Many studies have attempted to estimate the prevalence of adherence using different methods in a variety of patient samples. Early studies in primary care settings in the United Kingdom indicated that up to two-thirds of depressed patients who started courses of tricyclic drugs stopped taking them within a month (29). Peveler et al. assessed a large population of patients receiving tricyclic medication in primary care settings in the United Kingdom using EMS, and found that around 40% had discontinued treatment within 12 weeks (5). In 1990, McCombs et al. attempted to assess adherence in a large sample of depressed Medicaid-funded patients in California, United States, but found it difficult to separate patient's adherence to therapies from physician's adherence to treatment guidelines (6). Katon et al. assessed the extent to which patients of an HMO, on receiving prescriptions for antidepressant drugs, actually obtained supplies of medication. They reported that only 20% of patients who had been prescribed tricyclic drugs filled four or more prescriptions within 6 months, while 34% of patients who had been prescribed newer antidepressants did so (7). Lin et al. assessed a very large sample of HMO patients 6 - 8 weeks after starting treatment and found that 32 - 42% had not filled their prescriptions (8).

In a sample of patients with psychiatric disorders receiving prophylactic lithium treatment for unipolar and bipolar affective illness, Schumann et al. found that 43% of patients had discontinued their medication within 6 months (9). Ramana et al. interviewed patients discharged from hospital following admission for depression and found that at 18 months about 70% were "compliant", although this study also noted problems with physicians under-prescribing according to guidelines (10).

Gasquet et al. conducted a large telephone survey of the general population in France (11). He reported that 15% of the subjects admitted to early termination of their treatment, and 22% admitted to reducing their dose.