Systematic reviews and meta-analyses of observational studies (MOOSE): Checklist. MOOSE Checklist Infliximab reduces hospitalizations and surgery interventions in patients with inflammatory bowel disease: a systematic review and meta analysis. Criteria Brief description of how the criteria were handled in the meta analysis Reporting of background should include Problem definition Hospitalization and surgery are markers of inflammatory bowel disease (IBD) severity and significantly contribute to the high economic burden of IBD. Hypothesis statement Infliximab therapy was suggested to reduce the rate of serious complications in IBD (hospitalization and surgery rate). Description of study outcomes We aimed to perform a systematic review and meta analysis of all studies (observational and experimental) that evaluated patients with IBD treated with infliximab and incidence of hospitalizations and surgery. Secondary outcome described in Methods section. Type of exposure or intervention used Infliximab (any dose or regimen) Type of study designs used All studies (observational and experimental) Study population Adult patients (aged 18 years or older) with IBD, irrespective of IBD severity, baseline diseases and risk factors. Reporting of search strategy should include Qualifications of searchers The credentials of investigators are indicated in the author list. 1
Search strategy, including time period included in the synthesis and keywords PubMed, from inception April 2012. Cochrane Library, from inception April 2012. Web of Science with Conference Proceedings, from inception April 2012. Databases and registries searched Search strings are supplied in appendix Medline through PubMed, CENTRAL at Cochrane Library and Web of Science with Conference Proceedings. Search software used, name and version, including special features No software was involved in search method Use of hand searching We performed handsearch of references from obtained studies. List of citations located and those excluded, including justifications Method of addressing articles published in languages other than English Method of handling abstracts and unpublished studies Description of any contact with authors Present in the flowchart. No language restrictions were applied. Abstracts contained pretended data. We did not contact any author. Reporting of methods should include Description of relevance or appropriateness of studies assembled for assessing the hypothesis to be tested Rationale for the selection and coding of data Detailed inclusion and exclusion criteria were described in the Methods section. Authors extracted data from study design, location, timeframe of study, patients characteristics, drugs used and its assessment, studies primary outcome, data of required outcomes and estimates adjustments. Assessment of confounding We made subgroup analysis according to IBD type and we explored different sources of heterogeneity. 2
Assessment of study quality, including blinding of quality assessors; stratification or regression on possible predictors of study results Quality of reporting was analyzed using a qualitative classification according to risk of bias (high, unclear or low risk). For observational studies we used a 6 items classification based on MOOSE, QATSO and STROBE. This system was adapted from previous published systematic review. For RCTs we adopted Cochrane Collaboration s Tool for assessing risk bias. Assessment of heterogeneity Statistical heterogeneity was evaluated using I 2 statistics. Description of statistical methods in sufficient detail to be replicated Provision of appropriate tables and graphics Description of methods of meta analyses, subgroup analyses, NNT calculations and assessment of publication bias are present in Methods section. We provided 1 flowchart figure and 2 figures with forest plots of outcomes. For supplementary section we provided 4 figures with quality appraisal graphs (2 for RCTs and 2 for observational studies). We supplied 2 tables with studies characteristics: one for RCTs and one for observational studies. Reporting of results should include Graph summarizing individual study estimates and overall estimate Table giving descriptive information for each study included Results of sensitivity/subgroup testing Figure 2 and 3. Table 1 and 2. Figure 2 and 3. Indication of statistical uncertainty of findings 95% confidence intervals were presented with all summary estimates and I 2 values. Reporting of discussion should include Quantitative assessment of bias Quality of studies and the potential impact of bias in results were discussed. Subgroups analyses for surgery rate indicate that CD patients are more likely to benefit with infliximab 3
treatment. Justification for exclusion We excluded studies that did not evaluate pretended outcomes or infliximab therapy, studies that included pediatric population and case series studies or studies with a sample size population smaller that pre established. Assessment of quality of included studies Overall quality of included studies was considered to be good. The higher risk of bias was found for potential selective reporting and failure to describe withdrawals in RCTs and presentation of unadjusted risk estimates in observational studies. Reporting of conclusions should include Consideration of alternative explanations for observed results We discussed the limitations inherent to individual studies (selective reporting, unadjusted risk estimates) and metaanalysis (pooling data of studies with different designs, settings and baseline morbidities and heterogeneous risk for hospitalizations and surgery) that could bias observed results. Generalization of the conclusions Our results suggest an important role of infliximab treatment in hospitalization and surgery (at least for CD patients) risk reduction. Guidelines for future research Specific designed prospective long term effectiveness studies are required to establish definite conclusions and to better estimate the true magnitude of this impact. Future studies should also use active controls to inform comparative effectiveness. Disclosure of funding source Disclosure of funding source was made. 4
Systematic reviews and meta-analyses of randomized controlled trials (PRISMA): Checklist. PRISMA Checklist Infliximab reduces hospitalizations and surgery interventions in patients with inflammatory bowel disease: a systematic review and meta-analysis. Section/topic # Checklist item Page TITLE Title 1 Identify the report as a systematic review, meta-analysis, or both. Infliximab reduces hospitalizations and surgery interventions in patients with inflammatory bowel disease: a systematic review and meta-analysis. 1 ABSTRACT Structured summary 2 Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number. OBJECTIVE: To systematically review interventional and observational studies evaluating patients with inflammatory bowel disease (IBD) treated with infliximab to estimate their risk of hospitalizations and surgery. DESIGN: Systematic review and meta-analysis. DATA SOURCES: Medline through PubMed, Cochrane Library and Web of Science with Conference Proceedings from inception to April 2012. Systematic reviews and references of retrieved articles were comprehensively searched. STUDY SELECTION: Two reviewers independently selected clinical trials and observational studies evaluating IBD patients treated with infliximab and reporting on hospitalization and/or surgery rate, and retrieved studies characteristics and data estimates. DATA SYNTHESIS: Primary and secondary outcomes were incidence of hospitalization and surgery. Analyses were carried according to study design (randomized clinical trials RCTs, and observational studies) and IBD type (Crohn s 5 2, 3
INTRODUCTION disease [CD] and ulcerative colitis [UC]). Random-effects meta-analysis was used to derive pooled and 95% confidence intervals estimates of odds ratios (OR). Heterogeneity was assessed with I 2 test. RESULTS: Twenty-seven eligible studies were included (9 RCTs and 18 observational studies). Infliximab significantly reduced hospitalization risk, both in pooled RCTs (OR 0.51, 95% CI 0.40-0.65; I 2 =0%) and observational studies results (OR 0.29, 95% CI 0.19-0.43; I 2 =87%), without differences between CD and UC patients. Infliximab also significantly reduced surgery risk in pooled RCTs results (OR 0.36, 95% CI 0.18-0.71; I 2 =65%), both in CD and UC patients. Pooled estimate from observational studies favored infliximab for CD (OR 0.30, 95% CI 0.18-0.49; I 2 =78%), but not for UC patients. CONCLUSIONS: The best evidence available points towards a reduction of the risk of hospitalization and surgery requirement in IBD patients treated with infliximab. This impact is clinically and economically relevant because hospitalization and surgery are considered to be markers of disease severity and significantly contribute to the total direct costs associated with IBD. Rationale 3 Describe the rationale for the review in the context of what is already known. The requirement for hospitalization and surgery are both currently thought to be markers of IBD severity. Furthermore, medical and surgical hospitalizations positivity impact significantly on the globally high economic burden of IBD. In IBD, infliximab therapy has been claimed to be associated with a reduction of serious complications, such as hospitalization and surgery rate. However, contradictory findings have been reported and uncertainty exists regarding the true magnitude of this putative impact. 5, 6 Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS). We aimed to perform a systematic review and meta-analysis of all studies (experimental and observational) evaluating IBD patients treated with infliximab and providing data on the rate of serious complications (hospitalizations and/or surgery). PICOS: Participants: Adult patients (aged 18 years or older) with IBD, irrespective of IBD severity, baseline diseases and risk factors. 6 6
Interventions: Infliximab. Comparison: Infliximab vs. placebo, no treatment or other active non-biologic drug, adherence vs. non-adherence to infliximab therapy and schedule/persistence maintenance vs. episodic/nonpersistence infliximab therapy. Outcomes: Hospitalization related to IBD (primary) and global (abdominal and/or anal) surgery rate, colectomy rate and hospitalization length (secondary). Study designs: longitudinal studies (RCTs, cohort, case-control and pre-post exposure to infliximab studies). METHODS Protocol and registration 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number. Not a registered systematic review. - Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale. We considered both experimental (RCTs) and observational studies (prospective or retrospective cohorts including retrospective analyses of medical claims, case-control or pre-post exposure to infliximab). All types of studies participants were allowed irrespective of IBD severity, baseline diseases and risk factors. Only studies evaluating adult patients (aged 18 years or older) were considered. Accepted comparisons included infliximab (any dose and/or regime) vs. placebo, no treatment or other active non-biologic drug, adherence vs. non-adherence to infliximab therapy and schedule/persistence maintenance vs. episodic/nonpersistence infliximab therapy. Cohort studies could be based on ambulatory or institution/hospital population and had to follow patients to determine hospitalization or surgery outcomes. In case-control studies, cases had to be defined as IBD patients with pretended outcomes identified through clinical or by database codes. Controls should be matched to cases, and they should not have any of those outcomes. 6, 7 We included published studies retrieved in databases from inception until April 2012. Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched. 7
Medline through PubMed (inception-april 2012) CENTRAL at Cochrane Library (inception-april 2012) Web of Science with Conference Proceedings (inception-april 2012) 7 Handsearch of references from obtained studies. Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated. Appendix. Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis). Studies were identified through computerized database searches. Titles and abstract of obtained records were screened. Selected studies were assessed in full-text to determine its appropriateness for inclusion in the review. Flowchart presented in Figure 1. 7, Figure 1 Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators. Three authors independently extracted data from study. Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made. Investigators extracted data from study design, location, time-frame of study, patients characteristics, drugs used and its assessment, studies primary outcome, data of required outcomes and estimates adjustments. 7 6, 7, 9 Assumptions/simplifications: 8
Our primary outcome was hospitalization related to IBD, independently of its goal. Cases of hospitalization were extracted irrespectively of having been reported as pre-specified outcome or not. We did not aim to evaluate hospitalizations resulting from adverse effects, such as serious infections. Global gastrointestinal surgery rate was our secondary outcome. This was defined as abdominal and/or anal surgery. To perform meta-analysis raw data was firstly converted to OR through classic methods or through Peto s method if one arm had zero-count cell. Risk of bias in individual studies 12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis. For observational studies: We used a 6-items classification based on MOOSE, QATSO and STROBE. This system was adapted from previous published systematic review and took the following items into consideration: participants (if any justification was given for the cohort and study reported appropriate inclusion and exclusion criteria), intervention (if participants drugs use was adequately assessed, and not based on self-report), disease (if IBD was assessed by clinical, laboratory, radiologic and endoscopic criteria or through databases codes, and not based on self-report), outcome (if pre-specified) and adjustment for potential prognostic confounders. For RCTs: We adapted Cochrane Collaboration s Tool for assessing risk bias to evaluate reporting quality: randomisation method, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, selective reporting (if pneumonia was a pre-specified outcome) and description of withdrawals. 8 Risk of bias graphs were done using these tools. Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means). Forest plot were generated to show results of individual studies and combined analysis. These were expressed as OR with 95% CI and stratified according with studies design. 8, 9 9
Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e.g., I 2 ) for each meta-analysis. Forest plot were generated to show results of individual studies and combined analysis. These were expressed as OR with 95% CI and stratified according with studies design. Statistical heterogeneity was evaluated using I 2 statistics. 8, 9 10