Objective Ð To compare the safety and side-eect pro®les of the four selective serotonin reuptake inhibitor antidepressants (SSRIs), ¯uvoxamine, ¯uoxetine, sertraline and paroxetine.
Methods Ð The results from four observational cohort studies of the four SSRIs were compared. Each of these studies was conducted by Prescription-Event Monitoring (PEM). The exposure data were derived from general practitioner (GP) prescriptions con®dentially supplied by the Prescription Pricing Authority (PPA) in England. Outcome data were obtained from questionnaires (green forms) on which the prescribing doctor recorded event data. The main ®ndings comprised demographic information, including patients' date of birth and sex; the indication for prescribing the monitored drug; the eectiveness of the drug as perceived by the GP; the reasons for stopping treatment and all events recorded during and after treatment.
Results Ð The ®nal cohort for each of the four SSRIs exceeded 10,000 patients. The sex, age distributions and indications for prescribing the four SSRIs were very similar. Only 36% of the GPs expressing an opinion reported ¯uvoxamine as eective, compared with approximately 60% for ¯uoxetine, sertraline and paroxetine. Fluvoxamine was associated with a higher incidence of adverse events than the other three SSRIs. Nausea/vomiting was both the most frequent clinical reason for stopping all four SSRIs and the most frequently reported clinical event. Adverse events reported in patients aged 70 years and over were comparable with the events reported for the total cohorts. Dierences were identi®ed between the four SSRIs for less frequently reported adverse events. Withdrawal symptoms were signi®cantly more frequent with paroxetine than the other three SSRIs.
Conclusions Ð The data from the four studies were comparable in terms of age distribution, sex of patients and indication for prescribing the drugs. Fluvoxamine had a considerably higher incidence of side-eects associated with its use than the other three SSRIs. The side-eect pro®les of the four SSRIs were comparable for frequently reported events. Important dierences were identi®ed between the four SSRIs in respect of less frequently reported events. This study suggests that ¯uvoxamine compares unfavourably with ¯uoxetine, sertraline and paroxetine, both in terms of reported eectiveness and the incidence of adverse events. Biases possibly aecting the comparisons involved in this study are unlikely to account for the observed dierences between ¯uvoxamine and the other three SSRIs.