Predictors of male partners’ behavioral intentions and cognitive readiness toward maternal health service utilization in rural Tanzania: an application of the theory of planned behavior Authors: Fabiola V. Moshi, Walter C. Millanzi Published: 2026 ABSTRACT BackgroundMen are not only heads and decision-makers in most families, but they also heavily influence the healthcare access of their female partners and/or children. However, low male spousal accompaniment at health facilities remains a persistent public health challenge in rural Tanzania. While structural barriers are well documented, there is a lack of evidence regarding the behavioral intention factors that drive this men's physical disengagement from clinical spaces. Understanding these internal drivers is critical for designing effective targeted interventions to improve maternal service utilization in rural settings. Applying the Theory of Planned Behavior (TPB), this study assessed the predictors of male partners’ behavioral intentions and cognitive readiness toward spousal accompaniment in maternal health service utilization in rural Tanzania.MethodsA baseline community-based assessment from a longitudinal interventional study was adopted to evaluate 546 randomly selected male partners from June to October 2017 was conducted using a quantitative research approach. A multi-stage sampling technique was employed to select participants. While male accompaniment broadly encompasses emotional and domestic support, this study specifically operationalized male spousal accompaniment as men's physical accompaniment of their expectant mothers to the health facility for clinical maternal health services. Data were analyzed using IBM SPSS version 25. Bivariate and multivariable logistic regression analyses were performed to identify independent predictors (α = 0.05).ResultsParticipants’ mean age was 31 ± 7.726 years, and 71.8% were married. Findings revealed that 61.7% of male partners had never physically accompanied their spouses to health facilities for maternal services. High rates of negative behavioral constructs were observed, with 59.0%, 57.0%, and 57.0% of male partners reporting negative attitudes, demonstrating negative perceived subjective norms, and negative perceived behavior control, respectively. Multivariable analysis showed that positive attitude (AOR = 3.616; p < 0.05; 95%CI: 1.333, 6.138), positive perceived subjective norms (AOR = 2.374; p < 0.05; 95%CI: 1.669, 5.984), and positive perceived behavioral control (AOR = 1.759; p < 0.05; 95%CI: 1.004, 4.373) were significantly associated with increased prospective intention to accompany partners to care.ConclusionA significant proportion of male partners in rural Tanzania remain physically disengaged from clinical maternal spaces, driven primarily by negative cognitive constructs and traditional social norms that designate pregnancy care as an exclusively female domain. Policy and programmatic actions should prioritize community-based social and behavior change communications (SBCC) that challenge restrictive subjective norms, foster joint family decision-making, and reshape male attitudes toward spousal accompaniment.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12884-026-09427-0. Introduction Male partner accompaniment (MPA) is a pivotal determinant in Maternal Health Services (MHS) utilization and positive birth outcomes as well, yet its prevalence remains a significant global challenge [1]. While participation is nearly universal in the Global North, it remains unacceptably low across the Global South [2]. In Sub-Saharan Africa (SSA), male accompaniment in antenatal care (ANC) is often fragmented, with many men acting as financial providers but remaining physically absent from clinical spaces [3]. For instance, while some SSA regions report rising trends, others still see less than 30% of men accompanying their partners to health facilities [4]. In Tanzania, although institutional delivery has improved to approximately 78% nationally by 2024, male attendance in clinical settings lags significantly, particularly in rural areas where gender-defined roles are more rigid [5]. Moreover, in the Rukwa region, recent data suggest that skilled birth attendance and postnatal care utilization remain below national targets, often exacerbated by the lack of active male partnership [6]. Addressing this gap is central to achieving Sustainable Development Goal (SDG) 3.1, which aims to reduce the global maternal mortality ratio to less than 70 per 100,000 live births by 2030 by shifting from individual maternal focus to a couple-centered approach [7, 8]. Tanzania has made commendable strides, reducing its maternal mortality ratio (MMR) from 980 in 2000 to approximately 276 per 100,000 live births by 2023 [9]. The government has integrated male accompaniment into the National Road Map Strategic Plan to improve Reproductive, Maternal, Newborn, Child, and Adolescent Health, which explicitly advocates for male accompaniment [3]. Internationally, these efforts align with the World Health Organization (WHO) guidelines and the 1964 International Conference on Population and Development (ICPD) framework, yet the transition from policy to community-level practice in rural Tanzania remains low [10]. The risk is unevenly distributed; for example, institutional MMR in regions such as Mtwara and Iringa has reached as high as 174 and 105 per 100,000 live births in 2024 [11]. In the Rukwa region, where this study is situated, maternal service utilization is impacted by geographical and socio-economic barriers [6]. In rural Tanzania, traditional gender roles often define maternal health as a female domain, where a man’s presence at an antenatal clinic is perceived as a sign of weakness or a violation of cultural norms [12]. While existing literature has extensively documented structural barriers, such as distance to facilities, long waiting times, unfriendly healthcare staff, unfriendly health infrastructure, and the dominance of female healthcare providers, there is a profound gap in understanding the psychological and behavioral drivers of male disengagement in maternal health services [1, 13]. Most studies in the Rukwa region have focused on the “what” (prevalence) and socio-demographic indicators rather than the “why” (behavioral intent) [14]. Do men with access to services choose not to engage? There is an urgent need to understand the psychological drivers, attitudes, and social pressures that influence a man’s decision-making process during his partner’s pregnancy. This study is designed to fill this gap by examining the men's internal cognitive factors that precede the decision to participate in maternal care. The study was grounded in the Theory of Planned Behavior (TPB), which provides a robust framework for understanding health-seeking actions [15]. The TPB posits that an individual’s behavior is driven by their behavioral intention, which is shaped by three constructs, including Attitude (personal evaluation of the favorable or unfavorable feelings of being involved), Subjective Norms (perceived social pressure from peers or elders), and Perceived Behavioral Control (the man’s belief in their ability to overcome barriers such as work or clinic costs) [15]. In this study, the TPB is utilized to move beyond descriptive statistics, allowing the researchers to quantify how a man’s personal beliefs and his perception of community expectations directly influence his accompaniment in maternal services, providing a scientific basis for behavioral change interventions. Research in the Rukwa region is particularly timely as the government seeks to decentralize maternal services to the community level and to find cost-effective ways to increase maternal health services uptake [16]. By quantifying whether negative attitudes or restrictive social norms through the lens of TPB are the primary barriers, this study will provide the policy-relevant evidence needed to design targeted behavioral change communications. This study does not merely document low accompaniment; it provides a baseline for interventional strategies that can reshape the behavioral intentions of men, thus improving the utilization of maternal health services and reducing preventable deaths in rural Tanzania. Nevertheless, this study fills a methodological gap by providing a baseline for interventional studies aimed at transforming men from passive financial providers into active, informed partners in maternal health. Methods and materials Study design and target population This study presents the baseline assessment of the cross-sectional phase of a larger quasi-experimental interventional design evaluated through a quantitative research approach [17–21]. Data collection was executed between June 1 and October 30, 20,217. The target population comprised male partners of pregnant women residing within the selected clusters. To be eligible for inclusion, participants had to be male partners cohabiting with a pregnant woman whose gestational age was documented at ≤ 24 weeks at the time of recruitment, and both partners had provided informed consent. If a female partner reported amenorrhea for at least two consecutive months but had not confirmed her pregnancy status, a urine human chorionic gonadotropin (hCG) test was administered on-site; a positive result granted eligibility for household enrollment. Exclusion criteria applied to male partners who were transitionally resident in the village for less than six months, those with cognitive impairments that restricted communication, or couples where the female partner presented with immediate obstetric emergencies requiring urgent referral. Study setting The study was conducted across 45 randomly selected villages within the Kalambo and Sumbawanga Rural Districts of the Rukwa Region, situated in the Southern Highlands of Tanzania. The Rukwa Region is predominantly rural, with the economy anchored heavily on subsistence agriculture, livestock keeping, and small-scale fishing, occupations that separate men from their households for extended daytime hours. The region features a fragmented healthcare infrastructure, consisting of approximately two regional referral hospitals, three district hospitals, and a network of scattered rural health centers and dispensaries. On average, communities in these rural districts reside at a distance of 7 to 15 km from the nearest facility offering comprehensive emergency obstetric and newborn care (CEmONC). This geographical terrain, characterized by poor road infrastructure and seasonal sub-optimal transport coverage, severely limits access to maternal care. These structural factors, coupled with the dominance of female healthcare providers in maternal clinics and long facility waiting times (frequently exceeding three hours), heavily predispose men to avoid accompanying their partners to health facilities. Sample size determination The sample size for this study was calculated based on the estimated prevalence of skilled delivery utilization. A two-tailed formula comparing two independent proportions (intervention vs. control groups) in a quasi-experimental design by West et al., [22]. Fixed statistical parameters were used, including a 95% confidence interval/level (Z = 0.196), a 5% standard margin of error (e = 0.05), and a design effect of 1.5 to account for cluster sampling variations. The primary baseline parameter was drawn from the national prevalence of skilled delivery utilization in the Tanzania Demographic and Health Survey (TDHS) (NBS, 2016). Crucially, because the active fieldwork and data collection phase of this study was conducted from June to October 2017, the 2015–16 TDHS represented the most current, legally mandated, and contemporary national dataset available at the time of this study. Adjusting for a non-response safety margin of 10%, the final conservative sample estimate was established at 546 male partners. While recognizing that every arm of the maternal continuum of care is critical, skilled delivery was selected as the reference indicator for sample estimation because it represents the primary bottleneck in rural maternal health utilization. Because its baseline utilization rate is closer to the maximum statistical variance threshold (50%) compared to early ANC indicators, calculating the sample based on skilled delivery yielded a larger, more conservative sample size. This approach effectively ensured that the study remained fully powered to evaluate the secondary maternal health utilization metrics analyzed throughout the text. Sampling technique and procedures Sampling procedures employed in this study have been benchmarked from some previous scholarly works [23–29–30]. To minimize selection bias and ensure regional representativeness, a multistage cluster sampling design was executed across two purposefully selected districts within the Rukwa region, including Sumbawanga Rural District and Kalambo District (Table 1). These districts were selected as the primary strata due to their historically high rates of home childbirth and low male partner accompaniment as documented in regional health profiles [6, 14, 16, 31].Table 1Sampling technique and proceduresStageProceduresStage 1: Ward selectionUsing a simple random sampling approach via a computer-generated random number sequence, five (5) wards were selected out of 12 in the Sumbawanga district, and ten (10) wards were selected out of 17 in the Kalambo districtStage 2: Village/cluster selectionWithin the 15 total selected wards, all constituent villages were enumerated to serve as primary sampling units (clusters). Using a simple random lottery method, 15 villages were selected from Sumbawanga rural and 30 villages from Kalambo district, yielding a total of 45 clustersStage 3: Household and participant selectionAt the village level, a complete sampling frame of households hosting pregnant women was generated in collaboration with Village Executive Officers (VEOs) and Community Health Workers (CHWs). The index household in each village was selected using simple random sampling from the village registry. Subsequent households were selected using a systematic sampling interval(k), calculated by dividing the total number of eligible households in the village by the allocated sample size for that cluster. If a selected household did not meet the inclusion criteria or if the occupants declined participation, the data collection team skipped to the immediate next household in a predetermined clockwise direction until the cluster quota was achieved Table 1Sampling technique and proceduresStageProceduresStage 1: Ward selectionUsing a simple random sampling approach via a computer-generated random number sequence, five (5) wards were selected out of 12 in the Sumbawanga district, and ten (10) wards were selected out of 17 in the Kalambo districtStage 2: Village/cluster selectionWithin the 15 total selected wards, all constituent villages were enumerated to serve as primary sampling units (clusters). Using a simple random lottery method, 15 villages were selected from Sumbawanga rural and 30 villages from Kalambo district, yielding a total of 45 clustersStage 3: Household and participant selectionAt the village level, a complete sampling frame of households hosting pregnant women was generated in collaboration with Village Executive Officers (VEOs) and Community Health Workers (CHWs). The index household in each village was selected using simple random sampling from the village registry. Subsequent households were selected using a systematic sampling interval(k), calculated by dividing the total number of eligible households in the village by the allocated sample size for that cluster. If a selected household did not meet the inclusion criteria or if the occupants declined participation, the data collection team skipped to the immediate next household in a predetermined clockwise direction until the cluster quota was achieved Sampling technique and procedures Data collection instrument The primary data collection instrument was a highly structured, interviewer-administered questionnaire adapted from validated psychometric tools used in a similar Sub-Saharan African reproductive health context [2, 30–32]. To ensure cultural alignment and linguistic accuracy, the tool was translated from English into Swahili by an independent linguistic expert from the University of Dodoma (UDOM) and back-translated to English by an independent researcher to verify semantic equivalence. The questionnaire was segmented into four operational sections (Table 2).Table 2Data collection instrumentSectionVariableNo. of ItemsDefinition/DescriptionASocio-demographic Profiles16Captured variables include age, marital status, education level, primary occupation, income level, religion, parity, and distance to the nearest health facilityBAttitude15Utilized a 5-point Likert scale (ranging from 1 = Strongly Disagree to 5 = Strongly Agree) to assess prospective evaluations of physical, emotional, and financial male accompaniment in maternal healthCPerceived Subjective Norms15Measured perceived social pressures from peers, community elders, and religious leaders using a 5-point Likert scale (ranging from 1 = Strongly Disagree to 5 = Strongly Agree)DPerceived Behavioral Control16Evaluated perceived external or internal facilitators and barriers (e.g., clinic waiting times, financial constraints, work commitment) influencing the male partner’s capacity to participate Table 2Data collection instrumentSectionVariableNo. of ItemsDefinition/DescriptionASocio-demographic Profiles16Captured variables include age, marital status, education level, primary occupation, income level, religion, parity, and distance to the nearest health facilityBAttitude15Utilized a 5-point Likert scale (ranging from 1 = Strongly Disagree to 5 = Strongly Agree) to assess prospective evaluations of physical, emotional, and financial male accompaniment in maternal healthCPerceived Subjective Norms15Measured perceived social pressures from peers, community elders, and religious leaders using a 5-point Likert scale (ranging from 1 = Strongly Disagree to 5 = Strongly Agree)DPerceived Behavioral Control16Evaluated perceived external or internal facilitators and barriers (e.g., clinic waiting times, financial constraints, work commitment) influencing the male partner’s capacity to participate Data collection instrument Data collection procedures Data collection was conducted by the principal investigator and a team of six trained research assistants who held bachelor’s degrees in nursing or public health and were fluent in both Swahili and local vernacular languages. Research assistants underwent an intensive three-day training workshop covering ethical protocols, unbiased questioning techniques, and cluster management. To guarantee maximum visual and auditory privacy, data collection was conducted in neutral, unoccupied community spaces (such as designated classrooms or community centers/public offices) outside the earshot of spouses or peers. Participants were seated at a minimum distance of approximately 3 m apart to prevent peer dialogue or cross-contamination of responses. Before initiating the interview, brief standardized instructions were read aloud. The average time for questionnaire administration ranged between 30 and 45 min. Completed questionnaires were assigned unique alphanumeric codes rather than personal identifiers to maintain strict confidentiality, and physical copies were stored in secure, locked cabinets accessible only to the principal investigator. Validity To establish robust content and face validity, the initial draft of the tool was submitted to a panel of reviewers, including two biostatisticians, two public health specialists, and two senior academic experts in reproductive health. These experts evaluated each item for its thematic relevance, linguistic clarity, and structural alignment with the Theory of Planned Behavior (TPB) constructs. The Content Validity Index (CVI) was calculated at both the item level (I-CVI) and the scale level (S-CVI). Items yielding an I-CVI below 0.80 were eliminated or revised. Based on expert recommendations, item phrasings were adapted to suit the lower literacy levels characteristic of rural populations, and abstract theoretical jargon was replaced with concrete examples of maternal support behaviors (e.g., replacing “maternal service utilization” with “accompanying your spouse to the antenatal clinic”). Reliability To verify the psychometric stability and internal consistency of the translated instrument, a pilot pre-test was conducted among ten (10) couples residing in an independent ward exhibiting identical sociodemographic features but excluded from the final study sample [6, 14, 16, 31]. Inter-observer reliability was assessed among the data collectors, achieving a high consensus rate (> 90%). The pilot data were subjected to an Exploratory Factor Analysis (EFA) to ensure proper factor loading structure. The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy yielded a value of 0.78, and Bartlett’s Test of Sphericity demonstrated highly significant Sphericity (p < 0.001), confirming that the data matrix was appropriate for factor extraction. Internal consistency was calculated using Cronbach’s alpha (α) coefficients for each independent domain. For the actual field data, the instrument achieved highly dependable reliability scores: α = 0.81 for the Attitude scale, α = 0.79 for Perceived Subjective Norms, and α = 0.84 for the Perceived Behavioral Control scale, exceeding the internationally recognized threshold of ≥ 0.70 for behavioral sciences [33, 34]. Variable measurement and psychometric evaluation To capture the cognitive constructs of the Theory of Planned Behavior (TPB) before the completion of obstetric events, items were designed to measure prospective behavioral intentions and attitudes (e.g., “My intention to accompany my wife to the health center for her upcoming deliveries is …”). To ensure methodological rigor and respect temporal ordering, male accompaniment was operationalized into two distinct dimensions. These included: (1) Retrospective experience, assessing the male partner’s reported accompaniment in previous pregnancies as a baseline behavioral profile, and (2) Prospective cognitive readiness, measuring current attitudes, subjective norms, perceived behavioral control, and explicit behavioral intentions to accompany their partners during the current ongoing pregnancy and upcoming childbirth. The variable “Intention to accompany a female partner to childbirth” was assessed at baseline by asking male partners whose spouses were in their first trimester whether they intended or were willing to physically accompany their partners into the labor and delivery room during childbirth. This was measured as a binary baseline indicator (“Yes” vs. “No”) of future behavioral intention, rather than an observed retrospective practice. This dual-operationalization was opted for as it aligns directly with the predictive structure of the Theory of Planned Behavior. This aligns with the psychometric architecture of the TPB, which posits that prospective behavioral intention is the most proximate cognitive precursor to actual future behavior. Construct validation was executed using Exploratory Factor Analysis (EFA). Initially, each of the three behavioral domains (Attitude, subjective norms, and perceived behavioral control) comprised 25 raw items (Table 3). Following the initial iteration of the EFA, items with poor factor loadings (< 0.40) or high cross-loadings were systematically eliminated.Table 3Output findings of the explanatory factor analysisFactor analysisAttitudeSubjective normsBehavior controlInitial statements analyzed252525Initial eigenvalues (before extraction)Comp.1 = 36.072%; Comp. 2 = 14.290%Comp.1 = 34.452%; Comp.2 = 18.777%Comp. 1 = 37.207%; Comp. 2 = 15.216%Final questions statements with component matrix > 0.3151516Initial eigenvalues (after extraction)59.98%57.26%58.06KMO measure0.7260.7820.623Mean000 Table 3Output findings of the explanatory factor analysisFactor analysisAttitudeSubjective normsBehavior controlInitial statements analyzed252525Initial eigenvalues (before extraction)Comp.1 = 36.072%; Comp. 2 = 14.290%Comp.1 = 34.452%; Comp.2 = 18.777%Comp. 1 = 37.207%; Comp. 2 = 15.216%Final questions statements with component matrix > 0.3151516Initial eigenvalues (after extraction)59.98%57.26%58.06KMO measure0.7260.7820.623Mean000 Output findings of the explanatory factor analysis Ten items were dropped from the Attitude scale, and ten items were dropped from the subjective norms scale, leaving each with 15 verified items. For the perceived behavioral control domain, 9 poorly performing items were dropped, resulting in 16 retained items. To avoid the subjective biases and sample-dependent constraints of using empirical mean splits for dichotomization, scales were categorized based on an objective 50% maximum theoretical raw score cut-off. For example, on a 5-point Likert scale, scores below the absolute mathematical mid-point of the cumulative scale items were classified as “Negative/Poor,” while scores at or above the theoretical mid-point were categorized as “Positive/Good”. This ensured an objective, criteria-based classification independent of the sample distribution. Data processing and analysis The data were checked for completeness and consistency, and then it was coded and entered into the computer using the statistical package IBM SPSS version 23. Both bivariate and multivariable analyses were performed using binary logistic regression models to identify independent predictors for each behavioral domain [6, 14, 16, 31]. As recommended and used by some previous scholars [28, 35–37–38], to assess the combined influence of covariates on the dichotomous outcomes, the logit-transformed linear version of the multivariable logistic regression model was defined as follows:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned} In(\frac{P}{1-P}) &={\beta}_{0}+{\beta}_{1}{X}_{1}\\&+{\beta}_{2}{X}_{2}+{\beta}_{3}{X}_{3}\\&+\dots +{\beta}_{k}{X}_{k}\end{aligned}$$\end{document}where: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$In=(\frac{P}{1-P})$$\end{document}: Represents the log-odds (logit) of the binary outcome of interest (e.g., positive vs. negative behavioral domain). P: Is the predicted probability that a male partner exhibits as a positive construct. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\beta}_{0}:$$\end{document} Is the intercept parameter (constant). \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\beta}_{1},{\beta}_{2},{\beta}_{3}\dots ,{\beta}_{k}:$$\end{document} Represent the regression coefficients (slopes) assigned to individual independent variables. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${X}_{1},{X}_{2}, \dots ,{X}_{k}:$$\end{document} Represent the sociodemographic and structural predictor variables entered into the multivariable framework. Variables demonstrating a significance level of p < 0.20 during bivariate evaluation were simultaneously entered into the multivariable model. Statistical significance thresholds for the final adjusted odds ratios (AOR) were strictly constrained to p < 0.05 with corresponding 95% confidence intervals (CI). Results Socio-demographic characteristics profiles of the study respondents Findings in Table 4 show that 546 male partners of pregnant women were included in the study, with a response rate of 100%. The mean age of the study respondents was 31 ± 7.726. The majority (71.8%, n = 392) were married, 85.5% (n = 467) were monogamous, 70.0% (n = 382) used less than 1 dollar per day, 82.8% (n = 452) got basic obstetric care services from dispensaries, and 64.7% (n = 353) had completed primary school education. Refer to the table for other findings.Table 4Socio-demographic characteristics of study respondents (N = 546)CharacterFrequency (n)Percent (%)Age (years): M ± SD = 31 ± 7.726 15–24 yrs14626.8 25 to 34 yrs17031.1 35–49 yrs23042.1Ethnic group Fipa36767.2 Mambwe11821.6 Others6111.2Marital status Cohabit15428.2 Married39271.8Type of family Monogamous46785.5 Polygamous7914.5Education level Non-formal15528.3 Primary School35364.7 Secondary school or Higher387.0Income per day Less than 1 dollar38270.0 More than 1 dollar16430.0Own radio Yes30856.4 No23843.6Own mobile phone Yes23442.9 No31257.1Household size At least 27613.9 3 to 525045.8 6 or more22040.3Adult female in the family None31557.7 1 or more23142.3Covered by Health Insurance Yes17031.1 No37668.9Health facility Dispensary45282.8 Health center9417.2Distance to health facility (kilometers) Less than 1 km25947.5 1 to 5 km23247.4 More than 5 km5510.1 Table 4Socio-demographic characteristics of study respondents (N = 546)CharacterFrequency (n)Percent (%)Age (years): M ± SD = 31 ± 7.726 15–24 yrs14626.8 25 to 34 yrs17031.1 35–49 yrs23042.1Ethnic group Fipa36767.2 Mambwe11821.6 Others6111.2Marital status Cohabit15428.2 Married39271.8Type of family Monogamous46785.5 Polygamous7914.5Education level Non-formal15528.3 Primary School35364.7 Secondary school or Higher387.0Income per day Less than 1 dollar38270.0 More than 1 dollar16430.0Own radio Yes30856.4 No23843.6Own mobile phone Yes23442.9 No31257.1Household size At least 27613.9 3 to 525045.8 6 or more22040.3Adult female in the family None31557.7 1 or more23142.3Covered by Health Insurance Yes17031.1 No37668.9Health facility Dispensary45282.8 Health center9417.2Distance to health facility (kilometers) Less than 1 km25947.5 1 to 5 km23247.4 More than 5 km5510.1 Socio-demographic characteristics of study respondents (N = 546) Proportional distribution of male accompaniment in maternal service utilization Descriptive analysis was performed to determine the overall level of male participation in maternal health services. As illustrated in Fig. 1, the majority of the study participants had never been involved in maternal service utilization (61.7%, n = 337), against 38.3% (n = 209) who reported active accompaniment.Fig. 1Proportional distribution of male accompaniment in maternal service utilization (N = 546) Proportional distribution of male accompaniment in maternal service utilization (N = 546) Proportional distribution of the theory of planned behavior constructs regarding male accompaniment in maternal service utilization The study applied the TPB framework to determine the proportional distribution of its behavior-intention factors (constructs) on male accompaniment in maternal service utilization. Findings in Fig. 2 indicate that a similar pattern emerged across all three behavioral dimensions, with negative perceptions consistently outweighing positive ones. Specifically, 41.0% (n = 223) of male partners demonstrated a positive attitude toward accompaniment in maternal service utilization, compared to 59.0% (n = 323) who held a negative attitude. Regarding perceived subjective norms, 43.0% (n = 234) reported positive norms, while 57.0% (n = 312) reported negative subjective norms. Similarly, for perceived behavioral control, 43.0% (n = 234) exhibited positive behavioral control, whereas 57.0% (n = 312) reported negative perceived behavioral control.Fig. 2Proportional Distribution of Attitude and the Behavior-Intention-Factors of a Theory of Planned Behavior on Male Accompaniment in Maternal Service Utilization (N = 546) Proportional Distribution of Attitude and the Behavior-Intention-Factors of a Theory of Planned Behavior on Male Accompaniment in Maternal Service Utilization (N = 546) Behavior-intention factors of a theory of planned behavior and male accompaniment in maternal service utilization Findings in Table 5 indicated that the odds of male partners’ positive attitude (AOR = 3.616; p < 0.05; 95%CI: 1.333, 6.138); perceived subjective norm (AOR = 2.374; p < 0.05; 95%CI: 1.669, 5.984) and perceived behavioral control (AOR = 1.759; p < 0.05; 95%CI: 1.004, 4.373) on male accompaniment in maternal service utilization were significantly high contrary to their counterparts with negative attitude; perceived subjective norm and perceived behavioral control respectively.Table 5Behavior-intention factors of a theory of planned behavior on male accompaniment in maternal service utilization (N = 546)VariablesAOR95% CIp-valueLowerUpperAttitude Positive3.6161.3336.1380.019 Negative1Perceived Subjective Norms Positive2.3741.6695.9840.038 Negative1Perceived Behavioral Control Positive1.7591.0044.3730.033 Negative1 Table 5Behavior-intention factors of a theory of planned behavior on male accompaniment in maternal service utilization (N = 546)VariablesAOR95% CIp-valueLowerUpperAttitude Positive3.6161.3336.1380.019 Negative1Perceived Subjective Norms Positive2.3741.6695.9840.038 Negative1Perceived Behavioral Control Positive1.7591.0044.3730.033 Negative1 Behavior-intention factors of a theory of planned behavior on male accompaniment in maternal service utilization (N = 546) Bivariate analysis of factors associated with TBP constructs towards male accompaniment in maternal service utilization Cross-tabulation with Chi-square tests was performed to explore the relationship between male partners’ background characteristics and the three TBP dimensions, including attitude, subjective norms, and perceived behavioral control (Table 6). Out of the nine characteristics tested, three factors were significantly related to the partners’ attitude towards maternal health service utilization. The ethnic group showed a highly significant relationship (X2 = 32.177, p < 0.001), with the Mambwe ethnic group exhibiting a higher proportion of positive attitudes (55.9%) relative to their sample size than the Fipa or other groups. Walking distance to the health facility also significantly influenced attitudes (X2 = 11.786, p < 0.003), as did having ever heard of birth preparedness and complication readiness (X2 = 6.676, p = 0.010). Socio-demographic factors such as age group, education, and economic status did not reach statistical significance.Table 6Bivariate analysis of factors associated with TBP constructs towards male accompaniment in maternal service utilization (N = 546)VariableAttitudeX2(p-value)Subjective NormsX2(p-value)Perceived Behavioral ControlX2(p-value)Negative n (%)Positive n (%)Negative n (%)Positive n (%)Negative n (%)Positive n (%)Age Group 15–24 yrs. (n = 112)86(76.8)26(23.2)0.321 (0.988)48(51.6)45(48.4)2.375 (0.667)48(51.6)45(48.4)2.375 (0.667) 25–34 yrs. (n = 188)108(57.4)80(42.6)124(60.2)82(39.8)124(60.2)82(39.8) 35–49 yrs. (n = 246)129(52.4)117(47.6)140(56.7)107(43.3)140(56.7)107(43.3)Age at Marriage≤ 18 yrs. (n = 233)136(58.4)97(41.6)2.220 (0.329)128(54.9)105(45.1)0.825(0.662)128(54.9)105(45.1)0.825(0.662) 19–24 yrs. (n = 251)155(61.8)96(38.2)148(59.0)103(41.0)148(59.0)103(41.0)≥ 25 yrs. (n = 62)32(51.6)30(48.4)36(58.1)26(41.9)36(58.1)26(41.9)Education Status No Formal (n = 192)115(59.9)77(40.1)0.191 (0.909)113(58.9)79(41.1)0.359 (0.836)113(58.9)79(41.1)0.359 (0.836) Primary School (n = 327)193(59.0)134(41.0)184(56.3)143(43.7)184(56.3)143(43.7) Secondary/Higher (n = 27)15(55.6)12(44.4)15(55.6)12(44.4)15(55.6)12(44.4)Ethnic Group Fipa (n = 349)215(61.6)134(38.4)32.177 (< 0.001)209(59.9)140(40.1)28.586 (< 0.001)209(59.9)140(40.1)28.590 (< 0.001) Mambwe (n = 126)51(40.5)75(59.5)49(38.9)77(61.1)49(38.9)77(61.1) Others (n = 71)57(80.5)14(19.7)54(76.1)17(23.9)54(76.1)17(23.9)Economic Status< $1 per day (n = 395)231(58.5)164(41.5)0.271 (0.603)228(57.7)167(42.3)0.195 (0.659)228(57.7)167(42.3)0.195 (0.659)≥ $2 per day (n = 151)92(60.9)59(39.1)84(55.6)67(44.4)84(55.6)67(44.4)Own Mobile Phone Yes(n = 134)72(53.7)62(46.3)2.161 (0.141)74(55.2)60(44.8)0.267 (0.605)74(55.2)60(44.8)0.267 (0.605) No (n = 412)251(60.9)161(39.1)238(57.8)174(42.2)238(57.8)174(42.2)Walking Distance≤ 1 km (n = 258)143(55.4)115(44.6)11.786 (0.003)------ 2–4 km (n = 234)156(66.7)78(33.3)----≥ 5 km (n = 54)24(44.4)30(55.6)----Ever heard of BPCR Yes (n = 447)253(56.6)194(43.4)6.676 (0.010)------ No (n = 99)70(70.7)29(29.3)----Intention to Accompany a Female Partner to Childbirth No (n = 136)---93(68.4)43(31.6)9.342 (0.002)93(68.4)43(31.6)9.342 (0.002) Yes (n = 410)--219(53.4)191(46.6)219(53.4)191(46.6)BPCR Birth Preparedness and Complication Readiness, Em dashes (-) = indicate variables that were not entered into that specific relationship model based on bivariate screening procedures, Km Kilometer, n = Sample, X2 = Chi-square test value, P = Probability value Table 6Bivariate analysis of factors associated with TBP constructs towards male accompaniment in maternal service utilization (N = 546)VariableAttitudeX2(p-value)Subjective NormsX2(p-value)Perceived Behavioral ControlX2(p-value)Negative n (%)Positive n (%)Negative n (%)Positive n (%)Negative n (%)Positive n (%)Age Group 15–24 yrs. (n = 112)86(76.8)26(23.2)0.321 (0.988)48(51.6)45(48.4)2.375 (0.667)48(51.6)45(48.4)2.375 (0.667) 25–34 yrs. (n = 188)108(57.4)80(42.6)124(60.2)82(39.8)124(60.2)82(39.8) 35–49 yrs. (n = 246)129(52.4)117(47.6)140(56.7)107(43.3)140(56.7)107(43.3)Age at Marriage≤ 18 yrs. (n = 233)136(58.4)97(41.6)2.220 (0.329)128(54.9)105(45.1)0.825(0.662)128(54.9)105(45.1)0.825(0.662) 19–24 yrs. (n = 251)155(61.8)96(38.2)148(59.0)103(41.0)148(59.0)103(41.0)≥ 25 yrs. (n = 62)32(51.6)30(48.4)36(58.1)26(41.9)36(58.1)26(41.9)Education Status No Formal (n = 192)115(59.9)77(40.1)0.191 (0.909)113(58.9)79(41.1)0.359 (0.836)113(58.9)79(41.1)0.359 (0.836) Primary School (n = 327)193(59.0)134(41.0)184(56.3)143(43.7)184(56.3)143(43.7) Secondary/Higher (n = 27)15(55.6)12(44.4)15(55.6)12(44.4)15(55.6)12(44.4)Ethnic Group Fipa (n = 349)215(61.6)134(38.4)32.177 (< 0.001)209(59.9)140(40.1)28.586 (< 0.001)209(59.9)140(40.1)28.590 (< 0.001) Mambwe (n = 126)51(40.5)75(59.5)49(38.9)77(61.1)49(38.9)77(61.1) Others (n = 71)57(80.5)14(19.7)54(76.1)17(23.9)54(76.1)17(23.9)Economic Status< $1 per day (n = 395)231(58.5)164(41.5)0.271 (0.603)228(57.7)167(42.3)0.195 (0.659)228(57.7)167(42.3)0.195 (0.659)≥ $2 per day (n = 151)92(60.9)59(39.1)84(55.6)67(44.4)84(55.6)67(44.4)Own Mobile Phone Yes(n = 134)72(53.7)62(46.3)2.161 (0.141)74(55.2)60(44.8)0.267 (0.605)74(55.2)60(44.8)0.267 (0.605) No (n = 412)251(60.9)161(39.1)238(57.8)174(42.2)238(57.8)174(42.2)Walking Distance≤ 1 km (n = 258)143(55.4)115(44.6)11.786 (0.003)------ 2–4 km (n = 234)156(66.7)78(33.3)----≥ 5 km (n = 54)24(44.4)30(55.6)----Ever heard of BPCR Yes (n = 447)253(56.6)194(43.4)6.676 (0.010)------ No (n = 99)70(70.7)29(29.3)----Intention to Accompany a Female Partner to Childbirth No (n = 136)---93(68.4)43(31.6)9.342 (0.002)93(68.4)43(31.6)9.342 (0.002) Yes (n = 410)--219(53.4)191(46.6)219(53.4)191(46.6)BPCR Birth Preparedness and Complication Readiness, Em dashes (-) = indicate variables that were not entered into that specific relationship model based on bivariate screening procedures, Km Kilometer, n = Sample, X2 = Chi-square test value, P = Probability value Bivariate analysis of factors associated with TBP constructs towards male accompaniment in maternal service utilization (N = 546) 0.825 (0.662) 0.825 (0.662) BPCR Birth Preparedness and Complication Readiness, Em dashes (-) = indicate variables that were not entered into that specific relationship model based on bivariate screening procedures, Km Kilometer, n = Sample, X2 = Chi-square test value, P = Probability value Regarding subjective norms and perceived behavioral control, identical statistical patterns emerged across the sample. Two key factors were found to be significantly associated with both constructs, including ethnic group, which showed a highly significant association (X2 = 28.586, p < 0.001 for subjective norms; X2 = 28.586, p < 0.001 for the perceived behavioral control). Moreover, based on the intention to accompany their partners to childbirth, men who expressed a willingness to accompany their partners during labor and childbirth demonstrated significantly more positive subjective norms and higher perceived behavioral control (X2 = 9.342, p = 0.002 for both). None of the other structural or demographic characteristics, including age, marriage age, education, economic status, or mobile phone ownership, was significantly associated with subjective norms or perceived behavioral control (p > 0.05). Multivariable predictors of TPB constructs regarding male accompaniment in maternal health services utilization Binary logistic regression analysis was conducted to determine the independent predictors of positive attitudes, positive subjective norms, and positive perceived behavioral control regarding male accompaniment in maternal health services utilization (Table 7). After adjusting for confounding variables, ethnic group, having heard of birth preparedness and complication readiness, and walking distance remained significantly associated with a positive attitude. Male partners from the Mambwe ethnic group had more than twice the odds of holding a positive attitude compared to those from the Fipa group (AOR = 2.089; 95%CI: 1.360, 3.210; p < 0.05). Additionally, male partners belonging to “other” ethnic groups were less likely to have a positive attitude to accompany a female partner for maternal health services (AOR = 0.372, 95%CI: 0.198, 0.699; p < 0.05).Table 7Multivariable predictors of TPB constructs regarding male accompaniment in maternal health servicesVariablesPositive AttitudePositive Subjective NormsPositive Perceived Behavioral ControlAOR(95% CI)AOR(95% CI)AOR(95% CI)Ethnic Group Fipa1.00(Ref.)1.00(Ref.)1.00(Ref.) Mambwe2.089(1.360–3.210)*2.316(1.523–3.521)*2.326(1.523–3.521)* Others0.372(0.198–0.699)*0.540(0.296–0.984)*0.540(0.296–0.984)*Birth Preparedness No1.00(Ref.)-- Yes1.717(1.053–2.799)*--Intention to Accompany a Female Partner to Childbirth No-1.00(Ref.)1.00(Ref.) Yes-1.597(1.040–2.453)*1.60(1.04–2.45)Walking Distance≤ 1 km1.00(Ref.)-- 1 to 5 km1.210(0.652–2.245)--≥ 6 km0.609(0.417–0.889)*--AOR Adjusted Odds Ratio, CI Confidence Interval, Ref. Reference category, Em dashes (-) = indicate variables that were not entered into that specific regression model based on bivariate screening results*Represents categories that are significant at p < 0.05 Table 7Multivariable predictors of TPB constructs regarding male accompaniment in maternal health servicesVariablesPositive AttitudePositive Subjective NormsPositive Perceived Behavioral ControlAOR(95% CI)AOR(95% CI)AOR(95% CI)Ethnic Group Fipa1.00(Ref.)1.00(Ref.)1.00(Ref.) Mambwe2.089(1.360–3.210)*2.316(1.523–3.521)*2.326(1.523–3.521)* Others0.372(0.198–0.699)*0.540(0.296–0.984)*0.540(0.296–0.984)*Birth Preparedness No1.00(Ref.)-- Yes1.717(1.053–2.799)*--Intention to Accompany a Female Partner to Childbirth No-1.00(Ref.)1.00(Ref.) Yes-1.597(1.040–2.453)*1.60(1.04–2.45)Walking Distance≤ 1 km1.00(Ref.)-- 1 to 5 km1.210(0.652–2.245)--≥ 6 km0.609(0.417–0.889)*--AOR Adjusted Odds Ratio, CI Confidence Interval, Ref. Reference category, Em dashes (-) = indicate variables that were not entered into that specific regression model based on bivariate screening results*Represents categories that are significant at p < 0.05 Multivariable predictors of TPB constructs regarding male accompaniment in maternal health services AOR Adjusted Odds Ratio, CI Confidence Interval, Ref. Reference category, Em dashes (-) = indicate variables that were not entered into that specific regression model based on bivariate screening results *Represents categories that are significant at p < 0.05 Nevertheless, male partners who had heard of birth preparedness and complications readiness had significantly higher odds of a positive attitude compared to those who had not (AOR = 1.717; 95%CI: 1.053, 2.799; p < 0.05). Regarding spatial access, living a moderate distance (1 to 5 km) from health facilities was associated with higher odds of a positive attitude compared to living a nearby and higher distance (≥ 6 km) (AOR = 0.609; 95%CI: 0.417, 0.889; p < 0.05). The predictive models for positive subjective norms and positive perceived behavioral control revealed identical independent determinants. Mambwe ethnicity was strongly predictive of both positive subjective norms (AOR = 2.316; 95%CI: 1.523, 3.521; p < 0.05) and positive perceived behavioral control (AOR = 2.326; 95%CI: 1.523, 3.521; p < 0.05). Furthermore, the intention to accompany a partner to childbirth was a significant predictor of both outcomes. Men who intended to accompany their partners had 1.6 times higher odds of reporting positive subjective norms (AOR = 1.597; 95%CI: 1.040, 2.453; p < 0.05) and positive perceived behavioral control (AOR = 1.597; 95%CI: 1.040, 2.453; p < 0.05) than those who did not. Discussion According to the Theory of Planned Behavior (TPB), an individual’s health-seeking behavior is fundamentally driven by a cognitive constellation of personal attitudes, perceived social pressure, and self-efficacy [15]. This study utilized a consolidated TPB framework to map how structural barriers and deeply entrenched socio-cultural dimensions independently shape male partners’ engagement in maternal health service utilization. By analyzing these constructs simultaneously, the study successfully captured both the internal mindsets and external systemic realities that govern spousal accompaniment in rural reproductive health networks. The findings of the descriptive analysis of this study revealed a profound deficit in active male participation, with the significant majority of male partners reporting they had never been involved in maternal health services utilization. This behavioral passivity corresponds directly with a dominant negative trend across all three underlying TPB constructs, including attitudes, subjective norms, and perceived behavioral control [39]. This widespread negative orientation contrasts sharply with findings from similar maternal health evaluations in Iran and India, where male partners demonstrated overwhelmingly positive attitudes and routinely participated in pregnancy care without facing social stigma [15]. In the context of this rural study community, these negative cognitive constructs persist because pregnancy and childbirth remain culturally constructed as exclusively female domains [40]. While men are socially expected to act as financial providers, physically entering maternal health spaces is often locally interpreted as a violation of traditional gender roles [41]. Men who attempt to break this boundary often face subtle social sanctions, domestic ridicule, or peer disapproval, which systematically depresses their enthusiasm to participate [42]. To dismantle these entrenched normative barriers, the Ministry of Health should leverage these findings to design targeted social and behavioral change communication (SBCC) programmes that engage local traditional gatekeepers to reframe clinical spousal accompaniment as an essential aspect modern partner. Findings of the multivariable logistic regression models demonstrated that ethnicity acts as a powerful, cross-cutting independent predictor across all three TPB constructs. Specifically, male partners from the Mambwe ethnic group exhibited significantly higher odds of holding positive attitudes, reporting positive subjective norms, and demonstrating positive perceived behavioral control compared to their Fipa counterparts. This significant variation aligns with wider implementation science literature from Sub-Saharan Africa, which establishes that maternal health-seeking behaviors are rarely uniform across a single geographical region; instead, they are highly fragmented by sub-district tribal governance and lineage systems [32, 43]. In the study’s immediate context, this divergence suggests that Mambwe socio-cultural frameworks and contemporary community practices place a much stronger and more progressive emphasis on shared spousal responsibility during reproductive milestones than neighboring tribes. Fipa cultural structures, by contrast, appear to enforce a more rigid domestic division of labor that distances men from clinical spaces. Consequently, reproductive health planners should reject one-size-fits-all strategies and instead implement culturally tailored peer-education interventions that utilize traditional elders from progressive cohorts to diffuse positive behavioral models to adjacent, conservative communities [44]. Exposure to target health information emerged as a vital entry point for reshaping personal mindsets. Men who had previously been exposed to messages regarding birth preparedness had significantly higher odds of maintaining a positive attitude toward maternal accompaniment than those who had not. This finding strongly reinforces existing maternal health literature across East Africa, which consistently demonstrates that structured reproductive health literacy directly correlates with increased physical and emotional spousal support during the perinatal period [29]. Possible reasons for this are grounded in the fact that when men are exposed to Birth Preparedness and Complications Readiness (BPCR) frameworks, they undergo a cognitive shift [45]. They begin to view physical accompaniment not as an unnecessary intrusion into a female space, but as a functional, protective marital obligation designed to safeguard their family against unexpected obstetric emergencies [46]. Formal knowledge effectively demystifies the clinical environment, bridging the traditional gap between passive financial provision and active physical presence [47]. This clear link signifies an immediate programmatic need to transition to couple-friendly Antenatal Care (ANC) protocols where health facilities structurally integrate male partners into joint BPCR educational consultations. A robust, identical predictive relationship was established between a man’s explicit willingness to accompany his partner during childbirth and his positive subjective norms and perceived behavioral control. This finding provides empirical support for the core theoretical assumptions of the TPB, which indicate that strong internal self-efficacy and perceived social alignment are the direct precursors to behavioral intentions and ultimate practice [32]. In this rural environment, men who possess an intrinsic desire to witness childbirth are better psychologically equipped to withstand localized stigma or traditional taboos. Their high perceived behavioral control functions as a cognitive shield because they feel highly capable of navigating the healthcare system and believe their immediate inner circle approves, they actively choose to ignore broader community disapproval [48]. To reinforce this internal self-efficacy, hospital managers should remove institutional disincentives by redesigning delivery wards to feature private spaces that accommodate and structurally validate male presence during intrapartum care. Additionally, physical access to health infrastructure emerged as a critical structural bottleneck, whereas living a moderate distance away from a healthcare facility significantly reduced the odds of a male partner holding a positive attitude toward maternal accompaniment compared to those living in proximity. This finding aligns seamlessly with the geographical dimensions of the classic three delays model in maternal health, which identifies physical distance and transit logistics as primary structural impediments to service utilization [15]. When healthcare facilities require long walking distances over difficult rural terrain, the opportunity cost of male accompaniment skyrockets [49]. Men have to weigh the physical burden and the loss of a productive working day against a behavior that their culture already tells them is low-priority [42]. This finding highlights that as distance increases, men are highly likely to relegate pregnancy care back to the custody of women, viewing the journey as an inefficient use of male domestic labor [42]. To mitigate these spatial barriers and lower the threshold for male participation, local governments should prioritize decentralized infrastructure by expanding dispensary-level maternity wings closer to peripheral villages [15]. By applying the TPB framework, this study contributes an empirical model demonstrating how socio-cultural identities and structural barriers independently shape male partner configurations, providing a blueprint for shifting behavioral paradigms from passive financial providers to active spousal accompaniment in rural maternal healthcare [50]. These findings are highly instrumental for Tanzanian policymakers shifting focus from purely financial interventions toward behavioral barriers, offering regional health administrators a framework to design targeted, couple-centric, and culturally tailored health programmes. However, the contributions of this study must be considered alongside its limitations, including the cross-sectional design that restricts the ability to establish definitive causal relationships over time, and the reliance on self-reported data from a single rural district introduces potential social desirability bias and limits the immediate generalizability of the behavioral patterns of urban or vastly different socio-cultural regions across Tanzania. Conclusion The findings of this study uncovered that male partner accompaniment in maternal health service utilization remains profoundly low in this rural community, a behavioral deficit that is tightly mirrored by predominantly negative personal attitudes, oppressive subjective norms, and low perceived behavioral control. Guided by the Theory of Planned Behavior, this study demonstrates that male participation is not merely an individual choice but a complex intersection of sociocultural identity, structural access, and health literacy. Ethnicity stands out as a powerful independent predictor, with the Mambwe partners exhibiting far greater cognitive readiness and positive behavioral intentions toward maternal care than their Fipa counterparts. Furthermore, exposure to birth preparedness messaging and closer geographical proximity to health facilities significantly optimize positive spousal attitudes and self-efficacy. Based on these conclusions, the following recommendations are made:i.Implement culturally tailored SBCC campaigns: The Ministry of Health, in collaboration with local government authorities, should design targeted social and behavior change communication (SBCC) programs. Rather than using generic regional messages, these campaigns should engage Mambwe traditional elders and community champions to model and diffuse progressive norms of spousal collaboration to neighboring Fipa and other conservative ethnic groupsii.Institutionalize couple-centric ANC: Reproductive health facility management should transition away from female-exclusive care spaces. The ANC guidelines should be structurally updated to mandate joint, couple-friendly consultations where birth preparedness and complication readiness educational packages are delivered to both partners simultaneously, effectively converting passive financial providers into active physical participantsiii.Decentralize maternal health infrastructure: To mitigate the restrictive opportunity costs imposed by long walking distances, local health departments must prioritize the expansion of community dispensary-level maternal wings and optimize mobile reproductive health clinics. Reducing the physical distance barrier directly lowers the structural threshold for male accompanimentiv.Adapt clinical environments for delivery accompaniment: To support and reinforce the high perceived behavioral control found among those willing to witness childbirth, healthcare facilities should adapt labor ward infrastructures to offer private, role-specific, dignified delivery spaces that safely accommodate male presence. i.Implement culturally tailored SBCC campaigns: The Ministry of Health, in collaboration with local government authorities, should design targeted social and behavior change communication (SBCC) programs. Rather than using generic regional messages, these campaigns should engage Mambwe traditional elders and community champions to model and diffuse progressive norms of spousal collaboration to neighboring Fipa and other conservative ethnic groups Implement culturally tailored SBCC campaigns: The Ministry of Health, in collaboration with local government authorities, should design targeted social and behavior change communication (SBCC) programs. Rather than using generic regional messages, these campaigns should engage Mambwe traditional elders and community champions to model and diffuse progressive norms of spousal collaboration to neighboring Fipa and other conservative ethnic groups ii.Institutionalize couple-centric ANC: Reproductive health facility management should transition away from female-exclusive care spaces. The ANC guidelines should be structurally updated to mandate joint, couple-friendly consultations where birth preparedness and complication readiness educational packages are delivered to both partners simultaneously, effectively converting passive financial providers into active physical participants Institutionalize couple-centric ANC: Reproductive health facility management should transition away from female-exclusive care spaces. The ANC guidelines should be structurally updated to mandate joint, couple-friendly consultations where birth preparedness and complication readiness educational packages are delivered to both partners simultaneously, effectively converting passive financial providers into active physical participants iii.Decentralize maternal health infrastructure: To mitigate the restrictive opportunity costs imposed by long walking distances, local health departments must prioritize the expansion of community dispensary-level maternal wings and optimize mobile reproductive health clinics. Reducing the physical distance barrier directly lowers the structural threshold for male accompaniment Decentralize maternal health infrastructure: To mitigate the restrictive opportunity costs imposed by long walking distances, local health departments must prioritize the expansion of community dispensary-level maternal wings and optimize mobile reproductive health clinics. Reducing the physical distance barrier directly lowers the structural threshold for male accompaniment iv.Adapt clinical environments for delivery accompaniment: To support and reinforce the high perceived behavioral control found among those willing to witness childbirth, healthcare facilities should adapt labor ward infrastructures to offer private, role-specific, dignified delivery spaces that safely accommodate male presence. Adapt clinical environments for delivery accompaniment: To support and reinforce the high perceived behavioral control found among those willing to witness childbirth, healthcare facilities should adapt labor ward infrastructures to offer private, role-specific, dignified delivery spaces that safely accommodate male presence. Supplementary Information Supplementary Materials 1. Supplementary Materials 1.