Methods

We analysed data on the epidemiological risk profile, consultation time and STI diagnoses for all clients who attended the MSHC between October 2005 and October 2008. MSHC is the main public sexual health clinic in Victoria and provides over 25 000 consultations per year. Services are free of charge to the public and the clinic is centrally located. In 2008, MSHC diagnosed 7% of chlamydia cases, 21% of gonorrhoea cases and 27% of syphilis cases in the state of Victoria.

The centre operates a walk-in triage system in which patients are seen by practitioners in the order in which they present.7 MSHC utilises a relational database electronic medical record, the clinic practice management system, to collect demographic, clinical and epidemiological information routinely on all clients at each clinic visit.

To comply with Victorian law, FSW attending MSHC have monthly swab tests for gonorrhoea, chlamydia and trichomonas and have serology every 3 months for HIV and syphilis. MSM who attend the clinic are screened with a throat swab for gonorrhoea, a urine test for chlamydia, and an anal swab for chlamydia and gonorrhoea. Asymptomatic heterosexual men and non-sex-working women without specific risks such as overseas contacts are screened for chlamydia only. Testing for Mycoplasma genitalium is restricted to cases of urethritis and pelvic inflammatory disease.

We analysed our clinic database to determine the number of specific STI among all clinic attendees by risk group: FSW, MSM, heterosexual men and non-sex-working women. STI were defined as chlamydia, gonorrhoea, M genitalium, trichomonas, early syphilis and HIV.

The consultation time spent in each visit is collected automatically within the practice software. The time from the start to the end of the consultation was calculated for each risk group, and the rate of STI detection per 100 h of consultation time was calculated. The number of STI diagnosed over the study period was taken as the numerator.

Testing rates and 95% confidence intervals were calculated using the Poisson method. Analysis was conducted using SPSS version 17.

Results

There were 81 635 consultations during the 3-year period, of which 13 440 (16.6%) consultations were among FSW, of these, 7672 involved swab tests for gonorrhoea, chlamydia and trichomonas without serological testing (termed “swab testing only”). There were 20 471 (25%) consultations with MSM (table 1).

Of the total of 37 670 consultation hours during the 3-year period, 5722 (15.1%) were for FSW, and of these, 2896 h were used for swab testing only.

A total of 4208 STI (defined in the Methods section) were detected during the study period, and of these, 249 (6%) were among FSW; 133 of these 249 STI were detected during swab testing only. There were 1726 (41%) STI detected among MSM. Table 2 shows the particular STI detected in each group.

For every 100 h of consultation time, on average, four STI were detected in FSW, 19 in MSM and 12 in all clinic attendees other than FSW for swab-only testing.

We have estimated the number of extra cases of STI that could be potentially be detected at the centre if the 7672 consultations (ie, 2896 h) of FSW, in which swab-only testing was done were replaced by seeing other individuals. If this time was used to see clients other than FSW then 338 STI would have been detected compared with the 133 that were detected in FSW (ie, an extra 205; 95% CI 180 to 230). Similarly, if this time was used to see only MSM, then 550 STI would have been detected compared with the 133 detected (ie, an extra 417; 95% CI 386 to 448 STI). These extra cases of STI detection would have occurred at no additional cost to the centre.

Over a year these changes would result in the detection of an additional 40 cases of chlamydia, 14 cases of gonorrhoea, five cases of early syphilis, four cases of M genitalium, three cases of HIV and three cases of trichomonas if clients other than FSW replaced the FSW who attended for swab-only testing. Over a year, there would have been an additional 56 cases of chlamydia, 48 cases of gonorrhoea, 20 cases of early syphilis, five cases of M genitalium and 10 case of HIV if MSM replaced the FSW who attended for swab-only testing. These extra cases of STI detection would have occurred at no additional staff cost to the centre.

Discussion

Our centre could have potentially detected 205–417 more cases of STI (4.9–9.9% increase, respectively) over the 3 years depending on the risk profile of clients who gained access to the centre if FSW were tested 3-monthly rather than monthly. This benefit would have occurred without additional staff costs. Furthermore, the benefit we observed may well be felt in all health services providing testing for FSW, and over the entire state of Victoria the absolute cost savings could be very large. However, reducing STI testing in sex workers could result in a long duration of infection in these FSW, although it is unlikely that a significant number of transmissions would occur given the low incidence of STI including chlamydia8 in this group and the near 100% condom use.3 These data may assist in other countries considering the introduction of a regulated testing system. Such a system needs to balance the benefits of frequent screening on the impact it can have on access to sexual health services in high-risk individuals.

Our study has a number of limitations that need to be considered. First, our findings are from one sexual health clinic and may not be representative of the situation in other settings where services may have the potential to expand with increased demand. Second, the results differ depending on which risk groups replace the consultation time currently spent with FSW, so any programme change would need to ensure high-risk individuals accessed these additional consultation hours. Third, it represents findings relevant to sex workers working in a regulated system with low rates of STI, and may not necessarily apply to other settings particularly if the incidence of STI in FSW was high or condom use was low.5 It, nevertheless, provides useful real-life data that are relevant to other countries contemplating mandatory screening programmes if STI rates are low.

The main strength of our study is that it had a high degree of precision because of the large number of clients studied over a number of years. The increasing use of electronic medical records that record key epidemiological and clinical data in categorical data fields should allow similar analyses to be undertaken in different countries with different legislative structures.

We did not include male sex workers in our analysis. This is because relatively few of the sex workers were men and many (40%) of them had private male partners and were therefore part of the high-risk group of MSM. We do not collect the gender of the clients that sex workers have sex with, so it was not possible in our study to estimate the rates in male sex workers who provide services only to female clients. Our data suggest that any legislative change would need to address male sex workers who had sex with men differently.

To our knowledge, no other study has undertaken an analysis like this to estimate the potential public health harm associated with legislation that requires the frequent testing of sex workers with low rates of STI. It is important that governments considering introducing legislation regulating sex work consider the possible effects of frequent testing on the health services in the context of the local epidemiology. The intention to protect the public by mandating frequent testing of sex workers may have the opposite effect.