(P-448) Nsaids and Paracetamol-induced Hepatotoxicity: A Systematic Review and Meta-analysis of Observational Studies

Background: NSAIDs and paracetamol play role in DILI. Performing systematic reviews (SR) and meta-analyses (MA) of observational studies that detect serious and uncommon adverse events in real life compiles the present evidence for a specific question by reducing bias.

Objectives: To provide a comprehensive SR and MA of the observational studies about hepatotoxicity of NSAIDs and paracetamol with regards to their liver safety.

Methods: Methodology based on MOOSE and ENCePP Methods. Age was limited to adults (>18-years old), exposure limited to NSAIDs and paracetamol. Safety endpoint was liver safety. PubMed and Web of Science (WoS) were used without time limitation until 2020 for completed observational studies in cohort, retrospective or prospective case-control, case-population, and case series designs. To create a comprehensive search strategy, pre-defined MeSH words (hepatotoxicity, paracetamol, NSAID, and observational study and their synonym words or equal phrases) were used. Study selection was performed stepwise by 2 reviewers independently: 1) by title & abstracts, 2) by full text, 3) papers presenting data for NSAID- and paracetamol-exposed DILI were included in MA. Data from selected papers for SR were extracted on a pre-designed data collection proforma including the details of study design, outcome, results, and exposure. Quality of all included studies in the SR was assessed using Newcastle-Ottawa Scale (NOS) tool for observational studies. Each study that collects at least 7 out of 9 stars is considered of high quality. MedCalc Statistical Software version 19.1 and Cochrane Q and I2 test were used for MA and heterogeneity assessment consequtively. At I2 statistic, values >75% were considered a sign of high, 50-75% moderate, and <50% low heterogeneity.

Results: Final search on 18/11/2020 resulted in 1246 papers (715 PubMed, 531 WoS). Based on title & abstracts 481 papers selected. Case reports, conference, pediatric, review, and overdose papers and the ones with different objective or exposure were excluded. In final 20 papers were included in SR; only six had low quality. Among the remaining, 5 were included in MA. Included studies showed wide range of DILI proportion between 2.5% (95%CI: 1.6-3.5) and 58.7% (95% CI: 45.6-71). The I2 statistic was calculated as 99.3% and considered as high heterogenous, therefore, the random effect model was selected. NSAIDs and paracetamol-exposed DILI is calculated as 33.2% (95% CI: 9.9-62) among all 2220 DILI cases that were included in quantitative analysis.

Conclusions: Despite the high heterogeneity of papers, results of MA reveal a relatively high proportion of NSAIDs or paracetamol-induced liver injury.