Article Text
Abstract
Objectives Australia introduced a nationwide human papillomavirus (HPV) vaccination programme for women aged 12–26 years in 2007 and has implemented various chlamydia control strategies over the last 10 years. We compared reported diagnoses of warts and chlamydia in two national population-based samples surveyed 10 years apart.
Methods An Australia-wide survey of women aged 18–39 years was conducted by random-digit dialling mobile phone numbers in 2011. The proportion of women self-reporting genital warts and chlamydia was compared with that in equivalent-aged women interviewed in another national telephone survey conducted 10 years earlier using logistic regression adjusting for age, and other factors.
Results 2394 women were included in the 2011 survey. 88 (3.7%) reported a genital warts diagnosis, and 127 (5.3%) had at least one chlamydia diagnosis. For women with warts and chlamydia, the majority were seen in general practice (63.3% and 70.0%, respectively). Comparing women surveyed in 2011 to 4874 women surveyed in 2001, among women eligible for free HPV vaccine through the national programme (women aged 18–30 years), there was a 41% decrease in warts (adjusted OR 0.59, 95% CI 0.39 to 0.89) but among ineligible women (31–39 years) there was a 64% increase (aOR 1.64, 95% CI 1.05 to 2.54). Over the same period chlamydia diagnoses increased by 97% (aOR 1.97, 95% CI 1.15 to 3.37) in women aged 18–24 years.
Conclusions Our results add to evidence from clinic-based studies showing a significant reduction in genital warts among HPV vaccine-eligible women in the general community and increases in the proportion of Australian women diagnosed with chlamydia.
- Genital warts
- HPV vaccine
- chlamydia
- surveys
Statistics from Altmetric.com
Background
Genital warts and chlamydia are two of the most commonly diagnosed sexually transmissible conditions in Australia.1 ,2 They are also conditions which have been the target, over the last 10 years, of significant national investment in prevention strategies. For genital warts, Australia was the first country to institute a funded national human papillomavirus (HPV) vaccination programme. While the primary aim of this programme at introduction was to prevent cervical cancer caused by HPV, the use of the quadrivalent vaccine which protects against wart-causing HPV types 6 and 11, as well as cancer-causing types 16 and 18, has meant that the vaccine offers high levels of protection against genital warts. The HPV vaccine programme commenced in April 2007 with routine school-based vaccination of all girls aged between 12 and 13 years, and a catch-up programme between 2007 and 2009 to all girls and women aged 12–26 years, with those aged up to 18 years mostly vaccinated through schools and the rest through primary healthcare clinics, largely general practice. In 2013, the programme was extended to 12–13-year-old boys with a 2-year catch up for 14–15-year-old males.
For chlamydia, in the 10 years from 2001 to 2011, a number of public information campaigns and strategies have been implemented to increase awareness, testing and treatment. In 2005, the first national Sexually Transmissible Infections Strategy was launched with chlamydia control featuring as a priority.3 New initiatives included a chlamydia screening study4; enhanced surveillance strategies for chlamydia5 and a number of health promotion programmes to promote opportunistic testing and treatment of chlamydia in primary care settings. Over this period, national sexual health guidelines were also amended to include the recommendation that all sexually active adults aged less than 25 years be offered annual chlamydia testing.6
Evaluation of the impact of these control strategies has relied on routinely collected clinical and infectious disease reporting data. For chlamydia, national surveillance data, based in most jurisdictions on laboratory notifications, have demonstrated substantial increases in diagnoses across most age groups, but particularly among those aged 15–29 years.1 Although mostly due to increased testing, assessments among sexual health clinic attenders suggest that there is a real, albeit small, component due to increases in prevalence.7 For genital warts, evaluation has relied on a network of sexual health clinics which have reported large declines in first diagnoses of genital warts in the cohorts of young women eligible for free HPV vaccination.8 ,9
Recognising the need for alternative sources of data for evaluation of the HPV vaccination programme impact, as part of a larger survey on reproductive health in a general population sample of young women,10 we undertook to ask about genital warts diagnoses. The goal of this study was to compare data collected after the programme commenced to that from a broader survey which had been undertaken prior to the programme. Secondary aims of our survey were to examine changes in self-reported history of chlamydia diagnoses over the same period and to examine characteristics of women reporting either genital warts or chlamydia.
Methods
This study was a repeat cross-sectional analysis, of data from two population-based telephone surveys conducted 10 years apart in 2001 and 2011. The 2011 survey was conducted by the authors, and aimed to recruit a representative sample of approximately 2000 Australian women aged between 18 and 39 years, and interview them to estimate the proportion reporting genital warts or chlamydia infection, and other reproductive health factors. Following a pilot study of 268 women,11 which demonstrated that a much higher proportion of women in the target age range were contactable by mobile rather than fixed line telephones, the survey was conducted by random-digit dialling mobile telephone numbers. More detailed information on the methods are outlined elsewhere.10 Briefly, sampling was stratified by age, to recruit women aged 21–30 years and 31–39 years in a ratio of 3 : 1, and by state or territory of residence, to recruit women in proportion to the geographic distribution of young women in Australia. Eligible respondents who agreed to participate completed a brief interviewer-administered questionnaire that included sociodemographic information and questions on reproductive health, including HPV vaccination, Pap testing, contraception and pregnancy. They were also asked whether they had ever had genital warts, the age at first diagnosis, and where they were treated. For chlamydia, they were asked if they had ever been tested, where they were tested, if they had ever had a positive test, and the age when they first tested positive (see online supplementary appendix for full questionnaire).
The earlier survey was the Australian Study of Health and Relationships, conducted mostly in 2001. Detailed information on its methods have been published previously.12 In brief, about 19 000 men and women aged 16–59 years were surveyed using random-digit dialling of fixed-line telephones across Australia. Sampling was stratified by sex and 15 geographic regions. The response rate was 77.6% for female participants. Participants completed a comprehensive interviewer-administered questionnaire on their sociodemographic characteristics and sexual behaviours. Participants were also asked if they had ever received a diagnosis of one or more sexually transmissible infections drawn from a list that included genital warts and chlamydia. For each positive response participants were asked if they had had the condition in the last 12 months and where they were treated. Line-listed records from this survey are available for analysis from the Australian Social Science Archives,13 and were obtained by the authors to provide a basis for comparison with the 2011 survey.
Both surveys used similar protocols for establishing contact with participants, that is, a standardised script administered by trained interviewers, and a Computer Assisted Telephone Interview (CATI) programme. The 2011 survey and the comparison with historical data was approved by the University of New South Wales Human Research Ethics Committee (ref HREC 10 366).
Analyses
Data from the 2011 survey combined participants in the pilot study who were contacted by mobile phone (n=128),11 with the participants in the main study (who were all contacted by mobile phone). The proportion responding to the 2011 survey was calculated based on the number of eligible contacted women who agreed to take part, and details of how response rates were estimated are outlined in the description of the pilot study.11 Sociodemographic characteristics of participants were examined overall and then in groups of women classified by whether or not they reported ever having a diagnosis of genital warts or chlamydia.
To compare between the two surveys, data from the 2001 survey were restricted to women in the same age group (18–39 years) as the 2011 survey. We calculated age at the time of survey by subtracting the reported year of birth from 2001. We first compared sociodemographic characteristics between both surveys using unweighted and weighted proportions and Pearson χ2 tests. For the 2001 survey, weights were calculated as the inverse to the probability of selection and according to the 2001 estimated residential Australian population. Full details are described elsewhere.12 The 2011 survey weights were estimated as the inverse to the probability of selection, based on the number of mobile phone numbers the participant could be contacted on, with more mobile phones resulting in a lower weight. The 2011 estimates were additionally weighted according to the 2009 census estimates for the Australian population.
We made comparisons between the two survey populations of the proportion who self-reported ever having genital warts and chlamydia. The analyses were conducted overall and then stratified into age groups. For warts, the age groups were 18–30 years and 31–39 years, as this distinguished women in the 2011 survey who would or would not have been eligible to have received free HPV vaccine through the national programme, respectively. For chlamydia, the age groups were 18–24 years and 25–39 years to distinguish those in the recommended chlamydia screening age from those who were not. We used logistic regression to compare the proportion with a self-reported history of genital warts between the 2011 and 2001 populations. We then conducted similar comparisons for chlamydia. Analyses were adjusted for age (as a continuous variable), place of residence (the eight states and territories of Australia), country of birth (Australia vs overseas), Aboriginal or Torres Strait Islander status, education level (two groups based on whether a woman had completed education up to the highest level possible at her particular age or not). Analyses for warts were also adjusted for self-reporting of chlamydia and vice versa. We adjusted the warts analysis for chlamydia and vice versa as they are both sexually transmissible infections and potential confounders of each other in the analyses.
We conducted sensitivity analyses restricting analyses to women classified as sexually active (based on a response to the 2001 survey question asking if the woman had had intercourse, and a response in the 2011 survey to the question on whether a woman had ever had a Pap smear); and restricting the women in the 2011 survey to those who indicated they were resident in Australia and eligible for Medicare during the 2007–2009 HPV vaccination campaign. Finally, as a small but unknown proportion of women in the 2001 survey completed the survey in 2002, we reran our analyses calculating age in this survey sample using an interview year of 2002. All analyses were conducted using STATA V.12.1.
Results
Overall findings from the 2011 survey
Excluding those who did not provide exact age data (n=3), a total of 2394 women completed the 2011 survey. The overall response rate among eligible contacted women was 80.0%. Of the 2363 responding to the question on genital warts, 88 (3.7%) women had ever had a diagnosis. For chlamydia, 2244 responded to the question, with 975 (43.4%) reporting being tested at least once for chlamydia, 475 (21.1%) being tested in the last year and 127 reporting at least one positive test (13.0% of those tested for chlamydia; 5.3% of the entire study population). There were 11 women who reported both warts and chlamydia and 2144 (89.6%) reported neither. Additionally, 31 women reported their warts status as unknown or did not answer the question, and 21 women responded in this manner regarding chlamydia.
Women reported having a first diagnosis of warts and chlamydia at similar mean ages, 21.1 years and 21.4 years, respectively (see table 1). Of the 89.8% (n=79) with warts who specified where they were treated, the majority (n=50; 63.3% of total responding) saw a general practitioner, 15.2% were treated in hospital, 12.7% were treated in sexual health clinics, 5.1% went to a specialist and 3.7% went to a mix of providers. Of the 975 women tested for chlamydia, 96.2% (n=938) reported where they were tested, with 77.0% seen by a general practitioner, 6.5% in a sexual health clinic, 6.2% in hospital, 4.2% in a family planning clinic, 2.2% by a specialist and 3.9% in other settings.
Compared to women reporting neither warts nor chlamydia, those reporting having a diagnosis of warts were older at the time of survey, less likely to have completed education (up to the highest level possible for their age), more likely to report having had a Pap smear, and more likely to have been tested for chlamydia. More women with neither warts nor chlamydia reported having the HPV vaccine than women who reported warts, but this was not significantly different (55.6% vs 46.5%; p=0.1). Compared to women reporting neither warts nor chlamydia, those reporting having a chlamydia diagnosis were substantially more likely to be Australian-born, less likely to have completed education, and more likely to report having Pap smears (table 1).
Findings comparing the 2001 survey to the 2011 survey
The 2001 survey recruited 4874 women in the age range 18–39 years. The sociodemographic characteristics of women reported in both surveys are shown in table 2 with both unweighted and weighted proportions. The age distribution differed substantially between surveys due to the different sampling strategies. The 2001 survey had more women in the older age groups than the 2011 survey, and there were also some differences in the proportion recruited in each state between the survey populations. The 2011 survey population had a higher proportion of women who were of Aboriginal and Torres Strait Islander origin, even after weighting.
Table 3 shows the ORs of reporting ever having a diagnosis of genital warts or chlamydia comparing the women surveyed in 2011 with 2001 overall and stratified by age. ORs were adjusted for age, country of birth, state of residence, education, Aboriginality; ORs for warts were additionally adjusted for chlamydia while ORs for chlamydia were adjusted for warts. Both unweighted and weighted estimates are shown, but findings were similar between the two methods. Overall, there was no difference in reporting of genital warts between surveys (2011 vs 2001 survey, weighted adjusted OR 1.10, 95% CI 0.78 to 1.54), however, significant differences were seen when examined by age strata. For women aged 18–30 years in 2011, that is, those who had been eligible for the free HPV vaccine in 2007–2009, there was about a 40% lower probability of reporting warts (adjusted OR 0.59, 95% CI 0.39 to 0.89). However, for women aged 31–39 years in 2011 who would not have been eligible for the free HPV vaccine, the women surveyed in 2011 were more likely than those surveyed in 2001 to have reported a past diagnosis of genital warts (adjusted OR 1.64, 95% CI 1.05 to 2.54).
For reporting of chlamydia between 2011 and 2001 across the entire population, there was a significant increase of about 41% (adjusted OR 1.41, 95% CI 1.07 to 1.87). When stratified by age groups, there was an indication that for women aged 18–24 years, the difference was greater than among women aged 25–39 years, however, this was not significant (adjusted OR 1.97, 95% CI 1.15 to 3.37 vs adjusted OR 1.30, 95% CI 0.93 to 1.82; p value for heterogeneity 0.1). In the sensitivity analyses, we found that these relationships were consistent.
Discussion
By comparing data collected in two population-based surveys of young women in Australia conducted 10 years apart, we have identified a 40% lower prevalence in ever having genital warts among women aged 18–30 years. This finding is the first in a representative general population sample of Australian women, of a reduction in warts since the introduction of the large-scale HPV vaccination programme, and it confirms earlier reports from sentinel sexual health clinic populations.8 Conversely, there was a significantly higher level of reporting a diagnosis of chlamydia over the 10-year period. The strengths of our study include that it is the first comparison of national, general population data on genital warts and chlamydia with almost identical questions used between surveys. Given that our data and that of others show the majority of Australian women diagnosed with genital warts or chlamydia are seen in general practice,14 our results make an independent contribution to the evaluation of the HPV programmes’ impact, as the assessment of prevalence of warts has so far only come from more selected study populations, such as sexual health clinics, where only a small proportion of women attend for diagnosis and treatment. There were some differences in the sampling strategies between the 2011 and 2001 surveys including the age stratification and the use of mobile phones rather than fixed lines in the 2011 survey. However, for the purposes of comparing data, we used similar age participants and also adjusted for characteristics that were reported in both surveys.
The reduction in the proportion of women reporting genital warts was less than the 72.6–92.6% decline in vaccine-eligible women attending Australian sexual health clinics.8 However, as we measured a history of ever having warts rather than occurrence of first presentations, our actual estimates of change over time are not directly comparable with these data. Also, as the average age at first diagnosis of warts in this population was 21 years, a substantial number of the population surveyed, aged 18–30 years in 2011, would have already had warts before the vaccine was available. In women aged 31–39 years, there was a suggestion of a net increase in those reporting genital warts over the 10 years. The recent sexual health clinic data show, among the older vaccine-ineligible women, stable incidence rates over the last 5 years.8 ,9 However, other data, from Canada and Denmark suggest that prevalence and incidence of genital warts in unvaccinated populations may have been increasing over the last decade.15 ,16
Our observation of an increase in the proportion of women reporting a history of chlamydia diagnoses, with somewhat greater increases in the women aged <25 years versus those 25 years and older is consistent with national surveillance data, although the close to 100% increase we found in 18–24-year-olds was smaller in magnitude than the increase of more than 300% in chlamydia diagnoses in similar aged women using laboratory notifications over the same period.1 Reasons for the smaller estimated change over 10 years in our study compared to that in national notification data may be because we measured a history of ever having a diagnosis rather than notifications, and the possibility that the proportion of notifications arising from repeat chlamydia diagnoses in the same individual may be increasing over time.17 Also our estimates had wide CIs. For those aged 18–24 years, the 95% CI for the increase ranged from 15% to 338%.
With respect to chlamydia testing, we found that 43.4% of women aged 18–39 years had been tested at least once in their life for chlamydia, and 21.1% in the last year. There are no routine national surveillance data available to estimate chlamydia testing coverage. However, a recent population-wide Australian study of chlamydia testing based on national health insurance (Medicare) reimbursements for chlamydia tests estimated that in 2007/2008, among 16–29-year-old sexually active women, the proportion who had a test each year was 12.5%.18 Higher testing reported among those participating the 2011 survey and the estimate from the study using 2007/2008 data may be because only Medicare reimbursement data was used, and this does not include chlamydia testing occurring in sexual health clinics or hospitals. Other explanations include that testing rates in 2011 have truly increased compared to 2007/2008, or that the 2011 survey participants were more health conscious and, hence, more likely to have been tested than in the study based on Medicare reimbursements. The 2001 survey did not ask about the history of testing for chlamydia, which would have assisted in interpreting the changes in chlamydia prevalence between two study populations, and it is, therefore, difficult to extrapolate from our data what changes over the 10 years may have occurred in overall chlamydia burden.
Differences in selection of study participants between surveys may have affected our results. We had limited sociodemographic variables to compare study populations. It is possible that the higher testing rates but lower than expected rise in women reporting a chlamydia diagnosis compared to national surveillance data may indicate that the 2011 survey population is more health conscious and has less sexual risk behaviours than the 2001 population, and this may also explain the decrease in women reporting warts. However, the increase between the 2001 and 2011 surveys in warts diagnoses among 31–39-year-old women suggests that our observed decrease in warts in the 18–30-year-olds was not attributable to such selection biases alone. Another limitation in understanding what might have contributed to the observed differences in proportions of the population diagnosed with warts and chlamydia between surveys was the absence of any comparable data in the 10 years between surveys. We were only able to estimate the net difference in reporting after 10 years, and not any trends occurring over that time interval. Therefore, it is not possible to attribute the reduction in genital warts among vaccine-eligible women between 2001 and 2011 directly to the introduction of the female HPV vaccine programme in 2007, as we do not have data for that specific time period. Similarly, we cannot say whether the increased reporting of warts in older women, or the increases in chlamydia, occurred gradually over the 10 years, or if there were rapid fluctuations that coincided with particular public health interventions at particular points in time. The ecological nature of our study means that we can only hypothesise about the potential contribution of other factors that could affect changes in the reporting of sexually transmissible infections over the 10-year interval, such as increases in the numbers of sexual partners among young people19 or more awareness of particular sexually transmissible infections. For example, as we relied on self-report, it is possible that greater awareness of genital warts, through the HPV vaccination campaign, may have resulted in more women in the 2011 survey reporting warts than in 2001. This may explain both the increase in the proportion of genital warts observed among the 31–39-year-old women, and the reductions among the 18–30-year-old women being less than that described in Australian sexual health clinics.
In summary, our results are the first general population-based data to show that comparing populations surveyed from 2011 to 2001, reporting of genital warts by women aged 18–30 years has decreased. The reporting of warts may have increased in women aged 31–39 years, and the reporting of chlamydia has increased among women aged 18–30 years and 31–39 years. The findings are consistent with a significant and enduring effect of the HPV vaccine programme on disease in the general community.
Key messages
-
Between 2001 and 2011, reporting of genital warts has decreased among community-dwelling women eligible for the human papillomavirus vaccine programme.
-
Over the same time period, reported chlamydia diagnoses have increased among all women.
-
The trends in genital warts and chlamydia reported in studies in sexual health clinics are also being observed in the general population.
Acknowledgments
The Hunter Valley Research Foundation conducted the survey for the researchers. We thank the women who took part in the 2011 survey, the Australian Social Science Archives for access to the 2001 data, and Anthony Smith for guidance on use of the weights for the 2001 data.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Footnotes
-
Handling editor Jackie A Cassell
-
Contributors All authors conceived the study, oversaw the conduct of the study, and commented on analyses and drafts of the manuscript. BL wrote the first draft and conducted the analyses.
-
Funding This study was funded by the Australian National Health and Medical Research Council (NHMRC) grant number 568971 and the Victorian Cytology Service. BL, BD and JMK are supported by NHMRC fellowships.
-
Competing interests BL holds shares in CSL Limited, distributor of the HPV vaccine in Australia; BD receives genital wart surveillance funding from the Australian Government Department of Health and Ageing and bioCSL Pty Ltd, and has received support as a speaker from Merck and SPMSD; JMLB and MS were investigators on an ARC linkage grant in which CSL Limited was a partner organisation.
-
Ethics approval University of New South Wales Human Research Ethics Committee.
-
Provenance and peer review Not commissioned; externally peer reviewed.