The impact of approaches in improving male partner involvement in the prevention of mother-to-child transmission of HIV on the uptake of safe infant feeding practices by HIV positive women in sub-Saharan Africa. A systematic review and meta-analysis

Background The low level of male partner involvement in Prevention of Mother to Child Transmission of HIV services such as safe infant feeding practices poses a serious challenge to the implementation of guidelines on safe infant feeding and may undermine efforts towards elimination of mother to child transmission of HIV in sub Saharan Africa(SSA). We conducted a systematic review and meta-analysis to identify the approaches that have been utilized to improve male partner involvement in PMTCT services as well as their impact on the uptake of safe infant feeding practices by HIV positive mothers in SSA. Methods In this systematic review and meta-analysis, Ovid Medline, Embase, PsycINFO, Cochrane library, ClinicalTrials.gov, Web of Science and Current Controlled Trials were searched. Only studies performed in SSA that reported an approach that specifically involved male partners and its impact on the uptake of safe infant feeding practices (irrespective of the language and date of publication) were included. Odds ratios were extracted or calculated from studies and combined in a meta-analysis using the statistical package Stata version 11.0. Forest plots were generated using the random effect model. Results From an initial 2416 non-duplicate articles, 06 articles were included in the systematic review and meta-analysis. The overall pooled unadjusted OR was 3.08[95%CI: 2.58–3.68], while the effect sizes for interventions aimed at promoting male partner involvement such as verbal encouragement, complex community intervention and enhanced psychosocial interventions were 1.93[95%CI: 1.34–2.79], 3.45[95%CI: 2.79–4.25] and 5.14[95%CI: 2.42–10.90] respectively. Using only adjusted odd ratios, the pooled adjusted OR = 1.78[95%CI: 1.35–2.34]. The I2 = 60.1% p = 0.057 using adjusted ORs. Conclusion Interventions aimed at promoting male partner involvement such as enhanced psychosocial interventions, verbal encouragement and complex community interventions increase the uptake of safe infant feeding options. The heterogeneity was moderate among studies. More studies including randomised trials that will recruit larger, representative samples of patients are needed in the future. Prospero registration number: 42016032673


Conclusion
Interventions aimed at promoting male partner involvement such as enhanced psychosocial interventions, verbal encouragement and complex community interventions increase the uptake of safe infant feeding options.The heterogeneity was moderate among studies.More studies including randomised trials that will recruit larger, representative samples of patients are needed in the future.

Rationale
The unprecedented commitment of the global fight against mother to child transmission of HIV has resulted in an encouraging reduction in new paediatric HIV infections in sub Saharan Africa(SSA) by 56% since 2010 [1].However, the disproportionate burden of HIV infection in SSA is still invariably noticed with 220,000 new pediatric HIV infections and 2.6 million children below 15 years living with HIV accounting for nearly 91% of the global burden [1][2][3].Contemporary evidence to explain this persistent disproportionate burden point towards the complex socio-cultural context in the region that impedes the effective participation of communities in HIV prevention activities such as safe infant feeding practices by HIV positive mothers [4].The World Health Organization (WHO) guidelines on safe infant feeding practices for HIV positive mothers calls for exclusive breastfeeding by HIV mothers on antiretroviral therapy or opting out of breastfeeding (infant formula) with discouragement of mixed infant feeding [5].Despite this recommendation, mixed infant feeding is still largely observed in many settings across sub Saharan Africa mainly due to the sociocultural belief that breast milk might not be sufficient for the infant [6].Intricately linked to this sociocultural context is the role of male partner involvement in providing the necessary psychosocial support needed by HIV positive women to shun mixed infant feeding [7].
Male partner involvement in the prevention of mother to child transmission of HIV encompasses communication between spouses (including HIV status disclosure), antenatal care attendance and childbirth, antenatal testing and couples combination antiretroviral therapy(ART) and antenatal support during pregnancy and infant feeding decision support [8].Evidence suggests that majority of women are in favour of involving their male partners in PMTCT.In a study conducted in Malawi to investigate women's choices with regard to HIV testing, disclosure and partner involvement in infant feeding and care, it was shown that 87% of the women regardless of HIV status felt that partners should be involved in decisions on health care for their child [9].Furthermore, a systematic review to evaluate the effectiveness of interventions that aim to improve PMTCT service delivery and promote retention throughout the PMTCT steps has shown that the involvement of male partners PMTCT plays a crucial role in retention of pregnant women [10].
Studies in the area of male partner involvement have suggested a positive impact on the uptake of PMTCT services.Baiden et al in a cross-sectional study showed that women who demonstrated a strong sense of willingness to be accompanied by their male partners were five times more likely to accept HIV testing during antenatal care (OR 5.2; 95% CI 1.4 to 19.8) [11].In a prospective cohort study which involved the follow up of 456 pregnant women and their infants in Kenya, Aluiso et al showed that vertical transmission risk was lower among women with partner attendance compared with those without [adjusted hazard ratio (aHR) = 0.56, 95% confidence interval (CI): 0.33 to 0.98; P = 0.042] [12].Furthermore, using a randomised controlled trial (RCT), Kiarie et al showed that compliance to antiretroviral regimens in pregnancy was strongly associated with partner notification (OR 7.5; 95% CI 1.4 to 40) [13].
Despite the positive impact of involving male partners, there is evidence to suggest that the level of male partner involvement in PMTCT in SSA is low which may reduce the likelihood of HIV positive mothers adopting safe infant feeding practices and hence undermine the ongoing efforts towards elimination mother to child transmission of HIV(eMTCT) in SSA [14,15].Studies have reported approaches that can be used in involving male partners in PMTCT such as invitation letters and complex interventions that use more than one strategy to engage male partners in the community [16].It was therefore necessary to conduct this systematic review and meta-analysis to determine these approaches and their impact on the uptake of infant feeding practices in view of informing policy makers, governments, implementing partners (such as NGOs) and researchers on how the uptake of PMTCT services can be improved through increased engagement of male partners.Objectives 1.To determine the interventions/approaches used to improve male partner involvement within the context of PMTCT, specifically safe infant feeding practices by HIV positive mothers in SSA.
2. To determine the impact of the approaches used in (1) on the uptake of safe infant feeding practices by HIV positive mothers in SSA.

Protocol and registration
This systematic review with meta-analysis was conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic review and Meta-Analysis) statement [17].The protocol for this systematic review was registered in the international prospective register of systematic reviews (PROSPERO).The registration number is CRD42016032673.The protocol was published in the British Medical Journal (BMJ) Open [18].In the protocol we considered outcomes such as infant antiretroviral prophylaxis uptake, maternal ART uptake, safe infant feeding practices, condom use as well as family planning.However, we decided in this review to focus on the safe infant feeding practices outcome due to its significant role in reducing the vertical transmission of HIV.

Eligibility criteria
This review considered studies that were conducted in SSA.Studies conducted outside of the SSA region were excluded.No restriction was placed on the setting of the study and the language of study.Randomized controlled trials, prospective and retrospective cohort studies, and serial cross sectional studies were eligible for inclusion.Studies were included if they provided data on the impact of male partner involvement on the uptake of safe infant feeding practices.One time cross-sectional studies and case-control studies were excluded because they did not present any evidence of the impact of male partner involvement.The participants were HIV positive mothers.1.

Information sources and search strategy
The outputs of the search were exported to Mendeley desktop 1.16.1 and duplicates were removed.After removal of duplicates in Mendeley, the titles and abstracts of the studies were screened independently by NFT and JAA.The full texts were obtained from the screened abstracts after inclusion and exclusion criteria were applied.Authors of articles were contacted for further information on any publication.

Data collection process and data items
A data extraction spreadsheet was developed in excel version 2013.The data extraction sheet captured characteristics of the studies such as: authors, country of study, study design/method, study population, approaches/intervention used for PMTCT improvement, and ORs (odds ratios).
The outcome of interest was uptake of safe infant feeding practices.Odds ratios were extracted from individual studies.Relative risks and proportions were converted to ORs.Two reviewers (NFT and JAA) independently extracted these data from the included studies.Any disagreement was settled by a third reviewer (LNA).The actual number of participants that incorporated/utilized (or did not incorporate/utilize) safe infant feeding practices after the male involvement approaches were extracted from the studies and odds ratios were calculated from a 2x2 table and the 95% confidence interval (CI) of the odds ratios were also calculated.Also, adjusted odds ratios were extracted from the studies.These has been summarized in an excel sheet and submitted as part of the data extraction sheet (S2 Table ).The characteristics of included studies were summarized in Table 2.

Synthesis of results and data analysis
The studies that remained relevant following application of inclusion and exclusion criteria were used in the synthesis.Studies with data on impact of male involvement on the uptake of safe infant feeding practices were considered for a meta-analysis that was performed using statistical software Stata version 11.0.In this review the studies included varied significantly in terms of approaches and outcomes which suggests that the true effect sizes measured could also differ.These disparities could very likely introduce high heterogeneity.Therefore, the random effect model was used to pool the evidence from the studies.
Heterogeneity was assessed using the I squared statistic generated.Heterogeneity refers to the variation between the included studies and it was assessed as follows: if the I 2 = 25%-49% we considered a "low" heterogeneity, if the I 2 = 50%-74% we considered a "moderate" heterogeneity and if the I 2 �75% we considered a "high" heterogeneity [20].

Quality assessment of studies
The Newcastle Ottawa scale was used in assessing the quality of non-randomized studies [21].This scale captured 8 core elements divided into 3 broad elements related to the study quality.
The first element was to determine the representativeness of the exposed cohort.The second element was to determine if the study controlled for other variables.The Third element was to determine if there was bias in the measurement of the outcome.

Results
The Six studies were finally included in the systematic review and meta-analysis.All six studies were cohort studies.

Characteristics of included studies
Table 2 shows the characteristics of included studies.The studies included a total of 3598 HIV positive mothers ranging from 122 to 2141.
In four of the studies, HIV positive mothers were encouraged verbally through counselling to bring their partners for counselling to the clinic [12,[22][23][24].Kalembo et al, Msuya et al and Aluiso et al gave no description of the verbal message given to the women or the personnel responsible for providing/delivering the message while Brou et al actively encouraged the women through counselling to engage their male partners through HIV status disclosure.
One study by Farquhar et al used enhanced psychosocial intervention to involve male partners [25].It was termed "enhanced" because trained HIV-positive male or female peer counselors were responsible to perform the couples counseling.The couples decided on the gender of the HIV peer counsellor.These specifically trained peer counselors were well equipped with communication skills that could improve on the level of male partner involvement.37 sexual partner � .mp. [mp = ti, ab, hw, id, tn, ot, dm, mf, dv, kw, fx, dq, nm, kf, px, rx, ui, sy] (36091) 38 infant feeding.mp.[mp = ti, ab, hw, id, tn, ot, dm, mf, dv, kw, fx, dq, nm, kf, px, rx, ui, sy] (16098) 39 breastfeeding.mp.[mp = ti, ab, hw, id, tn, ot, dm, mf, dv, kw, fx, dq, nm, kf, px, rx, ui, sy] (47383) 40 infant formula � .mp. [mp = ti, ab, hw, id, tn, ot, dm, mf, dv, kw, fx, dq, nm, kf, px, rx, ui, sy] (13156) 41 infant bottle feeding.mp.[mp = ti, ab, hw, id, tn, ot, dm, mf, dv, kw, fx, dq, nm, kf, px, rx, ui, sy] ( 14 The study by Semrau et al involved the use of complex community interventions to improve on male partner involvement [26].These were complex community interventions because several approaches to involve male partners were coupled with other changes in health care delivery to improve on maternal and child care within the community.Community mobilization talks and dramas were carried out.During these community meetings, community members were educated on the relevance of early antenatal clinic attendance of couples and the role of male partners in supporting the uptake of PMTCT services by women.Semrau et al engaged community health workers to carry out mobilization talks and dramas in football events, market places, clubs, churches and work places.

Quality assessment
The results of the quality assessment are shown in Fig 2 .None of the final eligible studies selected a representative sample of patients which most likely resulted in selection bias.In three studies the cohort possessed special characteristics that did not represent the general characteristics in the setting [23][24][25].In the Farquhar et al study, the couples were older, more educated and more likely to engage in HIV prevention activities.Brou et al recruited a consecutive sample of older HIV positive women who were more likely to be in polygamous marriages.Msuya et al included women who were younger, less educated and less likely to engage in prevention activities.Semrau et al consecutively selected participants only from socioeconomic disadvantaged communities while Aluiso et al and Kalembo et al recruited consecutively from a single district hospital.
Two of the studies did not adjust for confounding [23,26].Furthermore in terms of followup, participants in all the final selected studies had long enough follow-up time to observe the outcome but only the study by Aluiso et al provided evidence to justify that the characteristics of participants lost to follow-up(LTFU) were similar to those retained [12].The other studies were prone to follow-up bias because the authors didn't investigate if the characteristics of patients lost to follow-up were similar to those retained.

Results of meta-analysis
Fig 3 shows the overall forest plot for the studies that reported the impact of male involvement on the uptake of safe infant feeding practices.The pooled unadjusted OR = 3.08[95%CI: 2.58-3.68]while the effect sizes for interventions aimed at promoting male partner involvement such as verbal encouragement, complex community intervention and enhanced psychosocial

Discussion
A meta-analysis of Odds Ratios from six studies showed a statistically significant increase in the uptake of safe infant feeding practices with male partner involvement.This increase was associated with enhanced psychosocial intervention, verbal encouragement and complex community interventions.These findings are in contrast to what Brusamento and colleagues reported [27].The systematic review by Brusamento et al suggested that male involvement had a negative impact on the uptake of PMTCT services by HIV positive women [27].Even though Brusamento et al thoroughly evaluated multiple databases with a comprehensive and systematic search strategy, independently assessing the quality of individual studies, the search strategy was biased towards identifying only randomized controlled trials which might explain why the authors failed to also include relevant cohort studies that could have improved on the evidence synthesis with possible meta-analysis.Another possible reason for the disagreement with the Brusamento et al study is that the authors drew their conclusion from a single study that used only invitation letters as an approach to involve male partners.Our study has shown that other categories of interventions exist such as enhanced psychosocial intervention, verbal encouragement and complex community interventions that have a positive impact on the uptake of safe infant feeding practices.In addition, by pooling together the evidence from different studies in a meta-analysis, our study provides more convincing evidence than the Brusamento et al study that drew conclusions from a single unblinded randomised trial.
Our study is also different from two reviews conducted in the domain of male involvement [28,29].The systematic review by Yargawa and colleagues focused only on maternal outcomes such as postpartum depression, utilization of hospital services and postnatal care [28].The postpartum care did not involve infant feeding practices.Compared with the Auvinen et al study [29], our study did not identify invitation letters as a strategy to specifically improve the uptake of safe infant feeding practices.However, it should be noted that the study by Auvinen et al was a narrative review with no systematic search strategy, no multiple database search, and no independent assessment of the quality of included studies.In addition, since the Auvinen study was a narrative review with no focus on the impact on outcomes, the authors could not provide data on the impact of male involvement on the uptake of safe infant feeding practices.
Our study has shown that enhanced psychosocial interventions improve the uptake of safe infant feeding practices.The use of HIV-positive peer counselors was an advantage because it appeared to encourage sero-concordant couples to more readily accept their status.Using well trained peer counsellors ensured that they were well equipped with communication skills that could improve on the level of male partner involvement [30].In addition, allowing the couples to choose the gender of the counselor improved the response rate and overall acceptance of counselling.This acceptance may have a positive impact on the subsequent lifelong behavioural adaptations that could have long-term benefits for the family [31].However, identifying these HIV-positive peer counselors may be challenging in communities within SSA where stigma and fear of status disclosure still form a major component of the social context [32].
In our study, the use of complex community interventions to improve the uptake of safe infant feeding practices was a strength for many reasons.We feel that four of the most important strengths include the following: 1) The large sample size correlated with robust statistical power to find statistically significant differences not due to chance.2)By involving communities the authors went closer to the male partner than in other approaches mentioned so far.This could reduce the chances of measurement errors seen in other studies that relied on what was reported from the pregnant women.3)Utilizing several approaches within the community meant that the authors were more likely to have explored diverse sociocultural context of these communities which could have accounted for the relatively high response rates [33].4)The diversity in terms of the location of the meetings (i.e.bars, social clubs, football matches, market places, and churches) may have reduced the selection bias and increased the response rate and coverage.Despite these strengths, it is very difficult to attribute the effect observed to any single type of community intervention mentioned within the complex intervention category.This is because the authors did not stratify the individual interventions and their effects.In addition, implementing these complex packages could be quite costly since they involve significant resource allocation in the form of personnel/providers.This means they may not be sustainable in resource poor settings which further highlights the need for authors to stratify the individual community strategies in the package to better inform heath programmes that may want to take up only certain community strategies based on resource constraints.
This systematic review and meta-analysis had several strengths and some limitations.The search was comprehensive because several databases and grey literature were searched and authors were contacted for any unpublished studies.There was also independent search and screening of articles by 2 reviewers which reduced bias.Unlike other systematic reviews on male partner involvement, the quality of studies included in our review was critically assessed.The heterogeneity was high when unadjusted odds ratios were used but adjusted odds ratios gave moderate heterogeneity.A limitation in the study was the high selection bias which resulted from the selection of a non-representative sample by the authors.This may limit the generalizability of the findings to certain groups of individuals.Despite this limitation, our study is the first meta-analysis to provide data on the impact of male involvement on the uptake of safe infant feeding practices and will therefore set the pace for more studies with improved recruitment of participants.Our study is also relevant in guiding policy and research in this very important area of male involvement in PMTCT.Another limitation is that no randomized trial has been conducted to investigate the impact of male involvement on the uptake of safe infant feeding practices.Even though randomised trials provide a higher level of evidence than cohort studies, the latter provide a more reliable picture of the real world than the former [34].The findings from our study using cohort studies may therefore present a better picture of real world implementation science than if only randomized trials were used.

Conclusion
Interventions aimed at promoting male partner involvement such as enhanced psychosocial interventions, verbal encouragement and complex community interventions increase the uptake of safe infant feeding options.The heterogeneity was moderate among studies.

Recommendations for future research
No randomized controlled trial has been carried out to investigate the impact of male partner involvement on the uptake of safe infant feeding practices.Randomized controlled trials are needed to add to the strength evidence available.Stratification of the impact of each category of intervention is needed in future to further inform policy and research.Finally, an economic evaluation is needed to adequately inform policy.
electronic search on Ovid Medline, Ovid Embase, Ovid Health and Psychosocial instruments, Web of Science, Cochrane library, ClinicalTrial.gov,Current controlled trials returned 3559 results and after removal of duplicates this reduced to 2416 results.The study selection process is shown on the PRISMA flow diagram in Fig 1.

Fig 5 .
Fig 5. Funnel plot for publication bias for studies reporting impact on safe infant feeding practices.https://doi.org/10.1371/journal.pone.0207060.g005 A literature search was conducted from December 1st 2015 through May 30th 2018(inclusive).A search strategy was developed by the principal investigator (NFT) with input from Coauthors JAA and LNA using evidence from a United States Centre for Disease Control and Prevention(CDC) study on how to carry out a detailed systematic search in HIV prevention[19].Six main databases were searched: Ovid Medline, Ovid Embase, Ovid Health and Psychosocial Instruments (HPSI), PsycINFO, Web of Science and Cochrane library.Current Controlled Trials and ClinicalTrials.govwere searched for ongoing and newly completed trials.A final search was conducted on May 30, 2018.A detailed search strategy is shown in Table