Best Practices: Best Practices in the Use of Involuntary Outpatient Treatment

Evidence for the efficacy of involuntary outpatient treatment (IOT) is equivocal (1–3). The conflicting findings may reflect differences in legal systems, in the patients for whom IOT is used, in the treatments delivered, and in outcome measures (4). Nevertheless, many professional psychiatric associations and many psychiatrists support IOT, and it is widely used in some countries (1).

In this column, we draw on quantitative research surveys, qualitative studies, and official guidelines to describe clinicians' views and their reported practices when using IOT in four Commonwealth nations—England, Canada, Australia, and New Zealand—where IOT is usually called a community treatment order. [Because of space limitations, additional information about use of and research on IOT, including references for statements made below, is available online as a data supplement to this column.]

Caution is necessary in extrapolating from Commonwealth nations to the context of the United States or continental Europe. Variations in legislation, health service structures, and financing clearly influence practice. However, in a report on international practices in the use of IOT, Churchill and colleagues (1) noted “remarkable consistency in the characteristics of patients on [IOT] across jurisdictions in very different cultural and geographic settings.” Thus it seems likely that there is some consensus among clinicians in various countries about the proper candidates and best practices for using this form of care.

Commonwealth clinicians' views of IOT

Purpose of IOT

Commonwealth clinicians consider IOT appropriate to address the clinical needs of individuals with serious mental illness by maintaining care, especially medication. Reduction of dangerousness is viewed as a secondary consequence of addressing this clinical need. The preferred regimen has a preventive focus and permits a longitudinal view of a person's illness and its likely prognosis. Depending on jurisdiction, there is an implicit belief or explicit judgment that the individual has impaired competence to refuse the treatment offered. Cases of refusal represent more than just a bad decision. Therefore, IOT is used when it can be reliably predicted that deterioration would occur if treatment should cease, producing serious risks or seriously diminished capacity for self-care, which the individual, when competent, would wish to avoid.

IOT ensures continuing contact with the person so that negotiation about treatment can proceed. It helps provide the structure for community care and helps garner social support needed for community tenure. IOT can focus attention on individuals at special risk who are difficult to engage, and it permits earlier intervention for those with a history of rapid or dangerous relapses.

Best-practice components of IOT

Preferred legal regimen.

Clinicians favor a regimen that gives them discretion about whether to accept an individual for treatment under IOT and that authorizes effective monitoring of that person's condition. The legislation should confer sufficient authority to give clinicians confidence to use IOT (5). If the person is noncompliant with treatment or follow-up, swift rehospitalization for reassessment should be authorized through a less onerous process than initial commitment (5). Police assistance should be available with recall, and brief rehospitalization should not revoke the person's IOT order (5).

Clinicians should consult fully with family members, other caregivers, and community providers and share necessary information. No unreasonable clinical requirements or liability for malpractice should be imposed on clinicians for the individual's conduct in the community, and review mechanisms should not be so onerous as to discourage proper use of IOT (5).

There is a general consensus that clinicians should not be authorized to use physical restraint to force individuals to comply with treatment in the community under IOT (5). No Commonwealth legal system authorizes use of forcible treatment, except in emergencies. If forcible treatment is required, the individual must be returned to hospital.

Selection of patients.

Many indicators for use of IOT originally proposed by Geller (6) would find favor with Commonwealth clinicians. International data suggest that IOT is used primarily for individuals who have schizophrenia or affective psychoses and more rarely for other conditions (1). Some evidence suggests that its use is more effective for individuals with schizophrenia than for those with primarily affective disorders (2).

Churchill and colleagues' (1) review found that individuals on IOT are typically men around age 40, with a long history of mental illness and previous hospital admissions and with a schizophrenia-like illness or a serious affective illness; they are likely to display psychotic symptoms, especially delusions. Criminal offenses and violence are not dominant features.

Nondiagnostic factors are critical in determining suitability for IOT. Lack of insight often leads to treatment nonadherence, and IOT can provide supervision for persons who lack insight into their illness (7). Impaired insight is frequently encountered among persons with schizophrenia and often persists after other symptoms have resolved. In contrast, although insight may be profoundly impaired during acute episodes of affective illness, most patients regain insight in the natural course of this condition. This difference may influence the finding that IOT appears more effective for persons with nonaffective psychoses (2).

A history of noncompliance with treatment is likely to predict future noncompliance, and noncompliance is another indicator for the use of IOT (6). Noncompliance may lead to repeated hospital readmissions, and recent studies suggest that IOT reduces time in the hospital.

Some Canadian provinces require a minimum period of hospitalization before use of IOT is authorized. Many clinicians consider that such an approach overly restricts use of IOT in early illness and underestimates the ability of IOT to reduce the duration or severity of a person's condition. Many clinicians would prefer to rely on a wider range of factors that indicate likely compliance with arrangements for community care.

A history of dangerous behavior usually indicates greater need for supervision. Risk associated with treatment nonadherence may be an indicator for use of IOT, but Commonwealth clinicians would also endorse Geller's (6) guideline that “the outpatient must not be dangerous when complying with the ordered treatment.” They would be reluctant to maintain a person on IOT who was at imminent risk of causing serious harm, which is one reason why risk of serious deterioration (not imminent dangerousness) should govern its use.

Geller's (6) guidelines require that the ordered treatment “have demonstrated efficacy.” IOT is almost always initiated after an episode of inpatient care, which provides an opportunity to assess the individual's response to a given treatment.

Finally, the individual must accept at some level the validity of the treatment order. IOT is not suitable for persons who are completely determined to avoid contact with treatment services or who are so opposed to treatment that use of IOT is counterproductive.

Treatment plans.

Guidelines in Commonwealth countries require a comprehensive treatment plan that often specifies the proposed medication and other treatment for the primary condition; likely variations when the individual's condition changes; place of treatment; services responsible; crisis intervention plan; monitoring and supervision arrangements; and additional services to be offered, such as addressing other health problems. The plan may require the individual to live in a specified form of accommodation that provides certain levels of support.

Clinical appointments.

Clinicians should give the person on IOT as much choice as possible in determining the location and time of appointments. Appointments often occur at a clinic, but when treatment is compulsory it is unreasonable to expect the person to commute long distances and absorb travel costs. Other arrangements, including home visits, need to be considered, but they may place clinicians at risk and raise legal issues about entry and invasion of privacy.

Frequency of appointments should depend on clinical needs and on the person's wishes. A study of IOT in North Carolina (2) found that a high intensity of services, averaging more than seven contacts per month, was associated with better outcomes. All appointments should involve assessment, formal or informal, of the person's condition.

Treatment and monitoring.

Success is more likely when treatment under IOT can be monitored (6). Most individuals on IOT receive antipsychotic medication, which can be monitored either by observing the patient taking oral medication or by administering long-acting injections. The latter are recommended in England when compliance is vital for clinical reasons (8). Some studies that have compared treatment with oral antipsychotics and with long-acting injections have suggested that fewer persons relapse with the latter approach, perhaps because of unrecognized noncompliance with oral medication. In Commonwealth countries, long-acting injections are frequently used to aid treatment adherence for individuals on IOT, although research evidence is limited and mixed. Long-acting injections are often depicted as more intrusive than oral medication, but their perceived intrusiveness varies between individuals. Daily monitoring of oral medication may be experienced as more intrusive, and reduction of unwanted contact with the clinical team may lead some people to prefer injections.

Little is known about the outcome of mandated psychotherapies. It is difficult to oblige patients to cooperate with interventions that require significant motivation. Psychoeducation may provide benefit, but many clinicians favor mandating only treatment that works and can be provided with minimal coercion.

Substance use.

Substance use can exacerbate both psychotic and mood disorders. When substance use is a major issue, random blood or urine screening may be imposed as a condition of IOT. However, detectable quantities of most illicit substances clear from the body within 72 hours, and delay in providing a sample may frustrate efforts to accurately monitor use.

Some individuals are willing to maintain abstinence from street drugs to avoid hospital admission. However, in many IOT regimens, rehospitalization occurs only when there is a genuine clinical need for inpatient care. In general, IOT may be more effective in obliging a person to do something, such as taking medication or attending appointments, than preventing behavior such as substance use.

Residence.

When a specific type of accommodation, such as 24-hour supervised care, is essential, this may be specified in the IOT order if it represents the least restrictive option. In such cases, IOT can stabilize a chaotic situation, facilitate treatment adherence, and ensure professional monitoring. Residential staff are often in a position to detect deterioration, evidence of substance use, or treatment nonadherence.

It may be necessary to restrict the area in which a person chooses to live to maintain the therapeutic relationship. Living at a specified address may be directed when only one local facility provides the necessary level of care, but the quality of care should be carefully monitored. It is difficult to justify confining people in substandard facilities.

In federal countries, individuals on IOT may move between states. Legislation in Australia includes provisions for an interstate agreement for transfers between Victoria and New South Wales. Transfers between Scotland and England are also permitted by law.

The service system.

Geller's (6) guidelines require the outpatient system to be able to deliver the treatment and to be willing to enforce compliance; the inpatient system must also be willing to provide hospital care. He accurately envisioned the practice whereby inpatient clinicians would initiate IOT and then refer the patient to an outpatient program. The outpatient team, however, may assess the patient differently or oppose IOT. Reluctance by outpatient clinicians may readily subvert this form of care. Similarly, lack of consequences for breaching the conditions of IOT may promote noncompliance. One solution is for inpatient and outpatient services to work closely together, and the ideal solution is for the inpatient and outpatient team to be the same.

The police service must be familiar with the IOT legislation and “buy into” it. Police personnel may be required to return the person to a hospital, even over a long distance, which involves commitment of considerable resources. Negotiated agreements between health services and the police to clarify responsibilities can be useful. In England, the police lend assistance only to ambulance teams and then only when the patient is actively disturbed.

A sufficient range of programs must be available in the community to meet the individual's treatment needs, including case management, supported accommodation for people with challenging behaviors, and programs for substance use.

Families are often the greatest ally of people with mental illness. IOT legislation should facilitate communication between families and clinicians when it is in the patient's best interests. In some legal systems, which otherwise limit sharing of health information with relatives, the law authorizes free exchange between all signatories to the treatment plan.

Above all, IOT is unlikely to operate successfully without a commitment from mental health administrations and leadership within the psychiatric profession to provide the necessary resources and support the regimen.

Duration of IOT.

There is considerable variation among countries and states in the average duration of IOT (1). Many serious mental illnesses tend to be lifelong, and there is evidence that longer-term use of IOT is associated with better results (2). There is no definitive clinical test to determine the appropriate time for discharge. Clinicians will confront the paradox that the longer IOT extends without the patient's readmission, the more successful and perhaps the more unnecessary it is. Positive factors suggesting discharge include enhanced insight, ability to continue monitoring the patient's condition, and a clear plan to address future relapse.

Conclusions

Commonwealth clinicians view IOT as a tool to be used when individuals will predictably not accept treatment and are likely to deteriorate and experience adverse consequences as a result of this nonadherence. We describe an emerging consensus about the appropriate candidates and best practices for IOT.