Repeat chlamydia screening among adolescents: cohort study in a school-based programme in New Orleans
- 1Division of Clinical Epidemiology and Biostatistics, Institute of Social and Preventive Medicine, University of Bern, Bern, Switzerland
- 2Department of Family Medicine, Faculty of Medicine, McGill University, Montreal, Canada
- 3Department of Medicine, Louisiana State University Health Sciences Center, New Orleans, Louisiana, USA
- Correspondence to Professor Nicola Low, Department of Epidemiology and Public Health, Institute of Social and Preventive Medicine, University of Bern, Finkenhubelweg 11, CH-3012, Switzerland;
Contributors NL had the idea for the study. NL, MF and MJN designed the study. MJN and SNT provided access to the data. MF conducted the statistical analyses with support from MJN. All contributors helped interpret the data. NL drafted the manuscript. All contributors helped revise the manuscript.
- Accepted 11 May 2012
- Published Online First 6 July 2012
Objectives To describe uptake of chlamydia screening, determine rates of repeated yearly screening and investigate determinants of repeated participation in an organised school-based screening programme.
Methods The authors analysed data from 1995 to 2005 from female and male students in up to 13 schools in New Orleans, Louisiana, USA. The authors calculated proportions of students tested among all enrolled students and among those with parental consent and the percentage of positive chlamydia tests in each school year. The authors used random effects logistic regression to examine the effect of past screening history on subsequent participation.
Results 35 041 students were registered for at least one school year. Overall coverage was >30% in all school years. Among all students registered for 4 years, 10.6% (95% CI 9.3% to 12.0%) of women and 12.7% (95% CI 11.2% to 14.2%) of men had a test every year. Among students with parental consent for 4 years, 49.3% (95% CI 44.6% to 54.1%) of women and 59.3% (95% CI 54.5% to 64.0%) of men had a test every year. Among students registered for 2 or more years, those with a previous positive chlamydia test were less likely to have a subsequent test (female adjusted OR 0.77, 95% CI 0.67 to 0.88 and male adjusted OR 0.84, 95% CI 0.69 to 1.02). Chlamydia positivity increased over time.
Conclusions High levels of uptake can be achieved in school-based chlamydia screening programmes, but repeated yearly screening is difficult to sustain over time.
- chlamydia infections
- STD surveillance
- epidemiology (clinical)
Repeat episodes of chlamydia infection are common, particularly in adolescents.1 ,2 This suggests that chlamydia control recommendations and levels of uptake might be suboptimal. Yearly opportunistic screening for chlamydia is recommended as a preventive strategy in several countries3–5 to control chlamydia transmission and prevent pelvic inflammatory disease, which can result in ectopic pregnancy and tubal infertility.3 Information about the uptake of repeated yearly screening is, however, scarce. The levels required to reduce the prevalence of chlamydia and its complications remain unknown.6 Longitudinal analysis of health insurance claims from women aged 15–25 years in the USA found that, among women enrolled in the same health plan for 5 years, only 0.1% received a test for chlamydia in each year and the uptake of chlamydia testing was lowest in adolescent women.7
School-based programmes have the potential to reach a high proportion of sexually active adolescent women and men and can offer services, such as chlamydia screening, repeatedly over a school career.8 ,9 Participation in chlamydia screening has been reported from a variety of school-based programmes, both campus wide8–14 and for users of school-based health centres.14–17 Most reports have focused on a single school year,11 ,14–17 so issues that are important for understanding repeated student participation in annual screening have not been researched in detail. The objectives of this study were to describe the uptake of chlamydia screening and chlamydia positivity over time, determine rates of regular yearly chlamydia screening and to investigate determinants of repeated participation by young women and men during 10 years of the New Orleans school-based chlamydia screening programme.
Details of the programme have been reported previously.8–10 In brief, screening and treatment began in the 1995–1996 school year in three New Orleans public high schools with school-based health centres and was conducted until 2004–2005, involving up to 13 schools with and without health centres (figure 1). School registers were used to determine the number of students eligible for screening in each school and school year. Each student had a unique identification number, which allowed follow-up between schools and subsequent tests over time to be tracked. Data about self-reported sexual activity were collected from students who had accepted the test offer for two school years 2000–2001 and 2001–2002. Consent was obtained in writing or by telephone from parents/guardians of students under 18 years. Until 1999–2000, all students in schools with health centres whose parents had signed for their child to receive clinic services were considered to have consent to participate in the yearly chlamydia screening because sexually transmitted infection testing was among the services provided by the school-based health centres. From 2000 to 2001 onwards, all students had to provide parental consent each year. Students aged 18 years or older gave their own written consent. The screening programme was approved by the Louisiana State University Health Sciences Center Institutional Review Board and the school district.
Testing and treatment procedures were identical in all participating schools. Testing was carried out on set days, during which classes were escorted to the testing area and each student was counselled individually about the opportunity to be tested. Students who had consent were asked to provide a first void urine specimen of about 30 ml. Specimens were refrigerated and delivered to the laboratory on the same day for testing by nucleic acid amplification test. PCR (Amplicor Chlamydia Test; Roche Molecular Systems, Branchburg, New Jersey, USA) was used initially, followed by ligase chain reaction (Abbott Laboratories, Chicago, Illinois, USA) from 1995–1996 to 1999–2000; strand displacement amplification (Becton Dickinson ProbeTec, Sparks, Maryland, USA) was used from 2000–2001 to 2004–2005.
Students with positive chlamydia test results were considered infected and were counselled and treated by trained nurses or physicians. Treatment was with a single oral dose of 1.0 g azithromycin. Infected students were asked to attend the city sexually transmitted disease clinic for further examination including HIV testing. They were also asked to refer their sex partner(s) for treatment to the city sexually transmitted disease clinic or to the school nurse if enrolled in school. Retesting 3 months after a positive test was not part of the screening protocol. Students were assured of confidentiality throughout the screening process.
We included only students aged 14 years or older in the analyses. The primary outcome was being tested for chlamydia. We first computed sex-specific proportions (with 95% CIs) of students taking part in the screening programme. Overall chlamydia screening coverage in participating schools in each school year was calculated as the proportion tested among all those registered and also among those with parental consent. We then determined the proportion having repeated screening tests in each school year for which they were registered using the unique identification number to track students between schools, which allowed all subsequent tests to be counted. We assumed that students present on the school register at the time of the screening remained in the school for the whole school year. For the school years 2000–2001 and 2001–2002, we computed the proportions sexually active among those who were tested.
We examined the effect of previous screening history and chlamydia test result on subsequent participation in screening among students registered in school for 2 or more years. For each student at each screening opportunity, screening status (‘tested negative’, ‘tested positive’ or ‘not tested’) was included and used to predict participation in the next year that the student was registered in a participating school. We used a random effects logistic regression model (xtlogit command in Stata) to account for the correlation between repeated observations; the model included a random intercept for each student. The results were presented as ORs with 95% CI, which were adjusted for age. Age was included as a continuous variable. A linear trend was observed when plotting proportions tested in each age group and confirmed in a model that included each age group as a dummy variable. We repeated this analysis using data for the school years 2000–2001, 2001–2002 and 2002–2003 only to examine the effect of recent sexual activity, which was measured in 2000–2001 and 2001–2002, on participation in the following year.
Our secondary outcome was chlamydia positivity among students tested. We computed the proportions with a positive test among those tested for each school year. All analyses were conducted using Stata, V.10.0 (StataCorp LP).
A total of 35 974 students was registered at all participating schools for at least 1 year between 1995–1996 and 2004–2005. The final sample analysed included 35 041 students (17 130 women and 17 911 men), after excluding 224 records with missing data on age or sex and 709 records from students who were under 14 years throughout the study period. Almost all students (97%) were African–Americans. The median duration of enrolment in school was 2 years (IQR 1–3 years) for both women and men.
Between three and 13 schools took part in the programme each year (figure 1). Overall screening uptake in any single year was higher in the earliest years when students enrolled to receive services at school health centres were considered to have consent to participate in screening. In 1995–1996, participation rates in three schools with health centres were 51.4% among young women (86.1% had parental consent and 59.7% of these had a test) and 61.5% among young men (88.4% had consent, 69.6% had a test). In nine schools taking part in 2004–2005, when enrolment in school health centres was no longer used as a basis for consent, participation rates were 32.3% among women (40.8% had consent, 79.1% had a test) and 33.3% among men (40.3% had consent, 82.6% had a test) (figure 1). Levels of participation and parental consent, by school year, are shown in the supplementary online figure.
Chlamydia positivity was higher in women than men in all years (figure 1). After declining slightly (in women) or remaining stable (in men) until 1998–1999, chlamydia positivity increased and was then stable in both sexes from school year 1999–2000 onwards.
Repeated participation in screening
Among all 35 041 students, 19 826 were registered in school for two or more consecutive or non-consecutive years. The proportion of students ever tested increased with the number of years registered in school (table 1). Among students registered for 4 years, over 80% of both women and men were tested for chlamydia at least once. The proportion of students tested at every screening opportunity decreased as the number of years they were registered in school increased. For students registered for only 1 year, 38.6% (95% CI 37.4% to 39.7%) of all women and 37.7% (95% CI 36.6% to 38.8%) of all men were screened. Among students registered for 4 school years, 10.6% (95% CI 9.3% to 12.0%) of women and 12.7% (95% CI 11.2% to 14.2%) of men were screened in each year.
There were 23 382/35 041 students with consent to participate in at least one school year (table 2). Among those with consent for 1 year, 74.5% (95% CI 73.5% to 75.6%) of women and 77.7% (95% CI 76.7% to 78.7%) of men were tested. For students who had consent for 4 years, 49.3% (95% CI 44.6% to 54.1%) of women and 59.3% (95% CI 54.5% to 64.0%) of men had a test in each year.
Determinants of repeated participation in screening
Among 19 826 students registered in participating schools for 2 or more years, women who had a positive chlamydia test at a previous screening round were less likely to be tested subsequently than those whose previous result was negative (age-adjusted OR 0.77, 95% CI 0.67 to 0.88) (table 3). There was a similar but weaker tendency in men (age-adjusted OR 0.84, 95% CI 0.69 to 1.02). Both women and men who did not participate in screening at a previous opportunity were less likely to be tested subsequently than those who had been screened. Among students with a previous positive chlamydia test result, the odds of obtaining parental consent for participation in the following year was lower for women (age-adjusted OR 0.81, 95% CI 0.70 to 0.92) but not for men (age-adjusted OR 0.97, 95% CI 0.83 to 1.14). Among students with a previous positive chlamydia test result and with a subsequent parental consent, the odds ratio of subsequent screening participation was lower for women (age-adjusted OR 0.72, 95% CI 0.60 to 0.88) but not men (age-adjusted OR 0.96, 95% CI 0.74 to 1.23).
Students who were registered in schools for consecutive years were more likely than those with non-consecutive registration to be tested more than once (table 3).
Among 3332 students who were registered in school years when data about sexual activity were collected, students with one or more partners in the previous 3 months were more likely than students who had never had sex or who did not have sex recently to participate in subsequent screening rounds (table 3).
Over 10 years of a school-based chlamydia screening programme, most students were tested at least once during their school career. Repeated participation in chlamydia screening decreased as the number of screening opportunities increased. Similar proportions of female and male students were tested repeatedly, both overall and among those with parental consent. There was some evidence to suggest that students with a positive chlamydia test result could be less likely to participate in subsequent screening rounds than those with a negative test, particularly among women.
The main strength of this study is the well-organised long-term nature of the screening programme.8–10 The use of school registers allowed the numbers of potentially eligible students in each school year to be ascertained, together with records of parental consent, screening and chlamydia test status at each screening opportunity. There are also weaknesses in the study. First, the total number of students was large but the numbers enrolled for multiple years was lower, so the precision of estimates of repeated participation in screening was limited. Nevertheless, the patterns of repeated participation over time and by sex are likely to be valid. We did not examine the effect of differences between schools on repeated testing because there were too few schools to be able to include them as a third-level variable in the regression model. Second, information about sexual activity was not available for all screening rounds and was only asked of students who were tested, so screening rates among sexually active individuals may be over or underestimated and we could not examine behavioural influences on repeat chlamydia testing in more detail. Third, we did not have complete information on the reasons for not being tested. A survey of reasons for not being tested was done in 2004–2005. Among 156 students surveyed, the most common reason for non-participation by men was simply not wanting to be tested and, for women, either not wanting to be tested or never having had sex.18 Further qualitative research about this would be useful to inform programme implementation. Finally, while the last year of data collection was in 2005,19 the patterns of uptake observed over the 10-year study period continue to provide useful data about school-based chlamydia screening programmes.
We believe that the New Orleans programme provides the only information about repeated participation in school-based chlamydia screening interventions8 and that this is the first report to examine determinants of repeated screening. There are other large school-based chlamydia screening programmes.11–13 In Philadelphia, chlamydia testing has been offered during educational sessions to groups of 9th–12th grade students since 2002–2003 in 69 of 71 schools.11 ,12 Participation in the first year was similar to that in New Orleans,11 but levels of repeated uptake have not been reported. In New York City, school-based chlamydia testing has been offered in areas with high reported chlamydia rates since 2006, reaching 110 of 405 public high schools in 2008–2009.13 Participation rates as a proportion of enrolled students have not yet been published. Chlamydia screening has also been offered in school-based health centres.14–17 In California in 2008, 19 educational institutions took part in an initiative to improve rates of chlamydia screening.16 Overall, 89% of sexually active young women accessing reproductive health services in school received a chlamydia test. Among young men, when chlamydia screening was offered opportunistically to those attending school-based health centres from 1999 to 2003, 56% of men in Baltimore and 17% in Denver accepted the offer.17 Participation in subsequent years has not been reported from these settings, to our knowledge.
The findings of this study suggest that the relationship between repeated uptake and impact in chlamydia screening programmes is complex. Repeated participation in chlamydia screening in the New Orleans school-based programme, which actively offered testing every year, was high in comparison with estimates from women receiving opportunistic screening in US commercial health insurance plans (<1% of women enrolled in their plan for 5 years).7 In participating schools, yearly repeat participation was >10% among all students enrolled for 4 years or more. The higher odds of subsequent participation among students with more recent sexual partners show that the programme reached students with higher levels of sexual risk behaviour. Over time, however, the percentage of students with positive chlamydia tests increased and overall participation fell, although it was >30% in all years. Additional analyses have shown high chlamydia incidence and repeat infection rates in these students.20 ,21 There are insufficient data to determine whether the behavioural risk profile of those being screened increased over time. Cohen19 has discussed in detail the complex mix of organisational, behavioural, economic and political reasons for this apparent programme failure. Chlamydia screening coverage levels of 30%–60% per year would be expected to result in a reduction in prevalence, based on the findings of several mathematical modelling studies.22–24 Fisman et al 24 found that declining participation over time, increasing risk behaviour and immunity following natural clearance of infection all resulted in an increase in chlamydia prevalence over time in their model of a school-based chlamydia screening programme. Part of the explanation is also likely to be that, unlike hypothetical model populations, students are not closed populations. The data from this study could help provide new insights in future modelling studies about the effects on chlamydia prevalence of different levels of sexual mixing between school students and the general population.
The results of this study have implications for future research and practice. The effects of positive test results on subsequent chlamydia testing merit further investigation. There was some evidence in this study that participation at the next screening round might be lower in those with a positive chlamydia test, particularly for women. This could not be explained entirely by withdrawal of parental consent among women with a previous positive chlamydia test because the finding was also present in women who had parental consent following a positive test. It is possible that those with positive screening tests had subsequent tests outside the school-based programme, but there is no information available to confirm or refute this. It is also plausible that the negative emotional effects of a positive chlamydia test result, particularly among young women,25 ,26 might put off young people from future participation in screening. Both qualitative and quantitative studies could help to investigate this issue further. Whether parental consent is collected actively or passively has implications for practice. The fall in yearly screening coverage over time was the result of a reduction in the percentage of students having parental consent rather than the percentage of students with consent who agreed to take part (supplementary online figure). It will be useful to compare patterns of repeated participation with data from school-based programmes that use opt-out systems for consent.13 In summary, the results of this study show that uptake of chlamydia screening of >30% per year can be achieved in school-based screening programmes, but repeated yearly screening is difficult to sustain over time.
In a school-based chlamydia screening programme, overall coverage remained >30% for 10 years.
The proportion of students screened at every yearly opportunity fell as the number of years they were registered in school increased.
The finding that students with a previous positive chlamydia test result were less likely than those with a previous negative test result to participate in subsequent screening rounds merits further investigation.
Repeated yearly chlamydia screening is difficult to sustain over time.
Part of this study was presented as a poster at the 3rd BASHH/ASTDA Joint Spring Meeting, New York, USA, 7–10 May 2008.
Correction notice This article has been corrected since it was published online first. The section ‘Methods’ has been updated to read ‘Statistical analysis’.
Competing interests In 2010, NL received fees from GlaxoSmithKline to attend a meeting about chlamydia vaccines.
Ethics approval Ethics approval was provided by Louisiana State University Health Sciences Center Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.