We found that opportunistically offering an online sexual risk assessment tool in general practice to young people did not significantly increase the proportion of young people tested for chlamydia, although a small increase in the proportion of 16 to 19 year old women tested was seen. While GPs generally had a favourable view of the tool, a number of barriers prevented them from referring young people to YCYR and consequently, less than a third of young patients were referred to the site, and of these, few accessed it.
The provision of onsite computers by practices did not seem to improve access by young people, and while no young people provided feedback about this, the location of the computers could have been a deterrent along with the fact that the only reason patients would access them was to use YCYR, hence potentially leaving young people feeling exposed or embarrassed. At a system level, practices may not be equipped to provide patients with a private area in which they can confidentially source online resources.
This study has some limitations. First, we do not know if the small increase in chlamydia testing among younger women was as a direct result of offering YCYR through general practice. In order to protect the confidentiality of patients accessing the website, it was not possible to track health service use of individual patients accessing the site, therefore we do not know how many of the patients who accessed the website actually returned to a GP for chlamydia testing. Nor can we determine which chlamydia tests were the result of the patient accessing the website and being exposed to the recommendations for chlamydia screening. It is possible that testing increased because of a greater awareness of chlamydia screening due to the practices' involvement in the study or because of exposure to the educational package provided to GPs as part of the intervention, rather than as a direct result of referral to YCYR. Secondly, although their views were sought, only a small number of YCYR users provided online feedback and none provided detailed interview feedback despite the incentive offered, so we do not know how most young patients felt about being referred to the website, its on-site access, use of the site or whether they followed up on the advice. Thirdly, as two of the practices were affiliated with universities and serviced high numbers of young people and international students, it is uncertain how generalizable the results are to other practice populations.
Reviews of interventions aimed at change in primary care show that while no intervention is effective in all circumstances, systematically developed, multi faceted interventions tailored to and engaging the target group and addressing barriers and facilitators to change are more likely to be effective [15–17]. A recent review by Ginige et al  specifically examining interventions aimed to increase chlamydia screening in primary care found that potentially effective strategies included enhancing GPs' communication skills, particularly around sexual history taking, and increasing GPs' knowledge and awareness around chlamydia, its associated complications, screening guidelines and non-invasive testing techniques. In addition, interventions that offer a combination of educational strategies, including interactive activities – as opposed to printed material only – are more likely to induce greater physician behavioural change .
It is possible that a more rigorous methodological approach in this study may have better addressed the existing barriers. In saying this, in a study by Merritt et al , despite the use of a simple screening protocol and extensively engaging and informing GPs in a multi-faceted, practice tailored intervention, increases in opportunistic chlamydia screening of young people were at most only moderate and were not sustained, with GPs reporting the same three barriers to testing.
While engaging and increasing GPs knowledge around chlamydia testing is likely to go some way towards facilitating chlamydia screening, it is unlikely to increase opportunistic screening to the levels required to have an impact on chlamydia transmission in the population. Enhancing GPs communication skills, particularly around sexual history taking, may also help, however, as previous research has shown, GPs often do not feel comfortable raising the issue of sexual health during unrelated consultations [8, 9] and are less likely to offer screening to patients who are asymptomatic or considered low risk .
Furthermore, GPs' perceptions that patient's are embarrassed discussing their sexual health is associated with a reduced likelihood of taking a sexual history and thus assessing a patient's risk . Recent work  examining women's attitudes to the introduction of chlamydia screening in general practice has shown that women did not want to be asked a sexual history when being asked to have a chlamydia test. Instead, they wanted to be offered testing based on age, rather than on GPs' assessment of their sexual risk. It could be argued that not taking a sexual history does not constitute best medical practice, however, given the high incidence of asymptomatic chlamydia cases, GPs' tendency towards screening only in high risk or symptomatic cases, and evidence that neither patients nor GPs feel comfortable with sexual history taking, further consideration needs to be given to routinely offering screening to young people according to their age rather than sexual risk. Testing is also more likely to be accepted by young women if it is normalised through wider community education campaigns that highlight the health benefits and destigmatise screening . The recent introduction of human papillomavirus (HPV) immunisation in Australian schools based on age rather than risk profile is an example of the success of this approach . Given the difficulties faced by GPs in remembering to refer patients for testing, further consideration should also be given to investigating whether alerts, programmed into electronic patient management systems, could prove successful in reminding GPs to screen patients presenting in the target age group. Following the success of the reimbursement programs for GPs to increase childhood immunisation rates in Australia , it may also be worth investigating whether a similar reimbursement program could prove equally as successful in increasing chlamydia screening rates.