Intended for healthcare professionals


Limits to demand for health care

BMJ 2001; 322 doi: (Published 24 March 2001) Cite this as: BMJ 2001;322:734
  1. Alan Maynard (, professor of health economics,
  2. Trevor Sheldon, professor
  1. York Health Policy Group, Department of Health Studies, University of York, Innovation Centre, York YO10 5DG
  2. The Mill House, Wantage OX12 9EH
  3. All Souls College, Oxford OX1 4AL
  4. Sandwell Health Authority, West Bromwich B70 9LD
  5. Department of Social Medicine, University of Bristol, Canynge Hall, Bristol BS8 2PR

    Rationing is needed in a national health service

    EDITOR—At the inception of the NHS its proponents asserted that, after the backlog of unmet need was met by the new service, demand would plateau in the 1950s. Instead demand grew rapidly and the cost containment crisis led to a royal commission and the financial stringency with which we are familiar. Now Frankel et al, the optimists in Bristol, are repeating the mistakes of the architects of the NHS in believing that demand is finite.1

    Can everything that results in some clinical benefit, and that patients want, be funded? Frankel et al's positive answer is based on studies that use expert opinion and research evidence to compare need and want for two elective procedures with the resources available. This ignores the fact that these criteria are themselves rationing devices that implicitly include notions of what is sufficient benefit. The authors provide estimates of demand given certain treatment (or rationing) criteria and argue that if demand, so defined, can be met then it is finite and requires no rationing. This reduces, absurdly, to “if you ration care using our criteria you don't need to ration care.”

    Treatment criteria are never static; technology changes, and what constitutes need and wants is socially determined—hence the huge variations in indications for elective procedures between the United States and United Kingdom. 2 3 New technologies do not automatically increase costs; they may do so if no intervention existed before (for example, interferon beta for multiple sclerosis and drugs for Alzheimer's disease) or if they lower the threshold (or extend the indications) for treatment. Even if technologies lower unit costs the increased numbers now eligible for treatment can lead to a disproportionate increase in the volume of activity and total spend. For example, the introduction of laparoscopic cholecystectomy resulted in an 11% overall increase in …

    View Full Text

    Log in

    Log in through your institution


    * For online subscription