Whereas Gaydos et al. invited all female army recruits undergoing a physical examination at a reception base over the course of 2-year period to participate in their study, our sample consisted of Israeli women actively seeking medical care. Specifically, participants in the current study were women serving in the Israel Defense Forces, who, over a 12-month period attended a women’s health clinic for a routine gynecologic examination, treatment of genitourinary symptoms, or contraceptive counseling. The clinic was associated with a major tertiary hospital in northern Israel. Of the 800 women we asked, 708 volunteered to participate, for a response rate of 88.5%. Participants were between the ages of 18 and 45 (median age 19, mean age 20). Most participants (77%) were born in Israel.

After signing an informed consent form, participants were asked to complete a short questionnaire (including demographic and sexual history items) and provide a urine specimen. We performed the polymerase chain reaction (PCR) of urine specimens to test for C. trachomatis and Neisseria gonorrhoeae (Roche Amplicor, Branchburg, NJ). The PCR results and questionnaire data were coded as dichotomous variables and examined on the basis of both univariate (chi-square and Fisher exact test) and multivariate (logistic regression) analysis.

Of the 708 participants, 23 (3.25%) tested positive for C. trachomatis, and 8 (1%) tested positive for N. gonorrhoeae. These rates are substantially lower than the 9.2% prevalence rate for C. trachomatis reported by Gaydos in the U. S. military study (4). The median and average age of those affected was 20. Most of those testing positive reported 1) having had sex in the past 3 months (86%), 2) having had more than one partner (45%) or having had a partner who had had more than one partner (57%), or 3) never having used a condom (70%) (Table). Nevertheless, none of these factors, identified by Gaydos as significant risk factors in the American sample, was found to be a significant predictor of C. trachomatis infection in the Israeli soldiers. Similarly, while most of those testing positive were born in Israel (81%), country of origin was not a significant predictor of infection. Neither were gynecologic symptoms found to significantly distinguish between C. trachomatis–positive and –negative soldiers. Indeed, the only factor that we found to significantly distinguish between Chlamydia-positive and -negative persons was urban versus rural residence. Specifically, rural residents were significantly more likely to be positive for C. trachomatis than urban residents (chi square=3.86, p<0.05). Similarly, in a logistic regression analysis, only rural residence was significantly associated with Chlamydia infection (p=0.043) (Table).

The relatively low rate of C. trachomatis prevalence was somewhat surprising for several reasons. First, since 1994, when Chlamydia infection became a nationally notifiable infectious disease in Israel, the number of cases reported annually has risen 10-fold. Secondly, since 2000, 14% of symptomatic soldiers attending the sexually transmitted disease clinic of another major medical center in northern Israel were diagnosed with Chlamydia infection. Finally, while Gaydos et al. found a 9.2% prevalence rate in a sample of soldiers reporting for a general physical examination, our sampling strategy focused on women who had a high probability of being sexually active (i.e., seeking contraceptives) and consequently, according to the previous literature, were at higher risk of contracting asymptomatic C. trachomatis infection (5).

However, oversampling those most concerned with their health (i.e., “the worried well”) may partially explain the lower than expected prevalence rate. Indeed, given the risk for a self-selection bias inherent in such a sampling strategy, one important limitation of the current study is that the sample population may not be entirely representative of the entire population of female soldiers in the Israeli Defense Forces. A second limitation is that the Roche PCR kit used in this study was not cleared by the Food and Drug Administration for testing female urine samples for N. gonorrhoeae because of its low sensitivity. Consequently, the actual prevalence of gonorrhea may in fact be higher than reported above.

Given the nearly universal draft of young Jewish females in Israel, however, our results may provide a strong indication of the prevalence of C. trachomatis in young, sexually active Jewish women in that country. Despite the relatively low rate of prevalence, based on a number of cost-efficacy studies (6,7) and consistent with the recommendations of the Centers for Disease Control and Prevention (5), focused screening of asymptomatic, sexually active female soldiers, by virtue of their age, may be a cost-effective mode for identifying those with C. trachomatis infection in Israel. Identifying gonorrhea in 1% of the women tested shows an important secondary benefit to such screening. Nevertheless, before recommending the adoption of a broader, more focused screening program of female soldiers, whether symptomatic or not, we recommend collecting additional data with a more randomized sampling strategy similar to that adopted by Gaydos et al. (4). Were an equivalent or higher prevalence rate to be found in a random sample of female soldiers during their service, then, according to recent estimates of cost-effectiveness (7,8), such a screening program would be worthy of consideration.