Association Between Psychotropic and Cardiovascular Iatrogenic Alerts and Risk of Hospitalizations in Elderly People Treated for Dementia: A Self-Controlled Case Series Study Based on the Matching of 2 French Health Insurance Databases

Lorene Zerah (1, 2, 3) , Jacques Boddaert (4, 2, 5, 3) , Armelle Leperre-Desplanques (6) , Dominique Bonnet-Zamponi (7, 3) , Marc Verny (4, 2, 5, 3) , Jean Deligne , Pierre-Yves Boelle (1, 8)

1 iPLESP - Institut Pierre Louis d'Epidémiologie et de Santé Publique
2 B2A-IBPS - Adaptation Biologique et Vieillissement = Biological Adaptation and Ageing
3 CHU Pitié-Salpêtrière [AP-HP]
4 Brain-C - Biologie et Pathologie du Neurone
5 DHU - FAST - Departement Hospitalo- Universitaire - Fight Ageing and STress [Paris]
6 HAS - Haute Autorité de Santé [Saint-Denis La Plaine]
7 OMEDIT Normandie - Observatoire du médicament, des dispositifs médicaux et de l'innovation thérapeutique de Normandie
8 CHU Saint-Antoine [AP-HP]

Background: Elderly people are at risk of repeated hospitalizations, some of which may be drug related and preventable. In 2011, a group of French healthcare experts selected 5 iatrogenic alerts (IAs), based on criteria identified in a literature search and from their professional experience, to assess the appropriateness of medication in elderly patients. Objectives: Our objective was to examine the association between hospitalizations and IAs in elderly patients treated for Alzheimer disease who are particularly sensitive to adverse drug events. Design: A 2-year (January 1, 2011, to December 31, 2012) longitudinal national database study, with a study design similar to self-controlled case series, was performed to analyze data on drug prescriptions and hospitalization. IAs were defined as (1) long half-life benzodiazepine; (2) antipsychotic drugs in patients with Alzheimer disease; (3) co-prescription of 3 or more psychotropic drugs; (4) co-prescription of 2 or more diuretics; and (5) co-prescription of 4 or more antihypertensive drugs. Data were obtained by matching of 2 French National Health Insurance Databases. Setting: France. Participants: All affiliates, aged >= 75 years, receiving treatment for Alzheimer disease, alive on January 1, 2011 were included. Measurements: We calculated the relative increase in the number of hospitalizations in patients with IAs. The analysis was performed over four 6-month periods. Results: A total of 10,754 patients were included. During the periods with IAs, hospitalization rates increased by 0.36/year compared with 0.23/year in the periods without for the same patient, and the number of hospitalizations doubled [proportional fold change = 1.9, 95% confidence interval (1.8, 2.1)]. We estimated that 22% [95% confidence interval (20%, 23%)] of all hospitalizations were associated with IAs, 80% of which were due to psychotropic IAs. Conclusions: The IAs could be used as a simple and clinically relevant tool by prescribing physicians to assess the appropriateness of the prescription in elderly patients treated for Alzheimer disease.

Mots clés: Iatrogenic alert; adverse drug event; self-controlled analysis

Domaines: Biologie cellulaire

Lorene Zerah, Jacques Boddaert, Armelle Leperre-Desplanques, Dominique Bonnet-Zamponi, Marc Verny, et al.. Association Between Psychotropic and Cardiovascular Iatrogenic Alerts and Risk of Hospitalizations in Elderly People Treated for Dementia: A Self-Controlled Case Series Study Based on the Matching of 2 French Health Insurance Databases. Journal of the American Medical Directors Association, 2017, 18 (6), pp.549.e1-549.e13. ⟨10.1016/j.jamda.2017.02.001⟩. ⟨hal-01545498⟩