Improving STD testing behavior among high-risk young adults by offering STD testing at a vocational school

The study population included students of the largest vocational school in the South of the Netherlands, covering four vocational branches (administration and business, health and well-being, technology and hotel and catering industry and tourism), attended by more than 9000 students mostly aged 18 to 20 years. Eighteen of 240 classes (n = 348 students) from three branches (health and well-being; hotel and catering industry and tourism; technology) were approached for participation.

All 348 students of the 18 classes, each including an average of 19 students, first received a 45-minute presentation by a trained nurse, concerning STD in general and CT and risk behavior in particular. Of these students 99% signed for informed consent, yielding 345 study participants. All participants were asked to fill in a questionnaire. Subjects were asked to submit a self-administered specimen for STD testing, after receiving instructions on how to take a self-administered vaginal swab (SVS) or first catch urine (FCU) (for men only). Prior studies showed that SVS were equivalent in sensitivity and reliability to traditional endocervical swab specimens and more sensitive and reliable than self-obtained urogenital specimen for the detection of CT and GO [16]. While the sensitivity of FCU may be somewhat lower than the sensitivity of SVS, data show that sensitivity may be no less than 87.7% [27]. This method is also widely used in both individual and population screening studies, and is now the CDC-recommended method of sampling for nucleic acid based testing [28]. Both SVS and FCU are used as standard STD clinic tests.

Submitting a specimen was voluntary. Students were free to take the test at home and return specimens the next day. To avoid peer pressure, all students were asked to return materials in the given coded envelope and drop it in a closed box installed at the school's bathrooms.

This study was approved by the Medical Ethical Committee of the University of Maastricht (MEC 07-3-089) and registered in the Dutch National Trial Register (NTR1410). Originally the study was conducted as a randomized controlled trial which was discontinued for lack of effect. Data from both the intervention (n = 186) and control group (n = 159) were pooled in current analyses since there was no intervention effect (OR: 0.731, p = 0.150), and intervention and control groups were largely similar, except that participants in the intervention group were somewhat older (41% versus 27% were 20 years or older, p < 0.01). In short, the intervention consisted of a tailored CT test advice which was the result of a personal score on a self-triage card. This card was provided before handing over the self-test materials but after students had had the sexual health education and had filled in the questionnaire. The questions on the triage-card, scoring method, and CT test advice was based on a Dutch population study which found that chlamydial infection was associated with high urbanization, young age, ethnicity (Surinamese/Antillean), low/intermediate education, multiple lifetime partners, a new sexual partner in the previous two months, no condom use at last sexual contact and symptoms of (post)coital bleeding in women and frequent urination in men [29]. A risk score based on an algorithm from these findings has the potential to guide individuals in their choice of participation when offered chlamydia screening [29]. Based on the score, the intervention group received a positive or negative CT test advice which was printed on the card. However, all students were free to submit a sample, even if they had a negative test advice.

The SVS and FCU samples were tested for the presence of CT and Neisseria gonorrhea (NG) by Nucleic Acid Amplification Tests (NAAT) with internal inhibition control: Strand Displacement Amplification assay (SDA) of Becton Dickinson Probe Tec ET system, Maryland, USA. Of note, none of the samples were positive for NG.

For feedback of testing results, standard STD center procedures were used, i.e., subjects were requested to provide a phone number or an e-mail address. Negative test results were returned in most cases by SMS or, in some cases, by e-mail. In case of a positive test result, STD nurses telephoned, offering treatment and partner notification. All CT-positive students were notified and were treated with one oral dose of 1000 mg azythromycin.

The questionnaire comprised the same items as the self-triage card supplemented with questions about partners, sexual behavior, drug use and sexual problems for adjustment of potential confounders. For participants who were sexually active a risk score was calculated from seven questions in the questionnaire. High risk was defined by an algorithm based on the association of chlamydial infection with high urbanization, young age, ethnicity (Surinamese/Antillean), low/intermediate education, multiple lifetime partners, a new sexual partner in the previous two months, no condom use at last sexual contact and symptoms of (post)coital bleeding in women and frequent urination in men. This dichotomous score (high versus low risk) has been shown to be predictive in discriminating CT-infected from CT-uninfected Dutch adults [29]. Predictive value based on a sum score of 6 was 3.5% [29].

Sexual mixing and concurrency of partners over the preceding six months were included as variables, based on self-reported information provided by the respondents. Partnerships that involve mixing behavior, as well as concurrent partners, have the potential to rapidly accelerate the spread of STD in a population [30]. Mixing by age was grouped into high and lower risk for STD; for women, high risk to acquire an STD was defined by at least one older partner, with a minimum age difference of one year; for men, high risk for STD transmission was defined by at least one younger partner, with a minimum age difference of one year [7, 31]. A variable concerning sexual orientation was included, based on self-reported same-sex activities (at least one reported same-sex partner) and opposite-sex activities. Partnerships were classified as concurrent if participants reported at least two partners simultaneously. Rates of reported sexual problems, risk behavior and CT testing were calculated and differences between men and women were assessed using chi-square testing. Sexual problems were categorized into three groups: problems with sex (problems with sex in general, problems with desire, arousal or orgasm, pain), problems related to teenage pregnancy (worried about being pregnant, use of morning-after pill, induced abortion, pregnancy test, pregnancy) and STD related symptoms (e.g. intermenstrual or post coital blood loss, abdominal pain, more genital discharge than normal).

Rate of CT positives was insufficient in our study to evaluate determinants of CT infection. Therefore, we considered high risk score, defined in accordance with findings from a population based screening, as a proxy for actual high risk. Determinants for high CT risk and CT testing were assessed using univariate and multivariate (using stepwise forward procedure) regression analyses. Determinants for high CT risk were assessed only among sexually active students because students who never had sex are not included in the prediction rule. Interaction terms were assessed between determinants and sex to explore differences between men and women. Since similar results were obtained from analyses adjusted and unadjusted for group (intervention and control), the results presented in this study are unadjusted. Prevalence of CT was calculated with 95 percent confidence intervals (95% CI).

A p-value < 0.05 was considered statistically significant. Analyses were performed with the SPSS package version 14.0 (SPSS, Inc., Chicago, USA).

Eighty five percent (n = 287) of the students reported they ever had sex; most had had two or more partners (n = 205). These participants had been sexually active for a median of 3 years (IQR: 3-5). Sixty five percent (n = 183) had not used a condom at last sexual contact. Men reported condom use more often than women (respectively 45% (n = 37) and 31% (n = 61), p < 0.05). A majority of the students (74%, n = 205) reported risky age mixing i.e. at least one partner older for women and at least one partner younger for men.

Eighteen percent (n = 51) had changed partners in the past two months. Men more often reported recent partner change than women (25% (n = 21) and 15% (n = 30), respectively, p < 0.05). Most students (91%) went to discotheques or bars (men 97%, women 89%, p < 0.05). Students frequently reported they had sex while going out, men more often with casual partners.

Of the sexually active men and women, 75% (n = 216) had high CT risk according to the predefined prediction rule (see methods) (Table 1). In both men and women, concurrency, mixing behavior by age, early sexual debut, drug use, sexual health problems, teenage pregnancy, STD-related symptoms and history of STD testing were univariately associated with having a high risk score (all p < 0.05). Early sexual debut, mixing behavior by age, being worried about teenage pregnancy and a history of STD testing were independently associated with having a high risk of CT (Table 2).

Of all participants, 53% (n = 183) submitted a self-administered CT test after this was offered to them at their school. Among sexually active students (n = 287), 61% (n = 176) had themselves tested for CT in the current study.

Of the sexually active participants, 22% (n = 64) had a history of STD testing; among the students with high CT risk (n = 216), this rate was 27% (n = 58) (Table 3).

Among sexually active students 36% (n = 102) were at high risk for CT, had no history of CT testing, but submitted a CT test in the current study. Thirteen percent (n = 37) were at high risk for CT, submitted a test in the current study, and also had a history of CT testing.

Ten percent (n = 29) were at low risk for CT, had no history of CT testing, but submitted a CT test in the current study. Two percent (n = 5) were at low risk for CT, submitted a test in the current study, and also had a history of CT testing.

Sexually active students with a high CT risk submitted a CT test more often (65%, n = 141) than those with a lower risk (49%, n = 35) (OR = 1.93, CI: 1.12-3.33, p < 0.05), but this was mainly true for women (OR = 2.41, CI: 1.35-4.29, p < 0.05) (Table 2). Risk score was not assessed as predictor for testing because all variables, including those comprised in the risk score, were assessed separately.

Independent determinants of CT testing among sexually active students in univariate analysis were female sex, no condom use during last sexual contact, reporting sexual health problems, teenage pregnancy, and STD related symptoms (Table 2). Following multivariate analysis, only STD related symptoms and no condom use during last sexual contact remained in the model (both p < 0.05).

Eight participants tested positive for CT; prevalence was 4.5% (n = 8/176) (C.I. 2.3%-8.7%) being 5.2% (n = 7) (C.I. 2.5%-10.3%) in sexually active women (n = 135) and 2.4% (n = 1) (C.I. 0.4%-12.6%) in sexually active men (n = 41) (p = 0.485). Number of CT infections was too low to perform risk factor analyses. All CT infections were found among those with high CT risk. In participants with high risk (n = 140), CT prevalence was 5.7% (n = 8) (C.I. 2.9%-10.9%), being 3.3% (n = 1) (C.I. 0.6%-16.7%) in men (n = 30) and 6.4% (n = 7) (C.I. 3.1%-12.6%) in women (n = 110).