healthmedicinet

Sociodemographic characteristics

Between July 2012 and February 2013, 300 eligible pregnant women were randomized and
assigned to home-based (?=?150) or clinic-based (?=?150) male partner HTC. One hundred and eighty-eight male partners of these women
were enrolled either at home (?=?133, 70.7 %) or antenatal care (ANC) clinic (?=?55, 29.3 %) (Fig. 1). This difference in male enrollment reflects the higher effectiveness of the HBHTC
strategy in this trial.

Fig. 1. Enrollment and follow-up of pregnant women and their male partners

Participating men had a median age of 29 years (interquartile range [IQR] 25, 35)
and 42 % of these men had secondary or higher level of education. Most men (87.8 %)
were in monogamous marital relationships. The majority of men (60.1 %) reported a
daily household income of??$2 USD. Thirty-one (17.0 %) men reported concurrency,
while 20 (10.6 %) men perceived their female partners as having concurrent partnerships.
During the preceding six months, 34 (18.1 %) men reported experiencing physical threat
from their female partner while 28 (14.9 %) received threats from other partner(s)
or family member. Few men reported experiencing forced sex over the preceding six
months either from their current sex partner (2.7 %) or another partner or family
member (1.1 %). HIV prevalence among the participating male partners was 16 %. Most
men reported that their female partners were multiparous (59 %) and desired more children
(84.0 %). About two-thirds of men (68.1 %) reported no contraceptive use by their
partner prior to the index pregnancy.

The baseline characteristics of women enrolled in the primary study have previously
been described and are summarized here (Table 1) 20]. Compared to male partners, women were younger with a median age of 22 years (IQR:20–26),
less educated, with only 32.7 % reporting secondary or higher level of education,
and had a lower income, with 74.7 % earning a daily household income??$2 USD. Almost
a third of women (28.7 %) perceived their partners were having concurrent partnerships.
During the preceding six months, a higher proportion of women (30.7 %) than men reported
experiencing physical threat either from their current partner (76.1 %) or another
partner or other person including family members (23.9 %). Overall, 16 % of the women
were HIV-infected. Over half (56.0 %) of the women had not used any contraceptive
method prior to the index pregnancy.

Table 1. Baseline characteristics of study population, both pregnant women and their male partners
(N?=?488)

Baseline preferences on male partner HIV testing setting

Most participants preferred home-based (59.4 %) over ANC clinic-based (28.3 %) and
VCT center-based (12.3 %) male partner HTC during pregnancy (Table 2). In addition, more men than women (68.1 vs. 54.0 %, p?=?0.002) preferred home-based
male partner HTC. Male partners were significantly more likely than pregnant women
to prefer home-based HTC. After adjusting for partner HIV status, male partners remained
significantly more likely than women to prefer HBHCT. Fewer men than women (19.2 %
vs. 34.0 %, p??0.001) preferred ANC clinic-based HTC. VCT center-based testing was the least preferred
setting of male partner testing by both men (12.8 %) and women (12.0 %) and the difference
was not statistically significant. The preferences did not vary by partner testing
status.

Table 2. Differences in baseline preferences on setting of male partner HIV testing by gender

Correlates of preferred settings for male partner testing

Men were more likely to prefer HBHTC testing if they desired more children [odds ratio
(OR) 3.47, 95 % confidence interval (CI): 1.53-7.89) (Table 3). This association remained significant after adjusting for partner HIV testing and
randomization arm (adjusted OR 3.51, 95 % CI: 1.54-7.97, P?=?0.003). Women who had
primary or lower level of education were more likely than those with higher education
to favor HBHCT for male partners and those who had a daily income??$2 USD were also
more likely to prefer HBHCT for male partner testing. Women who perceived their partners
as having concurrent partners were more likely to prefer HBHTC (OR?=?2.35, 95 % CI:
1.39-3.99) and those who reported physical threat were also more likely to prefer
HBHTC for male testing (OR1.44, 95 % CI: 1.06 to 1.95). HIV-infected women were more
likely than HIV-uninfected women to favor male partner HBHTC (OR 2.35, 95 % CI: 1.20-
4.60).

Table 3. Correlates of preferred setting of male partner HIV testing at baseline by gender

Changes in preferred settings for male partner testing

Overall, male and female participants were significantly more likely to prefer HBHTC
for male testing at follow-up (71.0 %) compared to enrollment (59.2 %), (OR 1.98,
95 % CI: 1.43-2.78) irrespective of the study arm (Table 4). Men were twice as likely to favor male HBHTC during follow-up (80.9 %) compared
to enrollment (68.1 %), (OR 2.20, 95 % CI: 1.27-3.94) (Table 4). Similarly, participating women were significantly more likely to favor male HBHTC
at follow-up compared to baseline, (OR 1.89, 95 % CI: 1.26-2.89).

Table 4. Changes in preferred model of male partner HIV testing by gender
^a

Men were less likely to prefer ANC clinic-based or VCT center-based testing at follow-up.
However, the decrease was not statistically significant. Women were significantly
45 % less likely at follow-up (25.3 %) than at enrollment (34.0 %) to favor ANC clinic-based
male testing. Fewer women preferred VCT center-based testing at follow-up, although
the difference was not statistically significant. Adjusting for partner testing status
and study arm did not significantly alter the changes in preferences for male partner
testing.