The search yielded 712 records from 3 databases (Embase = 323, CINAHL = 246, PubMed = 143), of which 187 duplicates were removed. Title and abstract screening of 525 unique records were done and 107 studies progressed to full text screening. Twenty-one peer reviewed studies were included in the analysis.

The key informant interviews yielded 6 studies of optimal quality that were included in the final analysis. (See PRISMA diagram) (Figure 1).

This is the most basic level of the SEM model and interventions at this level are defined as those that enhance an individual’s knowledge, capacity, agency or resources to take care of their health and well-being. Strategies at the ‘intrapersonal’ level of the SEM synthesized in this analysis can be grouped into 3 types—individual communication, group communication and practical training. These were reported across 14 papers and ranged from home visits by healthcare workers, individual counselling sessions, group sessions, mass media campaigns to practical training on infant care in neonatal intensive care units.

This is the second level of the SEM, interventions at this level foster interpersonal support and promote health and well-being. The interventions at the ‘interpersonal’ level synthesized in this review have been categorized into two types—namely, couples’ counseling, group education and couples’ education (see Table 2). These were reported in 4 papers included in this analysis.

The ‘community’ level of the SEM includes interventions that promote positive community norms and practices that positively impact the health and well-being of the community. We categorized interventions at this level into two types—service provider and stakeholder engagement and community norm building. Three papers reported ‘community’ level interventions (See Table 2 for details).

In addition to the 3 levels mentioned above, the SEM has 2 additional levels—service delivery, and socio-political context levels. In this analysis we did not encounter standalone interventions in the two other levels. Six papers described interventions across multiple levels of the SEM. The combinations of these multiple levels were—individual and interpersonal, community and enabling environment, and individual and community levels (see Table 2).

In this review, there were 17 quantitative, 6 qualitative, and 4 mixed methods studies. Only two studies reported immunization outcomes of ‘taking the baby to vaccination’ in a study that tested the implementation of kangaroo care by fathers (25), and age-appropriate compliance of childhood vaccines and knowledge of childhood vaccine timeliness in a study that implemented community workshops with male-specific messaging for fathers, respectively (26). Other quantitative child health outcomes included improvements in child feeding; improvement in child wellness (including lower likelihood of HIV infection, changes in development scores); improved care-seeking (including HIV treatment); and parental knowledge and behavior to promote better child health (uptake of institutional delivery, accompaniment to health facility, etc.). Broader men’s engagement outcomes were also reported, encompassing father’s knowledge on various aspects of child health; joint decision-making and maternal support; parental-child bonding and active parenting.

Eleven qualitative outcomes were reported within the following categories: health behaviors, in the form of facility births; self-reported men’s behaviors (including accompanying women to care visits, participation in MNCH, spousal and child relationships, couple decision-making, men health workers’ engagement with men); men’s knowledge (including understanding of danger signs, MNCH knowledge, different types of knowledge acquired by men through education initiatives); and perception about roles (including perception changes about masculinity and men’s role in fatherhood, and reasons for lack of men’s involvement in child health and care).

It should be noted that some men’s engagement frameworks do not consider increases in men’s knowledge as a relevant outcome (27), despite its consideration as such in the past, given that other factors can contribute to men’s knowledge (such as their literacy levels), and argue that increased knowledge only becomes relevant as it relates to how that knowledge is used, which then falls into other dimensions of men’s engagement. Improvement in knowledge levels is a formative step in changing men’s behavior and while it lies in the causal pathway of gender transformative behavior it cannot be considered one (28). However, given that multiple studies included in this review do include knowledge-related outcomes, they are discussed here as defined by those study authors.

Interventions demonstrating statistically significant findings (in the case of quantitative or mixed methods studies) or otherwise demonstrating positive changes (in the case of qualitative studies) in the above results areas are synthesized in Table 3 by SEM level and type of intervention. The same outcomes are described in more detail following the table.

Age-appropriate compliance of childhood vaccines in India was measured in a grey literature source included in this analysis which used program data to assess immunization uptake. It is not clear if a formal evaluation was done. The program implemented community workshop with male-specific messaging for fathers’ help overturn the gender norm that child health is the sole responsibility of the mother, to encourage improved interspousal communication, and to protect their children’s health and as a precursor to their child’s success. An increase in age-appropriate compliance was observed for critical vaccines: increase by 14% for pentavalent; 31.9% for rotavirus; 31% for measles and rubella vaccines (26).

Initiation of breast feeding was measured by Nasreen et al., through a cross-sectional survey in Bangladesh prior which activities were designed to improve men’s engagement in neonatal and child health such as birth planning in the presence of the husband and other family members, and orientation of religious leaders and village doctors to improve men’s participation in childcare activities. Breast feeding was initiated within 1-h of delivery by 65.5% participants in the intervention arm versus 61.3% in the control arm (29). Bich et al. in Viet Nam, and Su and Ouyang in China, also measured initiation of breast feeding in their respective interventions. Bich et al. used a multi-pronged approach of mass media communication, fathers’ group counselling, home visits for individual counselling, and community mobilization activities. The adjusted odds ratio (AOR) of initiating breast feeding in the intervention group was 7.64 compared to those in the control group (30). Su and Ouyang also used education approaches through the ‘father support model’ that imparted education and encouraged greater involvement of fathers to support mothers in feeding practices. Exclusive breast feeding at 6 months was 40% in the intervention arm compared to 17.6% in the control arm where only mothers received this education (31). Prelacteal feeding (AOR: 4.43 in control versus intervention arm), colostrum feeding (95.1% in intervention versus 90.3% in control arm) and formula feeding rate (20% in intervention versus 44.1% in control arm) were other nutrition-related indicators that were measured by Bich et al. (30), Nasreen et al. (29), and Su and Ouyang (31), respectively.

Sifunda et al., conducted a longitudinal study, part of an RCT in Malawi, measuring HIV infection among infants as the outcome. The intervention tested the efficacy of standard “prevention of mother to child transmission” (PMTCT) services plus the ‘protect your family’ program, which included all-mother or all-father groups and couples’ counselling sessions aimed at improving PMTCT uptake. Phase 1 of the study enrolled only mothers and phase 2 enrolled couples. There were increased odds (AOR: 4.55) of infants being infected with HIV in phase 1 (mothers only) compared to phase 2 (couples) of the study (32). Similarly, Kalembo et al., in a retrospective cohort study also measured the association between men’s involvement and completion of PMTCT follow up. Women were encouraged to come to the clinic with their husbands; couples were offered couple counselling and women who came alone received individual counselling. There were increased odds (AOR: 16.8) of completion of follow up among those who engaged as couples in the intervention (33). Even Lyatuu et al. in Tanzania, measured HIV testing uptake as part of PMTCT uptake in a multi-strategy study in six intervention districts that sensitized and engaged clinical and non-clinical staff, employed an integrated approach to improve men’s participation in ANC, created linkages between healthcare providers and community leaders, and recruited community champions to improve male involvement in ANC and PMTCT. The control arm had no intervention. Couples’ HIV testing uptake increased from 11.9 to 36% in the intervention arm as opposed to no significant change in the control districts (34).

Comrie Thompson et al., measured uptake of facility births in a study done in Bangladesh, Zimbabwe and Tanzania which tested multiple strategies across the various SEM levels—household visits and counselling by community health workers, group sessions with men, appointed gender equality champions/male role models, couples counselling on a host of issues including family planning, birth preparedness, etc. sensitization of health providers, sensitization of local religious leaders, and advocacy with government. Through qualitative enquiry, the study found these initiatives were linked to reduction in home births and increase in facility births (35).

Other studies that measured child health outcomes were Garcia et al. in Kenya, a sub-analysis of a cRCT that tested responsive parenting techniques. Arm 1 of the study did educational group sessions only and arm 2 mixed group sessions with home visits, and the control arm had no intervention. In arms 1 and 2, interventions were targeted towards mothers in half the sample and for both parents in other half of the sample. Garcia et al. focused the analysis on arms 1 and 2 of the study which involved fathers. Mothers in villages who participated in the intervention with fathers reported a non-significant trend of increased joint decision-making by the couple versus those where the mother was part of the intervention alone (36). Baheiraei et al. in Iran, an RCT that tested motivational interviewing techniques that included in-person counselling with mothers and telephone counselling with fathers to reduce child’s smoke exposure in households reported infant’s mean urine cotinine concentrations. Infant’s geometric mean urinary cotinine concentrations decreased from 48.72 ng/mg at baseline to 28.68 ng/mg at the 3-month follow-up in the intervention group. In the control group, mean urinary cotinine concentrations decreased from 40.43 to 36.32 ng/mg. The differences between the two groups varying over the course of the 3-month follow-up were statistically significant (p = 0.029) (37).

Knowledge about various child health and nutrition issues were measured in Cockcroft et al. and Rothstein et al. In Cockcroft et al., a stepped wedge RCT in Nigeria assessed knowledge on diarrhea and treatment after healthcare workers conducted home visits to all pregnant women and their spouses and counseled them on various safe practices of pregnancy and childcare. Thirty percent more men thought that children with diarrhea should be given extra fluids and 21% more men said they would not give a child medicine to stop diarrhea in the intervention group versus the control group (38). Similarly, Oguntunde et al. in a qualitative study in Nigeria reported men had a better grasp of danger signs for women and child health after the study implemented men’s support groups to raise awareness on women and child health issues (39).

Rothstein et al., a cluster RCT in Tanzania measured knowledge among men on breast feeding, child feeding and micronutrient, and hygiene across 4 study arms that tested communication through either interpersonal communication (IPC), SMS text messages (SMS), combination of both (IPC + SMS), or none of these interventions (control). The breast-feeding knowledge indicator was insignificant, but the child feeding and micronutrient knowledge scores were 0.048 and 0.090 points higher in the IPC and IPC + SMS intervention groups, and the hygiene knowledge score was 0.052 (SMS), 0.065 (IPC), and 0.073 (SMS + IPC) points higher in the intervention arms compared to control arm (40). This study also shed light on the types of knowledge men gained through the interventions such as the importance of seeking antenatal care, and barriers to men gaining knowledge, such as information provision being directed more towards women which made men feel inferior.

Parental-child attachment was measured using different scoring systems in Toprak and Erenel, Yildirim et al. and Rahimi et al. Toprak and Erenel and Yildirim et al. used the postnatal paternal-infant attachment score measured as a composite of three subscales—“patience and tolerance,” “pleasure in interaction” and “affection and pride.” Both studies used kangaroo care techniques and measured father-child attachment. Scores for paternal-infant attachment were higher in the intervention arm (73.53 and 80.57) versus the control arm (70.8 and 56.76) in both studies (25, 41). Rahimi et al. used active parenting techniques and engaged father in-wards to care for their infants. Mean paternal bonding scores were measured using a validated scale originally developed in 1977 that used 10 items to measure maternal-to-infant bonding. As per this scale, paternal-infant bonding improved more in the intervention arm compared to the control arm (42). Alemann et al. in Nigeria and Ghana also qualitatively ascertained improved father-child relationships because of implementation of a multi-pronged approach that included community-based group education session for father, social and behavior change communication, efforts to increase women and girl’s gender equality knowledge, and health system strengthening and capacity building of service providers to do the same (43).

Toprak and Erenal described above also measured other behavioral outcomes such as fathers taking baby to vaccination (60% in intervention versus 17.4% in control arm), fathers monitoring baby’s development (92% in intervention versus 52.2% in control arm), fathers changing baby’s diapers (36% in intervention and 8.6% in control arm), fathers bathing the baby (36% in intervention and 8.6% in control arm) and multiple others (25). Broadbent et al. in Tanzania also measured similar active parenting indicators of ‘likelihood of father naming objects for their child’ and ‘likelihood of fathers playing with their children’. The intervention used radio spots to reach both parents. Fathers who heard radio spots with the intervention messages were 1.42 times more likely to name objects for their child, and 1.58 times more likely to play with their child compared to those who were not exposed to the radio messages (44). Alemann et al. in Nigeria and Ghana qualitatively reported through focus group discussion that men’s participation in MNCH increased (43). The SPRING Project in Niger also showed a positive impact on men’s involvement in nutrition and hygiene, especially increase in percentage of men who supported food preparation and feeding for children after videos featuring local community members promoted high impact maternal, infant, and young child nutrition and hygiene behaviors (45).

Lusambili et al., qualitatively ascertained increase in men’s accompaniment to ANC and under 5 clinics after the study implemented interventions to educate and engage with men through dialogues and group sessions on the importance of men supporting RMNCH services (46). Comrie-Thompson et al. also qualitatively studied in three countries the impact of a range of interventions including training of male and female community health workers to better engage men in their activities, education, behavior change communication and capacity building initiatives for men, mobilizing male role models and traditional, religious and other community leaders, and advocacy with national and district-level authorities. The study found men were more likely to accompany their wives to the clinic in all three countries because of the interventions (35).

Mweemba et al. in Zambia and Alemann et al. in Lebanon reported positive changes in men’s perception of masculinity and how they should engage in child health and care. Both interventions took the couples education approach—Mweemba et al. selected eight educational films produced in the local language that covered a variety of health topics; and Alemann et al. conducted group education sessions with fathers/men and their female partners (43, 47).

Ten of 21 peer-reviewed studies, and 5 of 7 grey literature sources that demonstrated positive changes in health outcomes included interventions that aimed to actively reduce gender inequality between women and men regarding child health decisions and care. However, only 2 peer-reviewed studies and 4 grey literature sources measured gender transformative outcomes. All these peer-reviewed and grey literature studies measured the gender transformative outcomes of joint or shared decision making between couples. Garcia et al. found significant positive associations between men’s engagement (father’s interpersonal support to mother, and shared decision-making) and child development after interpersonal-level responsive parenting and couples communication interventions (36). Nasreen et al. conducted community-level interventions and found non-significant changes in joint decision-making at post-natal care (29). Comrie-Thompson et al. found improved couple communication and decision making related to MNCH and other household matters as a result of the multi-pronged approach that included multi-level holistic interventions including household visits and counselling by community health workers, group sessions, appointed gender equality champions, couples counselling on a host of issues including family planning, birth preparedness, etc. sensitization of health providers, sensitization of local religious leaders, and advocacy with government (35). Alemann et al. also reported improved shared decision making between couples because of community-based group education sessions in Bolivia and Lebanon, in addition to capacity building of community and facility-based providers in Ghana and Nigeria. An overview of these studies can be found in Table 4.