Skip to content

Essay on mental health bed provision in York and England

October 17, 2013

This is an essay I wrote for the Health Informatics module of my Public Health MSc, about the provision of mental health beds. I’m posting it as follow-up to yesterday’s BBC Breakfast appearance, because I didn’t get the chance to go into things in as much geeky detail as I’d have liked to. Sorry it’s a bit jargon-heavy in places, it was written with professionals rather than the public in mind.

Key points: home crisis care services don’t replace the need for beds as is often claimed. Where beds are cut, more people are sectioned.



How do we know if York has enough Working Age Adult Acute Inpatient Mental Health beds to meet population needs?






Acute inpatient mental health services in York are run from Bootham Park hospital, amongst the last ‘County Lunatic Asylum’ to still be used for its original purpose (National Archives, 2012), covering a population of 285 000 across an area of 400 square miles, in York and rural North Yorkshire (Foundation School, 2012). The number of inpatient beds at Bootham has been in decline for many years, with cuts in 2011 including Ward 3 for acutely unwell patients (Catton, 2011) and the Mother and Baby unit from Ward 1 (BBC News, 2011). It is now a matter of some concern amongst service users and staff as to whether Bootham has sufficient beds to meet population needs, for example in May then Ward 1 (Women’s acute ward) was at capacity of 13 beds, with an additional 8 women placed out of area (Personal communication, 2012).




What does need mean in this context?


Need is defined in the context of a Health Needs Assessment as the capacity to benefit from healthcare interventions. It should be differentiated from demand, since many patients in acute MH wards will not ‘express a felt need’ to make use of services. This essay is not an attempt to conduct a needs assessment; rather, it is a summary and evaluation of what information is available, and would be necessary, to conduct such an assessment. Stevens et al (2012) describe three methods of assessing healthcare needs: Corporate, seeking views of stakeholders and key informants; Comparative, where service provision and use across different areas is compared; Epidemiological, which extrapolates need from characteristics of the population to be served, treatments available, and current service provision.




What is the ‘capacity to benefit’ from MH beds?


A full review of the uses of acute MH wards is beyond the scope of this essay, but if there are such wards, they should effectively provide tertiary prevention and treatment. The Royal College of Psychiatrists (2011) is clear that bed occupancy should be below 85% to allow patients to be safely treated in a local bed without undue delays, but half English wards have occupancy rates above 100%.




Do ‘alternatives to admission’ remove the need for beds?


Intensive community treatment has been suggested by the Trust as an alternative to hospital admission, fulfilling the same need as inpatient beds at lower cost. However, most studies on the ‘intensive home treatment / crisis team’ approach were conducted decades ago, when such approaches were first introduced across England, against a background of higher inpatient beds than now exist.




York currently has 29 acute beds for a population of 285 000 (10 beds / 100 000 population). Minghella et al (1998) reduced bed days by 48% through intensive home treatment, but retained 25.6 acute beds per 100 000 population. Bracken & Cohen (1999) found that home treatment resulted in a 25% drop in admissions, against a background of 84 beds to a population of 380 000 (22 beds / 100 000 population), but stated that ‘[residential] places of genuine asylum are still needed in a crisis’, with 11% of the sample originally offered home treatment needing hospital admission. Tomar et al (2003) found 46% of patients with first-episode psychosis assessed by a home treatment team nevertheless required admission. Gould et al (2006) found that despite well-established home treatment services, by three months 72% of patients assessed for home treatment had been admitted.




For schizophrenia and related ‘severe mental illness’, a Cochrane metanalysis of eight controlled trials with total 984 patients (Murphy et al, 2012) found that 45% of those initially allocated to home ‘crisis care’ had experienced admission, with no significant difference between groups who did and did not receive crisis care found in number of admissions, nor in use of the mental health act, nor in days in acute care, at six months. For borderline personality disorder, Borschmann et al (2012), in a Cochrane review, were unable to find any RCTs of the benefits of any form of crisis intervention other than hospital. Burns et al (2001), in an extensive systematic review, found that home treatment reduced days spent in hospital by only about five per patient per month, a result that did not reach statistical significance.




Intensive home treatment has been rolled out across the whole of England since about 2000, allowing Jacobs et al (2011) a large epidemiological study. Whilst admission rates across England fell with declining bed availability, no significant difference in admission rates was found between areas which had and had not yet implemented crisis teams. Service provision may be a more important determinant of admission than need.




Keown et al (2011) found a strong association between bed closure and increased use of the Mental Health Act, such that for every two beds closed, one more person in the next year would be involuntarily admitted. Whilst an epidemiological study cannot show causation, Keown et al (2011) found a clear dose-response curve, across all beds and admissions in England for the twenty years prior to 2008, despite many ‘alternatives to admission’.




Intensive home treatment can at most reduce frequency and duration of hospitalisation, but that it cannot eliminate the need for inpatient beds for some patients. In many of the above trials, the patients most likely to need hospital admission were those with less secure living conditions, or more severe illness. Patients with dual diagnosis, or a history of violence, were often excluded from home treatment. A reduction in bed numbers below safe levels may exacerbate these health inequities.




How do we know how many beds are needed?




Screening population directly for need


Ideally, mental health service provision would be determined directly by need. The Adult Psychiatric Morbidity Survey (2007) is a household-based survey which attempted to screen for the presence of diagnosed and undiagnosed mental illness, thereby establishing need as well as demand. A household survey is likely to underestimate rates of severe mental illness, as people may be homeless or living in an institution, and differential rates of responding by people with a mental illness may also be problematic. As a member of the public I only have access to the final report for all of England, so the implications for York of this potentially useful data source cannot be discussed further here.




The Health Survey for England (Department of Health, 2011), another household-based survey, uses the General Health Questionnaire (GHQ), a depression screening tool. The North East Strategic Health Authority faces particular public health challenges, with a significantly higher prevalence (17.5%) of possible mental illness than the English average (13.2%), and also significantly more people drinking above recommended limits (32.3 vs 26.8).




Epidemiological – inferring need from related measures


Mental health needs are strongly predicted by social determinants, particularly deprivation (Wilkinson et al, 2007). The Jarman index (Jarman, 1983) was originally developed as a measure of need for primary health care, is often referred to as an index of deprivation, and is calculated from UK Census data. The Townsend index is a measure of material deprivation also reliant on Census data. Dependency on the Census means the underlying data are often years out of date, and it may under-represent groups commonly missed by the Census, who may be disproportionately likely to have mental illness.




The York Psychiatric Needs Index (Smith et al, 1994) is used to allocate funding within the NHS, and attempts to predict psychiatric need based on deprivation, using mathematical techniques to simulate iso-supply. It therefore tends to be more redistributive than demand-led models.


Mental Illness Needs Index (MINI) (Glover et al 2004) is based on annually updateable data of measures taken from the Index of Multiple Deprivation. MINI seeks to predict psychiatric admission rates, which may be determined by supply or demand over need. However, the model did not account for considerable residual patterning by administrative area, emphasising that bed use may be influenced by health system factors other than need.




An approach comparing York to areas of iso-need has several difficulties. The ecological fallacy means that area-based measures do not necessarily provide good information about individuals within that area, so people with high unmet need could be ‘hidden’ within an area with low need indicators. Nationally there is a considerable underprovision of acute mental health beds, with occupancy rates of between 100 and 140% (Royal College of Psychiatrists, 2011), so equal provision for iso-need may still be considerable underprovision.


Community Mental Health Profiles


The Mental Health Observatory (2012) provides ‘Community Mental Health Profiles’, which bring together a range of information about risk factors, prevalence, and service availability in each local authority, from a variety of sources. The local authority for York does not map precisely onto the catchment area for Bootham hospital, but there is considerable overlap. See Figure 1 for the Community Mental Health Profile for York. The utility for each item as a predictor of mental illness prevalence is discussed in the profile itself, and extensively discussed by Wilkinson et al (2007).


The Community Mental Health Profile shows York to be significantly better than the English average for most wider determinants of mental health, using data mostly from accurate and frequently updated sources, such as Department for Education records for Not in Employment, Education or Training, or unemployment from claimant count. However, using the Index of Multiple Deprivation to predict mental illness is problematic because the ‘Health Deprivation and Disability Domain’ is constructed using measures of mental illness.


York is at or better than the English average for risk factors. Whilst the physical activity levels for children are based on a survey of PE time so should cover all children in school and be reasonably reliable, the adult physical activity measure is based on a questionnaire which will have been subject to considerable respondent bias. Homelessness is a particularly interesting variable, because it is so closely linked to mental illness. One in three British patients with schizophrenia has been homeless, and between a quarter to a third of street homeless people have a severe mental illness (Rees, 2009).


For levels of mental illness, York has significantly more people than the English average on the GP depression register, which could reflect prevalence or diagnosis. Dementia and learning disability are unlikely to have a major impact on working age adult acute beds. It is unfortunate that levels of severe mental illness, which must also be kept on a GP register under QOF, are not given here, since these people will be heavy users of inpatient care. The GP severe mental illness register gives a prevalence for North Yorkshire & York PCT of 0.7%, not significantly different from an English average of 0.8% (NHS Information Centre, 2012).


For treatment, York has an average spend per head not significantly different from the English mean, though it is in the lowest quintile. Significantly more people use secondary care, although significantly fewer are on Care Programme Approach, and contacts with a Community Psychiatric Nurse or other professionals are significantly less frequent, suggesting perhaps that resources are thinly spread over many patients. York has significantly more beds days spent inpatient per head of population than the England average, although it is still within the central quintile. This statistic is prone to distortion by the small number of patients who have very long hospital stays.


Overall, although York has several socio-demographic features which should be protective against mental illness, then it has more people with depression, more people in secondary care, fewer contacts with services which might keep people out of hospital, and more bed days than the English average. It could be that York health systems use a lower threshold for detecting and referring mental health problems, such that those referred require less treatment. Perhaps over-reliance on protective socio-demographic features is an example of the ecological fallacy, so that whilst most people in York enjoy good mental health, then a substantial minority have considerable mental health support needs. It does seem that York could be providing more intensive support for those referred to secondary care, but this should not be at the expense of inpatient beds.



The Mental Health Minimum Dataset


The Mental Health Minimum Dataset (MHMDS) (NHS Information Centre, 2012) is a record of services provided in secondary care. This data is derived from routine care records in a similar way to Hospital Episode Statistics, but includes contacts with community services. Independent sector data is not included, which may particularly distort outcomes in areas with high bed occupancy and consequent high use of out of area placement, and for specialist services which are often privately provided.


Table 1. Comparison of service use in York and elsewhere, from the MHMDS.


York England ‘Prospering Small Towns’ (iso-need)


Rate of people accessing secondary care, 3186 2789 n/a
per 100000 population



Percentage of people accessing secondary


care who are admitted to hospital 9.2% 8.1% 7.7%



Percentage in hospital detained under


Mental Health Act 28.4% 40.9% 35.1%



Percentage staying in hospital longer than 34.1% 46.5% n/a


30 days



All figures in Table 1 are 2010-11 data unless stated otherwise. ‘York’ data is for the former North Yorkshire & York PCT. Rates would be preferable to percentages, but I do not have access to raw data to calculate these.


York has consistently had fewer Mental Health Act (MHA) detentions than England or similar towns. However, the average length of stay in hospital in York is consistently longer. This could suggest that hospital beds might be better used by shortening stay.





Whilst ‘corporate’ concerns from staff and service users reflect too few beds for patients referred, comparative and epidemiological data is more difficult to interpret. High rates of diagnosis, and longer inpatient stay, may reflect high need, better access, or different referral patterns. To assess York’s mental health bed needs as one part of a Healthcare Needs Assessment, I would need fuller access to underlying databases to better establish comparison with areas of iso-need, and also the ability to investigate causes of variations from comparators.



Word count: 2190







Adult Psychiatric Morbidity Survey (2007). National Centre for Social Research: London. [Accessed 21st June 2012].




BBC News (2011) Postnatal mental health unit at Bootham Park Hospital in York to remain closed. [Accessed 21st July 2012]




Borschmann R, Henderson C, Hogg J, Phillips R, Moran P. Crisis interventions for people with borderline personality disorder. Cochrane Database of Systematic Reviews 2012, Issue 6. Art. No.: CD009353. DOI: 10.1002/14651858.CD009353.pub2




Bracken, P., Cohen, B. (1999) Home treatment in Bradford. Psychiatric Bulletin 1999, 23: 349-352.




Burns T, Knapp M, Catty J, Healey A, Henderson J, Watt H, et al. (2001) Home treatment for mental health problems: a systematic review. Health Technology Assessment 5(15): 1-139.




Carr-Hill, R., Hardman, G., Martin, S., (1994)A Formula for Distributing NHS Revenues Based on Small Area Use of Hospital Beds. York: University of York.




Catton (2011, August 21st). Wards may merge at York’s Bootham Park Hospital. York Press: York.




Foundation School (2012) Job Description. [Accessed 21st July 2012]




Glover G., Arts G., and Wooff D. (2004). A needs index for mental health care in England based on updatable data. Social Psychiatry and Psychiatric Epidemiology 39:730-738.




Gould M, Theodore K, Pilling S, Bebbington P, Hinton M, Johnson S. (2005) Initial treatment phase in early psychosis: can intensive home treatment prevent admission? Psychiatric Bulletin 30: 243-6.




Department of Health (2011) Health Survey for England [Accessed 21st June, 2012]




Jacobs R, Barrenho E. Impact of crisis resolution and home treatment teams on psychiatric admissions in England. British Journal of Psychiatry 2011;199:71-6




Jarman, B. (1983) Identification of underprivileged areas. BMJ, 286, 1705 -1709.




Keown, P., et al. (2011) ‘Association between provision of mental illness beds and rate of involuntary admissions in the NHS in England 1988-2008 : ecological study’ BMJ (Clinical Research Edition) 343 (0959-8138




Mental Health Observatory (2012) Community Mental Health Profiles [Accessed 19th June, 2012]




Minghella, E., Ford, R., Freeman, T., et al (1998) Open All Hours: 24 Hour Response for People with Mental Health Emergencies. London: Sainsbury Centre for Mental Health.




Murphy S, Irving CB, Adams CE, Driver R. Crisis intervention for people with severe mental illnesses. Cochrane Database of Systematic Reviews 2012, Issue 5. Art. No.: CD001087. DOI: 10.1002/14651858.CD001087.pub4




National Archives (2012) Bootham Park Hospital [Accessed 21st June 2012]




NHS Information Centre (2012) QOF prevalence data tables 2010/11, [Accessed 20th June 2012].




NHS Information Centre (2012) Mental Health Minimum Dataset, [Accessed 22nd June 2012].




Rees, S. (2009) Mental Ill Health in the Adult Single Homeless Population. London: Crisis.[Accessed 19th June 2012].




Royal College of Psychiatrists (2011) Fair Deal campaign: In-Patient
[Accessed 11
th June 2012].




Smith, P., Sheldon, TA., Martin, S. (1996) An Index of Need for Psychiatric Services based


on In-patient Utilisation. British Journal of Psychiatry, 169:308-316.




Stevens, A., Raferty, J., Mant, J. (2012) Implementing Joint Strategic Needs Assessment. [Accessed 25th June 2012].




Tomar, R., Brimblecome, N. & O’Sullivan, G. (2003) Service innovations: Home treatment for first-episode psychosis. Psychiatric Bulletin, 27, 148– 151.




Wilkinson, J., Bywaters, J., Chappel, D., Glover, G. (2007) Indications of Public Health In the English Regions 7 – Mental Health. North East Public Health Observatory: Newcastle­. [Accessed 17th June 2012].



From → Uncategorized

One Comment
  1. Philip Owen permalink

    Capacity needs to meet peak demand not average demand so other factors also need to be considered. Ordinary variations in a normal distribution in particular. If the average demand is 10 then 13 beds will be required to meet peak demand of one standard deviation. Some mental illness is seasonal so the number of beds required to avoid crisis is the number for the peak season with at least one standard deviation of spare capacity for that time of year. Alternatively the x beds per 100,000 is not the right approach when the result is so low as 10 beds because one patient is significant. For low bed counts, some super regional capacity needs to be provided determined on the statistical needs of a much larger population. Only when the bed count under consideration reaches over 200 can reasonable planning assumptions based on a normal distribution be made.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: