Increased Use of Antipsychotic Long-Acting Injections with Community Treatment Orders

Abstract

Background:

Community treatment orders (CTOs) are increasingly being used, despite a weak evidence base, and problems continue regarding Second Opinion Appointed Doctor (SOAD) certification of medication.

Aims:

The aim of the current study was to describe current CTO usage regarding patient characteristics, prescribed medication and CTO conditions.

Method:

A 1-year prospective cohort study with consecutive sampling was conducted for all patients whose CTO was registered in a large mental health trust. Only the first CTO for each patient was included. Measures included sociodemographic variables, psychiatric diagnosis, CTO date of initiation and conditions, psychotropic medication and date of SOAD certification for medication. This study was conducted in the first year of CTO legislation in England and Wales.

Results:

A total of195 patients were sampled (mean age 40.6 years, 65% male, 52% black ethnic origin). There was significant geographical variability in rates of CTO use (χ² = 11.3, p = 0.012). A total of 53% had their place of residence specified as a condition and 29% were required to allow access into their homes. Of those with schizophrenia, 64% were prescribed an antipsychotic long-acting injection (LAI). Of the total group, 7% received high-dose antipsychotics, 10% were prescribed two antipsychotics and only 15% received SOAD certification in time.

Conclusions:

There was geographical and ethnic variation in CTO use but higher rates of hospital detention in minority ethnic groups may be contributory. Most patients were prescribed antipsychotic LAIs and CTO conditions may not follow the least restrictive principle.

Introduction

Community treatment orders

Since November 2008, supervised community treatment orders (CTOs) have been used to legally require patients to adhere to treatment in the community in England and Wales [Department of Health, 2007]. CTOs can only be applied after an initial period of detention in hospital but afford compulsory treatment in a less restrictive environment than in-patient care. Requirements or ‘conditions’ of the CTO may focus on aspects of treatment and risk management, including restrictions regarding place of residence. A further statuary condition is that the patient must meet with a Second Opinion Appointed Doctor (SOAD) for authorization of medication treatment within a given timeframe (usually 1 month). If the person fails to comply with the conditions of the CTO, they can be ‘recalled’ to hospital for up to 72 hours. The ability to recall a patient to hospital is generally seen to be the primary power of a CTO. If the patient continues to refuse treatment the CTO is ‘revoked’ and the patient is then detained in hospital under a hospital treatment order once more. Although the UK government anticipated that 450 people in England and Wales would be subject to a CTO in the first year of legislation, the actual figure was approximately 4000 people, and this resulted in a shortage of SOADs [Care Quality Commission, 2010; Gould, 2009]. The background, legislation, and clinical application of the CTO in England and Wales has been described elsewhere [Brookes and Brindle, 2010; Macpherson et al. 2010; Taylor, 2010b].

Efficacy and ethnicity

The high uptake of CTOs occurred despite the fact that neither of the two randomized controlled trials (RCTs) of CTOs in North America provided definitive evidence of efficacy [Steadman et al. 2001; Swartz et al. 1999] and both have methodological concerns [Swartz and Swanson, 2004; O'Reilly and Bishop, 2001; Szmukler and Hotopf, 2001]. A Cochrane review estimated that 85 people would need to receive a CTO in order to avoid one admission and that 238 people would need to receive a CTO in order to avoid one arrest [Kisely et al. 2005]. A third RCT is currently being conducted in England [Burns et al. 2008] and, hopefully, this will provide useful information relevant to the UK setting. This is pertinent as international generalizability for the efficacy of CTO is complicated by the widely differing clinical criteria and legal jurisdiction. Generalization of findings from international RCTs is complicated by the substantial differences in the structure and function of community services [Lambert et al. 2009]. Further, in the UK, there are concerns regarding institutional racism in psychiatry [McKenzie and Bhui, 2007] and patients of black ethnic origin are reported to be subject to more frequent re-admissions [Burnett et al. 1999] with a higher rate of compulsory detention [Morgan et al. 2005; Bhui et al. 2003]. Although differences in illness course and duration, acuity of presentation, insight, attitudes to treatment and family and social support [Fearon et al. 2006; Morgan et al. 2006] may be explanatory factors, it remains to be seen whether this difference by ethnic group holds true for CTO use.

Medication

In practice, probably the most common condition of a CTO is to compel the patient to continue with medication for symptom control and relapse prevention, particularly when insight into the need for continuing treatment is suboptimal. In Australia, a strong link has been identified in the use of CTOs combined with antipsychotic long-acting injections (LAIs), which are used to overcome covert medication nonadherence [Lambert et al. 2009; Patel et al. 2009; Patel and David, 2005]. In the USA, LAIs are more commonly used for patients of black than white ethnic origin [Shi et al. 2007; Valenstein et al. 2001] but this pattern is less evident in the UK [Connolly and David, 2008; Lloyd and Moodley, 1992; Chen et al. 1991]. Our objective was to quantify how CTOs were used in the first year of legislation, with particular regard to medication prescribing, in a prospectively ascertained cohort. The aims were: (i) to identify patient characteristics, including ethnicity, for those commenced on a CTO; (ii) to identify the nature of psychotropic medication prescribing at CTO initiation; (iii) to measure time taken for SOAD certification. It was hypothesized that patients with schizophrenia would have higher than average rates of LAI use as compared with national prescribing data [Barnes et al. 2009] but that there would be no ethnic bias.

Method

Design

Baseline data are presented for a 1-year prospective observational cohort study with systematic inclusion sampling of all patients initiated on a CTO within a large mental health trust in the first year (3 November 2008 to 31 October 2009) of CTO legislation in England and Wales.

Setting

This study was conducted in the South London and Maudsley (SLAM) NHS Foundation Trust in London, UK. The Trust serves a local population of 1.1 million and the proportion of people of black ethnic origin (the largest ethnic minority group) for the four local boroughs is as follows: Lambeth (25.8%), Southwark (25.9%), Lewisham (23.5%) and Croydon (13.3%) [Office for National Statistics, 2001]. During the first 6 months of the sampling period there were 667 acute inpatient beds, for which the average length of inpatient stay was 53 days and the ‘trimmed’ (for admissions of duration 3–90 days) mean length of stay was 28 days. Acute inpatient bed occupancy levels were 97% including patients on leave and 83% excluding patients on leave. Acute inpatient care is led by inpatient consultants and not community consultants. The Trust also includes specialist tertiary referral inpatient facilities (e.g. national psychosis unit) and forensic wards. This study was approved as a Trust-wide audit by the SLAM clinical audit and effectiveness committee. Ethical approval was not required and this was confirmed by the local research ethics committee.

Participants

Systematic consecutive sampling was conducted for all patients commenced on a CTO between 3 November 2008 (CTO legislation introduction in England and Wales) and 31 October 2009 whose CTO was registered at one of the five Mental Health Act offices within the Trust. Only the first CTO for each patient was included. There were no exclusion criteria.

Variables

A pro forma for data extraction at baseline was designed to enhance reliability and included the following variables.

Sample characteristics.

Sociodemographic variables included gender, date of birth, marital status, employment status, and ethnicity which was categorized using standard format from census data [Office for National Statistics, 2001]. Primary psychiatric diagnosis at CTO initiation was recorded as documented by clinicians (ICD-10) [World Health Organization, 1992].

Mental Health Act status.

Date of CTO initiation, reasons for CTO (protection of patient's own safety, health or others), preceding/parent section (sections 3, 37 or 25a) and CTO specified conditions were noted.

Medication.

Psychotropic medication prescribed at the time of CTO initiation (drug name and dose), history of previous clozapine (ever) use, and history of previous antipsychotic LAI (ever) use were recorded. SOAD assessment for use of psychotropic medication including date of SOAD (CTO11) form within first 6 months, medication approved and reasons for lack of SOAD form were also noted.

Data extraction

The primary sources of data included the Mental Health Act office registers, source legal CTO documents, prescription charts and electronic patient records including medication details, contemporary clinical notes and summaries. A baseline data pro forma was completed by clinical research staff.

Analysis

Statistical analyses with SPSS software included baseline descriptive analyses. Age at CTO initiation was calculated and categorized according to ‘Count me in’ census categories [Care Quality Commission, 2009]. Conditions stated on CTO were coded according to commonly occurring themes (as outlined in Table 3). Current antipsychotic medication was categorized according to class and formulation and antipsychotic polypharmacy was noted. Doses for antipsychotics were converted into percentage of maximum dose licensed according the British National Formulary (%BNF) [Royal Pharmaceutical Society of Great Britain, 2008]. Completion of SOAD certification was categorized according to time of completion and reason for lack of completion within 6 months (as outlined in Table 3).

Table 3.

CTO statutory reasons, conditions and medication.

	n	%
STATUTORY REASONS FOR CTO
Protection of patient's own safety	173	88.7
Protection of health	194	99.5
Protection of others	143	73.3
CONDITIONS
Attendance at appointments/assessments/reviews	178	91.3
Medication adherence	177	90.8
Residential address specified	103	52.8
Access to residence for clinical staff or associated carers	57	29.2
Urine drug screen sample	24	12.3
Blood samples for medication plasma levels or clozapine monitoring	16	8.2
Both urine drug screen and blood sample	5	2.6
MEDICATION AT CTO INITIATION
Antipsychotics	193	99.0
Mood stabilisers	54	27.7
Antidepressants	15	7.7
Benzodiazepines	9	4.6
Drugs for physical comorbidities	82	42.1
ANTIPSYCHOTICS AT CTO INITIATION
Second-generation antipsychotic oral	43	22.1
First-generation antipsychotic oral	3	1.6
Second-generation antipsychotic LAI	33	16.9
First-generation antipsychotic LAI	72	36.9
Clozapine	23	11.8
Polyantipsychotic use	19	9.7
No antipsychotics	2	1.0
COMPLETION OF SOAD CERTIFICATION
Completed within 6 months
Completed in time	29	14.9
– Time requirement of 1 month	(22)	(11.3)
– Time requirement of more than 1 month*	(7)	(3.6)
Completed late	107	54.9
– Time requirement of 1 month	(98)	(50.3)
– Time requirement of more than 1 month*	(9)	(4.6)
Not completed within 6 months
Event occurring within 6 months†	38	19.5
6 months period ended without SOAD	21	10.7

^*If patients are still subject to the 3-month rule of a section 3, then the SOAD review will take place before the end of that 3-month period (if that expires later than the CTO 1-month rule).

^†Events includes CTO revoked, CTO rescinded, patient transferred out of area.

CTO, community treatment order; LAI, long-acting injection; SOAD, Second Opinion Appointed Doctor.

Results

Sample characteristics

There were 195 patients initiated on a CTO in the first year of legislation: 65% were male, 52% were of black ethnic origin, 5.6% were legally married, and 3.1% were employed or were students (see Table 1). The mean age at CTO initiation was 40.6 years (SD 14.1, range 17.0–89.1 years). The most common diagnosis was schizophrenia (70.8%). Five potential cases were lost as the patients were entered into the nationwide RCT and randomized to the non-CTO arm of the trial (four cases were randomized to the CTO arm of the trial and are included here) [OCTET team, personal communication].

Table 1.

Sample characteristics.

	N	%
Gender
Male	127	65.1
Female	68	34.9
Age
≤24.99 years	25	12.8
25–49.99 years	134	68.7
50 + years	36	18.5
Ethnicity
White	68	34.9
Asian	13	6.6
Black	102	52.3
Mixed	7	3.6
Other	5	2.6
Marital status
Legally married	11	5.6
Not legally married	184	94.4
Employment
Employed/student	6	3.1
Not employed/student	189	96.9
Clinical service
Secondary care for four local boroughs	165	84.6
Forensic services	26	13.3
Specialist services	4	2.1
Main Diagnosis
Schizophrenia	138	70.8
Schizoaffective disorder	32	16.4
Bipolar affective disorder	18	9.2
Other psychotic disorder	4	2.1
Not psychotic disorder	3	1.5
Prior antipsychotic history
Clozapine naïve	144	73.8
LAI naïve	41	21.0

LAI, long-acting injection.

CTO use and ethnicity

For the 195 patients, 30 (15.4%) CTOs were conversions from former supervised discharge orders (section 25a), and 7 (3.6%) patients were previously on a criminal court appointed treatment order (section 37). CTOs were started more frequently in the first phase of the study (first cohort quartile (Q1): 64 (32.8%, including 10 section 25a conversions); Q2: 62 (31.8%, including 20 section 25a conversions); Q3: 39 (20.0%); Q4: 30 (15.4%)). Variation in CTO use was identified for secondary care patients across the four local boroughs in the Trust (N = 165, χ² = 11.3, df = 3, p = 0.012; see Table 2). For each individual borough, the proportion of patients of black ethnic origin in the sample was significantly greater than expected, based on population data [Office for National Statistics, 2001]. However, based on Trust Mental Health Act data (April 2007-March 2008) for section 3 (6-month hospital treatment order) patients, the proportion of patients of black ethnic origin in our sample was not significantly greater than expected (observed 52.3%, expected 50.2%, N = 195, χ² = 0.4, df = 1, p = 0.556).

Table 2.

Variation in CTO use by London borough and ethnicity.

	CTOs				Ethnicity
	Observed		Population*		Observed		Expected†		Statistic
Borough	N	%	N	%	N	%	N	%
Lambeth	31	18.8	266169	24.4
Black					22	71.0	8.0	25.8	χ2 = 33.0
Other					9	29.0	23.0	74.2	p< 0.001
Southwark	25	15.1	244866	22.5
Black					17	68.0	6.5	25.9	χ2 = 23.1
Other					8	32.0	18.5	74.1	p< 0.001
Lewisham	47	28.5	248922	22.8
Black					23	48.9	11.0	23.4	χ2 = 17.1
Other					24	51.1	36.0	76.6	p< 0.001
Croydon	62	37.6	330587	30.3
Black					20	32.3	8.2	13.3	χ2 = 19.3
Other					42	67.7	53.8	86.7	p< 0.001

^*Expected population % values calculated using population census data and rounded to one decimal point.

^†Expected N values calculated using population census % data and rounded to one decimal point.

[Office for National Statistics, 2001].

CTO, community treatment order.

Statutory reasons and conditions

Statutory reasons for CTO initiation and commonly stated conditions of CTOs are detailed in Table 3. Rarer conditions included a requirement to meet with a housing officer, eat lunch twice a week at an eating disorders clinic, attend drug counselling/rehabilitation/occupational therapy, and taking medication in front of staff.

Medication

At time of CTO initiation, 193 patients (99%) were prescribed an antipsychotic, and first-generation (typical) antipsychotic LAIs were most commonly used (see Table 3). Regarding antipsychotic doses, the mean %BNF was 61.6% (SD 37.1, range 2.5–183.3%). Of the total sample 7.2% had antipsychotic (combined) doses exceeding 100% BNF limits and 9.7% were prescribed two antipsychotics. The most commonly prescribed LAIs were risperidone LAI, pipothiazine palmitate and flupenthixol decanoate. Of those with a diagnosis of schizophrenia, 88/138 (63.8%) were prescribed an LAI.

SOAD involvement

Only 136 (69.7%) patients received SOAD certification within 6 months (see Table 3). Completion rates of SOAD certification within the required time-frame (usually 1 month) did not improve over the study duration: Q1, 16/64 (25.0%); Q2, 9/62 (14.5%); Q3, 3/39 (7.7%); Q4, 1/30 (3.3%); total, 29/195 (14.9%). The mean time from CTO onset to SOAD certification was 66.6 days (SD 40.8, range 1–175 days, N = 136); for those with a standard 1 month time requirement the mean was 67.4 days (SD 41.4, range 1–175 days, N = 120).

Discussion

Strengths, weaknesses and principal findings

This is the largest reported study on CTOs in England and Wales to date and comprised all 195 patients commenced on a CTO within the first year of legislation in a large mental health trust. A key strength of this study is that it used a systematic sampling strategy which included all patients, thus avoiding sampling bias. Weaknesses of this study include lack of detail regarding education, length of contact with psychiatric services, number of prior psychiatric hospital admissions (compulsory or voluntary) and duration of index admission (leading to CTO) and lack of a matched comparison group. We did not aim to definitively address the question of whether or not CTOs are beneficial or efficacious, which requires an RCT [Burns and Dawson, 2009; Churchill et al. 2007]; rather, we aimed to investigate how CTOs are being used with particular regard to medication use. Key findings included the considerable variability in CTO use across the four boroughs which the Trust serves. Over half of the patients were of black ethnic origin which is more than twice that suggested by the population census data for the four boroughs (13.3–25.9%) [Office for National Statistics, 2001]. Further, common CTO conditions included clinical assessment, medication adherence, specified place of residence and access to residence. Investigation of psychotropic medication revealed that 99% were prescribed an antipsychotic and 63.8% of those with schizophrenia on a CTO were prescribed an LAI. The majority was clozapine naïve and this was higher than anticipated but possibly reflects poor adherence by this patient population obviating the use of clozapine due to the requirement for weekly blood tests. A clinically important minority was prescribed two antipsychotics and 7.2% had (combined) antipsychotic doses exceeding 100%BNF dose limits. Only 14.9% of patients had timely medication SOAD certification.

CTO use and ethnicity

Reasons for the geographical variation in CTO use might reflect varying attitudes and beliefs of clinical staff regarding CTOs, perhaps stemming from the lack of definitive evidence of efficacy of CTOs, and lack of belief that the individual patient will comply with treatment despite the legal sanction. This may be further exacerbated by differences between inpatient and community consultant psychiatrists for the same patient and also influenced by additional services including home treatment and assertive outreach teams. Also, use of CTOs for patients of black ethnic origin appears to be more than twice that suggested by the population census data [Office for National Statistics, 2001] for the locality served by the Trust. However, this can probably be largely explained by rates of hospital detention for ethnic minorities [Eaton, 2010; Audini and Lelliott, 2002]. For this Trust, 43% of patients on acute inpatient wards were of black ethnic origin using ‘Count me in’ census data [Care Quality Commission, 2009] for the Trust, 50.2% of all patients detained with a section 3 hospital order were of black ethnic origin using Trust Mental Health Act data (April 2007-March 2008) and local antipsychotic prescribing data for inpatient wards showed almost identical proportions of ethnic diversity [Connolly and Taylor, 2008]. Hence, there does not appear to be any ethnic bias in the application of CTOs over and above the factors leading to ethnic differences in the current use of the Mental Health Act for hospital detention orders as shown by our nonsignificant finding (black ethnic origin: CTO, 52.3%; section 3, 50.2%). However, as with the early report on CTO use in Birmingham and Solihull [Evans et al. 2010], we lack the data necessary to demonstrate this using statistical modelling or more rigorously still, by comparing groups of differing ethnicities matched for illness severity and course. Future studies should quantify rates of CTO renewal, revocation, voluntary hospital admissions and with regard to differences by ethnic group.

Conditions

Conditions should only be applied to a CTO which are necessary for enabling treatment or for safety [Department of Health, 2008] and should be practical and enforceable. If the condition cannot be enforced or monitored then, critically, should the patient be at risk of being recalled to hospital for breach of that specific condition? Recall should be invoked only if it is deemed a proportionate action to the identified level of risk associated with the breach of the condition [Department of Health, 2008]. The Code of Practice highlights that conditions might include ‘where the patient is to live, and avoidance of known risk factors or high-risk situations relevant to the patient's mental disorder’ [Department of Health, 2008]. We wonder how many patients will be recalled to hospital specifically because a condition regarding place of residence (applicable to over half of our sample) has not been upheld, or a urine drug screen sample has not been consistently provided. Further, it is stated that conditions should only minimally restrict the patient's liberty while being consistent with achieving their purpose [Department of Health, 2008]. Requiring access to a patient's home is intrusive and may not be consistent with the ‘least restrictive principle’ governing compulsory treatment, as is suggested by the use of this condition for almost a third of our sample. Further monitoring of conditions, particularly regarding place of residence, seems appropriate and should include subsequent alterations to the conditions (after CTO initiation) for which there are no current additional checks in place. Also, the process by which a responsible clinician and approved mental health practitioner formulate the conditions to be imposed is in need of more consideration and structure by policymakers. This should also anticipate how these conditions might be challenged by the patient, potentially at the point of CTO renewal if not earlier.

Medication

Our finding of 88/138 (63.8%) of those with schizophrenia on a CTO being prescribed an LAI is double that reported by Barnes and colleagues who highlighted that in the UK, LAI use for patients with schizophrenia is 35% on acute wards, 36% for assertive outreach team cases, 28% on forensic wards [Barnes et al. 2009]. In 2002 in the state of Victoria, Australia, just under half of those with schizophrenia on CTOs were prescribed an LAI [Lambert et al. 2009]. Thus far, there has been no definitive evidence to suggest the use of a CTO with an LAI is more beneficial, in terms of long-term outcomes, than either an LAI alone or a CTO with an oral antipsychotic. Moreover the alleged superior long-term benefits in relapse prevention of LAIs over oral antipsychotics lack definitive evidence [Leucht et al. 2011; Rosenheck et al. 2011; Patel et al. 2009] although a large cohort study favoured depot antipsychotics over their equivalent oral preparations [Tiihonen et al. 2011]. That said, the CTO may grant access to treating the patient with an LAI and also provide sufficient time to establish a regular injection routine. In turn this allows patients a better chance of establishing some control over their illness and, with it, a measure of improved insight [Lambert et al. 2009]. This study also found that high-dose (>100%BNF) antipsychotic prescribing and polyantipsychotic use is evident for patients on CTOs, although to a lesser extent than that reported for other UK samples [Tungaraza et al. 2010; Paton et al. 2008], despite guidelines to the contrary [Langan and Shajahan, 2010; Taylor, 2010a; National Institute for Clinical Excellence, 2009]. Compelling any patient to have potentially dangerous or even fatal high doses of medication is extremely problematic. Indeed, a study of 93,300 patients treated with antipsychotics and 186,600 matched controls found that high doses (>300mg chlorpromazine equivalent) of first- and second-generation antipsychotics were associated with increased rates of sudden cardiac death [Ray et al. 2009]. Thus, whilst it is probable that the requirement for a SOAD to sanction treatment plans has already reduced the potential number on a CTO who would otherwise have been prescribed high doses, more needs to be done regarding prescribing monitoring for this patient group.

SOAD certification

SOAD certification involves an important safeguard for CTOs but this is subject to delay in our sample as highlighted by the low rate of SOAD certification within the specified time-frame (14.9%). This is probably in keeping with national trends and is undoubtedly due to higher than anticipated rates of CTO use across England and Wales. This may be further compounded by some patients failing to attend at the scheduled time but as this was not consistently documented, we were unable to accurately take this into account for this study. Owing to the debated legality of community treatment orders without completed SOAD certification, a temporary measure of using emergency treatment orders (section 64) was instigated but this has proved unpopular and is arguably also unlawful in this specific context. Until the shortage of SOADs is resolved, we are likely to continue facing the problem that many patients are being required to take medication that has not been approved nor had full legal process. As the system further adjusts to cope with the new Mental Health Act, it is hoped that several of these early problems for CTOs can be resolved.

Contributions

MXP designed the study, conducted the analyses, drafted the article and takes responsibility for the integrity of the data and the accuracy of the data analysis. JM, MKB, JG and KB contributed to the data collection, analyses, and interpretation. JB, FH, DT, GS, TL and ASD contributed to the data analysis and interpretation. All authors were involved in the drafting of this article and approved the final published version.