A Dynamic Clinical Prediction Model to Stratify the Risk of Preventable Drug-related Incidents in Infectious Diseases Ward Patients

Abstract

Background: Preventable drug-related incidents (PDRIs) remain a major challenge in patient safety, particularly among adults hospitalized with infectious diseases, where complex pharmacotherapy and frequent clinical changes increase the likelihood of harm. This study aimed to develop and validate a clinical prediction model to estimate the risk of PDRIs and support early, pharmacist-led interventions. Methods: A prospective cohort study was conducted among adult inpatients admitted to the infectious disease ward of a tertiary teaching hospital. Data were collected over nine months, from June 2019 to March 2020, and the study was concluded earlier than planned due to restrictions imposed by the COVID-19 pandemic. Information regarding demographic, clinical, and pharmacotherapeutic characteristics, as well as the quality of medication conciliation, was recorded. The model was developed using multivariate logistic regression and internally validated through cross-validation techniques. Results: A total of 212 patients were included. The incidence of PDRIs was high, especially among those with prolonged hospitalization and multiple comorbidities. Independent predictors of PDRI included the number of medications, the presence of comorbidities, infectious diagnosis, and the quality of medication conciliation. Notably, conciliation performed by non-pharmacists was associated with a significantly higher risk of PDRI compared with pharmacist-led conciliation. Conclusion: The proposed model provides a practical and reliable tool to dynamically stratify PDRI risk throughout hospitalization. Its implementation may enhance the ability of clinical pharmacists to prioritize high-risk patients, enabling earlier interventions and contributing to safer and more efficient pharmacotherapy in infectious disease wards.