(C11) Effect of Mobile Applications on the Spontaneous Reporting of Adverse Drug Reactions: A Systematic Review and Individual Patient Data Meta-Analysis
Professor of Pharmacy Discipline of Pharmacy, University of Technology Sydney, Sydney, NSW, Australia, Australia
Background: This study estimated the rate and quality of reports of mobile adverse drug reactions applications (MADRAS) for reporting adverse drug reactions (ADR).
Objectives: To explore the effect of MADRAS on the rate and quality of ADR reports.
Methods: The study’s protocol was registered with PROSPERO registry (registration number: CRD42021240555)
Design: Systematic review and individual patient data meta-analysis Search strategy and selection criteria We searched EMBASE, CINAHL, SCOPUS and MEDLINE, from inception to February 2022. Cluster randomised controlled trials (cRCT) and non-randomized intervention studies (nRIS) that reported rate or quality of adverse drug reactions reports (ADR) in English were included, regardless of time and location. Exclusions include: studies with no control, assessed MADRAS for treatment outcomes, diagnoses and treatment, rather than ADR reporting, studies without data (e.g., editorials and opinion papers) and studies by Pharmaceutical companies since they were not likely to be spontaneous. Risk of bias of included studies were assessed using the Cochrane Risk of Bias 2 (RoB 2) and Risk Of Bias In Non-randomised Studies of Interventions (ROBINS-I). Statistical analysis The primary outcome was the rate of ADR reporting and the measure of effect was the incident rate ratio (95% CI), while the secondary outcome was the quality of the reports. Individual patient data (IPD) meta-analysis was conducted for only the cRCT and the risk of variability or heterogeneity between the studies examined. A two-stage meta-analytic method adopted, using a generalised linear mixed model, with a Poisson distribution. The incidence rate ratio (IRR) estimated for each of the two cRCTs. All analyses were by intention to treat. The R statistical package version 4.21 (R Foundation for Statistical Computing, Vienna, Austria) was used for the IPD analyses.
Results: We identified 7,248 studies but included five in the systematic review and 2 in the IPD meta-analysis. The five studies presented five different MADRAS, including 415 participants, 25 hospitals/clinics and 11 health posts that reported 35,812 ADR reports. Patients and healthcare professionals reported more reports using the MADRAS compared with the conventional tools (57.79% (1,772/3066) vs (42.20% (1294/3066). MADRAS significantly increased the rate of reporting by approximately 19 fold, compared to the conventional methods (IRR 18.9, 95% CI: 5.8; 61.0). However, the overall quality of reports was lower with the MADRAS compared with the conventional reports.
Conclusions: MADRAS increased the rate of reporting but not the quality of the reports. Future research should evaluate the sustained effects of MADRAS.
