Article Text
PDF
Abstract
Objective England's National Chlamydia Screening Programme provides opportunistic testing for all 15–24-year-olds in healthcare and non-healthcare settings. The authors undertook an evaluation of a population service-based postal chlamydia screening and treatment service, with registration and access to the service via the internet, in order to target screening interventions more effectively in future.
Methods Individuals aged between 18 and 24 years, within the North East Essex Primary Care Trust, were identified for chlamydia screening by the service between 1 December 2008 and 31 January 2009. Associations between test uptake and positivity, and individuals' personal characteristics, were examined. The efficacy of partner notification was estimated, and the costs of screening were compared with the national average.
Results Uptake of chlamydia screening was 11.5%, with lower response rates among men, individuals older than 20 years and those living in more deprived areas. The chlamydia positivity was 4.4% and higher in those reporting multiple sexual partners, individuals older than 20 years and those living in more deprived areas. The proportion of partners notified per index case was 0.17. The service contributed to 3431 of the overall 11 209 annual chlamydia screens of the Primary Care Trust in 2008–2009, at a cost of £78 per screening test completed and £1764 per case detected.
Conclusions Our evaluation shows that this service model can contribute substantially to the overall coverage of chlamydia screening tests. However, the costs of service provision per case detected and treated, using this model, were high compared to the National Chlamydia Screening Programme.
- Chlamydia trachomatis
- health services research
- screening
- epidemiology (general)
- economic analysis
- service delivery
- epidemiology (clinical)
Statistics from Altmetric.com
- Chlamydia trachomatis
- health services research
- screening
- epidemiology (general)
- economic analysis
- service delivery
- epidemiology (clinical)
Introduction
Urogenital Chlamydia trachomatis is the most common treatable sexually transmitted infection in industrialised countries and the UK.1–3 Most infections are asymptomatic or subclinical and remain undiagnosed.4 Active case finding and early treatment are strategies to prevent the development of sequelae and to reduce transmission of chlamydia. In England, the National Chlamydia Screening Programme (NCSP) was established in 2003, and the roll-out of the programme was completed by 2008.5 ,6 The programme is currently under review to improve quality and cost-effectiveness but aims to ensure that all sexually active men and women younger than 25 years are aware of chlamydia and its effects and that they have access to services providing screening, prevention and treatment to reduce their risk of infection or onward transmission. The NCSP currently uses an opportunistic approach whereby testing is offered through a combination of health settings, such as general practice and sexual health services, and non-health screening venues, such as workplaces and educational facilities.
In 2008, North East Essex Primary Care Trust, responsible for overseeing the delivery of health services appropriate for its local population, commissioned a commercial clinical genitourinary (GUM) service provider to provide an innovative systematic chlamydia screening service, in order to increase its chlamydia screening coverage and to reach Department of Health targets7 (figure 1). The service provided postal testing kits for home sampling, online internet registration to a remote clinical service, access to results online and treatment sent by post if required. Partners could be anonymously notified online by index cases and had full access to treatment by the service.
Operational flowchart of the North East Essex (NEE) Primary Care Trust chlamydia screening service. GUM, genitourinary; SMS, short message service.
The aims of the service were to provide equitable and accessible information and advice on sexually transmitted infections and chlamydia screening tests; to promote sexual health and to detect and treat chlamydia infection, thus preventing sequelae and onward transmission.
In England, a National Audit Office review recommended in 2010 that the Department of Health should evaluate outreach events and ‘remote’ testing services, such as those provided through websites.8 The Health Protection Agency is now undertaking a review to identify where outreach work is relevant, particularly in providing access for hard to reach, young people and responding to local needs.9
The aims of this evaluation were (1) to quantify the uptake and the test positivity rate of a web-based and systematic postal population screening programme; (2) to identify personal factors associated with screening uptake and yield and (3) to compare costs with the NCSP, in order to target screening interventions more effectively in future.
Methods
Participant identification
Since most of the UK population is registered with a local family practitioner, individuals aged 18–24 years were identified using patient registers of all general practices within the geographical boundary of the Primary Care Trust (PCT) between 1 December 2008 and 31 January 2009. All eligible individuals were sent chlamydia screening kits by post from a commissioned central laboratory. Each kit included a uniquely numbered urine specimen container, instructions on how to take a urine sample, explanation of how individuals could register to accept screening, information about chlamydia and sexual health and a stamped addressed envelope for return of the urine sample for testing at an accredited central laboratory.
Screening registration
Individuals wishing to be screened were asked to register for the commissioned service using the unique test number supplied with their kit either via a bespoke secure website or a dedicated freephone number. Data were stored on a clinical database at a single remote clinic location, only accessible to commissioned clinical GUM staff who worked at this location. Respondents were asked to provide telephone contact information for result notification and the following minimum data set in line with NCSP data specifications: surname, forename, date of birth, address, postcode, gender, ethnicity, information on sexual partners (numbers of new partners in the last 3 months and whether they had had two or more partners in the previous 12 months) and whether they had symptoms such as pain while passing urine, vaginal or urethral discharge, testicular or lower abdominal pain and irregular vaginal bleeding.
Screening tests
The screening test detected C trachomatis DNA in urine by nucleic acid amplification testing using the BD ProbeTec NAAT platform.10 Results of screening tests were added to individuals' database records by commissioned clinic staff and individuals were notified, by mobile phone text or other preferred method of contact, that the result was available. Individuals then logged back into their record. If their result was positive, they completed an online clinical questionnaire and specified to which address they would like their treatment posted or to indicate that they would prefer to pick up the treatment from a local pharmacy. Those without computer access were asked to contact the service by freephone. A remote commissioned doctor reviewed the clinical questionnaire, then treatment, usually azithromycin, was prescribed and dispatched by registered post. If there was any concern about the clinical questionnaire, patients were contacted directly by the doctor to discuss treatment options. A single reminder letter was sent to individuals who had returned screening kits, but not registered, asking them to register online or to call the freephone number.
Partner notification
Partners, notified by an index case, were able to register, as described above. Partners were linked to cases via a unique code that was generated for each index case. In line with the UK guidance, they were offered empirical treatment for chlamydia.11
Costs
The costs of our screening service were calculated in 2008 and assumed a 10% uptake of screening tests (based on the experience of postal screening services in two neighbouring PCTs), a 12% positivity rate and one partner notification per index case (based on the previous experience of the commissioned provider). The cost agreed with the provider per posted kit was £6, cost of specimen return and testing was £12 and cost of treatment of index or partner was £19. The costs of our service were compared with the costs of the NCSP in 2008–2009.12
Data
Data collected on each individual to whom a test kit was sent included age, sex and socioeconomic status, which was determined by referencing their postcode against the Office of National Statistics Gridlink® postcode file to assign to administrative geographies called Lower Super Output Areas.13 Each Lower Super Output Areas was assigned a census-based Index of Multiple Deprivation (IMD) 2007.14
For those accepting screening, recorded data included the NCSP minimum data set, test results and whether treatment was accepted if they tested positive. The NCSP minimum data set, and treatment acceptance, was recorded for partners who had been notified and registered with the service.
The outcomes of interest were screening uptake, screening test positivity, treatment uptake and partner notification.
Uptake of chlamydia screening was defined as the proportion of individuals who registered and were tested after the offer of screening.
Test positivity was defined as the proportion of registered screening tests with a positive result recorded.
Partner notification efficacy was expressed as the number of partners treated per index case.12
Statistical analysis
Statistical analyses were performed using STATA V.11 software (StataCorp LP). Associations between outcomes (test uptake, online registration and test positivity) and individuals' personal characteristics were examined using logistic regression, adjusted for intracluster correlation of outcomes within general practices. For explanatory variables with more than two categories, the significance of the overall association between all categories of the variable and the outcome was estimated using Wald tests, with Stata's ‘testparm’ postestimation command. Multiple logistic regression was then used to identify factors independently associated with screening uptake and test positivity. For test uptake, potential explanatory variables were age, sex and IMD quintile. For test result, potential explanatory variables were age, sex, IMD quintile, ethnicity, new sexual partner in the last 3 months, two or more partners in last 12 months and whether symptomatic at time of screening. Backwards stepwise variable selection was done by first including all potential explanatory variables in each regression model and then removing the variables not associated with the outcome at the 10% significance level. However, variables that were substantial confounders (ie, their removal changed the ORs of any of the remaining variables by more than 10%) were retained in the model, regardless of their statistical significance. Unregistered but tested individuals were excluded from these analyses.
Ethics review and data protection
As this was a service evaluation, the National Research Ethics Service confirmed that ethics committee approval was not required. The study protocol was approved by North East Essex Primary Care Trust. Standard procedures for secure handling of data were followed, and data were cleaned and anonymised by PCT staff before transfer for statistical analysis.
Results
Screening uptake
The demographic details of each individual to whom a test pack was sent were entered onto the clinical database. Those sending in specimens but not registering could therefore be identified. Postal kits were distributed to 29 917 individuals. Three thousand nine hundred and nineteen individuals returned specimens, of whom 3431 (11.5%) registered for screening. Four hundred and eighty-eight (12.4%) of tested individuals did not register. These individuals tended to be younger than those who returned specimens and did register (mean age 20.2 (SD 1.8) vs 21.4 (2.0) years; p<0.001) and were less likely to be women (57.8% vs 43.1%; p<0.001).
Of 3431 registrations, the method of service contact was recorded in 3401. Eighty-two per cent registered and requested screening via the internet, and the remainder contacted the service by telephone.
Uptake of screening was less likely among men, among those aged older than 20 years than younger ages and among those living in more deprived areas compared to those living in less deprived areas. Sex, age and IMD were independently associated with uptake of screening after adjustment using a multiple logistic regression model (table 1).
Characteristics of people offered screening, who registered and accepted a Chlamydia trachomatis screening test; and independent predictors of test uptake: univariate and multivariable analysis
Test positivity
Of 3431 registered individuals tested, 152 (4.4%, 95% CI 3.8 to 5.2) had positive chlamydia test results, none of whom were symptomatic (table 2). Forty-two per cent reported having a new partner in the last 3 months, and 51% reported having two or more partners in the last 12 months. After mutual adjustment in a multiple logistic regression model, having two or more partners in the last 12 months was strongly associated with a positive test. Those older than 20 years compared with younger ages and those living in more deprived areas were also weakly independently associated with a positive test (table 2).
Characteristics of people who registered and screened positive for Chlamydia trachomatis infection and independent predictors of a positive test result: univariate and multivariable analysis
Of the 488 individuals who did not register to receive their results, 22 (4.5%) were test positive.
Partner notification
There were 26 partners of screen-positive individuals who registered for treatment, all via the internet. None were symptomatic. Eighty-one per cent were men, 23% reported a new sexual partner in the last 3 months and 35% reported two or more partners in the last 12 months. The partner notification efficacy (the number of partners treated per index case) was 0.17.
Treatment
Treatment was reported as accepted from the service by 157 individuals in total: 131/152 (86.2%) index cases, of whom 95.4% requested treatment through the post and 4.6% requested treatment from a local pharmacy; and 26 notified partners, all of whom accepted treatment via the post.
Discussion
Home testing for chlamydia, using postal kits, is an increasingly common approach used in England, but the evidence of the effectiveness of this approach is based mainly on sampled populations rather than entire populations included in screening programmes.15–21 To our knowledge, this is the first evaluation of an English population-based C trachomatis screening service using global dispatch of testing kits to eligible individuals of both sexes aged between 18 and 24 years, web-based data collection and test reporting, and treatment dispatch by post.
Main findings
Our population-based service evaluation showed that, for every 1000 individuals invited to take part, approximately 115 responded and five tested positive.
Screening uptake was 11.5%; lower in men, those aged 21–25 years compared with 17–20-year-olds, and in more deprived areas.
Chlamydia positivity was 4.4% overall and strongly associated with having two or more partners in the last 12 months. Positivity was weakly associated with being aged 21–25 years compared with younger ages and living in more deprived areas. Our web-based postal screening service contributed to 3431 of the total 11 209 chlamydia tests conducted as part of NCSP activity in 2008.22 The cost of our service was £78 per screening test completed and £1764 per case detected.
Strengths and limitations of this study
This study is based on an evaluation of variables routinely collected as part of the minimum data set for NCSP.
The strength of this study is that it is large, population-based and included both men and women. National general practice registration allows the population denominator to be estimated. Our analyses, however, were limited to data on sex, age and IMD for the predictors of test uptake. IMD scores are derived from postcodes, and therefore, any incorrect assignment of postcode to an individual could bias these analyses.
Comparison with other studies
Screening uptake, chlamydia test positivity and risk factors
Our screening uptake was lower than provisionally reported data on mail-out return rates from England's NCSP (13%–15%)23 and significantly lower than reported in the English ClaSS study (22% after a single postal invitation and rising to 34.5% after several reminders), in which men and women aged between 16 and 39 years were randomly selected from 27 general practices to take a home specimen and post it to a laboratory for testing.24 Unlike our study, where participants were required to register their details online to receive results, individuals in the ClaSS study with chlamydia received results and treatment at their practice, and those with negative results were informed by mail. Uptake was also much lower than that reported in studies of postal screening in Denmark (27%) and the Netherlands (41%), where results were received by post.15 ,16 It is therefore possible that requiring a participant to actively ‘register’ may have been a deterrent to taking up the offer of screening. However, a study conducted in Sweden, where participants received a home sampling kit and used the internet to receive their test results, reported an uptake of 36%.25 A more recent publication from the Netherlands on the first screening round of a comprehensive, register-based, Chlamydia screening implementation programme reported that after being invited to access a website to register and request a chlamydia screening test kit, 16% of the population accepted screening.26
We were unable to determine the true reasons for our lower uptake of screening, and further research on the barriers to screening uptake would be valuable. Some studies have reported that letters do not reach their intended addressees in 18%–33% of cases.21 ,27 This could account for some participants not taking part in our screening service, particularly as there is a large university with a very mobile population, situated within the PCT boundary. It has also been reported that the vast majority of participants who responded to a questionnaire, but who did not agree to chlamydia screening, indicated that they felt that they were not likely to be exposed to the risk of contracting a sexually transmitted infection, despite 36% reporting having had sex with one or more partners.28 Since chlamydia screening is offered opportunistically in England, it is also possible that individuals may have already been offered chlamydia screening in other settings and therefore declined further screening.
The risk factors for a low screening uptake in our study were similar to other studies in England24 ,29 ,30 but differed from reports of population screening in the Netherlands, where those <20 years old were less like to accept screening.16 Further qualitative research among those who did and did not respond would enable a deeper understanding of what influenced their decisions.
Chlamydia test positivity in our population was higher than that found in European studies where postal kits had been sent to a specified population (1.1%–2.8%)16 ,26 but slightly lower than that found in the ClaSS study (5.4%).24
The strongest risk factor for a positive test (two or more partners in the last 12 months) concurred with that found in the ClaSS study.24 The weak association that we found with age and deprivation has also been reported elsewhere.6 ,19 ,24 Older age groups (>20 years) were less likely to accept screening but more likely to screen positive and would be a suitable group for a more targeted approach in future.
Of concern were the 22 test positive individuals who did not register, despite a follow-up letter, and therefore remained unaware of their results. This gap highlights the paramount importance of specifications for newly commissioned services to include fail-safe mechanisms such as flagging screen-positive individuals and ensuring alternative methods of contact.
In 2008–2009, a median of 0.4 partners per index patient were confirmed treated by the NSCP (range 0.1–1).12 Our partner notification rate of 0.17 was within this range, but far below the recommended standard of 0.6 partners screened per index,31 and below the median number of partners tested for chlamydia reported in a study reanalysing data from a national chlamydia audit of GUM clinics in the UK (0.47–0.92).32
Costs
In 2008–2009, the period during which our screening service was delivered, the National Audit Office conducted a survey of all English PCTs and calculated that the average cost per test delivered under the NCSP was £56, including follow-up activities such as treatment of positive patients and partner notification, and local overheads. In PCTs with higher testing rates (17% and above), the average cost reduced to £45. The cost per infection treated was estimated as £506.8 ,12
The comparative costs of our service are shown in table 3. Even excluding set-up costs for the service, the costs per screening test and per positive diagnosis are 1.66 and 3.5 times higher than the NCSP average, respectively. Even assuming national average figures, with screening uptake of 15%, a partner notification rate of 0.4 and a 65% positivity rate among partners, the cost of our service would remain comparatively high. We were unable to calculate whether these costs may underestimate true costs since we do not know whether there was duplication of service provision to some individuals who may have been offered chlamydia screening in other settings.
Service costs reported by the National Chlamydia Screening Programme (NCSP) and North East Essex Primary Care Trust (NEEPCT): 2008–2009
The parameters against which the comparative costs of our service are calculated: screening service uptake; proportion of index cases screening positive; partner notification efficacy; and combined partner and index positivity, were all well below those reported by a study examining variations in positivity in different screening settings, as well as the average reported by the national chlamydia programme.6 ,12 Although our service was only evaluated over a short period, it would be unlikely that it would have become more effective, if repeated, since the proportion of the North East Essex population testing positive for C trachomatis had reduced to 3.3% in 2010–2011.33
Meaning of the study
Our study shows that home screening in combination with web-based communication and data capture is feasible and can contribute substantially to the overall coverage of chlamydia screening tests. However, the costs of service provision per case detected and treated, using this model, were high compared to the NCSP. To achieve the national average, the overall costs of delivering the service would need to be substantially reduced, and there would need to be a significant increase in uptake of screening and in the yield of positive cases.
Ongoing service-based evaluations of chlamydia screening intervention models, as demonstrated here, are required to assist policymakers and planners in ensuring that programmes deliver best value, by maximising reduction in infection and minimising costs.
Key messages
The evaluation of population-service-based postal screening for Chlamydia trachomatis is poorly documented.
This evaluation shows that home screening in combination with web-based communication and data capture is feasible and contributed to 3431 of 11 209 overall screening tests in North East Essex during 2008–2009.
Screening uptake was lower among registered men, those older than 20 years and those living in the more deprived areas. Reporting multiple sexual partners, older age group (>21 years) and living in more deprived areas were independently associated with positive tests.
Despite a significant contribution to screening coverage, the costs of service provision per case detected and treated, using this model, were high compared to the NCSP.
References
Footnotes
Competing interests None.
Ethics approval NRES advice: ethics approval not required as this was a service evaluation.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Standard procedures for secure handling of data were followed in line with the Data Protection Act and Caldicott guidance. Data were cleaned and anonymised by PCT staff before transfer to University of East Anglia for statistical analysis.