Background: A number of studies have examined the predictive utility and time to response of rating scales and demographic variables. Very few community samples have been examined in this way, and no studies examining the prognostic validity of early symptomatic response have been found in the literature.
Objectives: This study aims to describe how treatment response is reflected in rating scales in older community residents treated with sertraline and to explore the utility of these instruments in predicting response.
Methods: The study examines the open label therapeutic and continuation phases of a maintenance trial.
Results: 225 older depressed community residents were treated (openly) with sertraline. Fifty-three percent had a good outcome, 13% did not respond to sertraline and had a poor long-term prognosis. Increased age was associated with poor outcome and increased anxiety symptoms with a good outcome. In the compliant sub-sample, GMS/AGECAT schizophrenia symptoms were associated with poor response to treatment. Baseline HDRS items and related symptom clusters were not of predictive utility, however early changes in HDRS score (improvement from baseline of four or more by four weeks) was associated with good outcome. All symptom clusters improved within two weeks of treatment with sleep symptoms improving by six weeks. Optimum symptomatic improvement was achieved by eight weeks.
Conclusions: Clinicians in primary care can expect 53% response to treatment. In the absence of symptomatic improvement by one month (HDRS score of four or more) treatment should be reviewed. Optimum treatment response is usually achieved within eight weeks.
Copyright 2002 John Wiley & Sons, Ltd.