Development and validation of a clinical instrument to predict risk of an adverse drug reactions in hospitalized patients

• DOI: 10.1371/journal.pone.0243714
• Publication Date: 2020-12-11T18:46:38Z
• Authors: Sara Iasmin Vieir...; Rand Randall Martins; Valdjane Saldanha; Vivian Nogueira S...; Isabelle Cristina...; Ivonete Batista d...; Antonio Gouveia O...
• Keywords: Adult; Male; Prescription Drugs; Drug-Related Side Effects and Adverse Reactions; Science; Drug Prescriptions; Risk Assessment; Cohort Studies; 03 medical and health sciences; 0302 clinical medicine; Risk Factors; Patients' Rooms; Odds Ratio; Humans; Aged; Incidence; Q; R; Middle Aged; 3. Good health; Hospitalization; Logistic Models; ROC Curve; Case-Control Studies; Medicine; Female; Research Article

Objective Development and internal validation of a clinical tool for assessment the risk adverse drug reactions (ADR) in hospitalized patients. Methodology Nested case-control study an open cohort all patients admitted to general hospital. Cases ADR were matched two controls. Eighty four patient variables collected at time analyzed by conditional logistic regression. Multivariate regression with clustering cases random sample 2/3 respective controls, baseline odds-ratio corrected estimate incidence, was used obtain coefficients each factor develop score. The validated remaining 1/3 observations. approved institution’s research ethics committee. Results In 8060 patients, occurred 343 (5.31%), who 686 Fourteen identified as independent factors ADR: female, past history ADR, heart rate ≥72 bpm, systolic blood pressure≥148 mmHg, diastolic pressure <79 diabetes mellitus, serum urea ≥ 67 mg/dL, sodium ≥141 mmol/L, potassium ≥4.9 main diagnosis neoplasia, prescription ≥3 ATC class B drugs, R intravenous drugs 6 oral drugs. sample, based on those showed sensitivity 61%, specificity 73% area under ROC curve 0.73. Conclusion We report stratification wards 14 variables.