[Prospective multicenter study of adverse drug reactions in hospitalized elderly patients]

The incidence of adverse drug reactions (ADRs) in hospitalized elderly patients is an important clinical problem. The purpose of this study was to determine the incidence of ADRs in elderly in-patients, to analyze the factors involved in their presentation and to evaluate the reactions detected.

Prospective multicenter study in patients older than 65 years during their stay in hospital or nursing home. The assessment consisted of a complete geriatric evaluation and a protocol for collecting information on suspected ADRs during hospitalization or nursing home stay. A multivariate analysis by multiple logistic regression was performed.

The study included 865 patients: 185 (21%) from 5 hospital units; 325 (38%) from 8 convalescent centers; and 355 (41%) from 8 long-term centers. The incidence of ADRs was 9%. In acute units the incidence was 15%; in post-acute units, 5%; and in long-term centers, 10% (p < 0.004). In the global multivariate analysis, the risk of experiencing an ADR was associated with a greater use of medications (OR = 1.15; 1.07 1.23 for each additional drug), the presentation of delirium (OR = 3.6; 1.95-6.85) and the type of unit (acute OR = 2.6; 1.16-6.01; long-term OR = 3.3; 1.62-7.05). Fifty type A ADRs were detected (65%). With regard to severity, 36 (47%) were moderate, 27 (35%) were mild, and 14 (18%) were severe. Causality: 58 (76%) probable and 17 (20%) possible. Potential avoidability: 39 (51%) were unavoidable and 38 (49%) were totally or partially avoidable; of these, 15 (39%) were the result of an interaction with another drug.

The incidence of ADRs in our study was close to 10%. ADRs were associated with the frequency of use, presence of delirium and type of unit, and occurred most frequently in acute and long-term units. ADRs were principally type A, moderate severity, probable causality and partially avoidable.