Data collection and statistical analysis
Each patient chart was scored using the Naranjo scale (Fig 1) and the Liverpool causality assessment tool (Fig 2). For the questions addressing whether there has been any objective evidence for the adverse event, a score of 0 was assigned to all charts since this study aims to retrospectively assess the tools based only on history and clinical impression prior to any investigations as a screening method. Charts scoring “Doubtful” on Naranjo scale and “Unlikely” on Liverpool tool were considered negative for ADR, and any higher scores were considered positive for ADR screening. The results of the investigations were considered as true positive and true negatives. All data was organized in Microsoft Excel and calculations were done using its algorithm functions. Positive and negative predictive values as well as the sensitivity and specificity for Naranjo and Liverpool tools were calculated using appropriate formulas [20, 21]. The confidence intervals were determined using McCallum Layton confidence interval calculator for proportions with confidence level at 95%.