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


1-day azithromycin was as effective as 7-day doxycycline for nongonococcal urethritis syndrome in men

ACP J Club. 1996 Nov-Dec;125:82. doi:10.7326/ACPJC-1996-125-3-082

To the Editor

In his commentary on the article by Stamm and colleagues (1), Dr. Shekelle states that because the efficacy of the 2 treatments for nongonococcal urethritis is equal, compliance should be the primary reason that clinicians prescribe azithromycin rather than doxycycline. Given that the study on which he comments is an intention-to-treat analysis, his logic may be flawed. By design, an intention-to-treat analysis takes compliance rates into account. For the study in question, this means that the cure rates were equal regardless of whether the patients received all of the prescribed medication. Compliance with doxycycline in this study was presumably less than 100%. The question then remains: Why should we substitute the more expensive azithromycin for doxycycline when the study shows that they produce clinically comparable results?

Thomas D. Wendel, MD
St. Luke's Roosevelt Hospital Center
New York, New York

Response: Dr. Wendel is correct that Stamm and colleagues used an intention-to-treat analysis, showing no difference in cure rates between azithromycin and doxycycline; thus, any difference in compliance between patients who receive azithromycin or doxycycline are already accounted for in the analysis. But this study was a clinical trial with special resources for recruiting and following patients, and we know that patients in clinical trials are usually far more compliant than patients in real-life practice. In this study, for example, only about 10% of patients did not return for the follow-up appointment; this is a very low rate for persons with sexually transmitted diseases (STDs). In the medical center where I practice, a large proportion of patients with STDs do not get their prescription filled (let alone take pills every 12 hours for 7 days), and about 40% of patients do not return for the follow-up appointment. Thus, the effectiveness of doxycycline at my medical center is much lower than is observed in the trial. We have already switched to azithromycin.

The above discussion only highlights 1 important issue that clinicians will want to consider when applying the results of clinical trials to patients in their practice. Other issues may also exist. For example, a clinician who is confronted with a patient who has had a transient ischemic attack and who has > 70% stenosis of the internal carotid artery may search the literature and find 3 major randomized controlled trials (RCTs) that show a benefit for carotid endarterectomy in symptomatic patients with high-grade carotid stenosis (1-3). Before recommending surgery to the patient, however, the doctor should note that these RCTs only included surgeons and hospitals that had low perioperative complication rates—typically a 30-day death or stroke rate of under 8%. This incidence is much lower than the average rate in the only population-based study of carotid endarterectomy to date (4).

In addition, the patients in these RCTs were otherwise a relatively healthy group (i.e., patients with uncontrolled diabetes, congestive heart failure, renal failure, and other serious illnesses were excluded). To what extent does the patient with cardiac problems fit the criteria of those entered into the trial? For patients with comorbid conditions, the benefit of surgery may be larger, smaller, or nonexistent. Even within the trial, overall results may not apply to all patients. An analysis of 1 of the trials concluded that surgery primarily benefited patients who had the greatest risk for stroke and that patients with less risk were helped only slightly or even harmed by the surgery, although in aggregate the group as a whole did better with surgery (5). Finally, there is the question of patient preference. Surgery typically has a short-term disadvantage (i.e., increased perioperative morbidity and mortality), and it may take some time before surgery has a survival advantage over medical therapy. To what extent does the patient value short-term outcomes over longer-term outcomes?

Thus, although RCTs are our most powerful means of assessing the relative efficacy of different treatments, there are other important issues to be considered when applying their results to clinical practice.

Paul G. Shekelle, MD
West Los Angeles Veterans Affairs Medical Center
Los Angeles, California


1. Shekelle PG. Commentary on "1-day azithromycin was as effective as 7-day doxycycline for nongonococcal urethritis in men." ACP J Club. 1996 May-Jun;124: 70. Comment on: Stamm WE, Hicks CB, Martin DH, et al. Azithromycin for empirical treatment of the nongonococcal urethritis syndrome in men. A randomized double-blind study. JAMA. 1995 Aug 16; 274:545-9.

1. European Carotid Surgery Trialists Collaborative Group. Lancet. 1991; 337:1235-43.

2. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med. 1991;325:445-53.

3. Mayberg MR, Wilson SE, Yatsu F, et al. JAMA. 1991;266:3289-94.

4. Winslow CM, Solomon DH, Chassin MR, et al. N Engl J Med. 1988;318:721-7.

5. Rothwell PM. Lancet. 1995;3435:1616-19.