK.

Declarations of interest

Jing Lu: none known Ming Yang: none known Mei Zhan: none known Xuewen Xu: none known Jirong Yue: none known Ting Xu: none known

Notes

This protocol has been withdrawn as authors were unable to progress the review.

Withdrawn from publication for reasons stated in the review