Objectives: A quarter of people with schizophrenia may be classed as ‘treatment-resistant’. Clozapine is an antipsychotic that holds significant potential benefit for this patient group and has recently been recommended by the National Institute for Clinical Excellence. Early discontinuation is common. This study explores the factors that predict such discontinuation.

Method: This retrospective cohort design was carried out on two hundred and one people with treatment resistant schizophrenia who had commenced clozapine between 1990 and 1997 identified from the Clozaril Patient Monitoring Service (CPMS) in Northern Ireland. Clinical and socio-demographic data was collected for three years before and after commencement on clozapine. Kaplan Meier survival analyses were conducted to identify differences in discontinuation rates according to a range of variables.

Results: Forty-five per cent of patients had discontinued before three years. No difference in rates were found between men and women, nor whether side-effects were reported or not Patients who had commenced clozapine at an older age had significantly higher cessation rates. People who had started clozapine at age 50+ were four times more likely to stop taking clozapine within three years than people aged between 17-29 years.

Conclusions: Differences in continuation rates may be due to a number of factors. Treatment resistance may increase with age, side-effects may worsen or patient willingness to tolerate such effects may reduce. Alternatively, clinicians may be more anxious about prescribing clozapine as levels of co-morbidity increase. Further research is required to identify precise reasons and develop interventions to reduce discontinuation rates among older patients.