Treating Chlamydial Infection: Compliance Versus Cost

SINGLE-DOSE, directly observed therapy is a central strategy for the treatment and prevention of sexually transmitted diseases (STD) and has been the norm for gonorrhea and early syphilis for 5 decades and for chancroid for 10 years. Thus, the development of effective single-dose treatment with azithromycin for uncomplicated Chlamydia trachomatis infection was one of the most important recent advances in STD therapy.¹ Unfortunately, azithromycin is expensive. Whereas the standard doxycycline regimen (100 mg twice daily for 7 days) costs the pharmacy or clinic only $.50 to 2.00, the cost of 1.0 g azithromycin varies from $9.50 (for public clinics using the 1.0-g sachet) to $25.00 or more (for four 250-mg capsules). From this cost difference has arisen a modicum of controversy as to the regimen of choice for uncomplicated chlamydial infection. Because of the high frequency of serious, costly sequelae in women, computer-modeling studies of chlamydia treatment strategies suggest that routine use of directly observed treatment with azithromycin is likely to be more cost-effective than giving doxycycline.^2-4 However, these studies made assumptions that may or may not be valid about the risk of treatment failure because of partial compliance with doxycycline. Cost-effectiveness notwithstanding, the difference of $7.00 to $9.00 per patient may be very significant for poorly funded public clinics, which rarely benefit directly from the savings that accrue from preventing pelvic inflammatory disease, infertility, ectopic pregnancy, and neonatal morbidity.

Two studies^5,6 in this issue of the journal provide new information on this dilemma, but they support opposite conclusions. Hillis et al⁵ found no difference in the frequency of persistent chlamydial infection in women in several clinics (primarily public STD or family planning clinics) after treatment with single-dose azithromycin or the standard doxycycline regimen. Specifically, among women with cervical infection documented by isolation in cell culture, 93 (95%) of 98 women given azithromycin and 94 (96%) of 98 treated with doxycycline were free of infection 3 to 4 weeks after treatment, based on negative polymerase chain reaction tests on urine. Regardless of actual compliance with the prescribed regimen, the large majority of patients apparently took enough doxycycline to eradicate C. trachomatis.

Augenbraun et al⁶ examined compliance with the standard doxycycline regimen among patients attending either of two public STD clinics using a system whereby a computer chip in the cap records each opening of the medication container. Two hundred twenty-three patients with documented or presumptive chlamydial infection were enrolled, and 179 patients (80%) returned for follow-up. Twenty-four percent of the patients were judged to be noncompliant, failing to take at least one dose daily for at least 5 days or, in a few cases, delaying treatment for >48 hours; 25% took all or most of the drug on schedule; and 51% had intermediate levels of compliance. Compliance was not predicted by the patients' sex or, surprisingly, the presence or absence of symptoms. The authors conclude that compliance with doxycycline is very poor in STD clinic patients, even when the drug is given directly to patients without charge, and that at least 25% of patients take too little doxycycline to reliably cure their chlamydial infections. However, some patients might have removed more than one pill each time the container was opened, and the monitoring system may underestimate actual therapeutic compliance and the number of doses consumed.

Many readers will agree that the results of the Hillis study⁵ seem too good to be true; the use-effectiveness of doxycycline was similar to the biologic cure rate in controlled studies in which compliance is strongly encouraged and rewarded. Participation in a research study and the requirement for informed consent might have enhanced compliance with multiple-dose therapy. In addition, the study design required that if a potential subject's male sex partner presented before the patient herself, or if they attended the clinic together, both had to agree to the woman's participation in the study. Forty percent of those otherwise eligible for enrollment were excluded because of their partners' refusal. The effect of enriching the study population with women whose partners condoned their participation is difficult to predict, but one possibility is a bias toward women in steady partnerships. Involvement in a committed relationship might be a marker for personal responsibility and therapeutic compliance, and it appears likely that steady partners might encourage compliance in one another. Thus, the results may not apply to a substantial subset of patients with chlamydial infection, which the authors acknowledge.

The minimum dose and duration of doxycycline that reliably eradicates C. trachomatis is unknown, although it is likely that a full 7-day course often is unnecessary. Cerin et al⁷ conducted follow-up on 48 women with chlamydial infection using culture, enzyme immunoassay, and direct immunofluorescence tests during treatment with doxycycline (200 mg initially, then 100 mg once daily for 8 days). Among 33 women with positive cervical cultures at the start of treatment, C. trachomatis was re-isolated in 11 (33%) after 2 days treatment, and the urethral culture remained positive in one woman at 4 days. Several additional women had reactive antigen tests through four days' therapy, but all tests were negative in all women on day 6. In another recent study that included some of the subjects in the Augenbraun study, over 90% of STD clinic patients treated with doxycycline dispensed in containers with computerized caps said they had taken all 14 pills, but only about 40% had opened the containers ≥11 times; nevertheless, persistent chlamydial infection was documented in only 5 (6%) of 82 patients who were culture-positive at enrollment.⁸ Stamm and colleagues⁹ showed that in patients with both gonorrhea and chlamydial infection, C. trachomatis was eradicated in 27 (93%) of 29 patients treated with tetracycline hydrochloride (500 mg four times daily) for 5 days. However, not all of these studies were designed to detect transient suppression of C. trachomatis,¹⁰ which can only be determined with prolonged follow-up. There is an important need for well-designed, dose-ranging studies with doxycycline with several weeks follow-up, but until such data are available it seems prudent to assume that treatment with doxycycline for <5 days is associated with significant risk of treatment failure.

How should clinicians synthesize the available data? The Hillis study provides limited reassurance that, regardless of actual compliance, doxycycline usually is effective in women attending STD or family planning clinics.⁵ At the same time, the Augenbraun study results suggest that compliance with doxycycline in STD clinic patients is even worse than might have been predicted,⁶ although the available data suggest that even some of the least compliant patients might have taken enough drug to cure their infections. Neither study addresses an additional downside to incomplete compliance, that leftover drug may be used later without clinical guidance, contributing to delayed or missed diagnosis, selection of resistant bacteria, and all the other spin-offs of inappropriate antibiotic use.

The 1998 STD treatment guidelines of the Centers for Disease Control and Prevention¹¹ recommend directly observed, single-dose azithromycin (1.0 g) and 7 days of doxycycline (100 mg twice daily) as equal first-choice regimens for treatment of uncomplicated chlamydial infection and further state that "Azithromycin should always be available to providers to treat at least those patients for whom compliance is in question." However, the Augenbraun study suggests that compliance is "in question" for most patients, at least in some STD clinics. Instead of reserving azithromycin for unreliable patients, we favor starting with the assumption that single-dose, directly observed treatment will be used, then making exceptions for apparently reliable patients who are carefully counseled about compliance. Using this approach, at our STD clinic about half the patients with uncomplicated chlamydial infection, nongonococcal urethritis, or mucopurulent cervicitis currently are treated with azithromycin and the remainder with doxycycline or other multiple-dose regimens. One of us directs a municipal STD control program, and we are sensitive to the problem created by the high cost of azithromycin, particularly for public STD and family planning clinics. At the same time, all providers- and especially public health departments-have a responsibility to consider long-range and societal costs and to not use short-term costs as the primary basis for selection of routine treatment.