This is the first general population-based survey on sexual behaviours in young adults in Croatia that included willingness to test for C trachomatis. The response rate for collecting urine specimens was 32.5%, and the prevalence of C trachomatis infection in sexually active individuals was 7.3% in men and 5.3% in women. Women, those who perceived their socio-economic status to be lower than average compared to their peers and those who did not use condoms at last intercourse were more likely to provide urine, possibly indicating an awareness of greater vulnerability towards STIs.
The main limitation in interpretation of determinants for C trachomatis prevalence is a low response rate to urine specimen collection. This could be due to a lack of awareness of young adults of the importance of screening for C trachomatis, or alternatively of inadequate skills of interviewers and operational difficulties in the field implementation of such studies. Urine samples were transported from participants' homes to the Counties Institutes of Public Health (there are 21 counties in Croatia), and from there to the "Dr. Fran Mihaljević" University Hospital for Infectious Diseases in Zagreb. Such operational difficulties in carrying out integrated bio-behavioural surveys of STIs have been described elsewhere [13, 14].
The response rate achieved in our survey is similar to that achieved in population-based C trachomatis prevalence assessment surveys that used mail-delivered testing among young adults in Tartu, Estonia (34%), and the Netherlands (41%) [15, 16]. In Estonia, prevalence of C trachomatis among women and men aged 18-35 years was 6.9% and 2.7%, respectively . In the Netherlands, the prevalence of C trachomatis infection was 2.5% in women and 1.5% in men 15-29 years old . In Eastern Europe, a household-based probability sample survey that estimated prevalence of infection with C trachomatis was carried out in Slovenia in 2001 among 18-49 year olds . It found the prevalence of C trachomatis infection to be 4.1% among both men and women 18-24 years old, and the response rate to urine specimen collection was 50.9% among men and 60.0% among women. In Great Britain, prevalence of C trachomatis infection was 2.7% among men and 3.0% among women aged 18-24 in a household-based probability sample survey carried out in 2001 .
In the US, a number of household surveys have measured chlamydia prevalence. In three studies carried out amongst young adults, response rates of 61-81% were achieved [17–19]. The response rates in the studies carried out in the US are higher than in our study, possibly because of monetary and other incentives (food coupons) given for participating.
Compared to the European surveys that obtained similar response rates, the prevalence of C trachomatis found in this study is considerable, particularly in men. However, as the results suggest, those with higher levels of risk behaviours (not using condoms at last sexual intercourse) were more likely to provide urine samples, hence the prevalence observed in our study is likely to overestimate the true prevalence of C trachomatis in this population. Therefore, the prevalence of C trachomatis found in our study should not be generalized to the population aged 18-25 in Croatia due to greater willingness of those likely to be at higher risk to provide urine and the low response rate. Greater willingness of those with higher sexual risk behaviours to provide urine was also reported in the population-based survey in Great Britain .
Although findings were not statistically significant, in our study prevalence of C trachomatis was higher among those socially vulnerable, e.g., those who did not live with their parents up until the age of 18, unemployed and those with self-perceived lower socio-economic status. These findings provide strong evidence for the implementation of screening policies for C trachomatis in Croatia, and expanding access to STI prevention and control services to those socially and behaviourally vulnerable to this infection. Annual screening of all sexually active women younger than 25 years old is recommended in the United States, while in England, the National Chlamydia Screening Programme offers screening to men and women under the age of 25 years, either annually or when they change partners [21, 22]. Such guidelines or screening programmes for C trachomatis infection do not exist in Croatia.
Only a small proportion of those who gave urine called to obtain their test results. This is in contrast to the results of the bio-behavioural survey among men who have sex with men done at the University Hospital for Infectious Diseases "Dr. Fran Mihaljevic" in Zagreb in 2006 when 78% of respondents came to collect their test results . This might imply lower awareness among young adults of sexual health issues. Interviewers who also did urine specimen collection in our study reported that the vast majority of young adults whom they interviewed and asked to provide urine had low awareness of C trachomatis infection and were less likely to provide urine when parents were present at the household at the time of interviewing. They also reported that female and younger interviewers were more successful in obtaining urine specimens from respondents than their male and older counterparts.
In south-east Europe in particular, the absence of more comprehensive surveillance and over-reliance on routine case reporting contributes to perceptions that STIs are uncommon; this interferes with appropriate policy responses and presents a fundamental drawback in putting STI control higher on public health policy agendas. As mentioned earlier in the paper, C trachomatis case reporting rate in Croatia was 9.11/100,000 in 2007. Given the results of our survey, it appears that C trachomatis case reporting system substantially underestimates the burden of this infection. There is a pressing need to obtain a valid estimate of the burden of STIs, particularly curable bacterial STIs. Such estimates could lead to targeted responses, which would ideally be appropriate for and accessible to those at greatest risk, and likely driving disease transmission. The increased availability of nucleic acid amplification tests should be used to explore opportunistic screening and community-based testing in settings appropriate for young adults. In addition, such population-based surveys should be implemented in countries such as Croatia where STI case reporting systems substantially underestimates the true burden of disease.
In future surveys, efforts should be made to obtain a higher response rate. The bio-behavioural approach may be improved by adequate investments to ensure increased motivation of both interviewers and participants to provide samples and seek test results. In addition, efforts to increase knowledge about the consequences of untreated C trachomatis infection by carrying out targeted educational interventions might improve the uptake of home-based testing . Raising awareness of STIs among young adults is important in countries like Croatia where systematic education on sexual health in primary and secondary schools in lacking. Small financial incentives for providing specimens in bio-behavioural surveys, in this case urine, might also increase participation rate .