Identification of risk factors for adverse drug events in a general hospital.

Abstract

Adverse events (AEs), and particularly adverse drug events (ADEs), represent a health problem as they can cause permanent damage or death. Understanding the frequency, location, and causes of ADEs can prevent harm to patients. The Global Trigger Tool, produced by the Institute for Healthcare Improvement (GTT/IHI), is widely used to identify AEs. Recognizing the profile of patients who suffer ADEs can reveal clinical or individual characteristics that predispose to the occurrence of AEs. A cross-sectional study was carried out through a retrospective analysis of 120 medical charts of patients discharged from hospital between October 2020 and April 2021. Patients over 18 years old, with a length of stay of more than 24 h, were included. The list of triggers used was from the medication module of the GTT/IHI, which was adapted for use in the institution. Two primary reviewers and a medical reviewer applied this tool. The primary reviewers independently assessed the randomized charts. A meeting to achieve consensus among the reviewers was held every 2 weeks to validate the identified ADEs; classifications were based on harm severity. Multivariate logistic regression was utilized to assess the variables that predicted the occurrence of ADEs, using the backward stepwise method. A total of 43 ADEs were identified, with a frequency of 36 per 100 admissions (43/120). Of these, five ADEs (12%) were responsible for patients being admitted to hospital. In the case of in-hospital ADEs, there were 42.2 per 1000 patients/day. The clinical manifestation of altered kidney function (16%) and the anatomical drug group of the nervous system (33%) were the most frequent ADEs. The multivariate logistic regression model was significant (×2 = 44.960, P < .001), indicating that factors such as: known drug allergy [odds ratio 5.728; 95% confidence interval (CI): 1.249, 26.274, P = .025]; being clinically hospitalized (odds ratio 7.504; 95% CI: 1.654, 34.037; P = .009); number of medicines used (odds ratio 1.100; 95% CI: 1.054, 1.148; P < .001); and being under the care of internal medicine (odds ratio 3.633; 95% CI: 1.257, 10.511; P = .017) were predictor variables associated with the occurrence of ADEs. A significant percentage of hospitalized patients experienced at least one ADE, with rates surpassing those found in similar studies. The GTT/IHI effectively assessed medication-related harm, emphasizing the need for tailored triggers based on population characteristics. Predictor variables can inform preventive strategies. Overall, the tool facilitated a localized risk assessment of medication use.