Quality of mental health service care: the forgotten pathway from process to outcome.

The validity of the concept of outcome depends on a relationship between routine treatment and later health status. Outcome evaluations and audits are very rare in psychiatry. A substantial expansion in epidemiologically based, naturalistic, observational, process-outcome data collection in routine psychiatric practice is essential in order to identify treatment allocation biases and other reasons for unexpected outcomes. Identified causes of undertreatment should lead to locally agreed detailed clinical guidelines. Experimental evaluation should take place in routine clinical practice settings, with change in both process and outcome as the objective. Ultimately, the results of both experimental and observational outcome studies on representative service users should converge, permitting outcomes to be the ultimate arbitrator of quality.