Article Text

Download PDFPDF

Original article
Is the recall of men who have sex with men (MSM) diagnosed as having bacterial sexually transmitted infections (STIs) for re-screening a feasible and effective strategy?
  1. Derval Harte1,
  2. Danielle Mercey2,
  3. Jay Jarman1,
  4. Paul Benn1
  1. 1Department of Genitourinary Medicine, Mortimer Market Centre, Camden Provider Services, Central North and West London Foundation Trust, London, UK
  2. 2Research Department of Infection and Population Sciences, University College London Medical School, London, UK
  1. Correspondence to Dr Derval Harte, Bristol Sexual Health Services, Central Health Clinic, Tower Hill, Bristol BS2 0JD, UK; d.harte{at}


Objectives To assess the feasibility and outcomes of recalling men who have sex with men (MSM) diagnosed as having a bacterial sexually transmitted infection (STI) for re-screening.

Methods This evaluation was conducted from December 2008 for a 9-month period. MSM diagnosed as having a bacterial STI in that period were offered recall for re-screening 3 months after their diagnosis. Re-screening rates and infection incidence were calculated. Differences in baseline characteristics by re-screening status and factors predictive of infection at re-screening were assessed using the Mann–Whitney test, χ2 test and logistic regression.

Results Of the 337 MSM diagnosed as having a bacterial STI, 301 were offered recall. Of these, 206 (68.4%) re-screened after 3 months, 30 (10%) declined and the remainder did not re-attend despite giving verbal consent. Compared with those not re-screening, those re-screening were less likely to be HIV positive (p=0.001), but there was no difference in baseline risk behaviours. There were 15 diagnoses of bacterial STIs at re-screening (29 per 100 person-year follow-up (pyfu); 95% CI 14.3 to 43.7) and five new HIV diagnoses of whom three had a negative test at baseline, one tested negative 6 months earlier and one never tested. Among those testing at both time points, the HIV incidence was 8.3 per 100 pyfu (95% CI 0.0 to 17.7).

Conclusions This evaluation demonstrates a ‘recall for re-screening’ strategy is feasible in terms of high re-screening rates and incidence of new infections diagnosed. Experimental evidence is needed to assess cost-effectiveness and whether it achieves its aim of reducing transmission of STIs and HIV.

  • Homosexuality
  • male
  • sexually transmitted infections
  • bacterial
  • HIV
  • diagnosis
  • homosexuality
  • screening
  • diagnosis
  • infection

Statistics from


The incidence of most sexually transmitted infections (STIs) continues to rise among men who have sex with men (MSM) in the UK.1 New HIV diagnoses among this group, despite a recent modest decline, remain high and diagnoses are often made late, when the CD4 count is below the threshold at which treatment is recommended.2 3 Behaviours that put MSM at risk of HIV and other STIs are common with high proportions reporting unprotected anal intercourse (UPAI).4 The relationship between STI infection and increased HIV susceptibility and infectiousness, particularly for ulcerative STIs, has been long appreciated, and STI control remains an important strategy in the control of the HIV epidemic.5–7 Furthermore, there has been much attention focusing upon the early detection of HIV as a public health strategy, with ‘treatment as prevention’ resulting from the decline of the community HIV viral load8–10 and/or behaviour change which accompanies awareness of infectiousness11 12 contributing to a reduction in onward transmission.

MSM diagnosed as having an STI are at increased risk of future STIs.13 14 Recent risk behaviour and STI acquisition means that they are also at increased risk of HIV acquisition.15 Hence, early repeat STI screening of MSM who are diagnosed as having an STI could allow for early diagnosis of HIV and other STIs, with the potential to reduce HIV transmission through awareness of infection status and/or early treatment of STIs. There is a paucity of information on current re-screening practices among MSM diagnosed as having STIs in the UK, and there are no relevant recommendations in national guidelines. In Australia where guidelines recommend re-screening after 3 months of all MSM diagnosed as having Chlamydia trachomatis (CT) or Neisseria gonorrhoeae (GC),16 a recent study revealed that 3-month CT re-testing rates were <10%.17 A strategy of active recall for re-screening of MSM diagnosed as having a bacterial STI was introduced and evaluated in a central London clinic, and the findings are presented here.

The objectives of this evaluation were to assess the feasibility of a recall strategy for MSM in terms of re-screening rates and the number of incident infections diagnosed at re-screening.


Implementation of the intervention

This new service strategy was implemented in December 2008 in a central London sexual health clinic. Approximately 7000 MSM attend this service per year for sexual healthcare, which is free at the point of access. It is co-located with an HIV clinic to which newly diagnosed HIV-positive patients are referred. All MSM attending the sexual health service diagnosed as having a bacterial STI, including C trachomatis (CT), N gonorrhoeae (GC), syphilis and lymphogranuloma venereum (LGV), were identified. This was either at the baseline clinic visit or subsequently from a list of positive laboratory results that were reviewed by a ‘recall’ health advisor who had two sessions per week dedicated to running the recall for re-screening service. All identified were asked if they were willing to be recalled for repeat STI screening after a 3-month interval. Where consent was given, the patient was contacted by phone at 3 months and offered a new appointment for STI screening. Up to three phone attempts were made to contact each patient. Patients who did not attend a booked appointment were re-contacted and one further appointment was made. The ‘rescreening’ attendance was managed as per standard clinic protocol, which was in-line with national screening guidelines at that time.18 Screening tests offered to asymptomatic MSM included serological tests for HIV and syphilis, nucleic acid amplification tests (NAATs) using APTIMA Combo 2 (AC2) assay (Gen-Probe Inc., San Diego, California, USA) to detect CT and GC from a first-void urine or urethral specimen, and throat and rectal specimens for GC culture (Thayer Martin selective medium (Oxoid Ltd., Basingstoke, Hampshire, Hants.), incubated in 10% CO2, at 37°C for 48 h). Individuals reporting rectal symptoms were tested for rectal CT using AC2 assay, and positive samples were further tested for LGV serovars at the Sexually Transmitted Bacteria Reference Laboratory.

Evaluation of the intervention

The evaluation included all MSM diagnosed as having a bacterial STI during a 9-month period from December 2008. Ethics approval was not deemed to be required by the local research and development department as this was an evaluation of a service delivery. Details of all MSM diagnosed as having CT, GC, LGV or syphilis during this period, together with details of consent, attempt at contacting patients and clinic attendances were recorded on an Access database. Information relating to both the baseline and the re-screening appointment (where applicable), for each patient, was collected using clinic notes and a results database. Information recorded included baseline diagnosis and treatment, partner notification and treatment compliance, sexual behaviour at baseline and at re-screening, STI history, presentation type and diagnosis at re-screening appointment. For the purpose of this evaluation, patients were classified as having one of three outcomes: ‘attended for re-screening’ (AR), when the patient was contacted as part of the recall strategy and attended a re-screening appointment; ‘spontaneous attendance for screening’ (SA), when the patient re-attended within the recall period but not as a direct result of the recall strategy; and ‘failed to rescreen’ (FR), where the patient did not re-attend despite being offered recall for re-screening.


Analysis was performed using SPSS V16.0. The re-screening rate was calculated as the sum of individuals in the AR and SA groups as a proportion of all patients who were offered recall for re-screening. Differences in baseline characteristics between outcome subgroups (AR and SA vs FR) were examined using the Mann–Whitney test to compare medians for numerical data and the χ2 test to compare proportions. The 95% CIs for proportions were calculated using the simple method. The incidence of STIs diagnosed at re-screening was calculated as number of infections per 100 person-years follow-up (pyfu). Change in the median number of partners and the proportions reporting UPAI at re-screening appointment compared with baseline were investigated using Wilcoxin Ranks and McNemur tests, respectively. The association between baseline factors and infection at re-screening was investigated using Mann–Whitney test and χ2 test. Where an association approached statistical significance (p value <0.2), logistic regression was performed to calculate an OR.


Baseline characteristics

During the 9-month evaluation period, there were 337 MSM in whom a bacterial STI was diagnosed, see figure 1. Their median age was 32 years (range 19–79 years). The majority were of ‘white UK’ (43%) or ‘white non-UK’ (44%) ethnicity. UPAI in the preceding 3 months was reported by 51%. A history of previous STI was reported by 69%, and 7.4% (n=25) were known to be HIV positive, attending other HIV treatment centres. At baseline attendance, a further six patients (1.8%) were diagnosed as HIV positive, at which time 256 (76.0%) had a negative test and 50 (14.8%) did not have a HIV test. Of the latter, 34 (68%) had tested HIV negative within the preceding 6 months and 21 (42%) within the preceding 3 months. Attempts were made to obtain verbal consent from all patients (n=337) to recall for re-screening at 3 months.

Figure 1

Bacterial sexually transmitted infection STIs diagnosed at baseline appointment. CT, Chlamydia trachomatis; GC, Ngonorrhoeae gonorrhoeae and LGV, lymphogranuloma venereum.

Proportion attending for re-screening

Of the 337 patients who were eligible, 301 (89.3%) were offered the opportunity to be recalled for re-screening 3 months after their baseline appointment. This offer was declined by 30 patients (10.0%). Of those offered, 206 (68.4%) re-attended at a median of 3 months after their baseline attendance, 153 (50.8%) attended for re-screening as a result of the recall for re-screening strategy (AR) and 53 (17.6%) spontaneously re-attended for screening within the recall period but not as a direct result of the strategy (SA). Including those who declined the offer, 95 (31.6%) men failed to re-attend for re-screening (FR), and reasons are listed in figure 2. Those in the latter group (FR) were more likely to be HIV positive than those who were re-screened (AR and SA). Other differences in baseline demographics and risk behaviours were observed between subgroups but were not statistically significant (table 1).

Figure 2

Flow diagram demonstrating uptake of recall among eligible men who have sex with men (MSM).

Table 1

Comparison of baseline demographics, STI history and sexual behaviours by rescreening status

Symptoms at re-screening appointment

The majority of patients had no anogenital symptoms at re-screening appointment (156/207; 75.4%). Those in the SA group were more likely to be symptomatic at re-screening than those in the AR group: 49.1% and 11.8%, respectively (p<0.001).

Risk behaviour at recall

Among those re-attending, the proportion reporting UPAI in the 3 months preceding their re-screening appointment was significantly lower compared with their baseline appointment: 37.8% compared with 53.7%, respectively (p<0.001). The median number of partners reported in the previous 3 months declined from four at baseline to three at the re-screening appointment (p=0.056).

Bacterial STIs diagnosed at recall

There were 15 new bacterial STIs diagnosed at re-screening appointment (29 per 100 pyfu (95% CI 14.3 to 43.7)). Infections included GC (n=9), CT (n=5) and syphilis (n=1). There were nine and six bacterial STIs diagnosed in the AR and SA groups, respectively, giving incidence rates of 23.5 per 100 pyfu (95% CI 8.1 to 38.9) and 45.2 per 100 pyfu (95% CI 9.0 to 81.4), respectively.

Almost half (n=7) were infected with the same organism as at baseline (GC n=6, CT n=1), but only three of these had their infection diagnosed at the same site (GC n=2, CT n=1). Of those diagnosed as having GC at both time points, all were treated with first-line treatment regimens, as per national guidelines at that time, at baseline: four with cefriaxone 250 mg by intramuscular injection and two with cefixime 400 mg orally (neither had pharyngeal GC). In five of the six cases, the organism cultured at re-screening had a different resistance pattern, further suggesting re-infection with a new organism. The individual with CT diagnosed at both time points had been treated with erythromycin 500 mg twice daily for 2 weeks, a second-line treatment for CT shown to have 95% cure rate.19 All seven patients were confirmed to have been compliant and to have abstained from sexual activity for a prescribed time after treatment. Partner notification was completed for only three of the seven patients.

HIV diagnosis at recall

Of the 206 patients re-attending, nine were known to be HIV positive and 168 (86.0%) had an HIV test performed at re-screening. Of these, five had a positive HIV result. One patient had spontaneously attended (SA) due to symptoms of recurrent herpes simplex virus. The remaining four attended directly as a result of recall (AR). Among 144 patients who had a negative HIV test at baseline and a repeat HIV test at recall, three had a positive HIV result. Thus, the incidence of HIV acquisition among that group was 8.3 per 100 pyfu (95% CI 0.0 to 17.7). One patient had never previously tested for HIV, as he had previously declined testing, and the fifth (SA patient) had declined testing at baseline but had tested negative 4 months previously. None had experienced symptoms of primary HIV infection (PHI).

Factors predictive of infection at recall

No demographic or baseline behavioural factors were predictive of risk for STI diagnosis at re-screening. A diagnosis of N gonorrhoeae at baseline appointment was predictive of infection at recall, but the association did not reach statistical significance (table 2).

Table 2

Baseline factors predictive of infection at recall


This evaluation has demonstrated that recall for re-screening of MSM diagnosed as having a bacterial STIs is a feasible strategy both in terms of the high rates of re-screening achieved and the number of new diagnoses made. A minority (10%) of patients declined the service and 68% of MSM attended for re-screening a median of 3 months after their original diagnosis. There were non-significant differences between those attending for re-screening and those who failed to attend for re-screening, such that the latter reported less UPAI and a lower median number of partners at baseline. These differences were not statistically significant, but they do suggest that, while a larger sample may have allowed the determination of differences between these two groups, it is unlikely to demonstrate a positive association between lower risk behaviour and re-attendance for re-screening. Those attending for re-screening were less likely to be HIV positive, which could possibly be explained by individuals attending other treatment centres for their HIV and STI care.

High-risk sexual behaviours were common, with over a half of all MSM reporting UPAI at the baseline appointment. This proportion declined significantly by re-screening appointment. Explanations may include an STI diagnosis or interaction with a health advisor acting as a cue for behaviour change; compliance with instructions to abstain from SI for a period after treatment of baseline infection, reporting bias at re-screening appointment or selection bias at baseline appointment as all were diagnosed as having an STI, and thus, their preceding behaviour may have been was more risky than ‘normal’.

The rate of incident bacterial STIs diagnosed at re-screening was high: 29 per 100 pyfu. This is comparable to re-screening studies performed in the USA among general populations diagnosed as having CT or GC.20 21 The findings of these and similar studies led to a recommendation by the Centre for Disease Control that clinicians consider advising all women diagnosed as having CT, and all patients diagnosed as having GC, to re-screen 3 months after treatment.22 Data relating to UK populations and specifically to MSM are limited, and this is reflected by the lack of reference to re-screening in UK STI treatment guidelines.

The incidence of newly acquired HIV infection diagnosed at re-screening was high. Of the five new HIV diagnoses made, three had sero-converted within the preceding 3 months and a fourth within the preceding 6 months. Such early diagnoses may have important public health implications. Up to half of all onward transmissions are thought to take place during PHI,23 a risk that has been attributed to sexual risk behaviours and co-existing STIs at the time of PHI,24 in addition to high HIV viral load.25 26 A recall for re-screening strategy such as this could potentially be effective in reducing the incidence of HIV among MSM by reducing the risk of inadvertent transmission in early infection. This may be particularly relevant in the context of emerging HIV risk reduction practices among MSM such as ‘sero-sorting’ (UPAI with a HIV-negative partner).27 28 Early diagnosis of HIV through repeated testing has also been demonstrated to have important individual health benefits in terms of lower mortality and earlier starting of ART.29

Similar to the findings of other studies,14 risk behaviour at baseline among this cohort of MSM was not predictive of infection at re-screening. However, interpretation of this finding is limited by the small sample size of patients diagnosed as having STIs at re-screening, a failure to detect significant behavioural predictors of infection at re-screening may reflect a lack of power to do so. There was a non-significant association between N gonorrhoeae diagnosis at baseline and STI diagnosis at re-screening, again failure to demonstrate a significant association may be a reflection of the small numbers involved. The failure to identify predictive factors in this and other studies supports a strategy of recall for re-screening for all MSM diagnosed as having an STI.

An economic analysis has not been included in this evaluation, but evidence from elsewhere suggests that this is likely to be a cost-effective approach to STI control. A cost-effectiveness analysis in the USA which compared methods of encouraging patient return 3 months after a GC or CT diagnosis demonstrated that, while a telephone reminder system was more costly than other methods, it was the least costly method in terms of cost per new infection treated.30 Newer technologies being used in healthcare, such as short message service and email, may provide a less costly method of contacting patients, but similar high rates of re-screening may not be achieved. This would need further exploration. Use of short message service reminders for high-risk MSM in Australia resulted in a 64% re-screening rate, but this was measured over a 9-month period following initial presentation.31 Given that the estimated cost saving associated with averting a single HIV transmission is between £0.5 and £1 million,32 including the lifetime costs of ART (estimated to be between £280 000 and £360 000)33 and furthermore the potential annual saving of up to £13 000 for each individual who, through early recognition and treatment, does not progress to symptomatic HIV,32 a recall strategy such as this is likely be cost-effective or cost saving.

This evaluation demonstrates the potential benefits of a recall for re-screening strategy for MSM in the UK. The strategy resulted in high rates of re-screening with a high incidence of infections being diagnosed as a result. This allowed for timely provision of treatment and created the opportunities to interrupt further transmission of infection. However, there are some limitations to this evaluation; in the absence of a control group, this ‘rescreening strategy’ cannot be proven to be superior to ‘standard care’ in terms of increasing STI diagnosis. However, given the high rates of re-screening achieved, it is reasonable to hypothesise that this is the case. Subsequent to this evaluation being carried out, national guidelines have been modified such that rectal and pharyngeal CT and GC NAAT testing is now standard of care in asymptomatic MSM.34 35 Given the high prevalence (6–8%) of rectal CT infection demonstrated among asymptomatic MSM in the UK36–38 and the enhanced sensitivity of GC NAAT testing above culture, it is likely that the true incidence of bacterial STIs has been under-estimated in this cohort at baseline and re-screening. However, it is also plausible that the incidence of infection at re-screening has been over-estimated; the two cases where GC was diagnosed by culture at re-screening but not at baseline may represent persistence of infection where the baseline culture test yielded a false-negative result. Finally, the generalisability of this evaluation to other cohorts may be limited as it was conducted in a single central London centre.

This evaluation has demonstrated the feasibility of a ‘recall for re-screening’ strategy for MSM diagnosed as having bacterial STIs and has highlighted its potential public health benefits. There is now impetus to conduct prospective controlled studies to test the hypothesis that a strategy such as this is a cost-effective way of reducing the prevalence of HIV and STIs among MSM.

Key messages

  • A strategy of actively recalling all men who have sex with men diagnosed as having sexually transmitted infections after 3 months is feasible, with a re-screening rate of 68.4%.

  • A high number of new sexually transmitted infections and HIV infections were identified as a result of recall for re-screening.

  • Early re-screening of at risk men who have sex with men (those diagnosed as having sexually transmitted infections) may reduce transmission of sexually transmitted infections in particular HIV through early treatment and behaviour modification.


Staff at Mortimer Market Centre for contributing to the daily operation of the recall strategy.



  • Funding Camden Primary Care Trust was the funder of all services at Mortimer Market Centre. They provided extra funding for this new service strategy to be rolled out in the service. This extra funding was used to cover employ an extra health advisor to run the service.

  • Competing interests None.

  • Ethics approval Ethics approval was not deemed to be required by the local research and development department as this was an evaluation of a service delivery.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.