Table 2

Changes in reported vaginal cleansing practices and (sexual) behaviour between the enrolment and the M6 visit

Self-reported sociodemographic characteristicsEnr
(n=71)
M6
(n=65)
OR (95% CI)*
P value*
Reports using no products inside the vagina (other than for managing menses; all participants) n (%)35 (49.3)53 (81.5)5.2 (1.96 to 17.34)
<0.001
Reports using no products inside the vagina (other than for managing menses; controls and metronidazole users only)† n (%)15 (44.1)27 (79.4)13.0 (1.95 to 552.5)
0.002
Reports using water only n (%)23 (32.4)10 (15.4)0.37 (0.13 to 0.92)
0.029
Reports using water and soap n (%)3 (4.2)2 (3.1)0.67 (0.06 to 5.82)
1.00
Reports using paper, cloth or cotton wool n (%)9 (12.7)0 (0)0.13 (0.00 to 0.93)‡
0.008
Reports using traditional herbs, stones, powders as vaginal cleansing practice n (%)1 (1.4)1 (1.5)1.00 (0.01 to 78.5)‡
1.00
Mean weekly frequency of vaginal practices (95% CI)2.15
(0.97 to 3.34)
0.64
(0.18 to 1.11)
NA
0.328
Median no of sex partners in last month at baseline or per month during follow-up period (IQR)5
(3–16)
2
(1–4)
NA
<0.001
Any condom use reported in past 2 weeks (Enr) or since last study visit (M6), vs no condom use reported n (%)64 (90.1)60 (92.3)1.67 (0.32 to 10.7)
0.727
Reports exchanging sex for money/goods in past month (Enr) or since last study visit (M6) n (%)65 (91.5)58 (89.2)0.80 (0.16 to 3.72)
1.00
  • *McNemar’s OR and p value for binary variables and Wilcoxon signed-rank test p value for continuous variables, comparing the response at M6 with the response at Enr. ORs with 95% CI were also calculated for binary pre–post data.

  • †n=34.

  • ‡To enable calculation of effect measures, a zero value was replaced by 1.

  • CI, confidence interval; Enr, enrolment visit; IQR, inter-quartile range; M6, month 6 visit; NA, not applicable; OR, odds ratio.