Impact of Husbands’ Participation in Antenatal Health Education Services on Maternal Health Knowledge

Methods: Total 442 women receiving antenatal services at a hospital in Kathmandu, Nepal were randomized into three groups: women who attended education sessions with their husbands, women who attended education sessions alone, and women who attended no education sessions (controls). At baseline and after delivery, women’s maternal health knowledge and change in knowledge levels were compared between the groups.

husbands experienced increased modern contraceptive use and reproductive health knowledge levels, but these interventions' combinations of group and couples components made it difficult to discern which approach was most effective. 15,16 MetHODs Using a randomized controlled trial design, the impact of providing antenatal health education to pregnant women and their husbands in a couples format, compared to educating women alone or not providing education, was evaluated on women's maternal and reproductive health practices and knowledge levels in Kathmandu, Nepal from August, 2003-January, 2004. Findings related to maternal health practices are reported elsewhere. 17 With a total catchment area estimated at 1.1 million people, Prashuti Griha Maternity Hospital, Kathmandu (PGMH) has approximately 250 beds and has roughly 16,000 deliveries each year. 25 Approximately 40% of ANC clinic patients at PGMH are accompanied by their husbands, who normally wait on the hospital grounds during the time of their wives' visits. 26 All study procedures were approved by the Johns Hopkins Committee on Human Research (Baltimore, MD), the Nepal Health Research Council (Kathmandu, Nepal) and the Prashuti Griha Maternity Hospital Ethical Approval Committee.
Briefly, women who received education with their husbands were significantly more likely to utilize postpartum care than women who were educated alone or women who did not receive the education intervention. Higher levels of birth preparedness were seen among women who received education in either format (couples or alone), as compared to women who did not receive any education. 17 In order to better understand the impact of including male partners in pregnancy health services, this paper presents findings related to the secondary outcomes of changes in knowledge levels associated with the couples antenatal health education intervention.
Married women attending their first ANC visit at PGMH (gestational age 16-28 weeks) whose husbands were present at the hospital compound were eligible to participate in the trial. Systematic random sampling techniques were used to approach eligible women for recruitment. Using the statistical software program Stata 8.0 (Stata Corp., College Station, TX), a list was generated randomizing the sequence of recruitment of study groups for each day of the recruitment period. Upon administration of written, informed consent procedures, 442 participants were randomly assigned to one of three study arms ( Figure 1):  Upon enrollment, a baseline questionnaire was administered that collected information on women's socio-demographic background, reproductive history, and maternal health knowledge and practices. All women were requested to attend the hospital for a final ANC checkup at approximately 36 weeks gestational age. Additionally, all women were encouraged to deliver in the hospital and to attend a postpartum checkup within two weeks of delivery at PGMH. To minimize loss to follow-up, postpartum questionnaires that collected information on maternal health knowledge and practices were completed either immediately prior to discharge or at the postpartum checkup. For those not returning within two weeks of delivery, home visits were conducted during which research assistants administered the postpartum questionnaires.
The health education intervention consisted of two 35-minute sessions delivered in a private room in the hospital with pregnant women and their husbands when applicable. Participants in the intervention groups received the first health education session on the day of enrollment and were asked to return to PGMH for the second health education session approximately 4-6 weeks later.
The intervention was based on principles from the theory of reasoned action and the health belief model, and covered a number of maternal health topics, including nutrition during pregnancy, breastfeeding, pregnancy complications and family planning. 18,19 Female study nurses and male auxiliary health workers provided health information on each of the topics into talking points supplemented with culturally adapted, easy-to-follow graphic materials. The couples study arm consisted of one couple receiving a face-to-face education session administered jointly by one male health worker and one female nurse. Women educated alone received an individual face-to-face education session with a female worker. All health educators received a standardized training course and followed structured intervention protocols. Both intervention groups received the same curriculum messages.
Approximately 15% of Nepal's 27.5 million people live in urban areas; the largest of these is the capital city of Kathmandu (20). While ANC coverage is relatively adequate (82% of women attend at least 1 visit), other maternal health indicators are in need of improvement.
High rates of anemia and home delivery (30-40%), in addition to inadequate weight gain and low postnatal care utilization (approximately 12%), contribute to lingering poor maternal and infant health outcomes in urban Nepal. [21][22][23][24] Women were excluded if they were <18 years of age or lived >90 minutes away from PGMH (due to logistical constraints in tracking down participants).
Maternal and reproductive health knowledge levels were assessed through a series of seven questions (Table 1). Responses were coded and assigned: 1 point if completely correct, 0.5 points if partially correct (in questions 1, 3 and 6), and 0 points if either 'don't know' or an incorrect answer was given. Baseline and followup knowledge scores were created ranging from 0-7 points. A 'change in knowledge' variable was created based on the difference between each woman's baseline and follow-up knowledge score. Follow-up knowledge scores were also dichotomized into either 'high/low' or 'improve/did not improve.' Study participants were coded as lost to follow-up if no postpartum questionnaire was completed. Baseline variables were compared between groups using the student's t (continuous) and chi-squared (binary) tests. Similarly, differences between participants versus nonparticipants, as well as between those who completed the study versus those who did not, were examined. Univariate and multivariate linear regression analyses were conducted using the follow-up knowledge score and change in knowledge variables as continuous outcomes. Crude and adjusted relative risks with 95% confidence intervals were calculated for dichotomized study outcomes. Analyses were conducted with Stata 8.0 and followed an intent-to-treat approach.

ResULts
Consort diagram of study participants and their study participation were used in this study ( Figure 1). Attrition rates for each of the study points of contact were generally low and the levels of missing data for any one variable were negligible (<1%). Adherence to the intervention education sessions was high, with 92% and 88% of women from the couples and woman-alone groups, respectively, returning for the second education session. Between 77-86% of women in each study group returned for the 36-week visit, and follow-up rates for the postpartum questionnaire were also high (92% for couples group, 85% for woman-alone group, and 86% for the control group. All analysis including the follow-up knowledge outcomes was limited to the women who completed postpartum questionnaires (n=386).
Characteristics of the study population at baseline (n=442) were compared, and no statistically significant differences (P<0.05) between the three groups were found, suggesting successful randomization (17). Although educational status and caste were not significantly different at the P=0.05 level between study groups at baseline, these variables, as well as any other potential confounders, were included in all subsequent adjusted models in order to account for minor imbalances.
Unadjusted summary knowledge measures indicate significant differences between the groups in the follow-up knowledge scores as well as in the change in knowledge variable (Table 1). In addition, a general pattern of increased knowledge from baseline to followup is seen among all three groups, with the largest increase usually being seen among women in the couples group.
When entered into multivariate linear regression models, the majority of the differences among groups remained significant ( There were no differences between the woman-alone group and the control group. Upon dichotomizing the follow-up knowledge scores into either 'high/low' or 'improve/did not improve,' similar patterns were seen with respect to study group before and after adjustment for covariates (Table 3).
Women educated with their husbands were slightly more likely to have high knowledge scores at followup as compared to control group women (RR=1.14, 95% CI=0.99, 1.31) and to women educated alone (RR=1.18 , 95% CI=1.01, 1.38). In contrast, women in the woman-alone group showed no significant increases in having 'high' follow-up knowledge compared to the control group. Only the women educated with partners were significantly more likely to show improvement in knowledge from baseline to follow-up as compared to control group women (RR=1.25, 95% CI=1.04, 1.51) both before and after adjustment.

DIscUssIOn
Based on a randomized trial design, a generally consistent pattern was found in which women who received education with their husbands experienced the greatest gains in knowledge, followed by women who received education alone, and with the lowest knowledge levels seen among control group women. Although individual results were only marginally significant, women in the couples group had significantly higher knowledge as compared to control group women in all summary knowledge measures. In contrast, women educated alone had higher knowledge compared to control group women only in a few domains, and when they did so, the magnitude of the difference was less than that seen for couples group women. Despite apparent slight imbalances in background factors such as caste and educational status at baseline, the comparison of adjusted to unadjusted results indicates that improvements in knowledge levels were likely to have occurred as a result of the intervention.
Increased communication and interaction between spouses regarding health practices during or after the education sessions may have led to a greater understanding and/or retention of new information.
Although the premise for these findings is supported by several health education and behavior change theories, future in-depth qualitative research should examine what underlying mechanisms or motivations may be influencing the different learning (or retention) patterns seen in this study. Additionally, further intervention research is needed to replicate these findings in other settings. Involving sources of support from women's social networks other than male partners also merits further examination. The enhanced knowledge attainment seen in women receiving education with their husbands in this trial might, for instance, also be seen if another individual (who provides instrumental and emotional support to the woman) is present (e.g., mother-in-law, sister). In this setting however, designing a trial which included husbands was a logical choice since men tend to dominate decision-making and over 40% of pregnant women are accompanied by their husbands to ANC (compared to roughly 9% by motherin-law and <5% by friends/other relatives) (26).
This study had several limitations. The intervention targeted poor urban women seeking antenatal services, an important and growing population in Nepal. While women at PGMH are likely to be poorer than those seeking services at private hospitals or antenatal clinics in Kathmandu, they are also likely to be wealthier, better educated, and/or more motivated than women receiving no ANC. Even though health services for destitute individuals are free at PGMH, and only a small fee is charged for other patients, seeking ANC already makes these women 'special,' in the sense that they (or someone in their lives) had the motivation to take this preventive health measure. Since over 80% of urban Nepalese women receive at least one ANC visit, and the site of this study represents the primary ANC service provider for the population of Kathmandu Valley, the potential selection biases are likely limited (24).
In order to avoid resource-intensive home visits to recruit husbands not present at PGMH (reaching men in their homes would require a different strategy), women were only eligible for this study if their husbands were present at the hospital. Previous research at PGMH has shown that women accompanied by husbands to ANC are similar with respect to most background characteristics to women unaccompanied by husbands, with the exception of parity. Women with children are about half as likely to have their husbands accompany them to ANC as women without children. The formative qualitative research that preceded this study, however, suggested that most men who were not present were unable to attend ANC due to their job constraints, rather than childcare responsibilities. Women who were accompanied by their husbands to ANC were also slightly more likely to report sharing problems with their husbands, implying a level of emotional support that may be different from women who were unaccompanied by their husbands (26). Despite the potential limitations on the generalizability of these findings, it is important to bear in mind that this study population was representative of almost half of women who seek antenatal care services at the predominant maternal health facility in Kathmandu Valley (5).
Though not formally recorded in the current study, the additional resources required for including men in the health education sessions in this setting were relatively minimal. Only two male health educators were hired in addition to the female health educators, and the duration of the health education sessions administered to couples was the same as for sessions administered to women alone. Future studies should consider costeffectiveness analyses to assess to what extent the knowledge gain seen in women educated with husbands is worth the additional resources required as compared to educating women alone. In addition, Helzner has highlighted some of the potential risks of involving men in women's reproductive health services, particularly in terms of reinforcing patterns of male domination over decision-making (27). Although no negative effects of including men in antenatal care services were recorded in this study, caution must be exercised when exploring male involvement approaches in new settings.
While the 'couples' intervention design in this study required a relatively resource-intensive delivery format for health education messages, future interventions should explore other avenues for reaching wider audiences of males. Policy-makers and health care providers have traditionally implemented policies and procedures that are not conducive to males being included. Maternal health care services, for instance, are usually offered during hours when men are working, men are generally not allowed into checkup or delivery rooms of most hospitals and clinics, and providers often have a bias towards involving men in reproductive health services (28). In the setting for this study, although men regularly accompany their wives to the hospital, they are excluded from services. Such obstacles may be overcome through different approaches to health education. At this location, for example, general announcements could be made advertising health education sessions in the hospital cOncLUsIOn Though much research is still needed to examine the most effective ways of translating health knowledge into healthy behaviors, a fundamental first step to effective health promotion and behavior change is the acquisition of knowledge (29). The findings from this study show that educating women in conjunction with their partners may yield a greater impact on the attainment of maternal health knowledge than educating women alone or not educating women at all. Combined with results reported elsewhere on the improved maternal health behaviors seen in women who received education with their husbands compared to women who received education alone (17), the findings in this paper suggest that women may learn and retain more information when they are educated with their partners rather than being educated alone.  a Model adjusted for: woman's age, parity, educational status, baseline knowledge score, co-residence with motherin-law. a Models adjusted for: woman's age, parity, educational status, baseline knowledge score, co-residence with mother-in-law.