Objective Trichomoniasis vaginalis is a risk factor for the acquisition and transmission of HIV. The objective of this study was to determine the prevalence of T vaginalis (using culture) among HIV-infected women receiving gynaecological care at an university HIV clinic in Alabama in addition to predictors of infection.
Methods Electronic medical record review of women presenting to the clinic for gynaecological care during 2006–2012 was performed. Demographic and sexual history data was abstracted in addition to absolute CD4 cell count, HIV-1 viral load and sexually transmitted infection (STI) (including T vaginalis) testing results. Analysis was conducted using Stata V.12.
Results T vaginalis was prevalent in 17.4% (83/478) of HIV-infected women; other STIs were less prevalent. Among these women, 384 presented for routine STI screening, of which 12% (46/384) were T vaginalis-infected. Younger age, African-American race, lifetime history of tobacco and drug abuse, lack of HIV therapy, HIV-1 viral load >400 copies/ml, and report of seeking gynaecological care for reasons other than routine STI screening (ie, having symptoms) were significant predictors of T vaginalis in univariate analysis. Age, African American race, and report of seeking gynaecological care for reasons other than routine STI screening remained associated with T vaginalis in multivariable analysis.
Conclusions T vaginalis remains highly prevalent among HIV-infected women, a proportion of which may be asymptomatic. If left undiagnosed and untreated, these women may be more likely to transmit HIV. Increased emphasis on screening for high risk sexual behaviours, testing for T vaginalis, and risk reduction counselling is necessary for all HIV-infected women.
- HIV WOMEN
- CLINICAL STI CARE
Statistics from Altmetric.com
Trichomoniasis vaginalis is the most common non-viral sexually transmitted infection (STI) in the world.1 The annual incidence of T vaginalis in the USA has been estimated to be 3–5 million cases and the prevalence 3.1%.2 Risk factors for infection among women include increasing age, African–American race, greater numbers of lifetime sexual partners, sex with casual or older partners, lower educational level, poverty, injection drug use and crack cocaine use.2–6 However, despite being readily diagnosed and treated, T vaginalis is not a reportable disease and its control has not received emphasis from public health programmes.
T vaginalis is recognised as a risk factor for the acquisition and transmission of HIV.7–13 Among HIV-infected women, T vaginalis prevalence is estimated to be between 5–30%, depending upon geographical location.14–18 The US Centers for Disease Control and Prevention recommend routine screening for curable STIs (Chlamydia trachomatis, Neisseria gonorrhoeae, syphilis and T vaginalis) among HIV-infected women as a means of prevention of HIV transmission.19 The objectives of this study were to determine the prevalence of T vaginalis among HIV-infected women at the University of Alabama at Birmingham (UAB) HIV clinic during 2006–2012 in addition to predictors of infection.
This study was approved by the UAB Institutional Review Board. Electronic medical record review of HIV-infected women aged ≥19 years and presenting to the UAB HIV Clinic (a large outpatient university clinic) for gynaecological care (including routine annual visits and symptomatic visits) during 2006–2012 was performed using a standardised data extraction form. For simplification purposes, only data from the first gynaecological visit per patient during this timeframe was reviewed. Demographic and sexual risk behaviour data were abstracted in addition to patients’ use of highly active antiretroviral therapy (HAART) at the time of the visit. Patients’ absolute CD4 cell count (UAB Flow Cytometry),20 HIV-1 viral load (Roche Diagnostics, Indianapolis, Indiana, USA), STI testing results and Pap smear results (ThinPrep Pap Test, Marlborough, Massachusetts, USA) were recorded, if obtained at the time of the visit or within 3 months of the visit. At this clinic, women are tested for T vaginalis using wet prep and InPouch culture (BioMed Diagnostics, White City, Oregon, USA), C trachomatis and N gonorrhoeae using the Becton Dickinson ProbeTec C trachomatis/N gonorrhoeae amplified DNA assay (Franklin Lakes, New Jersey, USA) on endocervical and urine specimens, and syphilis using the Rapid Plasma Reagin card test (Becton Dickinson, Franklin Lakes, New Jersey, USA).
Extracted data were scanned into a Microsoft Access database using Teleform software. Analysis was conducted using Stata V.12 (College Station, Texas, USA). Percentages were calculated for categorical risk factors, with continuous variables dichotomised at their clinical cut-points for ease of interpretation. Unadjusted associations between each demographic/behavioural/clinical factor and T vaginalis were examined using logistic regression analysis. An additional adjusted model was then calculated to determine independent predictors of T vaginalis which included all variables significantly associated with T vaginalis in univariate analysis. Cases with missing values were excluded from the logistic regression analyses. For all analyses, a p value of <0.05 was considered statistically significant.
Between March 2006–February 2012, 487 HIV-infected women presented to the UAB HIV clinic for gynaecological evaluation. Data analysis was restricted to African American and white women (n=478) due to the small number of women of other races in the sample (n=9). Table 1 presents the demographic and clinical characteristics of this sample with T vaginalis prevalence noted for each characteristic. Age ranged from 19–75; mean age was 42.3 (SD±10.0). The majority of women were African-American (73%, n=350). Ever having smoked (40%, n=193) and using illegal drugs (26%, n=125) were common, though alcohol abuse was less so (13%, n=63). A large proportion of women reported not using condoms (36%, n=174), though the large amount of missing data on this variable resulted in it being excluded from further analysis. The majority of women were on HAART at the time of the visit (71%, n=339) and had absolute CD4 counts ≥200 cells/mm3 (75%, n=358) and HIV-1 viral loads ≤400 copies/ml (56%, n=266). Routine STI screening was the most commonly reported reason for seeking gynaecological care (80%, n=384) while 12% of women (n=56) sought gynaecological care for other reasons (the majority of which were symptomatic, n=52). It was unknown as to why the remainder of women sought care (8%, n=38). Data on the presence of vaginal discharge on exam, vaginal pH and wet prep results were missing for a large proportion of women and are not reported.
T vaginalis infection was common with 17.4% (83/478) of women testing positive. As depicted in table 1, 12% (46/384) of 384 women presenting to the clinic for routine STI screening and presumably asymptomatic tested positive for T vaginalis. Of the 56 women with known reasons for seeking gynaecological care, 41% (23/56) were found to be T vaginalis-infected. Of these 23 T vaginalis-infected women with known reasons for seeking gynaecological care, 91% (21/23) were symptomatic (data not shown). Finally, among women with an unknown reason for seeking gynaecological care, 37% (14/38) were found to be T vaginalis-infected.
Other STIs including C trachomatis, N gonorrhoeae and syphilis were less prevalent (data not shown). C trachomatis was found in 1.0% (5/466) of women, N gonorrhoeae in 0.2% (1/467) and syphilis in 17/231 (3.6%). Atypical squamous cells of unknown significance was noted on Pap smear in 17% (79/474) of women, 14% (68/474) had low grade squamous intraepithelial lesions and 3% (14/474) had high grade squamous intraepithelial lesions.
Numerous demographic and clinical factors associated with T vaginalis are presented in table 2. In univariate analysis, age was negatively associated with T vaginalis (OR 0.95; 95% CI 0.93 to 0.98; p=0.01). African-American women were more likely to have T vaginalis compared with white women (OR 2.2; 95% CI 1.2 to 4.1; p=0.01). Those who had ever smoked (OR 1.7; 95% CI 1.1 to 2.9; p=0.02) and who had ever used illegal drugs (OR 1.8; 95% CI 1.0 to 3.0; p=0.03) were more likely to have T vaginalis than women who had not. Women currently on HAART at the time of the visit were less likely to have T vaginalis (OR 0.5; 95% CI 0.3 to 0.8; p=0.01) while women with HIV-1 viral loads >400 copies/ml were more likely to have T vaginalis (OR 2.5; 95% CI 1.5 to 4.2; p=0.01). Women who sought gynaecological care for routine STI screening were less likely to have T vaginalis (OR 0.2; 95% CI 0.1 to 0.4; p=0.01) than those who sought care for other reasons (the majority of which were symptomatic).
An adjusted multivariable model (table 2) including all significant univariate predictors of T vaginalis revealed that age, African-American race and report of seeking gynaecological care for reasons other than routine STI screening remained strongly associated. Additional analyses (not shown) were conducted to explore the possible interaction between age and race in this sample. However, no evidence was found of an age-race interaction effect, which may be due to the limited number of cases available for analysis.
Among 478 HIV-infected women presenting to the UAB HIV clinic for gynaecological evaluation during 2006–2012, T vaginalis prevalence (based on culture) was 17.4%. Among 384 women presenting for routine STI screening, 12% had T vaginalis. Similar to prior studies,4 ,14 ,18 ,21 ,22 these data suggest that T vaginalis remains a highly prevalent STI among HIV-infected women, a proportion of which may be asymptomatic. Younger age, African-American race and report of seeking gynaecological care for reasons other than routine STI screening (ie, having symptoms) were independently associated with T vaginalis.
Although our study was not designed to examine the impact of T vaginalis screening among HIV-infected women at the UAB HIV clinic over time, it is interesting to note that the current prevalence of 17.4% is substantially higher than the prevalence of 5.3% (also based on culture) noted among HIV-infected women participating in a natural history study at this clinic between 1996–2001.18 Nevertheless, it is difficult to draw any meaningful conclusions from this observation as the study design of the two studies are different and the natural history study took place over a decade ago. A future study examining the prevalence of T vaginalis in the USA over time, similar to that which has been conducted in Australia,23 would be of considerable interest to determine the impact of recommended T vaginalis screening in clinic populations.
Similar to our study, others have documented T vaginalis to be highly prevalent among African-American women in major US urban centres.2 ,7 Explanations for this racial disparity are likely multifactorial and may include a high prevalence of T vaginalis among male sexual partners, decreased use of barrier protection, participation in high-risk sexual behaviours, lack of access to healthcare and/or distrust of the healthcare system leading to lack of STI screening, compliance with treatment and/or partner referral for therapy.7 In contrast to prior studies which found that increasing age was associated with prevalent T vaginalis2 ,24 or that age was not associated with T vaginalis,25 younger age was found to be independently associated with T vaginalis in our study. This finding is similar to the epidemiology of other curable STIs such as C trachomatis and N gonorrhoeae in which the prevalence is highest among adolescent and young adults26 and is corroborated by another study which found that prevalent T vaginalis was significantly more likely among HIV-infected women <22 years.22 This finding may reflect an increased likelihood of participation in sexual risk behaviours and/or inconsistent screening for STIs, including T vaginalis, among younger HIV-infected women. Additionally, rates of concurrent STIs including C trachomatis/N gonorrhoeae were low in our study compared with T vaginalis, similar to other studies of HIV-infected women.18 ,22 This may be due to the lack of primary prevention programmes for T vaginalis as it is not a reportable disease or due to the availability of more sensitive methods of C trachomatis/N gonorrhoeae screening (NAATs) compared with less sensitive methods of T vaginalis screening (culture). Finally, T vaginalis was not associated with lower absolute CD4 counts in this study, similar to prior studies,4 ,14 ,22 and does not appear to be an opportunistic infection in HIV-infected women.
To date, T vaginalis diagnosis has relied primarily on vaginal wet prep and culture, of which many clinics do not have access. The sensitivity of culture is 75–89% compared with the recently available T vaginalis nucleic acid amplification test (NAAT),27 a highly sensitive test which has been validated in asymptomatic and symptomatic women that can be combined with C trachomatis/N gonorrhoeae NAATs and used on multiple specimen types including first catch urine, vaginal and endocervical specimens, and endocervical ThinPrep specimens.28 It will be of considerable interest to re-examine the prevalence of T vaginalis among HIV-infected women at this clinic once implementation of T vaginalis NAAT testing occurs.
Our study has several limitations. First, data were collected retrospectively limiting the amount of sexual risk behaviour and clinical data (including rates of bacterial vaginosis) extrapolated. Second, the prevalence of T vaginalis, based on culture and not NAAT, is likely underestimated. Finally, as data from the first gynaecological visit per patient during the study timeframe was reviewed, we were unable to determine whether women with T vaginalis had a new infection versus a recurrent or persistent infection.
Despite these limitations, the results of our study have several public health implications. First, T vaginalis remains highly prevalent among HIV-infected women, indicating ongoing participation in sexual risk behaviours by these women and/or their partners. A proportion of women may be asymptomatic and should still receive the recommended screening for T vaginalis.19 ,29 If left undiagnosed and untreated, HIV-infected women with T vaginalis may be more likely to transmit HIV. Increased emphasis on screening for sexual risk behaviours, testing for T vaginalis and other curable STIs, and risk reduction counselling is necessary for all HIV-infected women. STI screening should occur at the initial HIV clinic visit and be repeated periodically (at least annually) if the patient is sexually active or earlier (3–6 month intervals) for women at higher risk.29 Second, implementation of T vaginalis NAAT testing into routine STI care for HIV-infected women is likely to improve the diagnosis of this common, curable infection and is pertinent to STI control efforts in the USA and elsewhere. Finally, additional interventions to diagnose, treat and prevent T vaginalis in HIV-infected women should be a high priority of routine care.
Prevalence of Trichomonas vaginalis among this cohort of HIV-infected women was 17.4%; other sexually transmitted infections (STIs) were less prevalent.
Age, African-American race, and report of seeking gynaecological care for reasons other than routine STI screening (ie, having symptoms) were independently associated with T vaginalis infection.
Increased emphasis should be placed on T vaginalis screening among HIV-infected women; if left untreated, these women may be more likely to transmit HIV.
The authors would like to thank Marga Jones, MSHI, for her assistance with data management during this study and Edwin Swiatlo, MD, PhD for helpful discussions on manuscript preparation.
Contributors Conception and design: CAM and JRS. Acquisition of data: CAM, CAR and JRS. Analysis/interpretation of the data: CAM, ELA and JRS. Drafting of the manuscript: CAM and ELA. Revising it for intellectual content: CAM, ELA, CAR and JRS. Final approval: CAM, CAR, ELA and JRS.
Competing interests None.
Ethics approval University of Alabama at Birmingham Institutional Review Board.
Provenance and peer review Commissioned; externally peer reviewed.
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.