N-acetylcysteine for non-paracetamol drug-induced liver injury: a systematic review protocol

Background Drug-induced liver injury (DILI) refers to acute or chronic liver injury that may occur as a consequence of using drugs and herbal or dietary supplements. Specific therapies for DILI are limited. There is considerable evidence for efficacy and safety of N-acetylcysteine (NAC) in management of paracetamol-induced liver injury. More recently, research has explored the use of NAC in non-paracetamol drug-induced liver injury. It is important to summarise the evidence of NAC for non-paracetamol DILI to determine if NAC may be considered a therapeutic option in this condition. Methods/design We will conduct a systematic review of the benefit and harm of NAC in non-paracetamol drug-induced liver injury. Primary and secondary outcomes of interest are pre-specified. Primary outcomes include all-cause mortality, mortality due to DILI, time to normalisation of liver biochemistry (e.g. return of alanine transaminase to <100 U/l and/or international normalized ratio (INR) <1.5) and adverse events. Secondary outcomes include transplantation rate, time to transplantation, transplant-free survival and duration of hospitalisation. We will include randomized controlled trials (RCTs) and prospective cohort studies. RCTs will contribute to the evaluation of safety and efficacy of NAC, whereas, the cohort studies will contribute exclusively to the evaluation of safety. We will search several bibliographic databases (including PubMed, Scopus, CINAHL, CENTRAL), grey literature sources, conference proceedings and ongoing trials. Following data extraction and assessment of the risk of bias, we will conduct a meta-analysis if feasible, as well as subgroup analyses. We will assess and explore clinical and statistical heterogeneity. Discussion The aim of this review is to provide evidence on the effectiveness and safety of NAC in non-paracetamol DILI. We anticipate that the results could aid health care practitioners, researchers and policymakers in the decision-making regarding the use of NAC in patients with non-paracetamol DILI. Systematic review registration PROSPERO CRD42014008771 Electronic supplementary material The online version of this article (doi:10.1186/s13643-015-0075-6) contains supplementary material, which is available to authorized users.


Is the source population (cases, controls, cohorts) appropriate and representative of the population of interest?
0 (high risk of bias) 1 2 3 (low risk of bias) Example of low risk of bias: A consecutive sample or random selection from a population that is representative of the condition under study.
Example of moderate risk of bias: A consecutive sample or random selection from a population that is not highly representative of the condition under study.
Example of high risk of bias: The source population cannot be defined or enumerated (i.e. volunteering or self-recruitment).

Domain of evaluation: Methods to control confounding (i.e. Performance bias)
Is the sample size adequate and is there sufficient power to detect a meaningful difference in the outcome of interest? 0 (high risk of bias) 1 2 3 (low risk of bias) Example of low risk of bias: Sample size was adequate and there was sufficient power to detect a difference in the outcome.
Example of high risk of bias: Sample size was small and there was not enough power to test outcome of interest.
Did the study identify and adjust for any variables or confounders that may influence the outcome? 0 1 2 3 0 = Definitely no (high risk of bias) 1 = Mostly no 2 = Mostly yes 3 = Definitely yes (low risk of bias) (high risk of bias) (low risk of bias) Example of low risk of bias: The study identified and adjusted for all possible confounders that may influence estimates of association between exposure and outcome (i.e. Was the patient being treated for a medical condition such as chronic pain and was being prescribed opioids while on methadone treatment?) Example of moderate risk of bias: The study identified and reported possible variables that may influence the outcome but did not explore the interaction.
Example of high risk of bias: The study either did not report any variables of influence or acknowledge variables of influence when it was clear they were present.

Domain of evaluation: Statistical methods (i.e. Detection bias)
Did the study use appropriate statistical analysis methods relative to the outcome of interest? 0 (high risk of bias) 1 2 3 (low risk of bias) Example of low risk of bias: The study reported use of appropriate statistical analysis as required (i.e. adjusting for an unbalanced distribution of a specific covariate among sexes, or correcting for multiple testing error) Example of moderate risk of bias: The study either used correct statistical methods but did not report them well, or used the incorrect methods but reported them in detail.
Example of high risk of bias: The study did not use appropriate statistical analysis as required (i.e. did not adjust for an unbalanced distribution of a specific covariate among sexes, or correct for multiple testing error when necessary) or did not report them adequately.
Is there little missing data and did the study handle it accordingly? 0 (high risk of bias) 1 2 3 (low risk of bias) Example of low risk of bias: The study acknowledged missing data to be less than 10% and specified the method of handling it.
Example of moderate risk of bias: The study either had greater than 15% but they specified the method they used to handle it.
Example of high risk of bias: The study had greater than 15% missing data and did not handle it at all.

Domain of evaluation: Methods for measuring outcome variables (i.e. Information bias)
Is the methodology of the outcome measurement explicitly stated and is it appropriate? 0 (high risk of bias) 1 2 3 (low risk of bias) Example of low risk of bias: The study provides a detailed description of the outcome measure(s) which are appropriate for the outcome of interest.
Example of moderate risk of bias: The study provides a somewhat complete description of outcome measurements and they are justified.
Example of high risk of bias: The study provides limited information on the methods of measuring the outcome and the measure is not appropriate considering the outcome.
Is there an objective assessment of the outcome of interest? 0 (high risk of bias) 1 2 3 (low risk of bias) Example of low risk of bias: The study used objective methods to discern the outcome status of participants (i.e. laboratory measurements, medical records).
Example of moderate risk of bias: The study relied on subjective data as the primary method to discern outcome status of participants (i.e. self-report).
Example of high risk of bias: The study had limited reporting about assessment of outcomes.