In the
current issue of
The American Journal of Medicine,
Lopes et al(1) report discouraging news. They examined a large cohort of patients taken from 3 large randomized double-blind clinical trials and asked the question: “Are patients in these trials who were in atrial fibrillation being treated with antithrombotic therapy in accordance with widely accepted guidelines?” Unfortunately, only a small percentage (13.5%) of the patients with atrial fibrillation in these trials were, in fact, receiving indicated prophylactic antithrombotic therapy with warfarin. This is not an unusual finding. Many other studies have documented that guideline recommendations are followed for a disappointingly small portion of inpatients and outpatients.(2, 3) It has become clear that in daily clinical practice, guideline advice is often ignored or overlooked. What can be the explanation for this lack of compliance with evidence-based counsel? Guidelines are written by acknowledged experts in the areas covered, and the published results of guidelines are widely disseminated. Why don't physicians follow this advice? Surely, it is not because as a group we are careless or don't care about the results of carefully conducted clinical trials.
I have given this topic a great deal of thought in recent months, and my personal opinion is that neither negligence nor hostility to guidelines underlies poor physician compliance. Rather, I believe that a number of other factors help to explain the observed failure to use evidence-based guideline information:
1. Currently, there are literally hundreds and hundreds of guidelines available. These have all been carefully prepared by a variety of different professional societies. Some of these guidelines are quite long, at times exceeding 300 pages in length. Considering the number of guidelines and their size, it is not surprising that busy physicians do not have time to peruse, no less absorb, the presented material. Furthermore, in a hectic physician work day with its constant interruptions, one can easily understand why guideline-recommended interventions might be overlooked.
2. Clinicians in practice develop a series of personal “tried and true” approaches for managing the disease entities commonly seen in their daily work. Once established and thought to be successful, the use of these personal approaches becomes second nature to the physician. Changing these personal protocols feels like an unnecessary imposition given the many demands already craving attention in the busy clinician's work day. Again, given these comfortable patterns of practice, it is not surprising that new evidence-based clinical approaches would not be adopted easily.
3. Physician training, particularly in the past, emphasized individual patient therapeutic regimens with an avoidance of rigid “cookbook”-style order sets. Thus, many physicians think that a rigid “cookbook” style of medical practice, including the use of standard order sets based on guideline recommendations, represents an improper approach to patient care.
4. With specific reference to preventive antithrombotic therapy in patients with atrial fibrillation, patients and doctors dislike warfarin therapy, the most widely recommended antithrombotic therapy for patients with atrial fibrillation. The many dietary and pharmacologic caveats associated with warfarin therapy, the need for frequent monitoring of drug effect, and the fear for both patient and physician of unexpected bleeding complications make this agent one of the least desired therapeutic recommendations in most practices, including my own.
5. Finally, large, randomized, global, double-blind trials produce results that reflect outcomes for the majority, but not all of the patients, in that particular trial. There are usually a substantial minority of patients in each of these trials, the results of which lead to evidence-based guideline recommendations, who do not behave like the majority. Physicians may think that their particular patients do not fit the pattern observed in the large clinical trials. To some extent they might be perfectly right in this assumption. This attitude might lead some doctors to advise a therapeutic protocol different from that suggested by the results of the large, “gold standard,” randomized, double-blind clinical trials.
Whatever reasoning affects clinical decision making in these patients, the end result is often nonadherence to guidelines. Unfortunately, guideline-directed therapy for a particular condition has been shown to lead to better clinical outcomes compared with “eminence-based,” personally derived, therapeutic strategies. In general, the majority of patients with a particular entity would almost certainly benefit if guideline-directed therapy were universally applied. This is the reasoning behind many quality initiatives. As noted above, research in this arena has supported the idea that guideline-based therapy produces better clinical outcomes compared with arbitrarily selected therapeutic regimens. If we accept the latter 2 statements, the question then arises, “how can we get doctors to adhere more closely to evidence-based therapy as suggested in clinical guidelines?”
Our experience at the
University of Arizona College of Medicine has suggested one potential solution to the conundrum just described, that is, more liberal use of standard order sets embedded in electronic medical records and computer order systems. An example of how such standardization might improve clinical quality is as follows. A few years ago, it was noted by our quality department at the University Medical Center that documentation of counseling concerning cessation of tobacco use was often lacking in the charts of our patients discharged after an acute myocardial infarction. To remedy this problem, a requirement was placed in our discharge order set for these patients to receive this counseling. The patient could not be discharged from the hospital until a doctor or nurse involved in the patient's care had checked a box in the electronic orders stating that such appropriate advice concerning smoking cessation had indeed been carried out. This small and simple alteration in our discharge order set guaranteed that all such patients received tobacco cessation recommendations.
This change in our approach to smoking cessation counseling resulted in essentially 100% compliance with the need to give this advice. Why not use the same strategy for other order sets with similar guideline-advised therapeutic orders placed in the discharge orders together with drop-down boxes detailing reasons why a particular intervention was not used? The full order set would not be accepted until all such queries had been satisfactorily managed. The introduction of evidence-based orders as well as statements detailing why exceptions to these orders were made would ensure that the clinician had been made aware of the importance of dealing with these interventions. In this manner it would be possible to ensure a very high rate of compliance with evidence-based guideline recommendations. Presumably, this would also lead to improved quality of care for these patients.
As always, I'd be interested in hearing your comments on this important topic. Feel free to post a comment on our blog.
References
1. Lopes RD, Starr A, Pieper CF, et al. Warfarin use and outcomes in patients with atrial fibrillation complicating acute coronary syndromes. Am J Med. 2010;123:134–140.
2. Cabana MD, Rand CS, Powe NR, et al. Why don't physicians follow clinical practice guidelines?. JAMA. 1999;282:1458–1465.
3. Bach DS, Awais M, Gurm HS, Kohnstamm S. Failure of guideline adherence for intervention in patients with severe mitral regurgitation. J Am Coll Cardiol. 2009;54:860–865.
-- Joseph S. Alpert, MD
Editor-in-Chief,
The American Journal of MedicineThis article originally appeared in the
February 2010 issue of
The American Journal of Medicine.
A 29-year-old woman with no significant medical history developed anasarca over the course of approximately 2 weeks. On the basis of renal biopsy, a diagnosis of minimal change disease was made. Her nephrotic syndrome was poorly responsive to steroid therapy, and despite high doses of prednisone of up to 120 mg/d, her serum albumin and proteinuria remained 1 g/dL and 10 g/d, respectively. Approximately 1 month after the onset of her symptoms, she developed transverse, nonpalpable, white lines on all her fingernails (Figure). These nail changes are consistent with a diagnosis of Muehrcke's lines.
