Retrospective study of the effect of enhanced systematic sexually transmitted infection screening, facilitated by the use of electronic patient records, in an HIV-infected cohort

Authors


Correspondence: Dr Gary Brook, Patrick Clements Clinic, Central Middlesex Hospital, Acton Lane, London NW10 7NS, UK. Tel: +44 0 208 453 2727; fax: +44 0 208 453 2224; e-mail: gary.brook@nhs.net

Abstract

Objectives

The aim of the study was to assess the impact of electronic checklists in enhancing sexually transmitted infection (STI) screening in routine HIV care.

Methods

This was a retrospective cohort study. In two HIV clinics, new STIs were recorded for three consecutive 12-month periods between 2009 and 2012 in a cohort of 882 HIV-infected patients. These three years coincided with the introduction of enhanced STI screening based on prompts within the electronic patient record (EPR) system.

Results

The number of diagnoses and the incidence of STIs more than doubled between 2010–2011 and 2011–2012 in both men who have sex with men (MSM) [from 18 of 115 (15%) to 42 of 132 (32%), a rise in STI incidence from 15.6 to 31.8/100 person-years; P < 0.001] and heterosexual patients [from six of 716 (0.8%) to 19 of 749 (2.5%), a rise in STI incidence from 0.8 to 2.5/100 person-years; P < 0.005]. The rise was significant in MSM for infections with chlamydia [from seven of 115 (6%) to 14 of 132 (11%), a rise in incidence from 6.0 to 10.6/100 person-years; P < 0.05], gonorrhoea [from five of 115 (4%) to 12 of 132 (9%), a rise in STI incidence from 4.3 to 9.1/100 person-years; P < 0.05] and early syphilis [from four of 115 (3%) to 13 of 132 (10%), a rise in incidence from 3.5 to 9.8/100 person-years; P < 0.001], but not for hepatitis C virus (HCV) and Lymphogranuloma venereum (LGV) infections. The rise was significant in heterosexual patients for infection with chlamydia [from four of 716 (0.6%) to 13 of 749 (1.7%), a rise in incidence from 0.6 to 1.7/100 person-years; P < 0.0001] but not for gonorrhoea, syphilis or Trichomonas vaginalis (TV).

Conclusions

These data show that implementing systematic, frequent and routine STI screening led to a large increase in detected STIs in this HIV-infected cohort. This process is greatly enhanced by the use EPRs.

Introduction

The diagnosis of sexually transmitted infections (STIs) in HIV-positive patients is important for several reasons. There is clear evidence that concurrent STIs increase the rate of transmission of HIV to uninfected sexual partners [1-4]. There is also some evidence that the complications of STIs can be worse in HIV-positive patients [1, 2, 5]. Some STIs, such as syphilis and hepatitis C, can be much more difficult to treat in HIV-positive patients, with lower response rates despite more intensive treatments, and there is also evidence of an adverse impact on HIV treatment response [5-8]. For these reasons and to ensure holistic care for HIV-positive patients, several guidelines recommend routine, ongoing STI testing in HIV-positive patients as part of their care [1, 2]. However, various audits in the UK and elsewhere have shown that comprehensive STI screening is often not achieved, with testing rates as low as 46% [9, 10].

The North West London Hospitals NHS Trust has provided care for its HIV-positive patients using an electronic patient record (EPR) system since 2008, and the use of this system has been shown to improve a range of health outcomes [11-13]. We introduced an ‘Annual Checklist’ in our EPR system at one of the clinics in 2009 and at the second clinic in 2010. The checklist is a way of prompting the doctors and nurses looking after the HIV-positive patients to ensure that all recommended interventions required for HIV-positive patients each year have been performed. The checklist prompts all health care workers involved in the patients' care to perform all recommended interventions each year. The effectiveness of this checklist in ensuring compliance with recommended interventions has previously been demonstrated [13]. After its first introduction in 2009, the success rate for ensuring that annual interventions took place was 84–90% and by the middle of 2011 94% of all patients were confirmed as having every intervention performed each year [13]. An annual STI screen for all HIV-positive patients, with more frequent screens as required by risk, is one of the key interventions on the checklist and so by 2011 both clinics were using this checklist. Also, in 2011 the extent of STI screens was further intensified by using proactive prompts for higher risk patients, such as men who have sex with men (MSM), to have more frequent and more comprehensive testing. We have therefore assessed how successful we have been in diagnosing STIs in our HIV-positive cohort since the introduction of the checklist and the proactive increased screening in higher risk patients.

Methods

Patients are managed using a paperless EPR system with consultations designed by ourselves using the Blithe ‘Lilie’ System (Blithe Computer Systems Ltd, Burton-on-Trent, UK). EPR enables large-scale review of any chosen parameters and thus a senior doctor reviews the STI risk assessments and test performance, including identifying patients at higher risk, and places prompts on the EPR for more frequent/extensive testing as required.

All HIV-positive patients with a concurrent acute STI were identified from our EPR system for the three years August to July 2009–2010, 2010–2011 and 2011–2012 and their demographic data, STI diagnoses and site of infection were recorded. During these three years our diagnostic test technologies were unchanged. Genital herpes and warts were not included as they are usually symptomatic and it can be difficult to distinguish between primary and recurrent infections. We also excluded STIs diagnosed within 6 months of the first HIV diagnosis as this was a study of new infection in HIV-positive patients in long-term care. The total number of HIV-positive patients in our care for each of the three years was taken as the number recorded at the mid-point of each year in question.

Statistics

The incidence of each STI was calculated as the number of STIs divided by the number of HIV-positive patients with that sexual orientation in the total cohort, assuming that the infections were acquired within 1 year.

Differences in proportions were measured using the χ2 test and with Yate's correction when necessary.

Results

The two clinics are based in north London, providing care for a cohort of just under 900 HIV-positive patients. The patients attending the clinics are largely from ethnic minorities and have mostly acquired HIV infection through heterosexual sex, with a significant minority of MSM. From our EPRs we know that, in 2009–2010 and 2010–2011, 716 of 696 (90%) and 806 of 831 (97%) patients, respectively, had at least one STI screen in that year, although the total number of screens was not recorded. In 2011–2012, 792 of 881 patients (97%) had at least one STI screen and STI screens were performed on a minimum of 1465 separate occasions in this cohort.

Table 1 describes the demographics of all patients in our HIV-infected cohort compared with the HIV-infected patients who were diagnosed with an STI within each of the three years in question. Those diagnosed with an STI were generally younger (in the first and third years), male and more likely to be MSM. The differences in the proportions of male and female patients (P < 0.05) and of MSM and heterosexual patients (P < 0.05) in those with and without an STI were statistically significant. There was a small rise in the number of patients with a detected STI between 2009–2010 and 2010–2011 but the increase between 2010–2011 and 2011–2012, from 21 to 41, was almost twofold (P < 0.001).

Table 1. Demographics of HIV-positive patients with and without a sexually transmitted infection (STI)
YearType of patientTotal numberAge (years) [median (range)]Gender [n (%)]Sexual orientation [n (%)]
  1. F, female; Het, heterosexual; M, male; MSM, men who have sex with men.
2009–2010Total cohort79641 (16–78)

399 (50) F

397 (50) M

106 (13) MSM

690 (87) Het

Patients with an STI1928 (20–55)

4 (22) F

14 (78) M

11 (61) MSM

7 (39) Het

2010–2011Total cohort83142 (18–87)

407 (49) F

424 (51) M

115 (14) MSM

716 (86) Het

Patients with an STI2140 (23–57)

5 (24) F

16 (76) M

15 (71) MSM

6 (29) Het

2011–2012Total cohort88143 (19–82)

427 (48) F

454 (52) M

132 (15) MSM

749 (85) Het

Patients with an STI4134 (19–57)

7 (17) F

34 (83) M

25 (61) MSM

16 (39) Het

The incidence of new STIs within the three years are given for HIV-positive MSM in Table 2 and HIV-positive heterosexual patients in Table 3. There was a small rise in detected STIs between 2009–2010 and 2010–2011 but the increase between 2010–2011 and 2011–2012 was more than twofold for MSM, from 18 to 42 STIs (P < 0.001), and more than threefold for heterosexual patients, from six to 19 STIs (P < 0.005). The rise was significant in MSM for chlamydia (from seven to 14; P < 0.05), gonorrhoea (from five to 12; P < 0.05) and early syphilis (from four to 13; P < 0.001), but not for hepatitis C virus (HCV) and Lymphogranuloma venereum (LGV). Similarly, the rise was significant in heterosexual patients for chlamydia (from four to 13; P < 0.0001) but not for gonorrhoea, syphilis or Trichomonas vaginalis (TV). In 2011–2012, 61 STIs were diagnosed in 41 patients, with 10 patients (mostly MSM) having multiple STIs of up to four infections.

Table 2. Acute sexually transmitted infections (STIs) diagnosed in HIV-positive men who have sex with men, by year
Year All STIsCTGCEarly syphilisAcute HCVLGV
  1. CT, chlamydia; EL, early latent; GC, gonorrhoea; HCV, hepatitis C virus; LGV, Lymphogranuloma venereum; NA, not applicable; Prim, primary; Rec, rectal; Sec, secondary; Th, throat; Ur, urethral.
  2. *Statistically significant rise from 2010–2011 to 2011–2012.
2009–2010Total1237200
SiteNA2 Ur, 1 Th7 Ur1 Prim, 1 SecNANA
Symptoms6 Yes, 6 No1 Yes, 2 No5 Yes, 2 No2 NoNANA
Incidence rate/100 person-years11.32.86.61.900
2010–2011Total1875420
SiteNA5 Ur, 2 Rec4 Ur, 1 Rec3 Prim, 1 ELNANA
Symptoms7 Yes, 11 No2 Yes, 5 No3 Yes, 2 No2 Yes, 2 No2 NoNA
Incidence rate/100 person-years15.66.04.33.51.70
2011–2012Total42*14*12*13*12
SiteNA13 Ur, 1 Rec7 Ur, 2 Rec, 2 Th, 1 Th+Rec5 Prim, 5 Sec, 3 ELNA2 Rec
Symptoms19 Yes, 23 No5 Yes, 9 No6 Yes, 6 No7 Yes, 6 No1 No1Yes, 1 No
Incidence rate/100 person-years31.810.69.19.80.81.5
Table 3. Acute sexually transmitted infections (STIs) diagnosed in HIV-positive heterosexual patients, by year
Year All STIsCTGCEarly syphilisTV
  1. CT, chlamydia; F, female; GC, gonorrhoea; Gen, genital; M, male; NA, not applicable; TV, Trichomonas vaginalis.
  2. *Statistically significant rise from 2010–2011 to 2011–2012.
2009–2010Total7 (3M 4F)6 (3M 3F)001 (F)
SiteNA6 GenNANA1 Gen
Symptoms1 Yes, 6 No1 Yes, 5 NoNANA1 No
Incidence rate/100 person-years1.00.9000.1
2010–2011Total6 (1M 5F)4 (1M 3F)002 (2F)
SiteNA4 GenNANA2 Gen
Symptoms7 Yes, 11 No1 Yes, 3 NoNANA2 Yes
Incidence rate/100 person-years0.80.6000.3
2011–2012Total19 (10M 9F)*13 (8M 5F)*2 (1M 1F)1 (1M)3 (3F)
SiteNA13 Gen2 Gen1 Prim3 Gen
Symptoms18 Yes, 24 No1 Yes, No 122 No1 Yes3 No
Incidence rate/100 person-years2.51.70.30.10.4

Discussion

Summary

This study demonstrates that the rate of newly diagnosed STIs has risen in this cohort of HIV-infected patients over the last 3 years at a time when we also introduced systematic proactive measures to increase STI testing. This rise was especially marked in the last 12 months (2011–2012), when we observed a more than doubling of detected STIs in MSM and a more than trebling in heterosexual patients compared with the previous 12-month period (2010–2011). During the last year, the total number of patients with at least one STI also doubled. The rise in STIs in this cohort was largely related to increases in chlamydia, gonorrhoea and early syphilis in MSM and in chlamydia in heterosexual patients. Although some of these increases may be attributable to a genuine rise in STI incidence, it is likely that better diagnosis is responsible as a result of the introduction of prompts on our EPR system to aid more systematic STI screening.

Interpretation of the results

We introduced the annual checklist for HIV-infected patients, an electronic ‘to do’ list, into the EPR system to facilitate various interventions, including STI screening, in one clinic in 2009 and across both clinics in 2010 [13]. Data published elsewhere showed that this led to successful routine STI screening of 94% of eligible HIV-infected patients in the first of our clinics to pilot it by 2010 [13]. In 2010 we also introduced a more proactive recall system whereby all HIV-positive patients at increased risk of STIs were flagged in their EPR to have more frequent and comprehensive STI testing. Coincident with these changes we have seen a large increase in diagnosed STIs.

In MSM the most significant rise was in chlamydia infection, which increased from three to 14 cases a year, equating to an incidence now of 10.6 cases per 100 patient-years of follow-up. The majority (nine of 14; 64%) were asymptomatic, showing that the diagnosis would not have been achieved without screening. Similar high rates of chlamydia have been seen in other HIV-infected cohorts [14-16], although in our patients the main site of infection was in the urethra as opposed to a higher prevalence of rectal infections seen in other cohorts [14, 15, 17]. The size of the increase in diagnosed chlamydia was also unusual in that we saw chlamydia infections increase by 100% between 2010–2011 and 2011–2012, which is far greater than the national trend of a 10% increase per year in MSM [18]. We diagnosed rectal LGV in two HIV-positive MSM, of whom one was asymptomatic and the other had mild symptoms. In most studies of rectal LGV, the infection is symptomatic in the majority [17, 19-21] and so it is of interest that one of our two cases did not have symptoms and was found through proactive screening. Another benefit of enhanced screening is that the anus is examined more frequently in a group of men at higher risk of anal cancer [22].

A large rise in gonorrhoea in HIV-positive MSM was seen in our cohort, from five to 12 cases in the last year, and reaching an annual incidence of 9.1 cases per 100 patient-years of follow-up. Again, a high proportion (six of 12; 50%) were asymptomatic. Gonorrhoea incidence in MSM has been rising by about 10% per year since 2009 in the UK [18] and we saw a rise of 140% in this cohort. The proportions with urethral, rectal and pharyngeal infections in our cohort were similar to those seen in other studies [23].

The rise in early syphilis diagnosed in our HIV-positive MSM was even more dramatic, from four to 13 cases in 2011–2012 at an annual incidence of 9.8 cases per 100 patient-years of follow-up. Asymptomatic infection was detected through routine screening in six of 13 patients (46%), although several of the symptomatic cases were also diagnosed at screening and only then did previously unidentified symptoms come to light. This 225% increase in incidence in our cohort is unusual in that early syphilis in MSM nationally has not been rising recently [18]. Several studies have shown the need for routine surveillance of HIV-positive MSM for syphilis because of the significant incidence in this group [24-27]. The high rate of syphilis in our patients might in part be explained by better testing, although we would expect to do an annual syphilis test in virtually all patients anyway and so the diagnosis would have come to light eventually. Therefore, this rise seems to reflect a real rise in local incidence, and the reason for this is not clear. However, we probably made the diagnosis several months earlier than would have happened previously and may have prevented progression to later stage disease [5].

Enhanced screening of heterosexual HIV-positive patients led to a rise in diagnosed STIs from six to 19 cases in the last year, although this was explained largely by an increase in detected chlamydia infection. The number of chlamydia infections seen in the last year was 13, rising by 225% from the four cases diagnosed in the previous year as set against an overall rise of just 4% nationally [28]. The majority (12 of 13; 92%) were asymptomatic and therefore this points to the rise in detected infection being attributable to more intensive screening. This relatively high incidence of chlamydia in our HIV-positive heterosexual patients contrasts with experience elsewhere where low prevalence has been argued as a reason not to screen this group [29]. We saw a small rise in detected TV infection, although the number of detected cases remains very low in this cohort. This is in contrast to other countries were TV is one of the commonest STIs in HIV-positive women [30, 31]. We do not offer vaginal swabs for TV in asymptomatic women, although one may be taken when we perform a cervical Pap smear and in women with vaginal discharge. It is therefore possible that we may have missed some cases. TV is the most common STI world-wide and there is evidence for TV enhancing the transmission of HIV [31]. There is therefore an argument for more enhanced screening, especially using newer diagnostic technologies [31].

In future research we need to look at further enhancing STI screening, for instance by using even more regular recall of patients for STI screening outside of their normal HIV attendances. Increased use of newer technologies such as TV polymerase chain reaction (PCR) is also likely to increase diagnostic accuracy.

Relation to other evidence

Sexual risk assessment and appropriate STI screening in HIV-positive patients are encouraged in many guidelines [1, 2] although audit elsewhere shows that screening can be patchy and suboptimal in many cases [9, 10]. What we have shown is that, with a systematized approach to STI screening using the facilities inherent in EPR, we have dramatically increased the rate of STI diagnoses in both MSM and heterosexual patients. This is likely to have been from a baseline screening rate similar to that in many HIV clinics in the UK. There is plentiful evidence that STIs enhance the transmission of HIV, can be more difficult to treat in HIV-positive patients and may adversely affect the course of HIV infection [1-4]. Therefore, by increasing the rate of STI diagnosis in this cohort, the risk of these problems will be reduced. It is also noteworthy that the median age of patients with an STI in this HIV cohort is much higher than in the HIV-negative populations [18, 28]. This association has been recognized before [21, 23, 26] and emphasizes the need to screen all sexually active HIV-positive patients irrespective of their age.

Strengths and limitations

The strength of this study is that it compares patient data from the same cohort over contiguous time periods, in a normal clinical situation, and is therefore a reflection of our standard clinical practice. It is also the first paper to demonstrate how EPR can help to improve STI screening within the setting of an HIV clinic. As we excluded STIs diagnosed within 6 months of the HIV diagnosis, this is a more valid reflection of the incidence of STIs in patients in an established HIV-infected cohort. It is likely that the incidence of STIs in this study is an underestimate, in that many of the patients may have gone to other services for the diagnosis and treatment of STIs during the year.

Conclusions

Our data suggest that systematic, frequent and routine STI screening is required for all HIV-positive patients and leads to a large increase in detected infections. This process is greatly enhanced by the appropriate use of technology, such as EPR.

Acknowledgements

We would like to acknowledge the efforts of the rest of the clinic staff who were key to the success of the EPR system and the implementation of STI screening.

Conflicts of interest:

GB has performed training and lecturing on the use of EPRs for Blithe Computer Systems Ltd in return for educational donations to the clinic. For the remaining authors no conflicts of interest are declared.

Funding:

No funding for this study was received.

Author contributions:

Contributors GB and JM planned the study and analysed the data. GB wrote the first draft of the paper and AS and JM contributed to the writing of subsequent versions.

Ancillary