The development of crisis resolution and home treatment (CRHT) teams has been a central tenet of English mental health policy since 2000. ¹ Although the policy implementation guidance ² was highly prescriptive in relation to staffing and operation of CRHT teams, there were relatively few such teams in operation in the UK at that time. Since then there has been unprecedented investment and growth in CRHT services, with the National Audit Office reporting a 400% increase in spend on CRHT between 2002 and 2007, with 343 teams in operation across England in 2006/2007. ³ These new teams have been intensively performance managed, with each expected to provide a threshold number of episodes of care as part of the local primary care trust annual targets.

Although there is good research evidence that well-resourced teams can reduce length of stay and hospital admission rates, ^{Reference Johnson, Nolan, Hoult, White, Bebbington and Sandor4,Reference Glover, Arts and Babu5} concern has also been expressed that the purely numerical activity targets can result in a shift away for the care of the most severely ill in order to maximise throughput. ³ The lack of medical involvement in services has also been cause for concern, with the National Audit Office ³ reporting that almost a third of teams have no dedicated psychiatrist and more than 50% have less than a half-time equivalent.

Manchester was one of the first centres to develop an acute home treatment service ^{Reference Harrison, Poynton, Marshall, Gater and Creed6} and repeated activity audits during the past 10 years have provided a unique opportunity to explore in more detail the impact of national targets. The home treatment service in central Manchester was established in 1997, with a focus on providing a 24-h alternative to in-patient care for those with acute or severe mental illness. ^{Reference Harrison, Poynton, Marshall, Gater and Creed6} The staffing levels and method of operation were very similar to those subsequently recommended by the Department of Health. ² The service has previously demonstrated an ability to care for people with acute symptoms comparable to those seen in in-patients but at a lower overall cost. ^{Reference Harrison, Marshall, Marshall, Marshall and Creed7}

Until 2008, the home treatment service only accepted referrals from within secondary care, with urgent primary care referrals triaged by other community mental health services. Data on referral source, patient demographics and diagnosis were collected over 6-month periods in 1999 ^{Reference Harrison, Allam and Marshall8} and 2005. From inception to 2008, all individuals taken on by the service were reviewed by a psychiatrist during their admission, and the majority underwent physical examination during their contact with the service. To meet local activity targets, referral routes into the service were extended to include primary care from early 2008. This was accompanied by a small increase in staffing and much greater managerial focus on throughput.

Method

A single data collection sheet was used to prospectively record information on all referrals to the service in a 6-month period from November 2008 to May 2009. Referral sources were grouped into five categories: accident and emergency (A&E), out-patients, community mental health teams, in-patient service and primary care. For individuals accepted by the service, an ICD-10 diagnosis ⁹ was taken from the clinical record. This was later grouped into five categories: schizophrenia and related disorders, bipolar affective disorder, psychotic depression, less severe depressive categories and other diagnoses. Individuals were also grouped into those already under secondary care, with an established care plan, and those not previously known to services. The data items and codings replicated those used in earlier service evaluations. The data were entered into SPSS (version 15 for Windows) for analysis and compared with the data-set from 2005. Diagnostic categories from the original 1999 sample were also included to allow a longer-term perspective.

Results

In total, 301 people were treated by the CRHT service during the 6-month period in 2008/2009, compared with 128 over a similar period in 2005. The mean duration of contact with the service fell from 69 to 24 days. Compared with 2005, a higher proportion of people were referred from primary care and A&E, and a lower proportion from all secondary care sources (Table 1).

Table 1 Source of referrals for individuals accepted by the crisis resolution and home treatment service over 6-month period

	2005	2008/2009
	n (%)
Accident and emergency	10 (8)	71 (24)
Community mental health teams	42 (33)	97 (32)
Out-patient department	44 (34)	16 (5)
In-patient unit	32 (25)	58 (19)
Primary care	n/a	59 (20)
Total	128	301

Only 39% of individuals in the more recent sample were already known to services and in receipt of a care programme approach (CPA) care package, compared with 70% in 2005 (P<0.005). This is reflected in the diagnostic groups, with a significant increase in the proportion of individuals treated for less severe illnesses (including mild to moderate depression, anxiety, adjustment reactions, personality disorders in 2008/2009 and substance misuse) and a reduction in the proportion of individuals with schizophrenia and related disorders compared with both 1998 and 2005 (Table 2). The proportion of people treated with severe mental illness (schizophrenia and related disorders, bipolar disorder and psychotic depression) fell from 75% in 2005 to 50% in 2008/2009, whereas the proportion with mild to moderate illness (less severe depression and other diagnoses) rose from 25 to 50% (P<0.001).

Table 2 Diagnostic groups for individuals accepted by the crisis resolution and home treatment service

	1998	2005	2008/2009
	n (%)
Schizophrenia and related disorders	47 (47)	63 (49)	84 (28)
Bipolar affective disorder	13 (13)	24 (19)	44 (15)
Psychotic depression	10 (10)	9 (7)	22 (7)
Less severe depression	24 (24)	28 (22)	116 (39)
Other diagnosesa	7 (7)	4 (3)	35 (11)
Total	101	128	301

Discussion

The original central Manchester home treatment service was one of only a handful of such teams in the UK. The primary aim of the service was to look after people with acute and severe mental illness at home as an alternative to admission. It was therefore agreed from the outset that the service would only take referrals from within secondary care, as would be the case for individuals requiring admission.

As far as possible the service mirrored the functioning of an in-patient unit: all service users had a full psychiatric history and regular medical reviews, most had a physical examination and investigations and psychotropic prescribing was taken over by the home treatment service using in-patient type prescription cards. Individuals were with the service long enough for a full needs assessment and could be offered short-term psychosocial interventions. Discharges were planned around CPA reviews and the team doctors provided discharge summaries.

The service had changed dramatically by 2008, with throughput more than doubling and 20% of referrals coming directly from primary care. The average duration of contact with the service also fell significantly. There have been benefits, including reduced pressure on community mental health teams, more rapid access to help for individuals not under care, and more individuals receiving care from the service. However, this has been accompanied by a shift in emphasis away from those with established mental illness and towards primary care referrals, with new challenges resulting. The team doctors are not able to see all patients, hence fewer have full histories, physical examination or investigations and prescribing arrangements are less clear. Although data on contact with doctors were not collected as part of this audit, a separate data collection has indicated that the proportion of individuals being reviewed by a psychiatrist has fallen to around 60%.

As individuals are with the service for such a short time, the main interventions for those with less severe illness are assessment, support and signposting to other services. The treatment paradigm for this work is unclear and there is little research evidence to assess its effectiveness.