Background: Analysis of self-collected swab samples from the genital tract could improve accrual and retention of women in studies of human papillomavirus (HPV) infection and precancerous cervical lesions. Self-collected vaginal swab specimens and physician-collected cervical swab specimens were compared for detection and typing of HPV DNA in 158 HIV-seropositive women.
Methods: Paired samples were collected for 157 participants. Beta-globin was not detected in 6 (3.3%) physician-collected specimens and 8 (4.3%) self-obtained specimens collected from 11 women, leaving 146 paired samples suitable for PCR analysis. HPV DNA was amplified with the HPV primers PGMY09 and PGMY11 and typed using the line blot assay.
Results: HPV DNA was detected more frequently in self-collected samples (95 [65.1%] of 146), compared with physician-collected samples (78 [53.4%] of 146) (P = .04). Self-collected samples contained a greater number of types (mean +/- SD, 1.60 +/- 1.80 types; 95% confidence interval [CI], 1.31-1.90), compared with physician-collected samples (mean +/- SD, 1.25 +/- 1.66 types; 95% CI, 0.98-1.52) (P = .04). A good agreement between sampling methods was achieved for detection of any HPV DNA (kappa = 0.73; 95% CI, 0.58-0.89), high-risk types (kappa = 0.84; 95% CI, 0.68-0.99), and low-risk types (kappa = 0.71; 95% CI, 0.67-0.75). Agreement between sampling methods for detection of HPV DNA was found for 24 (88.8%) of 27 follow-up samples collected from a total of 20 women. A comparison of samples collected at consecutive visits revealed agreements for detection of any HPV DNA, detection of high-risk HPV, and HPV typing results between visits of 88.9% (24 of 27 samples), 81.5% (22 of 27), and 55.5% (15 of 27), respectively, for physician-collected samples, and 96.3% (26 of 27 samples), 92.6% (25 of 27), and 55.5% (15 of 27), respectively, for self-collected samples.
Conclusion: Analysis of self-collected vaginal swab samples improved the detection rate of HPV, suggesting that such samples might be of greater value than physician-obtained samples in studies of HPV transmission.