Risk prediction models on adverse drug reactions: A review

Background: The risk prediction model has become increasingly popular in recent years in helping clinical decision-making. Existing models cannot be directly applied in Indonesia. Objective: To review the existing prediction models and their limitations. Method: A search related to the prediction of ADR risk was conducted using several journal databases: PubMed, Scopus and Google Scholar. Articles were screened to match specified criteria and further studied. Result: Nine articles met the criteria and were then analysed. Studies were carried out in various countries. The study population include; the elderly (>65 years, three studies), age (≥15 years, three studies), patients with Chronic Kidney Disease (CKD) (≥18 years, one study) and two studies in cancer patients. The outcomes were; ADR (five studies), ADE ( two studies), DRPs (one study), and cardiovascular effects (one study). The methods for determining the predictors of ADRs all used multivariable logistic regression. Conclusion: Each country has different treatment patterns, prescribing practices, traditions and drug distribution, so it is necessary to develop a prediction model for ADRs that is country-specific.