Abstract
The validity of diagnostic psychiatric instruments for depression and anxiety disorders may be compromised among patients with complex physical illness and disability. The objective of this study was to determine the effect on the prevalence rate of depression and anxiety in a nursing home population of attributing somatic symptoms of depression and anxiety to either somatic or psychiatric disorder. Symptoms of major depression (MD), generalized anxiety disorder (GAD) and panic disorder (PD) were measured using the Schedules for Clinical Assessment in Neuropsychiatry (SCAN). Somatic symptoms of MD, GAD and PD were attributed to somatic causes when the interviewer was not sure about a psychiatric cause. To analyse the effect of this attribution on the prevalence rate of MD, GAD and PD, a sensitivity analysis was undertaken in which symptoms that were attributed to somatic causes were recoded as symptoms attributed to psychiatric disorder. Prevalence rates of MD, GAD and PD were calculated before and after recoding. The prevalence of MD after recoding rose from 7.5% to 8.1%. The prevalence of GAD did not change. The prevalence of PD rose from 1.5% to 1.8%. Attribution of somatic symptoms to either somatic or psychiatric disorder when the interviewer was not sure about a psychiatric cause of the somatic symptoms had only a very modest effect on the prevalence rate of major depression, generalized anxiety disorder and panic disorder in a nursing home population.