Sir: We wish to reply to Moncrieff & Smyth's paper on community treatment orders (CTOs) (Psychiatric Bulletin, November 1999, 23, 644-646). We write as British psychiatrists working in Melbourne where CTOs have been in use since 1987. In our service 161 adults (18-65 years) out of a total catchment area population of 225 000 are subject to a CTO. CTOs are regarded as part of good clinical facilitating treatment in the least restrictive environment many lasting for only a few months after hospital admission.

The CTO enables the clinician to insist on clinic attendance and the patient's acceptance of oral or intra-muscular medication. If the patient refuses to comply then the CTO may be revoked and the patient admitted to hospital, usually for a very brief period. This ultimate sanction is rarely required. The CTO is only one part of a comprehensive biopsychosocial care plan. An order may alter the dynamics of care, but the clinician-patient relationship is usually remarkably well preserved. In our clinical experience, CTOs are most helpful if some sort of therapeutic alliance has been established. This alliance can be continually developed and improved particularly by psycho-education sessions.

We dispute Moncrieff & Smyth's surprising statement that CTOs would increase the amount of medication administered and, therefore, side-effects. In our experience, doses as low as 20 mg of flupenthixol depot monthly are sufficient to spare an insightless patient the indignity of regular compulsory admissions to hospital with all the extra medication and restrictions that these inevitably entail. A prolonged symptom-free period may demonstrate the benefits of psychiatric care to a person with a recurrent psychotic illness.

The introduction of CTOs in Victoria was an integral part of the state-wide development of community care. They are not now, and never were, intended to be a means by which psychiatry could control antisocial behaviour.

The efficacy of CTOs has not been established, but having practised with them and without them we feel that they have an important part to play in good clinical care. We agree with Burns (Psychiatric Bulletin, November 1999, 23, 647-648) and wish to move the debate away from the issue of prevention of violence towards the provision of humane, community-based care.