First hospitalization for transient ischemic attack in France: Characteristics, treatments and 3-year outcomes.

INTRODUCTION: Characteristics of patients hospitalized for transient ischemic attack (TIA) management before and during this hospitalization and follow-up are not well documented on very large populations. METHODS: Among the 51 million beneficiaries of the French national health insurance general scheme (77% of French population), those subjects hospitalized for a first TIA in 2010 were included using the national health insurance information system (SNIIRAM). The frequencies of comorbidities during the previous five years and drug treatments received during the previous year and the first month after discharge were estimated from the SNIIRAM and then compared to data derived from the permanent randomized sample of all health insurance beneficiaries based on standardized morbidity ratios (SMR). The three-year outcome and factors associated with at least one readmission for TIA or ischemic stroke during the three months following the first hospitalization were investigated. RESULTS: A total of 18,181 patients were included (mean age: 69 years, 55% of women). The crude incidence of hospitalized TIA was 0.36 per 1000. Before hospitalization, patients presented a significantly higher rate of carotid and cerebral atherosclerosis (2.4% SMR=1.4), atrial fibrillation (9.1%, SMR=1.3), ischemic heart disease (13.7%, SMR=1.3), valvular heart disease (9.7%, SMR=1.5), and treatment with platelet aggregation inhibitors (29%, SMR=1.4), antihypertensives (60%, SMR=1.2) and antidiabetics (16%, SMR=1.5). These SMR decreased with age. One month after discharge from hospital, 82% of patients still alive filled at least one prescription for antithrombotic therapy (platelet aggregation inhibitor: 74%, vitamin K antagonist: 12%), one class of antihypertensive in 57% of patients, an antiarrhythmic in 9% of patients, an antidiabetic treatment in 14% of patients and a lipid-lowering agent in 53%. During the month following discharge from hospital, 3.2% of patients were readmitted at least once for TIA, 1.9% were readmitted for ischemic stroke and 1.5% of patients died. These figures were 3.9%, 2.4% and 2.9% at three months, and 7.2%, 5% and 16.3% at three years, respectively. On multivariate analysis, factors associated with readmission for TIA or ischemic stroke were age ≥ 65 years and antidiabetic treatment before hospitalization. In contrast, male gender, admission to a stroke unit and length of stay were associated with a lower readmission rate. CONCLUSIONS: These results illustrate the value of administrative databases to study TIA. Hospitalizations for TIA were relatively frequent and the recurrence rate was similar to that reported in similar recent studies. Level of primary and secondary prevention must be improved.