The relationship between inequitable gender norms and provider attitudes and quality of care in maternal health services in Rwanda: a mixed methods study

Recruitment and data collection with RMNH providers took place from January to February 2018. We determined a minimum required sample size of 146 health providers from across all three districts (49 per district) using a conservative value estimate of p = 50%, desired precision of +/− 10, 80% power and alpha = .05, along with an adjustment for a conservative design effect of 1.5. The desired sample size was then rounded up to an even 50 providers per district (a total of 150). The sample size also met rule of thumb criteria of a minimum of 10 respondents per item for factor analysis [22]. Within each district, respondents were recruited from a total of 10 health facilities: the district/referral hospital and nine randomly selected health centers, selected from the full list of health centers in each district (excluding prison-based facilities) using a random number generator in STATA Version 14 (Stata Corporation, College Station, TX). On the day of the survey, convenience sampling was used to select eligible providers. Upon reaching each facility, the study team made separate, sex-disaggregated lists of providers who were present in the facility and met the eligibility criteria. Every other provider, alternating from each sex-disaggregated list, was then invited to participate (1 female, 1 male, and so on), until the desired sample size at facility and district level was reached. The number of providers recruited from each facility was capped at 10 respondents for hospitals and five for health centers, to ensure that providers were recruited from all 10 selected facilities per district. Equal numbers of male and female providers were sought to enable comparison between male and female providers’ responses, but the ratio of male to female providers surveyed was limited by the sex distribution of providers working in the district, and the number and sex of eligible providers present on the day the survey was administered. Ultimately 151 respondents participated (61 men and 90 women), out of a total of 295 eligible providers.

Recruitment and data collection for focus group discussions (FGDs) took place in February 2019. Convenience sampling was used to select respondents from a sub-sample of selected health facilities and their catchment areas in Ngoma district. The qualitative study was limited to one district due to financial and human resource constraints. Eligible RMNH providers were recruited from the district hospital and two health centers. Eligible clients were recruited from the catchment areas of two health centers. The health facilities selected were located in close proximity to each other, to minimize respondents’ travel time. The study team received lists of eligible RMNH providers (disaggregated by sex) from health facility managers, and of eligible female and male clients willing to participate in the study from local community health workers. The study team contacted every other individual on the respective lists by telephone to verify eligibility and enroll willing participants until the desired sample size was reached.

Three separate FGDs were conducted with 10 RMNH providers (five men, five women; five nurses, four midwives, one doctor), 10 female clients, and 10 male clients, respectively. The FGDs were conducted in Kinyarwanda by two Rwandan researchers (one male and one female) trained by the research team; FGDs with clients were conducted by sex-matched interviewers. The discussions were conducted in rooms with auditory privacy within public buildings such as sector offices, schools or health centers, and lasted on average one and a half to 2 h. All FGD participants were provided with a transportation stipend of 3000 Rwandan francs to enable their participation. Mothers who attended the FGD with infants were offered food/beverages.

Quantitative data collection consisted of a 32-question survey developed for this study (Additional File 1), which examined: providers’ attitudes related to gender roles and relations, the provision and quality of RMNH care, and provider-client interactions and power dynamics related to labor and childbirth. The surveys were self-administered by recruited health providers and then placed in sealed envelopes and collected by the research team. Fourteen questions on the survey examined gender attitudes related to gender roles, intimate partner violence, household decision-making, sex, and reproduction. The questions were inspired by or adapted from the Gender Equitable Men (GEM) Scale [23] and the International Men and Gender Equality Survey (IMAGES). Fourteen questions examined attitudes related to RMNH service provision and QOC, all of which were created or adapted for the study based on relevant literature describing dimensions of QOC and disrespect and abuse [16, 24], discussions with the study team, and knowledge of the local context. For the attitude questions, respondents were asked the degree to which they agreed with particular statements (1 = strongly agree, 2 = agree, 3 = disagree, or 4 = strongly disagree). Most statements were phrased such that disagreement with them indicated a more equitable attitude, with the exception of five statements for which agreement indicated a more equitable attitude. These five statements were reverse coded (1 = strongly disagree, 2 = disagree, 3 = agree, or 4 = strongly agree), such that higher scores indicate more equitable attitudes.

We also conducted an exploratory factor analysis on the responses to the 14 statements on RMNH quality of care to create an RMNH QOC attitude scale (including responses from 142 service providers with no missing values across the items). This resulted in a 12-item scale with a borderline acceptable alpha of 0.62. An RMNH QOC attitudes score was then constructed for each respondent across the 12 items, ranging from 1 to 4, with higher scores indicating attitudes more supportive of respectful or quality RMNH care. A multivariate ordinary least squares (OLS) regression model was then used to examine the associations between the scores on the two scales (gender attitudes and RMNH QOC attitudes), controlling for demographic factors such as age, sex, and type of service provided (i.e. provides both family planning (FP) and maternal or newborn health (MNH) services vs. providing only one or neither of these services). The full study explored the broader range of RMNH services, including family planning services. As such, our analysis included examining any differences in attitudes between providers who solely provided family planning services and those who also provided maternal and newborn health services.

Vignettes – or brief, evocative case studies – of hypothetical women and men attending RMNH services were utilized in the three FGDs (with providers, female clients, male clients) to explore respondents’ perceptions and experiences of accessing or providing RMNH care. Vignettes can be a useful tool for exploring social norms and sensitive topics, such as violence or abusive care, which respondents may otherwise feel uncomfortable to discuss [25, 26]. For example, discussing fictional scenarios can be a less personal or threatening way to discuss topics than responding to direct questions about their own opinions and experiences, and can enable respondents to decide if or when to share their own experiences. The vignettes in this study were developed based on team discussions, a review of relevant and country specific research, and situations and trends encountered by MCSP during project implementation in Rwanda. They included scenarios of fictional women attending ANC, maternity and family planning services – including one describing an instance of obstetric violence – which were tailored to each study population. The scenarios touched on issues of quality of RMNH care, provider-client interactions, disrespect and abuse, and men’s presence at RMNH services, for example:

Claudine is now pregnant for the first time. She has been married for about 18 months, and she and her husband Jean Marie have a small plot of land that they farm, which has allowed them to purchase health insurance. Let’s think about Claudine’s first visit to the antenatal clinic. Her husband joined her for this visit, as recommended.

Respondents were asked how they imagined or believed the characters in the situations would or should act, including how provider-client interactions might differ depending on a woman’s age, number of pregnancies or whether she was accompanied by a partner. Semi-structured discussion questions were also used to further explore themes of provider-client interactions, barriers to and quality of RMNH care, and men’s attendance at RMNH services.

The audio-recordings of the FGDs were translated and transcribed verbatim from Kinyarwanda into English by a professional translator. Three of the authors (SK, FS, MK) listened to the audio files and verified translation and transcription, after which the audio recordings were deleted to ensure confidentiality of respondents. The analysis included both deductive and inductive approaches. A coding structure was developed based on the FGD guides and existing research and further refined based on the data in the transcripts, to enable thematic analysis [27]. Transcripts were coded and analyzed in Dedoose software by a member of the research team and reviewed by the lead author. Emerging findings were discussed by the lead author with members of the research team to verify the analysis and interpretation of the data.

In the qualitative research, female and male clients described positive provider-client interactions as a defining aspect of quality RMNH care. Both female and male clients emphasized how the initial interaction with a provider shaped their perceptions of the service quality, and described the need for a friendly reception by the health provider. Several female clients described the importance of providers being kind, spending time, and listening attentively to women attending services, which they noted were important for developing trust and alleviating their anxieties. One female client suggested the health provider’s behavior determines how comfortable a woman is; another noted that when trust is not built, women may withhold vital information from the provider, which can undermine their care. Female clients also emphasized the importance of providers explaining the recommendations and care being given.

When discussing poor quality RMNH care, female and male clients described providers using harsh, unkind or “bad” language. Women and men both noted that clients can be afraid of health providers and suggested a fear of being spoken to badly can deter people from attending RMNH services. Male clients in particular complained about health providers who were distracted or delayed providing care, for example by being on the phone (for personal reasons). One male client stated that it can make a person feel “that there is something more important to the health provider.” Clients also described how negative experiences or interactions with providers can deter future care-seeking at the same facility or with a particular provider, for themselves, as well as their partners and friends. However, several clients suggested that most health providers treat people kindly these days, and negative interactions are due to a provider’s personality, as shared by this female client;

“Sometimes a health care provider behaves badly because that is her/his personality, and when you ask her/him something she/he responds to you in a bad way, but you don’t see many people like that. Now many are really trying … It is now rare to see a health provider who insults someone.” (female client)

Health providers similarly described the importance of positive, kind, and friendly interactions with women and men attending RMNH services as a critical component of quality care. Several providers emphasized the need for a “warm” welcome, to make clients comfortable and reduce their fears or worries. When discussing female clients, providers emphasized the importance of developing trust, by listening and providing explanations, as highlighted by this provider: “When they have understood well the explanations, in their hearts they start taking decisions based on what you explained.” Providers also acknowledged that a negative interaction with a provider or receiving poor-quality care could deter a client from future care-seeking. RMNH providers described how women’s personal experiences of RMNH services in particular informed their own future care-seeking as well as that of others, as shared by one provider: “If … we beat her, we disrespect her, she will not come back … even those who want to come, she discourages them.”

On the whole, both female and male clients showed support for men’s engagement in RMNH services, and described the added value of men’s participation, particularly at ANC. Female and male clients described men’s presence at the first ANC visit (which is strongly encouraged in Rwanda) as an opportunity for men to learn how to support a healthy pregnancy, and for couples to jointly prepare for the birth and decide on post-partum contraceptive use. Female clients described the importance of men learning this information directly from the health provider, rather than from the woman herself, as they felt men were more likely to trust and therefore follow the recommendations given. Some female clients also suggested that health providers treat women attending ANC with a partner with greater respect than women attending alone.

However, several providers admitted to using unsanctioned practices to encourage men’s antenatal participation – such as seeing couples faster than single/unaccompanied women seeking ANC, being overly kind to male partners, or publicly praising men for their participation. These practices were informed by providers’ beliefs that men needed to be seen faster because they are impatient, do not like to spend long waiting, or have other responsibilities they need to attend to. They also described these tactics as encouraging men’s (and their peers’) future ANC participation. In addition to these practices, one provider suggested that there may be bias against women attending the first ANC visit without a partner, which could undermine the quality of care:

“Sometimes you may not give her the services the same way you gave it to others because you were in a good mood with those couples, interacting with both the husband and the wife. So, you might give her the quality of service which is not as good as the one you gave others. We should admit that it can happen.” (RMNH provider)

When discussing childbirth, female clients appreciated providers who spoke kindly to them, made time for them and guided them throughout labor and childbirth. They noted the importance of providers explaining the care being given and emphasized that women need more interaction with providers during this time. Several female clients suggested that first-time mothers in particular required more attention, care and guidance; while multiparous women had a better understanding of what to expect, and therefore required less attention. A number of female clients noted the need for providers to show more respect and empathy to women during childbirth.

Female clients noted that verbal and physical abuse – such as slapping or hitting a woman during labor or childbirth – happened less frequently than in the past, when it used to be common. However, they acknowledged that it sometimes still occurs, particularly if a woman is seen to be misbehaving. Several female clients suggested that verbal or physical abuse from a provider was most likely to occur when the baby is crowning and a woman closes her legs instead of pushing. One female client described her own experience of how the health provider “shouts at you with an angry face so that you can push the baby”. Several female clients suggested that providers may slap or hit a woman to correct her misbehavior, particularly if it threatens the life of the baby:

“Sometimes she can behave badly and the health provider realizes that it can make the baby’s life in danger, so the health provider can slap her with the intention to bring her back to the right track in order to save the baby’s life.” (female client)

Several female clients suggested that women will forget about the abuse once they see their healthy baby. They also reported that providers sometimes apologize to a woman if they spoke harshly or slapped her during the delivery. One female client suggested that a woman will actually thank the provider “instead of accusing him/her of anything”. However, another female client stated that while an apology from the provider “reduces your trauma”, the experience is not easily forgotten.

Most providers also agreed that verbal and physical abuse during labor and childbirth were less common than in the past when some described it as “normal”. However, they also acknowledged that disrespectful or abusive interactions may still occur. Most providers suggested that experiences of verbal or physical abuse were linked to women’s perceived misbehavior – such as screaming loudly or not pushing at the right moment. Providers suggested that providers may become frustrated and react unkindly to women who are perceived as misbehaving, particularly if the woman’s behavior is seen as a threat to the child’s life. Most providers agreed that such abuse was not acceptable, but felt that when it was done, it was often done out of good intentions. As described by one provider:

“Most of the times it is done depending on the way a woman behaved. A health provider doesn’t beat her out of unkindness, she/he beats her to save the baby. Sometimes a health provider does it when he/she sees that if she does another step, it will be over for the baby! It doesn’t happen often, but it happens.” (RMNH provider)

Providers suggested that maternity services are more likely to include high stress situations that can lead to poor interactions between women and providers, compared to other health services. One provider noted that although midwives sometimes “punish” women, “a woman in labor can stress you more than another woman”. Providers highlighted several factors that may contribute to disrespect and abuse, such as exhaustion and stress (due to the limited number of providers compared to clients), poor stress management, personality, and the fear of potential consequences (such as imprisonment) for providers if negative birth outcomes (e.g. death of mother or child) are believed to be due to their negligence:

“It depends on the personality of the health provider. There are some who get stressed...So, one is stressed because the baby can die, and the stress of what will occur after the baby’s death and the imprisonment, so because of all that, you find yourself having beaten her already without even being aware of it.” (RMNH provider)

Several providers suggested that providers have less tolerance for misbehavior by multiparous women – especially those in their third or fourth pregnancy – whom they expect to know better. However, disrespectful interactions were also described for first-time mothers:

“It is not often, but it happens. I remember when I started working, there is someone who told a woman – you see when it is the woman’s first time to deliver, when you tell her to open up the legs, she tends to close them. So, she/he told her, ‘Why didn’t you close your legs when he was impregnating you?’ I am sure that woman didn’t forget that.” (RMNH provider)

A number of female and male clients expressed support for men accompanying their partners to labor and/or childbirth. One man suggested that a man will accompany his wife to, “know what is going on … and because you love her”. Overall, female and male clients reported that women were treated with greater respect by providers when accompanied by their male partner. One female client noted that a woman is received differently – better – by health providers when her partner accompanies her. One male client suggested that a man accompanies his wife in order “to make her respected”. However, several female and male clients reported that men are not always allowed to be present during labor or birth, which one female client said can cause anxiety. Several clients suggested that men’s ability to be present depends on the attitudes of the providers on duty.

Providers were generally supportive of men’s accompaniment during labor and childbirth and acknowledged men as a source of affection, comfort, encouragement, and support for women. The providers agreed that a woman is allowed to have a birth companion present. One provider stated that a woman’s husband has priority, but emphasized the need for a woman’s consent before allowing her partner into the delivery room. The provider also acknowledged that some women do not want their partner present for the birth. However, several providers acknowledged that men are not always allowed to be present, citing the limited number of private rooms in some facilities (where several women are together in one room) as a barrier.

Several providers suggested that when men are present, they are not always capable to handle the experience. They shared how some men are uncomfortable with their partners being physically examined by a provider or become distressed when witnessing a partner’s pain. A few providers said they have had to ask a man to leave the delivery room, but one provider suggested that you cannot ask a man to leave because it implies that he was not welcome in the first place. Others noted that it was usually men themselves who asked to leave. Providers also noted that men’s lack of experience with childbirth could lead to misunderstandings or accusations of poor-quality care, as described in this example from one provider:

“There was a man who decided to go out and called my supervisor. He told him, ‘I don’t understand what the health provider is doing.’ So, my supervisor came to help me, but he found that I am done with helping her to deliver. He told me, ‘There is a man who told me to tell you that he can see that his wife is about to die.’ Then I told him that there was no problem, that the woman was screaming, but that it was normal. She didn’t have any other problem, it is just that the baby was right at the exit.” (RMNH provider)

Several health providers suggested that women whose partners were present at labor or birth were less likely to receive disrespectful or abusive care. Providers said they sometimes request a man’s help to calm his partner, particularly if she is perceived to be “misbehaving”. One provider suggested that, “allowing the husband to be present in the delivery room, sometimes it is not because of reasons of comfort, sometimes it is because in their mind they are like, ‘please help us with your wife.’” However, some providers also suggested that in such a situation, a woman’s partner may also be a source of abuse in the delivery room:

“Even if they said that when a woman is accompanied she is not beaten, sometimes her husband also beats her! Her husband can be like, ‘You stupid, do you want us to lose everything in vain?’” (RMNH provider)