Calculation of cost-effectiveness of resident medical cover for psychiatric hospitals

Rudolf N Cardinal

BMJ, 24 August 2008
http://www.bmj.com/cgi/eletters/337/jul31_3/a942#201026

The forthcoming maximum working week of 48 hours [1, 2] and financial pressures from junior doctors' pay banding system [3] provide strong incentives to Trusts to reduce hours. Whereas general hospitals can pare out-of-hours cover down to a core Hospital At Night team [4], psychiatric hospitals often have medical cover provided by a single on-site doctor. Their urgent out-of-hours work comprises medical and psychiatric emergencies, some of which are life-threatening. Salary supplements cannot be reduced from 40% to 20% without making doctors non-resident [3]. Should mental health Trusts make junior psychiatrists non-resident? How should they make this decision?

National recommendations provide one source of guidance. The David Bennett Inquiry recommended that "there should always be a doctor in every place where a mentally ill patient is detained, or if that is not possible, foolproof arrangements should be in place twenty-four hours a day to ensure that a doctor will attend within twenty minutes of any request by staff to do so", adding that it is "highly unsatisfactory for doctors to be available in an institution where mentally ill patients were treated only if they came by car or taxi in a wholly unpredictable fashion" [5]. These recommendations are supported by the Mental Health Act Commission, who noted the "risk... [of spreading] available medical and other expertise too thinly, so that no inpatient units can realistically have immediate access to a doctor when emergencies arise... [w]here patients are detained for their own safety, such a lack creates an ethical dilemma, if not a legal one" [6, 7]. The National Institute of Clinical Excellence (NICE) guidelines on restraint recommend that that a doctor must be available within 30 minutes of being called to an alert, and add that "dialling for emergency services... is not sufficient in itself" [8].

An additional technique is to establish the cost-effectiveness of resident medical cover. This requires establishing (A) how many quality-adjusted life years (QALYs) need to be saved per year by resident medical cover for it to be cost-effective, and (B) whether resident medical cover does in fact provide this level of benefit.

Establishing (A) is simple. Making juniors non-resident saves an amount of money £a per year. To ensure that their rota complies with a 20%-or-less banding, it may be necessary to employ additional non-medical staff (e.g. to perform psychiatric assessments in emergency departments overnight). Suppose this costs £b per year. The annual cost saving is thus £(a - b). A price must then be put on a QALY. NICE currently values one QALY at £20,000 to £30,000 [9]. To be cost-effective by this standard, resident medical cover must save between (a - b)/20000 and (a - b)/30000 QALYs per year.

A QALY may be saved in many ways. Most simply, life may be saved. General adult psychiatric wards typically cater for patients aged 17-65. Across the population, males aged 41 have  a life expectancy of 44 years, and females 47 years [10]. Psychogeriatric wards typically cater for over-65s. At the age of 75, the population life expectancy for a man is 12 years, and for a woman 14 years [10]. So, saving the life of an inpatient on a general adult ward might save about 46 life-years, and on an old-age ward about 13 life-years. If the patient had an enduring poor quality of life, the QALY figures would be correspondingly reduced. On a scale from 0 (death) to 1 (perfect health), one study estimated the quality of life of outpatients with enduring mental illness at ~0.95 [11], while a population survey found that quality of life was reduced by psychiatric disease from 0.900-0.949 to 0.848-0.899 in the young and from 0.788-0.873 to 0.721-0.811 in the elderly [12]. Another way to save QALYs is to prevent enduring serious morbidity. A third way is to reduce the probability of death or serious morbidity. For example, reducing the probability of death from 30% to 20% is equivalent to saving one-tenth of a life.

Establishing (B) is difficult, because it involves the monitoring of low-frequency events, and using clinical judgement to guess what would have happened if a doctor had or had not been able to attend with the speed that residence brings. Suppose a Trust's calculations show that resident medical cover costs £60,000 net per year; at £30,000/QALY it would be cost-effective if it saves 2 QALYs per year. This could mean that saving one life in an emergency on a general adult psychiatric ward every 20 years would nonetheless make resident cover cost-effective. Distinguishing an average of zero lives saved from one life saved every 20 years is not easy to do with short-term monitoring (unless a serious emergency occurs during the monitoring, in which case it is very much easier); very long-term monitoring is difficult, and there is an important role for ad-hoc case reporting. Likewise, small short-term studies that fail to find a clear dangerous effect of removing resident medical cover may convey a misleading impression through extreme lack of statistical power [13-15] or through lack of appreciation that an apparently low adverse event rate may still far exceed the threshold for cost-effectiveness [16].

References

1. Council of the European Union (1993). Council directive 93/104/EC. (At http://eur-lex.europa.eu/LexUriServ/site/en/consleg/1993/L/01993L0104-20000801-en.pdf; accessed 23 August 2008.)

2. Her Majesty's Stationery Office (2003). Statutory Instrument 2003 No. 1684: The Working Time (Amendment) Regulations 2003. (At http://www.opsi.gov.uk/si/si2003/20031684.htm; accessed 23 August 2008.)

3. UK Department of Health (2000). Pay banding criteria. (At http://www.dh.gov.uk/assetRoot/04/05/38/77/04053877.pdf; accessed 23 August 2008.)

4. UK Department of Health (2005). The implementation and impact of Hospital at Night pilot projects: An evaluation report. (At http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_4117968?IdcService=GET_FILE&dID=1804&Rendition=Web; accessed 23 August 2008.)

5. Blofeld, J. et al., Inquiry to the Secretary of State for Health, and Norfolk, Suffolk, and Cambridgeshire Strategic Health Authority (2003) Independent inquiry into the death of David Bennett. (At http://www.nscsha.nhs.uk/resources/pdf/review_inquiry/david_bennett_inquiry/david_bennett_inquiry_report_2003.pdf; accessed 23 August 2008.)

6. The Mental Health Act Commission (2005). In Place of Fear? Eleventh biennial report 2003-2005. London: The Stationery Office. (At http://www.mhac.org.uk/files/MHAC%2011%20TEXT%20FA.pdf; accessed 23 August 2008.)

7. The Mental Health Act Commission (2007). Risk, rights, recovery. Twelfth biennial report 2003-2005. London: The Stationery Office. (At http://www.mhac.org.uk/files/pdf%2012th%20biennial%20report.pdf; accessed 23 August 2008.)

8. NICE (2005). Violence: the short-term management of disturbed / violent behaviour in inpatient psychiatric settings and emergency departments. National Institute of Clinical Excellence, Clinical Guidelines, 25 February 2005. (At http://www.nice.org.uk/nicemedia/pdf/cg025fullguideline.pdf; accessed 23 August 2008.)

9. NICE (2008). Measuring effectiveness and cost effectiveness: the QALY. (At http://www.nice.org.uk/newsevents/infocus/MeasuringeffectivenessandcosteffectivenesstheQALY.jsp; accessed 23 August 2008.)

10. Government Actuary's Department (2006). Cohort expectation of life, 1981-2056; principal projection; United Kingdom. (At http://www.gad.gov.uk/Demography_data/Life_Tables/docs/2006/wUKcohort06.xls; accessed 23 August 2008.)

11. Wilkinson G, Croft-Jeffreys C, Krekorian H, McLees S, Falloon I (1990). QALYs in psychiatric care? Psychiatric Bulletin 14: 582-585.

12. Saarni SI, Suvisaari J, Sintonen H, Koskinen S, Härkänen T, Lönnqvist J (2007). The health-related quality-of-life impact of chronic conditions varied with age in general population. Journal of Clinical Epidemiology 60: 1288-1297.

13. Nicholls JE (1992). Role of the duty psychiatrist. Psychiatric Bulletin 16: 218-219.

14. Mason J, Irani T, Fountoulaki G, Warwick S, Da Roza Davis J, Sudbury P (2006). Psychiatry at night: experience of the senior house officer. Psychiatric Bulletin 30: 329-333.

15. Palanisamy V, Agarwal V (2006). First on-call psychiatrist: resident or non-resident? Psychiatric Bulletin 30: 468.

16. Lawrie A, Serfaty M, Smith C (1996). Should the on-call psychiatrist be residential? Psychiatric Bulletin 20: 12-14.

Competing interests: The author is a junior psychiatrist on a resident rota.