Explicit versus implicit review to explore combination antipsychotic prescribing

To use structured implicit review following large-scale explicit audit of antipsychotic polyprescribing to: (1) determine the true rate of antipsychotic polytherapy that deviated from best practice for schizophrenia treatment; and (2) assess whether explicit antipsychotic polytherapy criterion was appropriate for identifying patients at risk for medication problems and assessing quality of care.

Antipsychotic prescribing was reviewed for outpatients in four public health services in Auckland, New Zealand on 31 October 2004 (T1). Schizophrenia patients in one service (n = 794) prescribed antipsychotic polytherapy (n = 84, 10.6%) were followed up 10 months later (T2). Historical medication summaries were prepared for those remaining on polytherapy, including diagnosis, clinical problems and treatment plan. Criteria for structured implicit review and rating form for quality of antipsychotic management were piloted. All medication summaries were independently rated by two reviewers, and a third independent rater reviewed summaries where disagreement was found.

Forty-nine patients remained on long-term polytherapy at T2 (6.2% of original population). All but two cases included a second-generation antipsychotic. At T2, average polytherapy duration was 35.8 months, and average antipsychotic dose was 699 mg day(-1) chlorpromazine equivalents. Two raters achieved agreement for 24/49 summaries, and the remaining 25 were rated independently by a third reviewer. Consensus agreement of antipsychotic management (by two raters) was reached for 44/49 cases (89.8%). Polytherapy was rated 'well-justified' in 32.7%, 'some justification' in 10.2% and 'lacked justification' in 46.9% cases. The final rate of polytherapy deviating from best practice reduced from 10.6% to 3.5% when short-term polytherapy was excluded, and details of the clinical situation and care plan were included in implicit review.

Audit of prescribing in routine practice using explicit guideline-based criteria may be a useful baseline performance indicator. It does not provide an accurate measurement of quality of care because it overestimates the deviation rate from good practice. It may also identify complex patients at risk for poor treatment outcomes who may benefit from structured treatment review.