Abstract
Community Treatment Orders (CTOs), in use in New South Wales, Australia since 1991, are intended to reduce the rehospitalisation rate for patients who have an established history of exacerbation of symptoms due to non-compliance with treatment, particularly treatment with antipsychotic medications. It is hoped that CTOs will encourage patients to comply with the maintenance phase of treatment and so remain outside hospital long enough to experience the positive aspects of social and community life that stability can bring [1]. This is in keeping with a core principle underlying community treatment that court-ordered treatment is the least restrictive option in that it ultimately reduces the total time spent in hospital. Other potential benefits that could accrue from CTOs include fewer episodes of violence and self-harm, less social and occupational disability and less family burden. Despite these potential benefits, many psychiatrists in the USA regard CTOs as unconstitutional [2] and the use of CTOs are in conflict with the guidelines of the American Psychiatric Association [3]. In the UK the inability to allow the enforced delivery of medications to patients is regarded by many as rendering CTOs ineffective [4].
Swartz et al. [5] reported an unpublished recent outpatient commitment study in New York City that found no significant differences between groups under voluntary or court-ordered treatment in hospital readmissions or other outcomes. The best quality study in the area of CTOs comes from North Carolina [5] where a random-allocation, controlled trial of Community Treatment Orders failed to demonstrate an overall impact on readmission. A post-hoc analysis found that CTOs maintained for at least 180 days when combined with a higher frequency of outpatient services (> 6/month) were associated with a reduced readmission rate. In that study, CTOs could not enforce medications against the patient's will and no information was provided on medication compliance or types of medications prescribed. The lowered readmission rate associated with a high frequency of outpatient services and a period of 180 or more days on a CTO does not establish causality. Patients on longer orders are a selected group and a high level of patient-staff contact could independently predict a better outcome. However, the study raises the possibility that prolonged increased contact may prevent readmissions by inducing a more robust remission or by earlier intervention during exacerbations.
Studies such as that of Swartz et al. [5] emphasise that CTOs are best seen as complementing case management, making the patient more accessible to counselling, psychoeducation and rehabilitation. Increased responsibility is also borne by the community services to provide adequate treatment. To be effective, CTOs may rely on the adequacy of the community services and their capacity to deliver increased community care: this may be a crucial variable to address when evaluating the efficacy of CTOs.
Community Treatment Orders in New South Wales
Community Treatment Order legislation in New South Wales (NSW) is as favourable and easy to implement as in any developed society [6]. To obtain a CTO, mental health services are obliged to present a magistrate or a Mental Health Tribunal with a community management plan, which usually includes requirements to accept medication and to attend outpatient appointments and sometimes rehabilitation programs. Should patients later refuse to comply with the requirements, the legislation allows the Health Care Agency to enforce CTOs. If an oral and written warning to patients of the consequences of breaching the conditions of the CTOs fails to ensure compliance the patients may be brought into hospital, if necessary by the police. There, if they continue to refuse medication following counselling, they may be detained as involuntary inpatients and forcibly given medication. Maximum duration of CTOs was 3 months prior to October, 1997, and 6 months subsequently. Applications for further CTOs can be made to the Mental Health Tribunal prior to their expiry. Patients have the right to appeal under certain circumstances. The powers to detain the patient as an inpatient and to administer medication forcibly following breach of the order, are the two unusual features of the NSW Mental Health Act that distinguish it from the legislation in most countries [7]. The present study attempted to investigate whether CTOs in NSW reduce hospitalisation and/or influence the behaviour and medication compliance of patients prior to readmissions during or after their receiving CTOs.
Method
Community Treatment Order group
Hornsby Ku-Ring-Gai Hospital (HKH) Mental Health Service, in the northern suburbs of Sydney, with a catchment area of 270 000 people, has a commitment to case management and community treatment. Hospital records were examined to identify all patients with a diagnosis of schizophrenia, schizoaffective disorder, schizophreniform disorder and atypical psychosis who received CTOs between 1 July 1994 and 1 July 1998. Mean follow up was 27.7 months (range = 12–60 months). Medication and rehabilitative treatment plans were determined from the CTO treatment plans and outcomes were determined from rates of rehospitalisation; instances of breaches of CTOs or changes to medication were determined from patients' records.
Comparison group
Each patient placed on a CTO was matched with a patient discharged without a CTO who resembled the CTO patient in gender, age within 5 years, and number of previous hospital admissions and who was hospitalised as close as possible within 12 months of the admission in which the counterpart CTO was made (referred to subsequently as the index admission). Features of both groups are shown in Table 1. Most members of both groups were male, single, on a disability allowance and living with their parents.
Demographic and clinical details of 123 patients who had at least one Community Treatment Order (CTO) and matched group
Readmission to hospital
The duration and level of behavioural disturbance of patients readmitted during or after discontinuation of a CTO were determined from case records and mental health schedules and compared for the periods prior to the index admission, and to readmissions during and following the CTOs. Periods of their medication non-compliance were also assessed from inpatient and outpatient notes. Non-compliance was examined both for the whole group and separately for patients receiving depot injections.
Mental health services before admissions
A rater blind to the purpose of the enquiry counted the number of outpatient services to patients or their families from doctors and case managers for each of the CTO and comparison patients, for each of 3 consecutive months prior to the index admission and prior to readmissions during or following termination of a CTO.
Results
In the 4-year period there were 969 patients admitted with a diagnosis of schizophrenia, schizoaffective disorder or schizophreniform disorder, 533 (55%) of whom were admitted as involuntary patients under the Mental Health Act. After discharge 74% were case managed by the HKH Mental Health Service with the remainder either returning to, or referred to other agencies usually outside of the HKH catchment area. A total of 133 patients (14% of 969) received an initial CTO, of whom 39 received a second, and 13, a third. Ten of the 133 were referred for management to other agencies and were excluded from further analysis.
The present cohort is composed of the remaining 123 patients. The mean continuous CTO length was 288 days (SD = 210 days). The matching procedure resulted in two groups equivalent in age, gender, marital status, living circumstances, employment, diagnosis and number of previous admissions. However, the index admission of the 123 CTO patients was more frequently involuntary (χ2 = 66.74, p < 0.001) and longer (t = 7.73, p < 0.001).
Readmission in the Community Treatment Order and comparison group
Timing of readmissions to hospital of patients who received CTOs and comparison group is reported in Table 2.
Timing of readmission in the Community Treatment Order (CTO) and comparison groups after CTO is made
Fifty-nine patients (48%) were readmitted in the CTO group and 45 (37%) in the comparison group (χ2 = 3.26, p = 0.07). Almost half of the readmissions in both the CTO and comparison groups occurred in the first 3 months (see Table 2). Thirty-six (61%) readmissions in the CTO group were involuntary compared with 15 (33%) involuntary admissions in the comparison group (χ2 = 7.8, p = 0.005).
The length of the initial readmission in both groups was similar: a mean of 18.4 days (SD = 17.9) in the CTO group and 19.04 days (SD = 20.9) in the comparison group. Total length of hospitalisation of the CTO patients in the one year prior to and following the making of the CTO was compared in two ways. The number of days the patients spent in hospital during the year that followed the date the CTO was issued and the patient discharged (mean = 12.6 days, SD = 27.9) was significantly less (t = 8.7, p < 0.001) than the number of days (mean = 44.5 days, SD = 29.6) they spent in hospital during the year prior to that date. The latter number of days included the duration of the index admission. When the length of the index admission was excluded, by determining the mean number of days of hospitalisation in the year immediately prior to that admission, the number was substantially reduced (mean = 9.2, SD = 20.5), but did not differ significantly (t = 1.22, p = 0.23) from the number of days they were hospitalised in the year that followed the awarding of the CTO.
Readmission during and after discontinuation of Community Treatment Orders
In the CTO group 38 patients were readmitted during the CTO and 21 of the remaining 85 were readmitted after its elective discontinuation. No admissions occurred during the first 3 months after the discontinuation and only two within 3–6 months. Ten (12%) occurred within 6–12 months of discontinuation. Five (6%) occurred within the second year and 4 (4%) were after 2 years. Perhaps the low readmission rate (two patients) in the first 6 months after elective discontinuation of CTOs implies that some benefits of the CTO are slow to accrue, but may persist beyond the length of the CTO. After the 6-month period following discontinuation, the readmission rate climbed again.
Readmission and medications
At the start of their period on a CTO, 76 patients were prescribed depot neuroleptics and 47 oral medications (see Table 3). During the course of the CTO 18 (24%) of those receiving depot medications were subsequently readmitted to hospital compared with 20 (43%) of the 47 patients who were prescribed oral medications (χ2 = 4.84, p = 0.03). There was no significant difference in the readmission rate of those receiving typical as compared with atypical oral antipsychotics: 7 (50%) of the former and 13 (39%) of the latter.
Readmission rate in relation to medication prescribed for the 123 patients during the Community Treatment Order (CTO) or after termination of CTO
Within the comparison group, only 13 patients (11%) were prescribed depot medications of whom six (46%) were readmitted. This was a similar readmission rate to that of patients given oral medications (39 admissions, 35%). The trend for a higher readmission rate with depot medications in the comparison group compared with the CTO group failed to reach statistical significance (χ2 = 2.84, p = 0.09).
Readmission and medications after discontinuation of Community Treatment Orders
Of the 58 patients on depot medications at termination of the CTO, seven patients moved out of area, 12 patients were transferred to atypical antipsychotic medications, and two patients suicided. Of the 37 remaining on depot medication, 14 (38%) were readmitted to hospital. Seven of these readmissions occurred within 8 weeks of abrupt cessation of medications against medical advice. Of the 27 patients on oral medications not readmitted during the CTO, seven (26%) were readmitted following it. Although 13 (39%) of the 33 patients receiving atypical antipsychotics were readmitted while on a CTO, only four (20%) of the remaining 20 were readmitted after discontinuation of the CTO. This could indicate a better longer-term outcome with these medications once treatment becomes established. Readmission occurred only in one of eight patients taking clozapine after their CTOs were terminated and it followed immediately after the clozapine was discontinued because of neutropenia.
Adherence to Community Treatment Orders
Compliance with depot medications while patients were on CTOs was high. Of the 76 patients prescribed depot medication, only four were taken off depot medication. All four were re-admitted. In two patients, discontinuation of depot medications was medically supervised; one was changed to clozapine to combat negative symptoms, and haloperidol decanoate was stopped for another patient who developed thrombocytopenia. Two further patients moved out of area to avoid the CTO. Fifty-eight patients were still on depot medications at the termination of their CTO. Compliance with oral medications could not be accurately determined from the case notes. When patients' symptoms increased, clinicians tended to query compliance, but patients usually reported they were continuing to take their medication.
Community Treatment Order breaches
Orders were breached for three patients during the course of the CTO. All three were on a depot medication and although their condition had not deteriorated, they were brought to the hospital on one occasion, where they were given a depot injection, but were not re-admitted.
Level of disturbed behaviour before and during hospitalisations
The level of disturbed behaviour before hospitalisations was difficult to reliably quantify from case notes, but may be reflected in the need for involuntary admission. The readmissions of the 38 patients while they were on CTOs compared with their index admissions were more likely to be voluntary (19 vs 0, χ2 = 22.9, p < 0.001), and showed a trend to require less police involvement (10% vs 29%, NS). They were also more likely to be of shorter duration (means of 19.4 days compared with 34.1 days, t = 2.78, p = 0.007).
Length of noncompliance and disturbed behaviour associated with the Community Treatment Order
The mean number of days of behavioural disturbance and of non-compliance with medication before the index admission and the readmissions during and following the CTO are reported in Table 4.
Length of medication non-compliance (MNC) and behavioural disturbance (BD) of readmitted patients on Community Treatment Orders (CTO) prior to index admissions (pre-CTO), readmissions during the period of the CTO (CTO) and readmissions after discontinuation of the CTO (post-CTO)
Prior to the index admission, non-compliance with medication significantly preceded the onset of disturbed behaviour. The duration of both non-compliance and disturbed behaviour occurred for significantly longer periods prior to those admissions than prior to the readmissions occurring during the CTOs. In re-admissions occurring during the CTO, non-compliance and disturbed behaviour commenced around the same time. Prior to the readmissions after discontinuation of the CTO, the period of non-compliance and the period of disturbed behaviour increased, with the onset of non-compliance again significantly preceding that of disturbed behaviour.
Because the assessment of non-compliance using oral medications depends greatly on patient and family reports and so could be invalid, the same data as reported in Table 4 for all 59 readmitted patients were investigated for the 32 patients readmitted while receiving depot medication. In patients admitted during the CTO, behaviour disturbance preceded the date where the depot injection was due. They showed a mean of 12.8 (SD = 16.6) days behaviour disturbance before admission compared with 6.3 (SD = 12.6) days non-compliance. Duration of their medication non-compliance was much greater before the CTO (mean = 117.7, SD = 125.9 days, p < 0.001), as was duration of behaviour disturbance (mean = 38.2, SD = 47.4 days, p < 0.01). After discontinuation of the CTO the duration of their non-compliance again significantly increased (mean = 58.9, SD = 50.3 days, p < 0.01) and the duration of their behaviour disturbance showed a trend in the same direction (mean = 25, SD = 27.6 days, NS).
Frequency of consultations prior to hospitalisations prior to, during and following Community Treatment Orders
Table 5 shows that there was a statistically significant increased mean number of services in the 2 months before hospitalisations during CTOs as compared with those before or after CTOs. The CTO group showed a significant increase in the number of consultations received for each month from the third to the first month before readmission. Trends for an increase in number of consultations in the 3 months before the index admission and that following a CTO did not reach statistical significance.
Frequency of services (mean) during the 3 months preceding hosptialisations, before, during and after Community Treatment Orders (CTOs)
Frequency of outpatient services in Community Treatment Order group compared with comparison group
There was a trend for CTO patients to receive a lower number of consultations in the 3 months before their index admission compared with the number received by the comparison group in the equivalent period (mean = 2.1, SD = 3.3 vs mean = 3.9, SD = 6.89 services monthly, t = 1.7, p = 0.08). The number of services received by patients monthly for the 3 months prior to hospitalisations while on CTOs were non-significantly greater than those received by the comparison group prior to the equivalent hospitalisations (mean = 5.6, SD = 4.4 vs mean = 3.9, SD = 6.8, t = 1.77, NS). The level of service when readmission was not imminent was assessed for the first 3 months of CTOs compared with the corresponding period for patients not on a CTO. After patients who were readmitted within the first 6 months were excluded, the mean number of consultations received monthly by the remaining patients in the first 3 months of CTOs was 4.8. The mean number received monthly by the comparison group in the 3 months following their discharge was 5.2 (NS).
Discussion
Community Treatment Orders were issued for 14% of patients hospitalised with a schizophrenia-related diagnosis. They were used mostly for male, unmarried patients. Of 123 patients given CTOs in the present study, 31% were readmitted while the CTOs were active and a further 17% following their termination. Community Treatment Orders were maintained on average for just over 9 months. During the CTO very high compliance was found with depot medications. Formal breaching was seldom necessary suggesting that the threat of breaching was sufficient to promote compliance in most cases. Depot medications conferred significant advantage in terms of rehospitalisation compared with oral medications during the CTO. However, during the post-CTO period this advantage was lost due to a lower rate of relapse in the oral medication group.
Rehospitalisations occurring during a CTO were shorter, with less police involvement and involuntary admission than the index hospitalisations, suggesting the patients were admitted at an earlier stage of relapse when they were more amenable to treatment. This conclusion is supported by the most significant finding of the study that duration of non-compliance and disturbed behaviour was reduced in the period prior to hospitalisations during CTOs in the comparable period prior to the index admissions. After termination of CTOs admissions once again reverted to the pre-CTO pattern of a longer period of non-compliance and disturbed behaviour prior to hospitalisation. In the 2 months prior to hospitalisation, the number of psychiatric services used increased to a significant extent for patients when on CTOs compared with prior to or after CTOs. It would seem likely that CTOs enable closer monitoring of patients and establishment of more clinical contact as symptoms of relapse became apparent. In the patients on CTOs receiving depot medication who were readmitted, disturbed behaviour preceded medication non-compliance, indicating the latter was not a factor in relapse.
Although the comparison group was well matched to the CTO group in terms of demographic variables, they differed in regard to important clinical variables. Their illness was presumably less severe, in view of the shorter duration of their index admissions. The much higher percentage having voluntary status at that admission, consistent with the decision not to apply for a CTO for their management after discharge, indicated that they appeared more insightful and accepting of a therapeutic alliance.
There was a trend for the CTO group compared with the comparison group to receive less care in the 3 months prior to the index admission, but not in the 3 months prior to the following readmission when the former group were on CTOs. This difference possibly resulted from the non-compliance of the CTO group prior to receiving CTOs. Interestingly, of patients not at imminent risk of relapse, those on CTOs received slightly less services than the comparison group, at least in the first 3 months after discharge from the index admission.
The 6-month readmission rate of 27% in the group given CTOs was comparable to the 23% in the comparison group and also to the readmission rate of 28% in 71 consecutive patients with chronic schizophrenia recently reported by Bergin et al. [8]. However, 48% of patients were hospitalised at least once during or following CTOs as compared with 37% of the comparison patients. It is possible a randomised control group not given a CTO would have a significantly higher rate of rehospitalisation than the lessseverely ill comparison group. The procedure [9–11] of evaluating CTOs by comparing the period patients spent in hospital following a CTO with the equivalent period immediately prior to receiving a CTO appears questionable methodologically. A long period of hos-pitalisation is a likely factor determining the seeking of a CTO. In the present study the mean period spent in hospital by patients who subsequently were given a CTO was nearly three times longer than that of the comparison group and twice as long as the mean period of their readmissions following the CTO. The longer hospitalisation indicates a more severe phase of illness, so that some regression towards the mean in the severity of a subsequent episode is possible. In the present study when the duration of the index admission was included in the duration of hospitalisation in the year prior to the CTO, the total duration was significantly greater than the duration of hospitalisation in the year following the CTO. When it was excluded, there was no significant difference.
Community Treatment Orders in this study appear associated with possibly contrary effects: hospitali-sations appear to have been reduced by the CTO ensuring medication delivery particularly of depot medications, but may have been brought forward or even increased by earlier intervention during exacerbations in disturbed behaviour. A great advantage of CTOs is their ability to reduce the period of the patient's disturbed behaviour as it is not necessary to wait until the patient is sufficiently ill to justify involuntary admission. Shorter periods of disturbance before readmission may provide benefit by reducing the associated level of psychosocial havoc including damage to key relationships, demonstrated to be a major determinant of expressed emotion [12], an established risk factor for future relapse.
Increased delivery of prophylactic medications may not be the only mechanism that could allow CTOs to reduce rehospitalisations. Remissions may be made more robust by encouraging patients to accept social skills training, psycho-education, family therapy, personal therapy, assertive training, stress management, or improved problem solving and other interventions that enhance recovery.
It could be argued that this investigation simply demonstrates the benefits associated with assertive case management. But this study suggests that before CTOs, patients reduced the frequency of, or withdrew from, seeing their case managers and doctors, so reducing the efficacy of assertive case management. When patients refused to see mental health workers, they were obliged to withdraw until the level of the patients' disturbance resulted in a risk or harm to themselves or others such that they required involuntary admission under the Mental Health Act.
This study has all the limitations of a retrospective analysis from case notes and replication in a prospective study would seem of value. The influence of CTOs on rate of rehospitalisation could not be determined due to the lack of comparability of the comparison group. It would seem this determination would require a prospective study using a random allocation design.
Footnotes
Acknowledgements
We are grateful to Tom Burns for his very helpful comments on an earlier draft of this paper and for the statistical advice from Kar Kiat Yeo.