To the Editor: What would you say if an asymptomatic 24-year-old woman who tested positive for chlamydia asked you how long she has had it? We say several “years.” In fact, in the absence of antibiotics prescribed for other reasons or prior testing for chlamydia, we would argue that there is about a 25% chance the infection has been present for 2 or more years! We base our argument on 2 published cohort studies.1,2 The largest was undertaken in the early 1990s in Bogata, Columbia, where a random community-based sample of women had cervical specimens collected over a period of up to 5 years as part of a natural history study of human papillomavirus infection.1 Women were included if they tested positive for chlamydia and had follow-up samples available. Two women were excluded because they received chlamydia-sensitive antibiotics during the study. The study showed that about 50% of infections had cleared by 1 year, 80% by 2 years, and 90% by 3 years. Similar findings were reported in other cohort studies, while some other studies suggest clearance rates may be higher.2–6 Using the data from the study by Molano and colleagues,1 we can use Bayes’ Theorem to generate a probability distribution for the duration of infection via the formula: is proportional to In this equation, we approximate the probability that a woman is tested and test positive given she contracted the infection x years ago by the proportion of woman with unresolved infection at x years (Molano et al.1) and assume that for an arbitrary infected woman, who may or may not have been tested, that the infection could have been contracted at any time in the recent past. We used this to calculate the probability that a woman has retained the infection for more than 1 year, 2 years, and so on. We have provided 2 estimates; the first assumes that no infection lasts longer than 3.5 years (the longest follow-up time recorded by Molano et al.1) whereas the second, more reasonable scenario assumes infections continue to resolve for up to 5 years, when the proportion still positive is 3%, but all resolve at this point (Table 1). These calculations assume that an infected asymptomatic woman is equally likely to present for testing at any stage during her infection. They are clearly not valid if, for example, the women has presented for testing because she has symptoms or if she has a partner who recently tested positive for chlamydia. They can, however, be applied to any asymptomatic woman who is screened for chlamydia as part of a screening program, providing she has not had antibiotics for another unrelated condition or has not been tested for some years. These data suggest that there is reasonable probability that she may have had the infection for at least 18 months.TABLE 1: The Probability That the Infection Has Persisted for a Specific Length of Time, Given That an Individual Has Tested Positive for ChlamydiaThese data are particularly valuable when discussing positive results with a woman, largely because of the implications for her sexual relationships. Younger people are more likely to have more than one new sexual partner in the last 12 months and also to have relationships of shorter duration.7,8 Therefore, it is quite possible that a young woman may have acquired her infection from a previous sexual partner two or even three years ago. This information should be passed on to asymptomatic women diagnosed with chlamydia, as it could help prevent the potential destructive repercussions with current sexual partners that may arise if she believes that her infection had been recently acquired. This information should be included in any chlamydia education material targeting both clinicians and the general public and may help reduce any harm associated with screening.