Question
Which types of specimens are important for the detection of T. vaginalis infection in men, and does the testing of specimens from multiple sites increase the sensitivity of testing?  

Design
Multiple genitourinary specimens were cultured for T. vaginalis from men in Malawi to determine the relative sensitivities of different specimens and to examine the consequences of using a single specimen for testing.  

Participants
Subjects (n = 1361) who were enrolled in a clinical trial of the addition of metronidazole to syndromic treatment of urethritis in men or a longitudinal study of HIV and T. vaginalis infection in men from the STD (n = 929) and dermatology clinics (n = 432) of Lilongwe Central Hospital in Malawi were tested.   Subjects ranged in age from 18 to 63 years (mean = 26.2 years).  

Description of Tests and Diagnostic Standard
At enrollment, blood was obtained for HIV and syphilis serologic testing.  Urethral swabs were obtained for T. vaginalis culture and Gram stain.  Urethritis was defined as >4 white blood cells per high power microscopic field on Gram stain.  Twenty to 30 mL of first void urine was collected after the physical examination for T. vaginalis culture.  Men who were HIV seropositive with no diagnosis of any other STD and men infected with T. vaginalis provided follow-up specimens at 2 or 3 visits within the next 4 weeks.  These specimens included urethral swabs for T. vaginalis culture and Gram stain, and first void urine and semen for T. vaginalis culture.  

The swabs and sediment, obtained by centrifugation of the urine and semen specimens, were inoculated into the InPouch TV culture system (Biomed).  The cultures were examined on days 2 and 5.  A positive culture was defined as the visualization of parasites with morphology and motility characteristic of T. vaginalis.  Infection with T. vaginalis was defined as the presence of motile trichomonads in at least 1 of the specimen cultures.   

Main Outcome Measures
The sensitivity of each specimen type was calculated relative to the other specimen types.  

Main Results
T. vaginalis was detected in 98 (7.2%) of 1361 men enrolled at both clinics.  The results of T. vaginalis cultures performed on 1791 pairs of urine specimens and urethral swabs obtained at enrollment and follow-up visits are shown in Table 1.  The sensitivities of T. vaginalis cultures of urine and urethral specimens were 66.4% and 66.3%, respectively.  The results of T. vaginalis cultures performed on 288 sets of urine, urethral swabs, and semen samples obtained at follow-up visits are shown in Table 2.  The sensitivities of T. vaginalis cultures of urethral swabs, urine specimens, and semen specimens were 62.2%, 61.5%, and 66.7%, respectively.  For subjects from whom all 3 specimens were cultured, urethral swabs and urine specimens were more sensitive among subjects with urethritis than among subjects without urethritis, while semen appeared to be more sensitive among subjects without urethritis compared to subjects with urethritis.  Among subjects who were diagnosed with T. vaginalis infection on at least 2 visits, the type of specimen that tested positive changed at least once for 59.5%.   

Table 1.  Results of T. vaginalis cultures on 1701 paired urethral and urine specimens from 1361 men attending STD and dermatology clinics in Malawi.  

Table 2.  Results of T. vaginalis cultures on 288 sets of urethral, urine, and semen specimens from 146 men attending STD and dermatology clinics in Malawi.  

Authors’ Conclusions
Urine specimens, urethral swabs, and semen specimens together identified the greatest number of cases of trichomoniasis.  The use of multiple specimens increased the number of cases by 31%.    

Source of funding:  National Institutes of Health, University of North Carolina Fogarty Center, National Institute of Diabetes and Digestive and Kidney Diseases, UNC HIV Prevention and Treatment Network, and UNC STD Cooperative Research Center.

For correspondence:  Marcia M. Hobbs, Department of Medicine, Division of Infectious Diseases, CB 7031 UNC School of Medicine, Chapel Hill, NC 27599.  E-mail address:  HYPERLINK "mailto:mmhobs@med.unc.edu" mmhobs@med.unc.edu.