METHODS

The Well Person’s Health Check (WPHC) is described in detail elsewhere and was a cross sectional survey conducted between March 1998 and December 2000 in 26 rural and remote Indigenous communities in northern Queensland, Australia.⁴ Survey participants completed a comprehensive general health checkup that included a first void urine specimen which underwent polymerase chain reaction (PCR) testing (Roche Amplicor CT/NG, Branchburg NJ) for Chlamydia trachomatis (chlamydia) and Neisseria gonorrhoea (gonorrhoea).

RESULTS

Most infections were identified among people aged 15–24 years (table 1⇓). More than 90% (92% female and 96% male) of the infections detected in this survey were asymptomatic at the time of screening.⁸

In the unadjusted analysis (table 2⇓), younger age, female sex, alcohol consumption, smoking, a lower SEIFA CD score, the absence of a hospital, and the absence of a resident medical officer were significantly associated with the presence of an STI. The absence of a resident medical officer and hospital in the community were highly correlated (p<0.001).

In the adjusted analysis (table 2⇑) younger age, female sex, alcohol consumption, smoking, and a lower SEIFA score were significantly associated with the presence of an STI. The variables hospital and resident population were also included in the final regression model; however, these factors were not statistically significant.

A wave analysis was used to examine prevalence of infection by week of attendance to assess possible participation bias. Data from screens that ran for 3 weeks (three communities, n=499) and 4 weeks (two communities, n=490) were used. The prevalence of bacterial STI was calculated by week of attendance. A difference in prevalence for each week was observed in 3 week screens (week 1 = 9.9%, week 2 = 5.2%, week 3 = 19.7%; p=0.001) and 4 week screens (week 1 = 5.2%, week 2 =12.1%, week 3 = 7.7%, week 4 = 12.2%; p=0.118). For both 3 and 4 week screens prevalence was significantly less in the first compared to the final week (3 week p = 0.009, 4 week p = 0.049). The prevalence in the final week of both screens was greater than the overall prevalence (3 week 11.1% overall and 19.7% in final week, p = 0.012; 4 week 8.8% overall, and 12.2% in final week, p = 0.33).

DISCUSSION

This study is one of the first to analyse both individual and community factors associated with STI prevalence among individuals living in remote communities.

Young age, socioeconomic disadvantage, and the structure of locally available health services emerge as strong predictors of bacterial STI prevalence. These factors provide opportunities to reduce prevalence by focusing programmes on those individuals at highest risk, and by considering community factors associated with a higher STI prevalence. This study demonstrates that screening programmes should focus on younger individuals and that such programmes cannot rely on the presence of symptoms, as most infections detected were asymptomatic.

The association between STI and substance use has been noted in a number of studies and in Australian Indigenous communities petrol sniffing and alcohol abuse have been demonstrated as risk factors for STI.⁹ We found the majority of survey participants drank alcohol and smoked tobacco, behaviours which are highly socially patterned. In this environment, where STI prevalence is also high, associations between the use of these substances and STI would be expected.

A number of community factors were associated with a higher prevalence of STI.

Most communities in this study occupied the SEIFA quintile of greatest disadvantage and within this group of poor communities there exists a gradient of risk that increases with increasing disadvantage.

Improving the socioeconomic status of these communities is a highly complex problem requiring long term strategies to which health services may only make a small contribution. Other factors such as those reflecting reduced access to health care are more amenable to short term solutions. Ease of access to a hospital and doctor appeared to be associated with lower STI prevalence and may indicate community confidence in accessing a particular type of health service. Issues related to perceived skills of local practitioners, confidentiality, family connections, and how health is prioritised often result in delayed presentations by Indigenous people with known conditions.¹⁰ Consideration of these factors in the design of STI screening programmes is likely to improve access to health care and to reduce prevalence.

Possible participation bias is a limitation of this study. We used a wave analysis and assume later attendees, as a group, were less likely to access health services routinely, and more likely to have a disease prevalence that closest reflects non-attendees. A second comparative method used was an analysis of clinic PCR results over a 1 year period (2000–1) in 17 communities that participated in WPHC (1998–2000). The proportion of tests found positive through clinic testing was approximately twice that of WPHC testing (data not shown). The above estimates suggest that the prevalence reported here is probably an underestimate of the true prevalence of chlamydia and gonorrhoea in these communities.

Continuing high rates of pelvic inflammatory disease (PID) among north Queensland indigenous women¹¹ and the emerging HIV epidemic in nearby Papua New Guinea¹² highlight the need to reduce STI prevalence in these populations where HIV is currently uncommon.

The provision of accessible, acceptable primary healthcare services will help; however, solutions to the underlying social and economic issues that facilitate the relative disadvantage of remote Indigenous communities are also required if the differential between the health of Indigenous and non-Indigenous Australians is to be addressed.