METHODS

This study involved two postal questionnaire surveys: (i) a four page questionnaire to GPs in the North Thames area of the United Kingdom (a demographically mixed urban and suburban area close to and including central London). The questionnaire requested tick box and free text responses to demographic questions and investigation and management questions stemming from the case of a 20 year old woman on the combined oral contraceptive pill (COCP) who complained of vaginal discharge with a new sexual partner. (ii) A three page questionnaire to lead consultant microbiologists at the hospitals identified by the GPs as those to which they send their samples. This questionnaire requested tick box and free text responses, and focused on those tests which the laboratories would undertake on an HVS sample from the same patient both routinely and given clinical information. A questionnaire was sent to the first two out of every three GPs in an alphabetised list of those currently registered in 13 health authorities of north London. GP responses were anonymised. A single mailing was used for GP questionnaires. GPs were asked to pass the questionnaire to the practice nurse if he/she saw most of the women presenting with vaginal discharge. Microbiologists were sent a second questionnaire if no response was received to the first, and in some cases contacted by telephone.

The GP questionnaire was revised in line with comments from a pilot study of 10 GPs outside the geographical area of study.

Ethical approval was obtained from the Northern and Yorkshire Multicentre Research ethics committee.

Data were recorded and analysed in an Access database.

RESULTS

Demographics

In all 2146 GPs were identified and completed responses were received from 553 (25.8%). Characteristics of responding GPs are shown in table 1.

Seventy two per cent of GPs reported that women with vaginal discharge sometimes present directly to the practice nurse.

Completed responses were received from 14 (63.6%) of the 22 laboratories which were sent questionnaires.

Clinical management

GPs’ management of a young woman with symptoms and a new sexual partner is summarised in table 2. The case presented was as follows: “Laura, 20 years old, attends your surgery complaining of 2 weeks of increased vaginal discharge. She has a new male sexual partner and is taking the combined oral contraceptive pill.”

Of the 397 GPs who would take an HVS 298 (53.9%) would also routinely test for Chlamydia trachomatis and 71 (17.9%) would not (missing 27, 6.8%). Of those requesting C trachomatis routinely, 280 (94.0%) would use material from an additional endocervical swab, 18 (6.0%) would request C trachomatis on the HVS and one GP (0.3%) would test for C trachomatis on a urine sample.

Requesting, testing, and reporting of HVS samples

GPs were asked about how they request tests. When performing an HVS, 412 (75%) GPs would write “M,C&S” (microscopy, culture, and sensitivity) on the request form, with or without additional requests for culture of specific organisms. Four per cent (20) of GPs would not write “M,C&S” but would request culture for at least one of Candida spp, Trichomonas, Gardnerella, or anaerobes. Three per cent (14) of GPs would request culture for Neisseria gonorrhoeae on the HVS material.

Although there are no universal standards, as a point of reference, the PHLS SOP⁶ recommends an HVS sample from this patient be Gram stained for bacterial vaginosis (BV) and cultured for Trichomonas and yeasts. Eleven (79%) laboratories would perform microscopy and/or culture for candida, 13 (93%) would perform wet smear, six (43%) would culture for Trichomonas, and 10 (71%) would perform microscopy for clue cells. One laboratory (7%) routinely performs “full culture” on all HVS samples. Full culture varies between laboratories but usually includes aerobic and anaerobic culture. Other laboratories perform full culture on the basis of further clinical information—for example, presence of abdominal pain or postpartum state, or initial microscopy findings, but such criteria differ between laboratories (data not shown).

Most (303, 54.8%) GPs preferred to be informed about specific pathogens only and their respective sensitivities rather than all micro-organisms found, and 12 (85.6%) laboratories report in this way. One hundred and thirty GPs (23.5%) wanted to be informed about all micro-organisms isolated (as reported by two (14.3%) laboratories) and 120 GPs (21.7%) expressed no preference.

Four hundred and thirty (77.8%) GPs would like the HVS report to include a suggested diagnosis (1.3% wouldn’t, 116 (20.1%) missing) but only six laboratories (42.9%) ever provide this. Laboratories commented that they only provide a diagnosis on the basis of clinical information given by the GP, and several stated that they think it is inappropriate to give a clinical diagnosis on the basis of microbiological findings. Seventy four per cent (409) of GPs would like a suggested treatment to be included on the HVS report (4.3% would not, 120 (21.7%) missing) but only two laboratories (14%) provide this.

DISCUSSION

Our study confirms that vaginal discharge is a common presentation in general practice. Almost half of our sample is seeing 1–5 women per week with vaginal discharge, and one third of our sample of GPs are prepared themselves to manage a young woman at risk of STIs, without reference to specialist services. Almost three quarters of GPs surveyed would take an HVS.

Encouragingly, GPs are screening for STIs and over half would test appropriately for C trachomatis.

Just under half of the GPs offered empirical therapy, and most commonly treated for candidiasis first line. However, bacterial vaginosis is at least as common a cause of vaginal discharge as candidiasis.^7–9

GPs wanted as much information as possible from an HVS report, including direction from their laboratory with regards to diagnosis and treatment. Most laboratories do not provide this, and some expressed the opinion that this is outside their role.

We have shown that there is wide variation in the processing of HVS samples by microbiology laboratories in north London, and it is reasonable to assume that this is the case throughout the United Kingdom. In view of the workload that HVS samples represent to microbiology laboratories and the potential increase in samples received as GPs increase their involvement in sexual health¹ a national guideline on processing HVS samples is likely to be beneficial.

Despite a disappointing GP response rate, we are able to report data from over 500 GPs. Selection bias is likely given that female GPs are over-represented in our sample,¹⁰ single handed GPs are under-represented (YH Carter, personal communication), and GPs who see fewer women with vaginal discharge were less likely to respond. Almost three quarters of GPs felt that women with vaginal discharge might present directly to their practice nurses but we only had one returned questionnaire clearly completed by a practice nurse. Further work is needed to look at how practice nurses manage women presenting with vaginal discharge. As with all questionnaire surveys our study is likely to include reporting bias.

GPs in the United Kingdom are facing ever increasing clinical and administration pressures. The emphasis of the National Strategy for Sexual Health and HIV¹ in widening access to sexual health care by increased provision in primary care settings is likely to be a considerable challenge to GPs who, while welcoming increased sexual health education, would have limited time in which to access it.

The HVS remains the mainstay of GPs’ management of vaginal discharge. Three quarters of GPs request non-specific processing (“M,C&S”) whereas most would prefer specific and directive reports, including suggested further investigation and treatment. Almost two thirds of GPs would do an additional C trachomatis test but in the specialist setting all such women would be tested for C trachomatis. GPs would appear to be receptive to guidance on STIs (C trachomatis) screening included on HVS reports.

CONCLUSIONS

Although many GPs are appropriately managing young “at risk” women with vaginal discharge our study highlights the potential for laboratory GP liaison in terms of HVS processing and reporting, which could greatly and cheaply assist GPs’ decision making when faced with a woman with vaginal discharge.