Abstract
Objective To measure the coverage and uptake of systematic postal screening for genital Chlamydia trachomatis and the prevalence of infection in the general population in the United Kingdom. To investigate factors associated with these measures.
Design Cross sectional survey of people randomly selected from general practice registers. Invitation to provide a specimen collected at home.
Setting England.
Participants 19 773 men and women aged 16-39 years invited to participate in screening.
Main outcome measures Coverage and uptake of screening; prevalence of chlamydia.
Results Coverage of chlamydia screening was 73% and was lower in areas with a higher proportion of non-white residents. Uptake in 16-24 year olds was 31.5% and was lower in men, younger adults, and practices in disadvantaged areas. Overall prevalence of chlamydia was 2.8% (95%confidence interval 2.2% to 3.4%) in men and 3.6% (3.1% to 4.9%) in women, but it was higher in people younger than 25 years (men 5.1%; 4.0% to 6.3%; women 6.2%; 5.2% to 7.8%). Prevalence was higher in the subgroup of younger women who were harder to engage in screening. The strongest determinant of chlamydial infection was having one or more new sexual partners in the past year.
Conclusions Postal chlamydia screening was feasible, but coverage was incomplete and uptake was modest. Lower coverage of postal screening in areas with more non-white residents along with poorer uptake in more deprived areas and among women at higher risk of infection could mean that screening leads to wider inequalities in sexual health.
Introduction
Randomised trials show that systematic screening for genital Chlamydia trachomatis might reduce the incidence of pelvic inflammatory disease by about 50%.1 We investigated the coverage and uptake of systematic chlamydia screening by post, estimated prevalence of chlamydia, and explored factors associated with these measures.
Methods
The chlamydia screening studies (ClaSS) project has been described previously,2 and the full protocol is available at www.chlamydia.ac.uk.
We invited men and women aged 16-39, randomly selected from 27 general practices in the West Midlands and Avon, to collect their own specimens (urine in men, urine and vulvovaginal swab in women) and post them to us, and to complete a questionnaire on risk factors. Invitations and testing kits went to patients' registered address, and we confirmed residence at this address. We used postal reminders, telephone calls, home visits, and “flagging” of records at study practices to encourage participation.
Our colleagues in the reference laboratory (SS and AHe) used nucleic acid amplification tests to detect C trachomatis and confirmed positive results. Postal delays did not affect positivity rates (data not shown). Participants with chlamydia received results and treatment at their practice. Notification of partners was undertaken at either the practice or a genitourinary clinic. Statistical analyses of coverage of chlamydia screening (proportion receiving an invitation to be screened), uptake (proportion returning a specimen), and prevalence took into account clustering at the practice level and sampling probability.
Results
Of 19 773 people aged 16-39 who had been invited, we contacted 73% (14 382, see figure on bmj.com) successfully. Uptake of screening was 34.5% (95% confidence interval 31.2% to 38.0%) overall, and 31.5% (28.6% to 34.6%) in 16-24 year olds (table 1). After a single postal invitation, uptake was 22.2% (18.6% to 26.2%) in 16-24 year olds and was higher in women than men (table 1). A postal reminder and face to face contact (in a home visit, or “flagged” patients being invited to participate when they attended their practice) each increased uptake by around 5%. Screening coverage and uptake varied by practice (see table A on bmj.com). Coverage was lower in areas with higher proportions of residents from minority ethnic groups, and uptake was lower in practices with higher deprivation scores (see table B on bmj.com).
Table 1.
Uptake of screening and prevalence of chlamydia according to effort needed to secure participation
| Groups by sex and age | Effort needed | No of people who received this form of invitation to be screened | No of people who responded after each form of invitation* | % uptake (95% Cl) | % cumulative uptake (95% Cl) | No of cases | Prevalence (95% Cl) | P value for linear trend calculated from Wald test | Cumulative prevalence (95% Cl) |
|---|---|---|---|---|---|---|---|---|---|
| All | Single postal invitation | 14 382 | 3239 | 25.0 (21.7 to 28.6) | 25.0 (21.7 to 28.6) | 137 | 3.0 (2.3 to 3.9) | 0.014 | 3.0 (2.3 to 3.9) |
| Two postal invitations | 11 143 | 683 | 5.3 (3.9 to 7.1) | 30.3 (26.5 to 34.3) | 32 | 3.4 (2.1 to 5.4) | 3.1 (2.3 to 4.0) | ||
| Phone call | 10 460 | 53 | 0.5 (0.1 to 2.2) | 30.8 (26.9 to 35.0) | 2 | 2.0 (1.1 to 3.8) | 3.0 (2.3 to 3.9) | ||
| Home visit or flagging† | 10 407 | 765 | 3.7 (2.0 to 6.8) | 34.5 (31.2 to 38.0) | 48 | 5.0 (3.4 to 7.2) | 3.2 (2.5 to 4.2) | ||
| All 16-24 | Single postal invitation | 11 245 | 2436 | 22.2 (18.6 to 26.2) | 22.2 (18.6 to 26.2) | 133 | 5.4 (4.3 to 6.6) | 0.054 | 5.4 (4.3 to 6.6) |
| Two postal invitations | 8809 | 507 | 4.3 (3.0 to 6.2) | 26.5 (22.6 to 30.8) | 29 | 6.5 (4.7 to 8.9) | 5.5 (4.6 to 6.7) | ||
| Phone call | 8302 | 39 | 0.5 (0.1 to 2.2) | 27.0 (22.9 to 31.5) | 2 | 4.4 (2.8 to 6.8) | 5.5 (4.6 to 6.7) | ||
| Home visit or flagging† | 8263 | 627 | 4.6 (2.0 to 10.0) | 31.5 (28.6 to 34.6) | 45 | 7.4 (5.5 to 9.9) | 5.8 (4.9 to 6.9) | ||
| Men 16-24 | Single postal invitation | 5454 | 1017 | 18.9 (16.0 to 22.3) | 18.9 (16.0 to 22.3) | 46 | 5.4 (4.4 to 6.5) | 0.768 | 5.4 (4.4 to 6.5) |
| Two postal invitations | 4437 | 196 | 3.57 (2.6 to 4.9) | 22.5 (19.3 to 26.1) | 9 | 5.3 (2.8 to 9.8) | 5.4 (4.3 to 6.6) | ||
| Phone call | 2241 | 21 | 0.49 (0.1 to 2.5) | 23.0 (19.6 to 26.8) | 0 | 0.0 (0.0 to 0.1) | 5.2 (4.2 to 6.5) | ||
| Home visit or flagging† | 4220 | 243 | 3.61 (1.5 to 8.3) | 26.6 (24.0 to 29.4) | 12 | 5.4 (3.3 to 8.8) | 5.3 (4.4 to 6.3) | ||
| Women 16-24 | Single postal invitation | 5791 | 1419 | 25.3 (20.9 to 30.3) | 25.3 (20.9 to 30.3) | 76 | 5.3 (3.9 to 7.3) | 0.033 | 5.3 (3.9 to 7.3) |
| Two postal invitations | 4372 | 311 | 5.1 (3.4 to 7.7) | 30.4 (25.7 to 35.6) | 20 | 7.3 (4.8 to 11.0) | 5.7 (4.3 to 7.5) | ||
| Phone call | 4061 | 18 | 0.4 (0.1 to 2.0) | 30.9 (26.0 to 36.2) | 2 | 9.5 (5.8 to 15.2) | 5.7 (4.4 to 7.4) | ||
| Home visit or flagging† | 4043 | 384 | 5.5 (2.5 to 11.9) | 36.4 (33.1 to 39.8) | 33 | 8.7 (6.8 to 11.1) | 6.2 (4.9 to 7.8) |
Includes nine people who returned a consent form but no specimen.
A “flag” was attached to patients' notes. Patients attending the practice during the study period were invited again to take part if they had not already participated, declined to participate, or been confirmed not resident at their address.
The overall prevalence of chlamydia was 2.8% (2.2% to 3.4%) in men and 3.6% (3.1% to 4.9%) in women, but it was higher in people who were younger than 25 (men 5.1%, 4.0% to 6.3%), women 6.2%, 5.2% to 7.8%). Prevalence was below 1% in men older than 24 and women older than 29 years. Prevalence was higher among the subgroup of 16-24 year old women who only participated after repeated contacts (table 1). Having one or more new partners in the past 12 months was the strongest predictor of infection (table 2). The prevalence of chlamydia was also higher in single compared with married women and in men aged 20-24 than in people aged 16-19 years (tables C and D on bmj.com).
Table 2.
Prevalence of infection according to number of sexual partners, and association between prevalence and sexual behaviour, in men and women
|
Odds ratio (95% confidence interval)
|
|||||||||
|---|---|---|---|---|---|---|---|---|---|
| Variable | Group | No of participants* | No of cases | Prevalence† | Crude | Adjusted‡ | P value‡ | Adjusted odds ratio (95% Cl)§ | P value§ |
| Total partners | 0/1 | 941 | 38 | 4.1 (3.0 to 7.0) | 1 (reference) | 1 (reference) | 0.018 | 1 (reference) | 0.149 |
| 2 | 226 | 15 | 8.8 (5.4 to 13.4) | 2.3 (1.1 to 4.7) | 2.0 (1.0 to 4.0) | 1.0 (0.4 to 2.4) | |||
| 3 | 139 | 15 | 11.7 (7.2 to 15.5) | 3.1 (1.5 to 6.3) | 2.9 (1.4 to 6.2) | 1.6 (0.7 to 3.8) | |||
| 4 or more | 200 | 27 | 12.5 (8.7 to 18.6) | 3.3 (1.5 to 7.3) | 3.0 (1.4 to 6.5) | 1.9 (0.9 to 4.2) | |||
| New partners¶ | 0 | 697 | 23 | 2.9 (1.9 to 5.7) | 1 (reference) | 1 (reference) | <0.001 | 1 (reference) | <0.001 |
| 1 | 417 | 24 | 7.4 (5.0 to 11.1) | 2.7 (1.3 to 5.5) | 2.7 (1.3 to 5.6) | 2.6 (1.4 to 4.8) | |||
| 2 | 166 | 21 | 13.9 (9.3 to 20.7) | 5.5 (2.7 to 11.2) | 5.1 (2.5 to 10.3) | 4.2 (2.0 to 9.1) | |||
| 3 or more | 226 | 27 | 11.0 (7.7 to 15.6) | 4.2 (1.8 to 10.1) | 4.1 (1.7 to 9.9) | 2.4 (1.1 to 5.3) | |||
Logistic regression models include 1506 people aged 16-24 who had ever had sex and who responded to questions about numbers of sexual partners.
Prevalence weighted for selection probability and adjusted for clustering at practice level by using inverse probability weights and robust standard errors.
From model additionally adjusted for age, marital status, and ethnic group.
From model additionally adjusted for age, marital status, and ethnic group and both sexual behaviour variables.
Number of new sexual partners in the past year.
Prevalence did not vary substantially between practices (I2 = 34% of variation attributable to heterogeneity between practices, tables A, C, and D on bmj.com). We found weak evidence that chlamydia was more common in practices in more deprived areas (adjusted odds ratio for a 10 point increase in deprivation score, 1.2, 1.0 to 1.4, P = 0.077).
Discussion
A single round of systematic, postal screening for chlamydia was feasible, but coverage was incomplete and uptake was modest, particularly in more deprived areas. The prevalence of chlamydia was 5-6% in men and women younger than 25 years. Having one or more new sexual partners in the past year was the strongest determinant of infection. In young women the risk of chlamydia was higher among those who were harder to engage in screening.
Strengths and limitations of the study
The strengths of this study are that it was large, population based, included both men and women, and tested a screening strategy shown to be feasible and effective in other settings.3 Limitations were that screening coverage was incomplete, reflecting inaccuracies in practice lists, and uptake was lower than expected,4 so the power of the study was reduced and prevalence estimates might be affected by selection bias.5 Repeated reminders had little impact on increasing uptake and would be unfeasible in routine practice.
Comparison with other studies
Our estimates of chlamydia prevalence were broadly comparable with other population based studies showing that 2-6% of men and women under the age of 25 are infected.1,3,6 The variability might reflect differences in study populations or selection biases since participation in all studies was incomplete.
Meaning of the study
Chlamydia risk was most strongly associated with having a new sexual partner in the past year, rather than the total number of partners. Although being from a black minority ethnic group has been found to be associated with chlamydia,7 we did not find a strong association with ethnicity. Our findings indicate that chlamydia screening should be offered to all men and women under 25 years because selection based on sexual behaviour could be difficult.
Effectiveness of chlamydia screening
Opportunistic chlamydia screening in selected settings, focusing on women younger than 25, is now being introduced in England,8 even though this approach has not been evaluated in randomised trials1 and has not controlled transmission of chlamydia in Sweden.1 In contrast, randomised trials in Denmark have found postal screening, with uptake similar to this study, to be effective and cost effective.9 The cost effectiveness of postal chlamydia screening in the ClaSS project will be reported separately. If postal screening were introduced, publicity and increasing familiarity may lead to higher uptake. Postal screening could also have a role as an adjunct to opportunistic screening since we also found that we reached a proportion of individuals who had not visited their practice in the past year.10
Effective health technology can contribute to widening health inequalities if access to care varies with social position. We found lower coverage of chlamydia screening in areas with more residents from minority ethnic groups, lower uptake in more deprived areas, and a trend towards higher prevalence in more deprived areas. Additionally, young women with the highest prevalence of chlamydia participated only after repeated invitations. Further research is needed to clarify the impact of chlamydia screening on sexual health inequalities.
Conclusion
Evidence for the long term effectiveness and impact of chlamydia screening programmes remains limited.1 Randomised trials to determine the most effective strategy are required. In England these could be incorporated within the phased introduction of the national screening programme.
What is already known
Rates of infection with genital Chlamydia trachomatis are increasing across Europe and in the United States
Opportunistic chlamydia screening is being introduced in England, although there is no high quality evidence of its effectiveness
Systematic population screening for chlamydia has been shown to be effective but factors affecting the coverage and uptake are poorly understood
What this study adds
Postal screening was feasible, but coverage and uptake were incomplete
Young women at higher risk of infection were harder to engage in screening
Undiagnosed chlamydia is common in both men and women under the age of 25 in the general population
An age younger than 25 and having a recent new sexual partner were the only important risk factors for chlamydia identified
Supplementary Material
Additional tables and a figure are on bmj.com
Contributors: JM wrote the first draft of the paper, which was revised by JM and NL following comments from CS, JACS, AMcC, GDS, and ME. All authors commented on the final draft. ME, JM, CS, PH, OC, GDS, NL, and RH devised the ClaSS prevalence survey and helped obtain funding. JM, CS, NL, AMcC, JS, PH, ME, and the ClaSS study group undertook strategic management of ClaSS. AMcC was also responsible for day to day project management. AHo, RP, and AM recruited practices and undertook the survey under the supervision of JM and CS. SS and AHe supervised sample testing. JS devised the analysis strategy with JM, CS, and NL. ES undertook statistical analyses under the supervision of JS and NL. ME is lead investigator of the ClaSS study and is the guarantor.
Funding: ClaSS is funded through the NHS R&D HTA programme. Participating individuals and institutions can be viewed at www.chlamydia.ac.uk. JM and NL are funded by NHS Career Scientist Awards.
Competing interests: None declared.
Ethical approval: South West Multicentre Research Ethics Committee and all relevant local research ethics committees.
References
- 1.Low N, Egger M. What should we do about screening for genital chlamydia? Int J Epidemiol 2002;31: 891-3. [DOI] [PubMed] [Google Scholar]
- 2.Low N, McCarthy A, Macleod J, Salisbury C, Horner PJ, Roberts TE, et al. The chlamydia screening studies: rationale and design. Sex Transm Inf 2004;80: 342-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Andersen B, Olesen F, Møller JK, Østergaard L. Population-based strategies for outreach screening of urogenital Chlamydia trachomatis infections: a randomized, controlled trial. J Infect Dis 2002;185: 252-8. [DOI] [PubMed] [Google Scholar]
- 4.Macleod J, Rowsell R, Horner P, Crowley T, Caul EO, Low N, Smith GD. Postal urine specimens: are they a feasible method for genital chlamydial infection screening? Br J Gen Pract 1999;49: 455-8. [PMC free article] [PubMed] [Google Scholar]
- 5.Low N, Macleod J, Salisbury C, Egger M, and the ClaSS study group. Bias in chlamydia prevalence surveys. Lancet 2003;362: 1157-8. [DOI] [PubMed] [Google Scholar]
- 6.Fenton KA, Korovessis C, Johnson AM, McCadden A, McManus S, Wellings K, et al. Sexual behaviour in Britain: reported sexually transmitted infections and prevalent genital Chlamydia trachomatis infection. Lancet 2001;358: 1851-4. [DOI] [PubMed] [Google Scholar]
- 7.Low N, Sterne JAC, Barlow D. Inequalities in rates of gonorrhoea and chlamydia between black ethnic groups in south east London: cross sectional study. Sex Transm Inf 2001;77: 15-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Department of Health. Choosing health: making healthy choices easier. London: DoH, 2001.
- 9.Østergaard L, Andersen B, Møller JK, Olesen F. Screening for klamydia med hjemmetest—en medicinsk teknologivurdering. Medicinsk Teknologivurdering—puljeprojekter 4(2). Copenhagen, Center for Evaluering og Medicinsk Teknologivurdering, 2002.
- 10.Salisbury C, Macleod J, Egger M, McCarthy A, Patel R, Holloway A, et al. Opportunistic or systematic screening for chlamydia? A study of consultations by young adults in general practice. Br J Gen Pract 2005. (in press). [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
