Current issues of ACP Journal Club are published in Annals of Internal Medicine


Transforming evidence into practice: evidence-based consensus

ACP J Club. 1993 Jan-Feb;118:A16. doi:10.7326/ACPJC-1993-118-1-A16

Related Content in the Archives
• Editorial: The science and practice of continuing medical education: a study in dissonance

As clinicians, we should never underestimate the importance of experience and clinical judgment in our decision making, but these indispensable components of medical practice cannot be applied to best advantage unless they are coupled with up-to-date knowledge of the scientific basis for our clinical actions. A mission of ACP Journal Club is to afford readers easier access to important evidence published in the medical literature, but the road from this evidence to clinical application is often not smooth. For example, evidence from a study will cover no more than a fraction of clinical situations that arise in practice, and even when there is compelling evidence about the value of a procedure or intervention, we must exercise clinical judgment in tailoring decisions to fit each patient's circumstances. Even so, we often move slowly to incorporate new evidence into practice.

Advances in the prevention, diagnosis, and treatment of venous thrombosis highlight some of the problems with evidence dissemination and illustrate some of the reasons for these problems. Within the short span of about 25 years, clinical researchers have documented the gains in patient care that can be achieved by replacing the time-honored but nonspecific features of clinical diagnosis with venography (1), serial plethysmography (2) and B-mode imaging for deep-vein thrombosis (3), and ventilation-perfusion scanning (4, 5) and pulmonary angiography (4, 5) for pulmonary embolism. To prevent thromboembolism in high-risk patients we can use low-dose heparin (6), vitamin K antagonists (7), low-molecular-weight heparins (8), and external pneumatic compression (9). Finally, for the treatment of thromboembolic disorders, we can help with full-dose heparin (10), moderate-dose coumarin (11), and low-molecular-weight heparins (12).

Despite the disability and death that application of these recent advances can prevent, studies have shown less than optimal implementation. For example, Anderson and colleagues (13) found in 16 hospitals that less than a third of patients at high risk for venous thrombosis and pulmonary embolism received prophylaxis, with the variation in implementation in the hospitals ranging from a distressing 9% to a disappointing 56%.

Why has there been reluctance to use, for example, perioperative heparin prophylaxis when it is clearly effective and cost effective (14)? The reasons are many, but three stand out. First, surgeons did not place much credence in reports in the late 1960s and early 1970s of a high incidence of postoperative venous thrombosis detected by routine postoperative screening with 125I-fibrinogen leg scanning (15). Because very few thrombi detected by fibrinogen leg scan are symptomatic and even fewer lead to fatal pulmonary embolism, surgeons did not view postoperative thrombosis as a problem in their individual practices. Second, they feared excessive operative and postoperative bleeding as complications of thrombosis prophylaxis and were reluctant to use low-dose heparin despite the results of double-blind, randomized controlled studies that showed that this form of prophylaxis is safe (16, 17). Third, many advocates of prophylaxis were physicians with an interest in thrombosis. Their opinions lacked the weight of the surgical societies, which took a passive stand on prophylaxis and, in general, did not recommend its routine use.

The probability of acceptance of new, important, and valid information is increased if endorsed by consensus panels of nationally recognized groups of experts. In 1986, a consensus conference organized by the National Heart, Lung, and Blood Institute in Bethesda recommended that thrombosis prophylaxis should be used in all high-risk groups (17). Acceptance of these particular recommendations was also spurred by lawyers who continue to use these recommendations to the advantage of their clients in malpractice suits. The medico-legal implications of consensus recommendations have caused justified caution among those involved in the process. These concerns can best be countered by clear statements about the evidence used to reach the recommendations, including the strength of this evidence. Thus, if the evidence is from properly designed and executed studies, the recommendation should be firm and definitive, but if the evidence is weak because it is based on nonrandomized trials or anecdotal reports, the recommendations should reflect the uncertainty of the evidence.

The Working Group on Anti-thrombotic Therapy of the American College of Chest Physicians (ACCP) has provided an excellent model for an explicit evidence-based approach to forming recommendations. The ACCP Working Group met on three occasions over the past 7 years to critically review evidence and develop guidelines for the management of thromboembolic disorders. They graded their recommendations according to the scientific rigor of the published studies. A comparison of the recommendations made in 1986 with those proposed in 1992 is most heartening because there has been a marked increase in the number of recommendations based on scientifically sound evidence (18). Of special importance from the perspective of the practitioner, we need no longer place our “faith” in consensus groups if they use this approach: The process of considering evidence is explicit, and we can judge for ourselves the strength of the evidence backing the recommendations. Unfortunately, most consensus statements and practice guidelines published by expert groups and professional bodies either lack a strong attachment to evidence or do not provide enough details of evidence handling to permit clinicians to critically appraise the quality of the recommendations.

It is clear for thromboembolic disorders, and for many others, that research has markedly improved our ability to detect, prevent, and treat problems. The effect of this on patient care, however, has been blunted by the slow translation of the evidence into action. The development of consensus recommendations by experts may help us to interpret and act on evidence from clinical trials, but the process of handling evidence must be explicit. Furthermore, the consensus process must be followed by broad-based efforts to publicize the recommendations according to their merit and gain the acceptance of practitioners who did not participate in the initial review of evidence and formulation of recommendations. Individually, we can accelerate this process by looking for practice recommendations that clearly do justice to current evidence.

Jack Hirsh, MD
R. Brian Haynes, MD, PhD


1. Lensing AW, Büller HR, Prandoni P, et al. Contrast venography, the gold standard for the diagnosis of deep vein thrombosis: improvement in observer agreement. Thromb Haemost. 1992;67:8-12.

2. Hull RD, Hirsh J, Carter C, et al. Diagnostic efficacy of impedance plethysmography for clinically suspected deep-vein thrombosis: a randomized trial. Ann Intern Med. 1985;102:21-8.

3. Lensing AW, Prandoni P, Brandjes D, et al. Detection of deep-vein thrombosis by real-time B-mode ultrasonography. N Engl J Med. 1989;320:342-5.

4. Hull RD, Hirsh J, Carter CJ, et al. Diagnostic value of ventilation-perfusion lung scanning in patients with suspected pulmonary embolism. Chest. 1985;88:819-28.

5. The PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA. 1990;263:2753-9.

6. Collins R, Scrimgeour A, Yusuf S, et al. Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin. N Engl J Med. 1988;318:1162-73.

7. Sevitt S, Gallagher NG. Prevention of venous thrombosis and pulmonary embolism in injured patients: a trial of anticoagulant prophylaxis with phenindione in middle-aged and elderly patients with fractured necks of femur. Lancet. 1959;2:981-9.

8. Hirsh J, Levine MN. Low molecular weight heparin. Blood. 1992;79:1-17.

9. Turpie AG, Delmore T, Hirsh J, et al. Prevention of venous thrombosis by intermittent sequential calf compression in patients with intracranial disease. Thromb Res. 1979;15:611-6.

10. Hirsh J. Heparin. N Engl J Med. 1991;324:1565-74.

11. Hull RD, Hirsh J, Jay R, et al. Different intensities of oral anticoagulant therapy in the treatment of proximal-vein thrombosis. N Engl J Med. 1982;307:1676-81.

12. Hull RD, Raskob GE, Pineo GF, et al. Subcutaneous low-molecular-weight heparin compared with continuous intravenous heparin in the treatment of proximal-vein thrombosis. N Engl J Med. 1992;326:975-82.

13. Anderson FA Jr, Wheeler HB, Goldberg RJ. Physician practices in the prevention of thromboembolism. Ann Intern Med. 1991;115:591-5.

14. Salzman EW, Davies GC. Prophylaxis of venous thromboembolism: analysis of cost/effectiveness. Ann Surg. 1980;191:207-18.

15. Kakkar VV. The diagnosis of deep vein thrombosis using the 125I-fibrinogen test. Arch Surg. 1972;104:252.

16. Kakkar VV, Corrigan TP, Fossard DP, et al. Prevention of fatal postoperative pulmonary embolism by low doses of heparin. An international multicentre trial. Lancet. 1975;2:45.

17. Prevention of venous thrombosis and pulmonary embolism. NIH Consensus Development. JAMA. 1986;256:744-9.

18. Dalen JE, Hirsh J; eds. Third ACCP Conference on Antithrombotic Therapy. Chest. 1992;102(Suppl).