Intended for healthcare professionals


Clinical guidelines in 1994

BMJ 1994; 309 doi: (Published 03 December 1994) Cite this as: BMJ 1994;309:1457
  1. Gene Feder

    Let's be careful out there

    Clinical guidelines based on the systematically analysed results of research and carefully introduced to doctors can improve clinical practice and outcomes. This is the main message of the current issue of Effective Health Care,1 which is based largely on an update of Grimshaw and Russell's landmark review2 and now covers 91 rigorous evaluations of the use of guidelines.

    Most of the guidelines in Grimshaw and Russell's review were not based on systematically reviewed evidence; however, it is a reasonable assumption that guidelines that are accurately based on evidence of effective treatment will benefit patients more than guidelines developed in an ad hoc manner or through informal consensus. The patchy nature of evidence, even in the best researched subjects in clinical practice, means that all guidelines in the conceivable future will be hybrid documents, with recommendations based on varying degrees of evidence and consensus. Good guidelines will clearly label recommendations according to strength of supporting evidence.

    Effective Health Care highlights those factors that increase the likelihood that doctors will adhere to guidelines—for example, active educational interventions to make doctors aware of the content of guidelines and patient specific reminders to prompt doctors to use them. It challenges the notion that “ownership” by doctors is a prerequisite for adherence in practice. Using Hurwitz's incisive analysis of the legal background to guidelines,3 it is reassuring about the medicolegal consequences (despite disturbing trends in the United States4).

    The authors do not shy away from some difficulties with guidelines, such as the effort needed to develop recommendations based on evidence, our ignorance about the relative strength of different implementation strategies, are the paucity of research on which to make judgments about cost effectiveness.5 These are technical problems amenable to more research and sufficient investment in guidelines programmes nationally and locally. This bulletin projects a bright future for clinical guidelines, leaving a darker side unexplored.

    Although the role of guidelines in the commissioning of health care is discussed, the connection between clinical guidelines and growing managerial interest in clinical practice is seen only as an opportunity and not as a problem. A related and equally awkward issue is the applicability of scientifically based knowledge (as represented in “valid” guidelines) in the encounter between doctors and individual patients.

    Why are governments, health service commissioners, and health care insurance companies so interested in clinical guidelines? As Woolf points out, “Although everyone wants practice guidelines, only patients care exclusively about clinical outcomes.”6 Governments and third party payers want tools for controlling the allocation of health care resources through containment or redirection of costs, although guidelines will not necessarily reduce overall costs7 even if they reduce unacceptable variations in practice.8 Guidelines are being put forward as a basis for shifting the purchase of health care away from activity towards protocols of care.9 Translating guidelines into protocols is problematic and is a method of enforcing guidelines that cuts across their voluntary adoption by doctors. Many American doctors have already been moved down this road with rigid protocols of managed care, which they have to follow for reimbursement.10

    Can they bear the weight?

    Can guidelines bear the weight of decisions about the allocation of resources? Even guidelines based on evidence are currently too weak to support this order of decision making,6 and their enforcement in practice misconstrues the nature of the scientific evidence underpinning them,11 simultanteously obscuring political questions about priorities and patients' choice. Of course decisions by commissioners of health care need to be informed by the relative effectiveness of treatments. Guidelines are a source of this information, and commissioners have a role in encouraging their implementation in clinical practice, but in most clinical areas their translation into rigid, enforceable standards of quality would be foolish.12

    If we remain cautious about clinical guidelines in the services of managerial control and resist their metamorphosis into standards and contracts can we otherwise relax about the spread of guidelines? Not quite. There is also a philosophical problem about guidelines, which is part of the “great debate” over individual judgment versus agreed standards.13 The problem has been raised by McCormick in a discussion of the role of judgment in clinical decision making14 and further articulated by Charlton in a discussion of limits to the application of research evidence to the care of individual patients.15 There is a danger that the knowledge represented in clinical guidelines will be given priority and will squeeze out other modes of knowledge.

    A trenchant exposition of these dangers comes from the United States, the birthplace both of guidelines based on evidence and of the assault on variation in medical practice.16 The critique is not based on the shortage of good clinical trials, particularly in primary care,17 or on the methodological problems of research into outcomes.18 It addresses the complex nature of medical (and nursing) knowledge.19 The 20th century project of transforming medicine from a craft into a science has given priority to objective, explicit understandings, which have undoubtedly made health care more effective. Yet clinical and moral knowledge of individual patients is often subjective and implicit.20 This will remain the case, as will the intrinsic uncertainty of individual clinical decisions,21 no matter how many clinical trials are performed. These aspects of practice are not beyond research, but the methods are qualitative and do not fit the paradigm of “effectiveness” that generates clinical guidelines in their present form.22 Nor can the artistic or intultive dimension of medicine be commissioned in a “package of care,” although it could be overwhelmed by the discrete interventions and measurable outcomes that dominate current discussion in the health service.

    We can be enthusiastic about guidelines based on evidence as tools for decision making at the coal face of clinical care. They give us access to knowledge, which can help us to practice more effectively. But we need to remain cautious about guidelines' role in commissioning of health care and recognise that explicit guidelines address only an aspect of good clinical practice.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.
    19. 19.
    20. 20.
    21. 21.
    22. 22.
    View Abstract