Psychiatric in-patient hospital activity in England: Time series analysis

 

 

Details of all authors and affiliations:

 

Dr Inderpreet Singh (Corresponding Author): Speciality Registrar, Greater Manchester West NHS Foundation Trust, UK

MBBS, MRCPsych, DMH

I_sohi@yahoo.com

 

Ms Bindweep Kaur: Research Analyst, National Institute for Health and Care Excellence, UK

BHMS, DHM, MPH

 

  Please consider our petition

 

Summary

Aims and Method: To identify the current trend of psychiatric hospital admissions and produce forecasts to estimate trends till 2015. Data for psychiatric in-patient hospital activity in England was obtained from Hospital Episode Statistics and the Information Centre. Forecasts were drawn by using the holt-series method of forecasting through R-software, to obtain estimated figures till 2015.

Results: The bed-days for General adult and Old age psychiatry are likely to remain similar; however, the bed-days for Learning Disability and Forensic Psychiatry are expected to increase substantially by 507% and 254% respectively. The bed-days for Child and adolescent psychiatry have increased by 59% since 1998 and are likely to increase further till 2015.

The number of bed-days for schizophrenia and related disorders have increased by 52% since 1998 and are likely to increase further by 27% till 2015 and bed-days for learning disability have doubled in the last decade and are likely to increase by a further 273% in the next 5 years. The detentions under section-136 have increased dramatically by 157% in England since 2005 and are likely to increase further by 242% till 2015.

Clinical Implications: Despite continuous efforts, number of bed days for specific specialities and psychiatric conditions are increasing and more needs to be done to provide efficient care. Clinicians need to be aware of the changing trends to be able to adapt their practices for the Quality, Innovation, Productivity and Prevention (QIPP) agenda.

 

 

 

 

Introduction

 

Mental health is the largest single cause of disability and represents around one fourth (23%) of the total burden of ill health in the UK (HM Government, 2011, WHO, 2008).  England’s secondary care services spend nearly 11% of their annual budget on mental health (DH, 2009). It has been suggested that the cost of treating mental health problems could double over the next 2 decades (Mccrone et al, 2008). Psychiatric secondary care clinical services including secure and high dependency service and psychiatric intensive care units cost 37% of the total reported investment in direct services in 2010/11 (DH, 2011). Since 2002/03, the cost of secure and high dependency services and clinical services have increased by 156% and 19% respectively and currently cost £1.9 billion (DH, 2011). Gap in the actual and expected spent on inpatient services by Mental Health trusts in England in 2009/20 varied widely, ranging from overspending £20,318,820 to under spend of £6,684,292 (DH, 2011).

 

The Department of Health suggests improving the acute care pathway as one of the means of driving up quality, efficiency and productivity and have estimated potential national gross savings of around £224 million annually by 2014/15 (DH, 2011). It suggests minimising hospital admissions through effective community care and more efficient inpatient care that avoids unnecessary long stays (HM Government 2011, DH 2011). Crisis Resolution (Baker V et al, 2011), Home treatment teams (HTT (Tan E et al, 2012), assertive outreach team (Mortimer et al 2012) and early intervention teams (Dodgson et al, 2008) play a major role in preventing hospital admissions.

 

In light of Department of Health’s suggestions, principles of the quality, innovation, productivity and prevention (QIPP) initiative and the recent focus on community oriented services, this project aims to identify the current trend of mental health hospital admissions and produce forecasts to estimate the trend till 2015.

 

 

Methodology (including data collection and analysis):

 

Data for in-patient hospital activity including total number of admissions, emergency admissions, full consultant episodes (FCEs) and bed days for patients with a primary diagnosis of mental illness in England was identified from Hospital Episode Statistics (HES) online. In-patient activity data for all the sub-specialities of psychiatry i.e. general adult, childhood and adolescent, forensic, learning disability, psychotherapy and old age was also obtained from HES. Data for in-patients formally detained in hospitals under the Mental Health Act 1983 in England was obtained from the Information Centre. All the data obtained from HES was available from 1998/99 to 2010/2011 and the data from Information centre was available from 2005 to 2010. Firstly, the number of admissions and FCEs were converted into rates by utilising the population data available on Office of National Statistics.

 

Secondly, the data was forecasted by using the Holt series method of forecasting by utilising the R software, to obtain estimated figures for 2015. Exponential smoothing refers to a method of forecasting which assigns exponentially increasing weights to more recent observations. In other words, this method of forecasting assigns less weight to older observations compared to recent observations. It is this aspect of not assigning similar weights to all observations that makes it a robust method of forecasting compared to moving averages. Holt’s method of double exponential smoothing accounts for any trend in the data whereas Holt-Winter method or triple exponential smoothing method accounts for trend and seasonality or seasonal patterns. For this study we utilised Holt’s double exponential smoothing method as our data showed linear trend and no seasonality. There are 2 equations associated with double exponential smoothing method (Holt’s method) (Prajakta SK, 2004):

 

• F_t = α*O_t + (1-α) (F_t-1+β_t-1)
• Β_t = γ* (F_t-F_t-1) + (1-γ)*β_t-1

 

Where:

• Ot is the observed value at time t
• Ft is the forecast at time t
• Βt is the estimated slope at time t
• α is the first smoothing constant, used to smooth the observations
• γ is the second smoothing constant used to smooth the trend

 

Thirdly, %change was calculated by using the following formula ((variable₂₀₁₀ – variable₁₉₉₈)/variable₁₉₉₈)*100

 

Results

 

The results produced by R software were compiled into tables and graphs for ease of interpretation.

 

 

Estimated change in admissions from 2010-2015 by mental health speciality

 

The rate of FCEs per 10,000 population and emergency admissions per 100,000 population for learning disability (LD), general adult psychiatry (GAP) and Old age psychiatry (OAP) consistently decreased from 1998 to 2010 and is expected to continue decreasing till 2015 for learning disability and old age psychiatry. However, the rate of FCEs per 10,000 population for general adult psychiatry is expected to increase by 18% (from 37 per 10,000 pop to 44 per 10,000 pop) from 2010 to 2015, although the rate of emergency admissions is likely to fall. The rate of FCEs per 10,000 population for child and adolescent (CAP) and Forensic psychiatry (FP) have consistently increased from 1998 to 2010 by 81% and 60% respectively and are expected to further increase by 13% and 109% till 2015. The rate of emergency admission per 100,000 population for child and adolescent psychiatry has increased by 86% respectively from 1998-2010 and it is expected to further increase dramatically by 321% over the next 5 years. The rate of emergency admission for forensic psychiatry has decreased from 1998 by 39% and is not likely to vary significantly. Although we have witnessed a decrease in the rate of FCEs for LD, GAP and OAP, the number of bed days have significantly increased for LD (148%) from 1998-2010 and likely to further increase by 507% till 2015, whereas the bed days for GAP and OAP are not likely to change much. Seemingly consistent with the increase in rate of FCEs, the number of bed days for CAP and FP have increased by 59% and 414% respectively from 1998-2010 and are likely to increase further by 16% and 254% respectively will 2015. (Table 1)

 

 

 

		Learning Disability	General Adult	Child and Adolescent	Forensic	Psychotherapy	Old Age
Rate of FCEs per 10,000 population	1998	7.928825127	53.78803609	1.74739451	0.34268	0.011060905	66.82917392
	2010	3.56	37.4	3.16	0.55	0.02	34.16
	2015	2.285859823	43.95453583	3.584127505	1.14909	0.02600747	29.03580329
	%change (1998-2010)	-55%	-30%	81%	60%	81%	-49%
	%change (2010-2015)	-36%	18%	13%	109%	30%	-15%
Rate of Emergency admissions per 100,000 population	1998	9.221517146	368.6628845	5.705777992	1.25767	0.002048316	286.1159476
	2010	2.87	187.3	10.6	0.77	0.052	150.73
	2015	1.354323432	136.4706312	44.67680531	0.80104	0.341542906	106.3526165
	%change (1998-2010)	-69%	-49%	86%	-39%	2439%	-47%
	%change (2010-2015)	-53%	-27%	321%	4%	557%	-29%
Number of bed days	1998	583877	5310720	112619	217001	6293	2079125
	2010	1449118	5444949	179031	1114803	3881	2246098
	2015	8795406	5793432	208042	3949923	520	2242822
	%change (1998-2010)	148%	3%	59%	414%	-38%	8%
	%change (2010-2015)	507%	6%	16%	254%	-87%	0%

Table 1: Forecasts estimating the number of admissions and bed days from 1998 to 2015 by mental health sub-specialities in England

 

 

Estimated change in admissions from 2010-2015 by mental health ICD codes

 

The rate of FCEs per 10,000 population have consistently increased for mental health disorders due to organic causes (F04-F09), mental and behavioural disorders due to psychoactive substance use (F10-F19) by 105% and 57% respectively from 1998 to 2010 and are likely to further increase by 124% and 19% respectively till 2015. Similarly, the rate of emergency admissions per 100,000 have consistently increased by 59% for mental health disorders due to organic causes (F04-F09) and by 4% for mental and behavioural disorders due to psychoactive substance use (F10-F19). The rate of FCEs per 10,000 population, and rate of emergency admissions per 100,000 population have remained fairly constant from 1998 to 2010 for neurotic, stress related and somatoform disorders, behavioural disorders associated with physiological disturbances and physical factors and disorders of adult personality and behaviour (F40-69); however they are likely to increase by 71%, and 42% respectively in the next 5 years. Similarly, rate of FCEs and emergency admissions for disorders of psychological development (F80-F89) has only slightly decreased or increased from 1998 to 2010 but is likely to increase by 18% and 12% over the next 5 years. The rate of FCEs and emergency admissions for schizophrenia, schizotypal and delusional disorders (F20-F29), mood [affective] disorders (F30-39) and learning disability (F70-F79) has decreased from 1998 to 2010 and is likely to decrease further till 2015. Consistent with the increase in rate of FCEs, the number of bed days for mental health disorders due to organic causes (F04-F09), neurotic, stress related and somatoform disorders, behavioural disorders associated with physiological disturbances and physical factors and disorders of adult personality and behaviour (F40-F69) are likely to increase further till 2015. The number of bed days has increased for disorders of psychological development (F80-F89) have increased from 1998 to 2010 by 71%, however they are likely to decrease by 16% till 2015 despite the increase in rate of admission. Although the rate of FCEs has decreased for schizophrenia and related disorders and for learning disability, the number of bed days has been increasing since 1998 and are likely to increase further till 2015. Also, despite the increase in FCEs and admissions for mental and behavioural disorders due to psychoactive substance (F10-F19), the number of bed days has decreased from 1998 to 2010 and are likely to further decrease by 9%. (Table 2)

 

 

		F00-F03	F04-F09	F10-F19	F20-F29	F30-F39	F40-F69	F70-F79	F80-F99
Rate of FCEs per 10,000 population	1998	6.52	1.034	8.690	7.889	11.752	6.842	4.643	2.414
	2010	4.52	2.12	13.65	7.51	7.85	6.78	1.03	2.1
	2015	4.832327	4.753271	16.24245	7.8596012	6.687355	11.57982	0.471334	2.47673
	%change (1998-2010)	-31%	105%	57%	-5%	-33%	-1%	-78%	-13%
	%change (2010-2015)	7%	124%	19%	5%	-15%	71%	-54%	18%
Rate of Emergency admissions per 100,000 population	1998	30.69	6.58	60.28	53.82	81.99	42.52	5.29	9.33
	2010	18.93	10.45	88.7	35.48	45.26	40.16	1.03	10.08
	2015	19.181993	19.15242	92.57527	28.452364	34.128625	57.15015	0.4075358	11.27334
	%change (1998-2010)	-38%	59%	47%	-34%	-45%	-6%	-81%	8%
	%change (2010-2015)	1%	83%	4%	-20%	-25%	42%	-60%	12%
Number of bed days	1998	1,206,255	116,540	437,541	2,050,881	1,958,744	589,553	356,914	313,017
	2010	861723	172305	385709	3113738	1397600	698875	725747	535886
	2015	704296	195314	350038	3968164	1133450	716210	2705818	448153
	%change (1998-2010)	-29%	48%	-12%	52%	-29%	19%	103%	71%
	%change (2010-2015)	-18%	13%	-9%	27%	-19%	2%	273%	-16%

Table 2: Forecasts estimating the number of admissions and bed days from 1998 to 2015 by ICD codes of mental health illness in England

 

Estimated change in formal admissions under Section 2, 3 and 4 from 2005 to 2015 in England

 

Number of formal admissions under the Mental Health Act has increased by 7% from 2005 to 2010 and is likely to increase further by 17% till 2015. Of these, formal admissions under section 2 have increased by 26% whereas formal admissions under section 3 have decreased by 17%. Same trend is likely to follow till 2015, with formal admissions under section 2 increasing by further 53% and formal admissions under section 3 decreasing by further 45%. (Figure 1)

 

Estimated change in place of safety detentions in England from 2005 to 2015

 

The place of safety detentions under section 136 have increased dramatically by 157% in England since 2005 and are likely to increase further by 242% till 2015. The place of safety detentions under section 135 have decreased since 2005 but are likely to remain quite similar till 2015. (Figure 1)

 

 

Figure 1: Trend of Number of formal admissions and place of safety detentions in England 2005-2015

 

 

Discussion and Clinical Implications

 

The rate of FCEs for Learning disability, General adult and Old age psychiatry have consistently decreased and are likely to decrease further till 2015. However, the rate of FCEs for Child and Adolescent and Forensic psychiatry has increased in the last year and are likely to increase further. The bed days for GAP and OAP is likely to remain similar in the next 5 years, nevertheless, the bed days for LD and FP are expected to increase substantially by 507% and 254% respectively. The bed days for CAP have increased by 59% since 1998 and are likely to increase further till 2015.

 

The decrease in FCEs and relative stability in bed days for GAP and OAP could be attributed to range of interventions implemented over years including effective utilisation of various community teams including crisis resolution, home treatment and early intervention teams, new ways of working as well as increased emphasis given to this group of patients in the last decade. However, the projected increase in FCEs for CAP may reflect increased awareness among professions, paediatricians and GPs. The increase in bed days for FP and LD needs careful explorations as these are generally expensive and long term placements; so may have greater financial impact.

 

The rate of FCEs have decreased for schizophrenia and related disorders (F20-F29) however the number of bed days for schizophrenia and related disorders have increased by 52% since 1998 and are likely to increase further by 27% till 2015. It represents a changing trend of prolonged admission which needs to be elaborated further with regards to cost to the NHS. A plausible explanation of prolonged admissions or increase in the bed days for mental health illnesses such as schizophrenia can be attributed to the practice of admitting only seriously ill patients. In recent times, crisis resolution teams and community mental health teams are being increasingly involved in managing patients with mild to moderate mental health conditions, who pose no risk to self and others. As a means to reduce the number of admissions, only severely ill patients who cannot be managed in the community are admitted for treatment. It can be hypothesised that these patients take longer to recover and therefore the number of bed days for these mental health conditions is increasing.

 

The rate of FCEs for learning disability (F70-79) have decreased since 1998, however the bed days have doubled in the last decade and are likely to increase further by 273% in the next 5 years. There is a likelihood that patients suffering from severe learning disability are difficult to manage in the community and take longer to recover, thereby increasing the number of bed days for this speciality of psychiatry. Perhaps, there is a need to explore options of learning disability care homes where these patients can be managed effectively and consequently be less reliant on an NHS bed.

 

This study has also established other important findings which need further study to explore reasons and identify practical solutions. These include

 

• Increase in bed days for other organic, including symptomatic mental disorders (F04-F09) and Neurotic, stress related and somatoform disorders (F40-F69)
• Increase in rate of admissions for mental and behavioural disorders due to psychoactive substance use (F10-F19) although bed days are decreasing.
• Increase in formal admissions under section 2  and a decrease in formal admissions under section 3

 

There has also been a massive increase in the place of safety detentions under section 136 since 2005 and these are expected to dramatically increase further by 242% by 2015. These detentions take up a lot of resources, including healthcare, social care as well as the judicial system. Identifying reasons for the increase in section 136 detentions is outside the scope of this paper; however it is a very important question which needs to be answered sooner rather than later.

 

Comprehensive utilisation of early intervention teams, crisis resolution teams, assertive outreach teams and home treatment teams is an effective step in this direction. This study adds to the evidence base in favour of community teams as it can be deduced from the results that, perhaps, the decrease in FCEs in GAP and OAP and relative stability in bed days for these psychiatric specialities can be attributed to the community teams. Is it time that we aim to utilise these teams more effectively? Should we be investing more in these community teams? According to the national survey of investment in adult mental health services, the real term investment in community oriented teams has effectively remained similar or has decreased in the last year. The evidence regarding the impact of various community teams, including crisis resolution, home treatment, assertive outreach and early intervention teams is inconsistent. Various studies have concluded a positive impact of community teams in reducing hospital admissions (Baker V 2011, Tan E 2012, Mortimer 2012, Dodgson 2008) whereas others have found no evidence of the efficacy of community teams (Jacobs and Barrenho, 2011). In addition to the community teams, targeted in-patient services for specific disorders which utilise massive resources such as personality disorders are also effective in reducing bed days (Jones B et al, 2012). There is also a need to explore other alternatives to in-patient admissions such as voluntary sector-run crisis houses and investment in peer support workers (HM Government 2011, Howard LM et al 2012) as well as timely assessment as diversion to relevant services.

 

Moreover, there has been a significant decrease in the number of NHS provided mental health beds since 1997(Green B, 2009). Since 2010/11 quarter 1 to 2012/13 quarter 3, the number of beds available for mental health patients in England has decreased from 23,515 to 22,496 (DH, 2013). However, it remains to be seen whether the decrease on availability of beds has had any impact on bed days. Despite the decrease in bed numbers, the bed days for conditions such as schizophrenia have increased whereas bed days for some mental health conditions such as affective disorders (F10-19) and mental and behavioural disorders due to psychoactive substances (F30-39) have decreased. The number of beds available for mental health patients seems to have a very complex relationship with the number of bed days.

 

Minimising hospital admissions and avoiding long stays is a means of driving up quality of care of patients while maintaining high productivity and this research provides us an insight of current practice. Sharing best practice, process mapping and lean management can be useful techniques to achieve this (Isles and Sutherland, 2001). Other alternatives can be reducing delayed discharges through integration of housing pathways, effective use of supported housing and better access to social housing. Clinicians need to be aware of the changing trends to plan ahead particularly when we are going through an aggressive efficiency saving programme of deficit reductions.

 

Further exploration of the findings of this study would require collaborative working with various professionals including social care and police for determining the rationale of the increase in place of safety detentions, exploring the options of reducing delayed discharges as well as establishing the patient’s satisfaction with services. In addition, further research is required to ascertain the impact of various re organisations of community teams on the continuity of care as well as staff and patient morale to ensure any changes do not impact the best interests of our patients.

 

Limitations of the study: Although the projections were drawn by utilising the holt-series method of forecasting, which is a robust process of estimating time trends, the projections can only be as good as the data they rely on. Therefore caution must be exercised in basing service delivery solely on these predictions but this piece of work should be seen as an eye opener so that further exploration can be undertaken to understand the full extent of changes in mental health admission landscape. Moreover, this analysis has been undertaken with national level data and therefore we must use these results with caution for local hospital providers and commissioners. Also, this study only looks at the inpatient admissions and bed days in NHS hospitals in England. It does not include figures for the independent sector, which has recently expanded significantly as a provider of psychiatry beds.

 

Acknowledgement: We would like to thank Dr David Hughes and Dr Gerard McDade for their input.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

References

• HM Government (2011) No health without mental health: a cross government mental health outcomes strategy for people of all ages. Department of Health
• World Health Organisation (2008) The Global Burden of disease. WHO available at www.who.int/healthinfo/global_burden_disease
• Department of Health (2009) The Health and Personal Social Services Programmes, DH available at www.official-documents.gov.uk/document/cm75/7593/7593.pdf
• Mccrone P, dhanasiri s, Patel A et al (2008) Paying the Price: The cost of mental health care in England. London: King’s Fund, 220–226.
• Department of Health (2011) Mental Health Strategies 2010/11 National survey of investment in adult mental health services, Department of health
• Department of Health (2011) NHS Reference Costs, DH available at http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_123459
• Department of Health (2011) No health without mental health: a cross government mental health outcomes strategy for people of all ages Supporting document – the economic case for improving efficiency and quality in mental health, DH
• Baker V. et al (2011) Impact of Crisis resolution and home treatment services on user experience and admission to psychiatric hospital. The Psychiatrist. 35:106-110
• Tan E et al (2012) Crisis plans in a home treatment team before and after a quality improvement programme. The Psychiatrist, 36: 331-334
• Mortimer AM et al (2012). Assertive outreach: mirror-image study with contemporaneous controls. The Psychiatrist, 36: 245-248

• Dodgson G et al (2008). Early intervention in psychosis service and psychiatric admissions. The Psychiatrist, 32: 413-416
• Green B (2009) The decline of NHS inpatient psychiatry in England. Priory journal. Available at http://www.priory.com/psychiatry/Decline_NHS_Inpatient_Psychiatry.htm
• Hospital Episodes Statistics available at www.hesonline.nhs.uk
• Information centre available at www.ic.nhs.uk
• Office for National Statistics available at www.ons.gov.uk
• Prajakta S. Kalekar (2004) Time series forecasting using Holt Winters exponential smoothing, available at www.it.iitb.ac.in/~praj/acads/.../04329008_ExponentialSmoothing.pdf
• Jacobs R and Barrenho E (2011). Impact of crisis resolution and home treatment teams on psychiatric admissions in England. British Journal of Psychiatry, 199(1); 71-6
• Jones B et al (2012). A two-model integrated personality disorder service: effect on bed use. The Psychiatrist, 36: 293-298
• Howard LM et al (2010). The effectiveness and cost-effectiveness of admissions to women’s crisis houses compared with traditional psychiatric wards – a pilot patient preference randomised controlled trial. British Journal of Psychiatry, 197; s32-s40
• Isles V and Sutherland K (2001). Organisational Change – a review for healthcare managers, professionals and researchers. NCCSDO

 

 

Copyright Priory Lodge Education Ltd. 2013 onwards

First Published May 2013.

All pages in this site copyright ©Priory Lodge Education Ltd 1994-

Online Journals

Vet On-Line

Pharmacy On-Line

Anaesthesia On-Line

General Practice On-Line

History of Medicine On-Line

Psychiatry On-Line

Medicine On-Line

Family Medicine On-Line

Chest Medicine On-Line

Dentistry On-Line

Surgery On-Line