METHODS

This qualitative research was essentially exploratory, aimed at obtaining information from all members of the primary healthcare team about their genital chlamydia management strategies in genitourinary and sexual health.

Participants and recruitment

We recruited general practices served by the Bristol, Gloucester, and Hereford microbiology laboratories, which represent urban and rural communities in the South West Region. Practices in Bristol that were involved in the Health Technology Assessment (HTA) study of different chlamydia testing methods were excluded, as these would not be typical of most UK general practices. Laboratory data and practice population data from April 2000 to March 2001 were used to determine GP requests/1000 patients for C trachomatis testing (range 0.6–23/1000 population/year). As there was significant geographical heterogeneity in test submission the practices in each locality were stratified into rural or urban and then ordered by number of chlamydial specimens sent to each laboratory per 1000 practice population (table 1). Practices in the highest and lowest 10 percentiles were randomised and approached, initially by telephone and then by letter, in order from these lists, with representation from each laboratory locality. This resulted in four strata (rural high and low testers, urban high and low testers). Twelve of 15 practices that were approached agreed to take part. Within a practice, all healthcare staff who may be involved in the management of chlamydia were invited to take part in a focus group, including practice managers, GPs, practice nurses, midwives, and district nurses. At every focus group at least one nurse participated and in the larger practices there were two or three nurses. Each focus group only consisted of staff from a single practice; therefore the focus groups varied in size from two (a single handed GP and nurse) to eight (four GPs, two practice nurses, district nurse, and midwife). At the focus group visit practices were asked for their practice age/sex breakdown. Townsend deprivation scores were similar and there was no significant difference in the percentage of 17–24 year olds in high and low testing practices (table 1).

Development of topics for focus groups

A group of microbiologists, GPs, epidemiologists, and gynaecological clinicians developed a series of open questions about diagnosis, testing, and management of genital chlamydia within the practice setting for symptomatic and asymptomatic patients. Questions were also asked in relation to testing asymptomatic patients, particularly women attending for cervical screening and family planning services. Practices were asked about their staff training, expertise, and interest in STIs and whether they provided specific cervical screening, family planning, and well men or women clinics. All practices were asked whether they thought they were testing appropriately and what would encourage them to send more specimens.

Conduct of focus groups

Twelve focus groups were held in the practice premises between April and August 2002. Practices were informed of their sampling rate and the purpose of the study at the beginning of the focus group. The focus groups were moderated by a researcher with extensive qualitative research experience (EF) and observed either by a microbiologist (CMcN 6), an obstetrics and gynaecology registrar (JS 4), or other researcher (JB 2). At the end of each focus group the points raised were summarised by the moderator (EF) and verified by the group. Following the focus groups the practices were presented with UK guidance on the management of chlamydia.^4,6 Notes of key issues raised and non-oral signals were made by the observer (during the focus groups) and by the moderator (immediately after the focus groups). Focus groups were tape recorded. Memos were used to capture pertinent issues after listening to the tape recordings. These data were appended to the transcripts for further analysis.

Ethics approval

Ethical approval was obtained from the local research ethics committees in each area (Gloucester No 01/152G, Hereford No C7b(i), Bristol No E5241). All participants gave written informed consent for the focus groups to be tape recorded and transcribed.

Data analysis

Data were collected and analysed concurrently until saturation occurred.⁷ All the tapes were transcribed and data checked by listening to the tapes for accuracy by the moderator (EF). These transcribed data were analysed using a modified grounded theory approach utilising the constant comparative method.^7,8 The transcripts were scrutinised independently (by EF and CMcN). A number of major themes and ideas emerged from the data, which were discussed and confirmed by the research team. A coding frame was developed which identified common codes, categories, and themes that explored the differences in sampling rates and the management strategies of the participants. Themes raised in the different focus groups were compared and contrasted as an iterative process. A high level of consensus was achieved between research team members in interpreting the data. Saturation of data occurred relatively quickly and the last four focus groups served to enrich the data.

RESULTS

There were major differences between high and low chlamydia testers (table 2). This included their knowledge of chlamydia and sexual health, threshold for testing for chlamydia in men and women, their staff training, and information available for patients.

Threshold for testing for chlamydia

The threshold for testing for chlamydia in women with vaginal discharge was very different between low and high tester practices. The low tester practices did not consider testing for chlamydia until women had re-consulted several times and after bacterial swabs were reported to be negative. In contrast, high testers always considered chlamydial infection in women with genitourinary symptoms and routinely sent a chlamydia swab whenever they took a high vaginal swab. All primary care staff were less confident in their management of men presenting with genitourinary symptoms. All low testers referred men directly to the GUM clinic. High testers were more likely to manage men with dysuria but thought they were unable to test for the full range of STIs and had no resources to do any contact tracing. Practices were unable to check whether patients attended the GUM clinic, as anonymity of GUM clinics usually meant that even if a letter of referral was sent with the patient no correspondence was returned. Practices felt that this lack of communication placed a barrier between GUM and primary care services.

Awareness of chlamydia and sexual health issues

Low testers knew very little about the epidemiology and presentation of chlamydia infection and did not consider testing for chlamydia in women with genitourinary symptoms attending routine appointments or for a family planning check or cervical smear. Low testers did not know that chlamydia was an intracellular organism and all staff were unsure how to take an endocervical swab. Many low testers believed that chlamydia tests were inaccurate when taken after a cervical smear.

High tester practices either had a GP staff member with an interest in sexual health or practice nurses who had completed specialist training in family planning (four of the six high testing practices had both). In contrast, only two of the low testers had a staff member with an interest in sexual health and these were both new to their practices. High tester practices were more likely to have patient information leaflets on genital chlamydia infections. Even so, only four of the 12 practices (three high testers) had patient information leaflets on chlamydia infection.

Perception of chlamydia prevalence in the catchment population

The low testing practices felt that their practice population had a low number of patients in the age group at risk of chlamydia. Although these practices perceived that their practice population was elderly, with low numbers of young adults, statistical analysis of their age, sex, and population data did not show a significant difference from high testers (Mann-Whitney U test comparing percentage of 17–24 year age groups p = 0.57). Four of the six low testing practices also felt that their practice population was not at risk of chlamydial infection, which was perceived as an inner city problem. As they had such a low testing rate, the low tester practices had very few positive results and they used this as evidence that their practice population was at low risk. High testers were aware that patients in their practice were at risk of chlamydia. Their testing policy meant they received positive results and this further reinforced their awareness of risk groups and value of testing for chlamydia in the primary care setting.

DISCUSSION

This study confirmed the wide variation in chlamydia testing and the lost opportunity in some practices to reduce morbidity caused by this infection within general practice.⁹ There was clear evidence of differences in professional awareness about the natural history, clinical presentation, and epidemiology of genital chlamydia. Low testing practices had little knowledge of who and when they should be testing.

Evidence based clinical practice is highly dependent upon staff awareness, training, confidence, and possession of the appropriate skills to undertake the work. Our study indicated that staff from high testing practices had attended either continuing education sessions about chlamydia or had received specialist family planning training or had an interest in sexual health. If we are to detect more chlamydia infections in primary care, practices that have a low testing behaviours will need to be identified within the framework of continuing medical education, continuing professional development, and clinical governance.

UK guidance on the management of genital infections recommends diagnostic testing should be performed in women with symptoms and signs suggestive of C trachomatis.^4,5 This includes vaginal discharge and post-coital, intermenstrual, and breakthrough bleeding. The high testing practices were aware of chlamydial infection and were testing appropriately in symptomatic women. However, our routine laboratory data indicate that only 44% of women under 25 years who have a high vaginal swab sent to the laboratory for genitourinary symptoms also have a chlamydia swab. We should be encouraging GPs always to take an endocervical swab for chlamydia when taking a high vaginal swab. The routine addition of a comment “have you considered chlamydia” on genital swab bacteriology laboratory reports in the at risk age group may help to reinforce the message.

Few practices tested men for chlamydia. A delay in the investigation and treatment of symptomatic men occurs as most are often referred directly to GUM clinics. This may not be considered ideal management, as waiting times for GUM clinics are extending which increases the possibility of further cross infection in the target population. Furthermore, because of desire by GUM clinics to maintain patient confidentiality, GPs have no way of knowing whether patients attend, as they often received very little, if any, correspondence about patients referred to GUM clinics. This lack of communication contributes to the lack of knowledge about STIs by primary care staff, as such correspondence has an educational role. In the 21st century surely it is time to break down this confidentiality barrier and, with patients’ permission, write to the primary carer? These issues need to be addressed during the implementation of the sexual health strategy.

The heterogeneity in chlamydia testing that currently exists across GP practices is partly due to the lack of insight by practices of their own testing behaviours. Laboratory use figures are not made available routinely to practices, unlike antibiotic use.¹⁰ This study underscores the importance of having good performance measures so that practices are aware of where they lie on the normal distribution of testing patterns and whether they test in the appropriate at risk age group.

In a qualitative study of patients, most women recently diagnosed with genital chlamydial infection had not previously perceived sexually transmitted infections as being personally relevant.¹¹ This is mirrored by the perceptions of healthcare staff in our low testing practices who felt that their patients were not at risk of chlamydial infection. These perceptions of risk will need to be addressed if we hope to increase testing within primary care as C trachomatis infection is broadly distributed geographically and by age, sex and behaviour group.^12,13

Strengths and weaknesses of this qualitative research

The data generated are from a much smaller sample than is possible from quantitative research, but the sampling strategy used meant that the findings may be considered to be representative of rural and urban general practices under similar pressures. The open questioning style revealed attitudes to chlamydia infection that have not been ascertained from previous questionnaire surveys. It was possible to explore in considerable depth the reasons why tests are performed and attitudes towards testing. By selecting practices at the extremes of the testing distribution we may have selected practices which are atypical in other ways. However, as we randomly selected practices we have avoided any systematic error in their selection within these broad groups. The testing data were for the whole practice; therefore data from large practices may have masked individual variations in practice that we were not able to explore. However, individuals were able to express their own practice beliefs and knowledge of sexual health and testing strategies. Qualitative data can be observer biased. However, the moderator was aware of the importance of not influencing participants’ responses and was not a microbiologist or sexual health specialist.

Implication for sexual health policy and planning

The Department of Health has announced the roll-out of the chlamydia screening programme across the United Kingdom.^2,14 Although screening is currently concentrated in GUM and family planning clinics, testing in general practice is planned in the future. Any moves to roll-out sexual health service provision in primary care must be accompanied by adequate investment in human and capital infrastructure to support this work. This funding must include continuing education about the prevalence, pathophysiology, epidemiology, and most appropriate diagnosis of chlamydia. Low testing practices will need to be targeted specifically in any strategy and this laboratory use data gathering exercise will also need resourcing.