Education and debate

Clinical guidelinesthe hidden costs

Alan Haycox, senior research fellow,  Adrian Bagust, senior research fellow,  Tom Walley, director. 

Prescribing Research Group, Pharmacology and Therapeutics, University of Liverpool, Liverpool L69 3GF

Correspondence to: Dr Haycox AHAY{at}liv.ac.uk

The publication of guidelines for statin treatment in the United Kingdom attracted criticism, particularly that no attempt had been made to assess the implications for resources associated with the guidelines. ^{1 2} The need to evaluate cost and clinical effectiveness in real terms before the widespread use of a treatment is frequently acknowledged.³ Unfortunately, these issues are rarely addressed in clinical guidelines, and there is little evidence that cost effectiveness in practice is considered in drawing up guidelines. ^{4 5}

Surprisingly little consideration has been given to the role of guidelines in a healthcare system in which resources are constrained. ^{6 7} Cholesterol lowering drugs epitomise the problems that affect clinical guidelines in many areas. We consider here some unintended consequences from the development of guidelines (whether regulatory or ... supposedly ... advisory) that arise for those funding health care, doctors, and patients.

	Nature and role of guidelines

All clinical guidelines aim to promote "best practices" that improve the outcomes of treatment.⁸ This rests on two assumptionsfirstly, that outcomes identified in clinical trials are reproducible in normal practice and, secondly, that rapid and universal adoption of an effective treatment leads to optimal treatment for the whole population. Guidelines place most weight on evidence generated in randomised controlled trials. Unfortunately, real life clinical practice, where resources are more restricted, patient compliance less reliable, and treatment cannot be limited to a narrow patient group, does not provide such a controlled environment. Consequently, the anticipated benefits are rarely fully realised in an everyday setting,⁹ and true therapeutic effectiveness is established only through the accumulation of practical clinical experience. ^{10 11}

The cost of haste
Guidelines may allow the introduction of new treatments to be fast tracked, thus bypassing traditional processes of adoption. However, the time "gained" is not without a cost. This period previously allowed treatment to be refined and true effectiveness to be evaluated. Costly new treatments may gain widespread acceptance through over hasty use of guidelines, when in practice their results may be poorer than expected.

View larger version (15K): [in this window] [in a new window]   Incentive regions for the introduction of new treatment. (The nature of the incentives is given in the table)

Modelling cost and outcome
Optimal decision making about treatments requires doctors to balance incremental changes in patient outcomes with incremental changes in cost (compared to current treatment). The six distinct combinations of cost and outcome that may be presented by a new treatment are illustrated in the figure. In region A, the new treatment has been proved to worsen the outcome appreciably, in region B it improves the outcome and reduces the cost of treatment, while in region E it improves the outcome but increases the cost. In regions C, D, and F, the impact of the new treatment in practice is still uncertain. In regions A and B the interests of doctors and health commissioners will coincide. In regions C and D, given the uncertainty over the impact of treatment, the most powerful opinion will tend to dominate. In practice, most conflict arises in region E, where improved outcome is achieved only at increased cost. Here strong clinical support exists for a treatment that may shift priorities away from less fashionable but more cost effective treatments. The natures of the incentive operating in each case are outlined in the table.

Interested parties
The role of guidelines is complex, inevitably reflecting the contrasting agendas of various interested partiesprofessional groups and clinical enthusiasts are keen to extend their influence, pharmaceutical suppliers to capture and exploit market share, while local and national authorities seek to promote public health while restraining public expenditure. It is naive to suppose that any formulating committee will be immune from these concerns, and we must therefore take into account the source of a guideline when interpreting its broader implications for the health service.

	Problems with guidelines

Curtailing discretion A new guideline presents the doctor with an explicit threshold for treatment, where he or she previously relied upon professional discretion drawn from accumulated experience. In an increasingly litigious environment, the doctor may opt for safety first, perceiving the guideline as a minimum level of treatment to be given. Introducing a guideline will therefore "ratchet up" the volume of treatment prescribed, overriding professional discretion and available evidence.

Constraining local flexibility

Pre-empting resources

Vested interests

	Developing better guidelines

What value does disseminating clinical guidelines have for the health service? There is a case to be made for stemming the flood of advisory documents and pursuing instead improved professional continuing education and audit based on outcomes. Where guidance is mandatory, the least dangerous form merely cautions against new treatments with limited cost effectiveness and existing treatments that no longer provide adequate value for money.

Any guidelines must start from a patient perspective while encompassing the needs of the whole community. Thus, in diabetes, benefit from reducing lipid concentrations must be compared with other methods of benefiting diabetic patients. The only secure basis for guidelines is a comprehensive assessment of all consequences (direct and indirect) of a new treatment. This requires analysis of the lifetime healthcare experience of the affected group of patients together with any "knock on" consequences for other patient groups. Such an approach requires that the entry of new treatments into the healthcare system should be managed.³ While this may delay access to beneficial innovations, it would ensure that both the costs and consequences of new treatments are fully explored before they become widely available.

Guidelines generate systematic and paternalistic pressure for the many to conform to the views of the few. They replaces the gradual and time honoured process by which non-specialists learn from experience with an attempt to force the pace of disseminating "good practice," measured by narrow criteria. The unquestioning pursuit of clinical guidelines without regard to the added problems and pressures they may engender poses dangers for us all, representing as it does the exercise of power without responsibility.

	Acknowledgments

Contributors: Each author contributed fully at all stages in the development of the core ideas. The final draft was prepared by AH, who is also the guarantor.

Conflict of interest: None.

Funding: None.

	References

1. Standing Medical Advisory Committee. The use of statins. London: Department of Health , 1997(11061 HCD Aug 97(04).)
2. Freemantle N. The use of statins: a case of misleading priorities?. BMJ 1997; 315: 826-828[Free Full Text].
3. Dent TH, Hawke S. Too soon to market. BMJ 1997; 315: 1248-1249[Free Full Text].
4. Thompson R, Lavender M, Madhok R. How to ensure that guidelines are effective. BMJ 1995; 311: 237-239[Abstract/Free Full Text].
5. Mann T. Clinical guidelines: using clinical guidelines to improve patient care within the NHS. Leeds: Health Care Directorate, NHS Executive , 1996.
6. Grimshaw JM, Russell IT. Effect of clinical guidelines on medical practice: a systematic review of rigorous evaluations. Lancet 1993; 342: 1317-1322[Medline].
7. Davies J, Freemantle N, Grimshaw J, Hurwitz B, Long A, Russell I, et al. Implementing clinical practice guidelines. Effective health care bulletin no 8. University of Leeds, 1994.
8. West E, Newton J. Clinical guidelines. BMJ 1997; 315: 324[Free Full Text].
9. Fayers PM, Hand DJ. Generalisation from phase III clinical trials: survival, quality of life and health economics. Lancet 1997; 350: 1025-1027[Medline].
10. Spearman ME, Summers K, Moore V, Jacqmin R, Smith G, Groshen S. Cost effectiveness of initial therapy with 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors to treat hypercholesterolemia in a primary care setting of a managed care organisation. Clin Ther 1997; 3: 582-602.
11. Brown NY, Gray D, Skene AM, Hampton JR. Inpatient deaths from acute myocardial infarction 1982-92; Analysis of data in the Nottingham heart attack register. BMJ 1997; 315: 159-164[Abstract/Free Full Text].
12. Axon ATR, Bell GD, Jones RH, Quine MA, McCloy RF. Guidelines on appropriate indications for upper gastrointestinal endoscopy. BMJ 1995; 310: 853-856[Free Full Text].
13. Jones R, Lydeard S. Prevalence of symptoms of dyspepsia in the community. BMJ 1989; 298: 30-32.
14. McCormick A, Fleming D, Charlton J. Morbidity statistics from general practice: fourth national study 1991-1992. In: London: HMSO , 1995(OPCS Surveys; series MB5 No 3.)