Surveys.

Women ≤24 weeks gestational age (GA) attending their first ANC visit were screened for eligibility, consented, and enrolled into the study. Additional eligibility criteria included minimum age of 15 years; ability to provide locator information; and plans to stay in the area for the following year. Surveys were administered to all enrolled women in the health facility at enrollment and in participant’s homes postpartum (3–6 weeks and one year). Three ANC providers per intervention site were chosen by site administration staff and invited to be trained as G-ANC facilitators. Participating providers completed surveys before and after completing all ANC meetings for study cohorts (i.e., minimum of four). All survey data were directly entered into REDCap^™ on tablets by trained research assistants.

Focus groups and interviews.

Focus group discussions (FGDs) and in-depth semi-structured interviews were conducted with G-ANC providers and a sub-set of recently delivered women enrolled in the intervention arm after cohorts finished their last G-ANC meeting. A purposeful sample of women who had participated in one of the last two cohorts and attended at least one meeting were invited to participate in focus groups specific to primiparous women; multiparous women; or adolescents. Further sampling strategies were employed to ensure representation from each level of health facility, urban and rural locations, and various ethnic and linguistic groups. A total of eighteen focus groups were planned to meet this criteria with ten to fifteen women invited to each. All G-ANC providers were invited to one of four provider-specific FGDs. Interviews were planned for deeper exploration of specific topics, without a priori sampling targets (See Table 1 for details). Separate but similar facilitation guides were used across FGDs and interviews which probed the experience of providing and receiving G-ANC, perceived effects, and recommendations for future implementation (S3 Appendix). All participarts were reimbursed for transportation, and no other incentive was provided. Experienced qualitative researchers who spoke local languages and were not otherwise involved in study implementation were trained and initially supervised by author PW to conduct FGDs and interviews using semi-structured guides alongside study research assistants trained to act as notetakers during FGD sessions. Debriefs were conducted at the end of each day to draft cover sheets summarizing key themes for each session. PW also participated in FGDs, and interviews conducted in English or Kiswahili. All FGDs and interviews were recorded and later transcribed and translated by native speakers present at the time of recording.

Qualitative analysis.

We adopted a multi-phased iterative approach to framework analysis to develop an empirically based theory of change for G-ANC. Qualitative data were managed and analyzed using both ATLAS.ti and Excel software. Phase 1) Familiarization: Initial coding was completed by BO, EO, and LC utilizing broad descriptive codes generated by review of three transcripts by each researcher and discussed in a four way conversation with PW. The full study team then identified the subset of descriptive codes related to experience and effects of G-ANC. Phase 2) Theme identification: Examining the subset of data within those codes, LG then used semi-directed content analysis informed by field observations, published literature, and discussions within the Global G-ANC Collaboration, to generate an expansive set of codes potentially related to a TOC based on open, in vivo coding. 3) Indexing: Codes were reviewed and refined with the larger research team before returning to the raw transcript data to recode using the new schema. 4) Charting: The resulting subset of data was organized within a matrix to visualize the relevance of codes across countries, providers, and recently delivered women. Codes were grouped into larger categories at this stage. 5) Interpretation/Mapping: diagram drawing was used to further refine and collapse subthemes within broader categories and place relational concepts within a TOC.

Quantitative analysis.

Next, as a validity check, the quantitative data were reviewed for content relevant to the individual concepts included in the TOC, and findings were compared (Table 1). We applied generalized estimating equations with generalized linear models clustering at the facility level to assess women’s perceptions of quality, satisfaction, and ANC model preference; health knowledge and uptake of health promoting behaviors; and communication with providers outside scheduled ANC by study arm. Bivariate difference in difference analysis was employed to assess comparative changes from enrollment to 3–6 weeks postpartum between study arms for women’s confidence in communication, skillful decision making, and locus of control related to maternal, newborn, and reproductive health. Descriptive statistics were performed for provider data (due to small sample size) and data collected only in the intervention arm (due to lack of comparator). Analyses were completed in R statistical software (R Core Team, 2017) and Stata15.

Experience of G-ANC: Forming supportive relationships and open communication.

Supportive relationships among and between women and providers in G-ANC emerged as highly valued by both parties and a defining characteristic distinguishing G-ANC from individual ANC. These relationships, in turn, provided the scaffolding for improved communication.

Women reported feeling supported, loved, cared for, respected, and unjudged in G-ANC. As one woman described, “they really cared and were patient with me…they were making way for me…I felt I was in a home, I was loved” (Woman, NG). In both countries, women used the term “one-ness” to define the sense of community they felt within their group, where “we love each other” (Woman, KE) and “live as friends, as if we come from one family” (Woman, KE). As one woman explained:

The service providers were approachable, if you have a problem, they could be free to help you, they never judged anybody, they could just be with us as fellow mothers. They were free to laugh with us…they could dance with us, laugh with us.

(Woman, KE)

In contrast, women expressed that individual ANC was a place where “there is no one-ness like this one” (Woman, NG), “you feel lonely” (Woman, KE), and fear “they [providers] will not keep your secrets” (Woman, NG).

The mutual trust created by this experience of “oneness” provided the foundation for “being free” within the group. Women described having the ability and courage to ask questions, share problems and ideas, and make mistakes:

Being in the group we had time to interact with [service providers]…You understand they are not bad people, when you tell them your problems they will care, so as we understood them, we opened up. Because when in a group there is that freedom…[In individual ANC]…you look at [the service provider], you fear even the looks, how will I tell her my problem? But when in a group you understand each other, you laugh together, you sit together, you are free.

(Woman, KE)

Provider’s echo these feelings, taking enjoyment from the closeness among women, and their own improved relationships and communication with their clients.

With group ANC I have seen a very improved relationship, client health worker relationship. As interactive as it is we have been able to bond. You know with the individual one is like we acted the boss…but now we interact at the same level, we bond with each other, we talk on phone, they consult anytime of the day, and by the end of the day we find some kind of job satisfaction as health workers

(Provider, KE)

What I enjoyed most is the relationship between the care giver and the client. At some point you will get the client calling you by name, she forgets that you are the sister then she starts telling you “[provider’s first name], you know this this this and that” and then you will also know them by name

(Provider, KE).

Women and providers linked the interactive nature of G-ANC meetings, decreased hierarchy, community forming activities, collective norm-setting (e.g., for confidentiality and respect of all opinions), continuity of care, and increased touch time to the development of supportive relationships, mutual trust, and open communication.

Experience of G-ANC: Becoming empowered partners in learning and care.

Providers and women describe a shift from viewing the provider as the sole source of information and care in individual ANC to viewing these as shared responsibilities in G-ANC. This new partnership was characterized by a shared workload (e.g., self-assessments); collective problem solving where, “everyone is embracing each and everyone’s problem” (Provider, KE); and an understanding that everyone is a teacher, and everyone is a learner. Both parties valued this new partnership and voiced advantages for themselves and others.

Women felt pride in their ability to assess themselves and each other, highlighting “The difference is, the things they will do for you in individual ANC, in group ANC we will do it for ourselves” (Woman, NG). Providers note how this contributes to increased partnership in their own care. As one gives by way of example, “When they discover that their weight dropped, they themselves, they will notice and they will call our attention “ah, ah, Aunty Nurse, why is it that my weight dropped today?” (Provider, NG).

Multidirectional collective problem-solving and learning was described—between women, between women and providers, and between providers:

We would share experiences. Someone would say that which she has undergone, another person also says, and then we see how we could help. We were working together. Everything we were doing we were coordinating. For example, like measuring weight, [or] teaching each other…Like, there was a day someone asked a question concerning whether someone can have sex during pregnancy. Someone said “no”, others, “It is normal you can just have sex”, Another said, “No you shouldn’t have sex until after delivery.” Then we later found a solution

(Woman, KE).

Providers found this collective learning and problem solving not only beneficial for the women, but for themselves as well:

The clients are empowered…in normal ANC we assume that the clients do not know. But we realize that they know so much more than even what we know, so when we empower them and allow them to give us their ideas first then talk about ours, they get so much help—both the clients and ourselves

(Provider, KE)

Self-assessments, actively seeking knowledge and ideas from women (based on the structured discussion methodology), and components which decrease hierarchy (such as sitting in a circle) were viewed as aspects of G-ANC which contributed to these new partnerships. Women and providers also described a feedback loop between feeling like partners in care and their improved relationships with each other.

Experience of G-ANC: Receiving and providing meaningful clinical services and information.

The experience of G-ANC was most succinctly summarized by women and providers as they contrasted it to individual ANC, which both parties characterized as rushed, incomplete, impersonal care focused on “clearing the queue” (aka “finishing” the waiting line of women). In contrast, G-ANC was credited with providing a supportive environment capable of responding to women’s needs.

Women cited three key advantages of G-ANC compared to the “clearing the queue” model of individual ANC: 1) being asked about their problems and given the opportunity to ask questions; 2) having their health status and/or treatment explained to them; and 3) being taught how to care for their and their baby’s health.

There is a difference between this and the general antenatal, because here we discuss in detail, we ask ourselves questions, we are given manuals and we discuss as one family, but it is not so with the general one because people are too much, you are not opportune to ask questions, also we don’t discuss in detail.

(Woman, NG)

[In response to what care they would recommend to a friend] I would like to tell her to come to the group where there are many teachings, but in the other one they just examine you and you are not taught. Even if something is happening to you, you don’t know

(Woman, KE)

Providers highlighted how G-ANC helped them to “get out of poor-quality clearing and forwarding routine ANC” (Provider, KE) and “provide holistic care to mothers” (Provider, KE), in part through feeling more invested in the care they were providing:

[In individual ANC] you want to finish your work; you just do it quickly. You finish your work and forget about them. But here [G-ANC] you are—you want to finish, you want to talk to them, you want to assess them, you want to interact with them, so it is a different perception.

(Provider, KE)

Providers outlined a number of specific ways G-ANC supported provision of meaningful care: scheduled meetings allowed anticipatory staffing and relieved provider pressure to “clear the queue”; the structured materials, methodology, and increased touch time allowed them to cover topics deeply while responding to women’s interests and concerns; and strengthened provider-patient relationships meant women were more likely to share specific concerns and providers were more motivated to respond with individualized care.

Effect of G-ANC experience: Self-reinforcing cycles of and more and better care.

Qualitative data indicates that the experience of care in G-ANC contributed to mutual satisfaction and a self-re-enforcing cycle of more and better care. This cycle was characterized by higher rates of satisfaction, preference for the G-ANC model in future care, and perceptions of improved quality of care, utilization of care and follow-up. Quantitative data support these qualitative findings.

Women credited both the value and the enjoyment they derived from G-ANC with an increased desire to return for additional ANC. As a woman who was initially skeptical of G-ANC explained:

When I came for the first time, I really enjoyed it. In fact, I felt as if we should not close for the day. When I went back home that day, I told my husband that I don’t mind we should be going for group ante-natal every day…[it] is very special. There you meet new people and learn a lot from them too.

(Woman, NG)

Women further highlighted how their experience of care improved their ability to access appropriate care. For example, one woman described how in past pregnancies she had been too scared to explain her problems to her provider and had suffered multiple losses. After gaining an understanding of the importance of her symptoms (through collective learning) and having the comfort and courage to discuss with her provider (through forming supportive relationships, and becoming an empowered partner in care), she was able to get the additional care she needed to prevent another loss.

Women in G-ANC were more likely than women in individual ANC to agree with statements of quality care and to report being “very satisfied” with the ANC received. Over 95% would prefer it over individual care in a subsequent pregnancy (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. G-ANC effect on client perceptions of quality, satisfaction, and ANC model preference.

https://doi.org/10.1371/journal.pone.0265174.t004

Providers universally felt they were able to provide higher quality care through G-ANC, highlighting their improved ability and motivation to 1) provide in-depth consistent health education; 2)identify and manage complications; and 3) provide “consistent” and “comprehensive” follow-up. They cited the structured meeting materials and decreased need for repetition, improved relationships, open communication with women, continuity of care, and G-ANC cohort specific registers as key drivers:

In group ANC the staffs are focused, the clients are focused. So, we all know what to expect for every meeting. To me it saves time, and you give quality, just like [R8, R3, R2] have said. You give quality because, ahhh, if I start talking about what to expect—nutrition, HIV, danger signs—to individual clients, and I want to see 30, by the time I see the sixth I am tired. So, the rest will just… I will just be clearing them

(Provider, KE)

Early identification of conditions and managing them promptly—we’ve realized this can be achieved so well through group ANC. In individual ANC when I see a mother who has come with an STI I tell them to go and get a drug, they do not come back, I do not even realize that they didn’t come, unlike group ANC which now I have these mothers at my fingertips.

(Provider, KE)

In group ANC you will be able to know your target number and if they are not there you call. But in other ANC, whoever is present, you don’t even know who doesn’t come or who comes…With this group ANC, it motivates us as a duty to do, but in other ANC, nobody cares whether they come or they don’t come, you don’t really bother to locate them.

(Provider, NG)

As compared to individual care, more providers reported feeling they were able to meet their own goals of quality care with G-ANC (Table 5). The number of providers who reported their ANC related job satisfaction as “extremely satisfied” doubled in both countries, and none chose individual care over G-ANC when asked their preferred model (Table 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. G-ANC effect on provider perceptions of quality, satisfaction, and ANC model preference.

https://doi.org/10.1371/journal.pone.0265174.t005

Effect of G-ANC experience: Linked improvements in health knowledge, confidence, and healthy behaviors.

Women placed great value in the exchange of relevant, actionable information to support healthy pregnancies and the transition to motherhood. They expressed feeling supported and empowered through G-ANC to access information, make decisions for themselves, and take action:

I feel good because now I can assist another person, and if I am pregnant, I am able to take action. That is a very great achievement and I respect my service provider for empowering me

(Woman, KE)

Discussions related to health knowledge and behaviors were often anchored in concrete examples of how women translated knowledge to action in the service of themselves, their children, and their greater communities. Most stories related to 1) improved recognition and response to danger signs; 2) better understanding of and ability to enact health-promoting behaviors; or 3) the ability to advocate for behavior change and care-seeking (see S2 Table for examples of each within and beyond women’s households). Many women reported becoming teachers and sources of health advice in their communities, from which they derived pride, “It is something that one is proud to have…with that confidence you can help the society” (Woman, KE).

Providers likewise perceived and valued a shift in understanding and engagement among women in G-ANC as compared to individual care:

What I enjoyed most about the group is the way the women are understanding what they are being taught and are expected to do…they are much committed to what they should do for themselves and are more empowered in participating towards their care

(Provider, NG)

Women and providers explicitly linked the changes in health knowledge and behaviors to G-ANC’s structure, content, and experience. They emphasized the importance of women having the ability to ‘be free’ and providers the time to ‘go deep’ alongside the meeting activities which actively encouraged women to learn from each other. Women also highlighted the role of provider attitudes in their ability to learn and trust what providers said: “they were friendly, polite and caring and as such we understand what they teach us easily”(Woman, NG). Take home booklets provided to G-ANC participants were frequently credited with easing the task of sharing information and changing the perceptions of family and community members.

While qualitative findings consistently linked participation in G-ANC to increased confidence with health-related communication, knowledge, and decision making, quantitative analyses show more variability across these areas (Tables 6 and 7). Quantitative findings align with qualitative findings in indicating that women in G-ANC were more likely than those receiving individual ANC to report uptake of health promoting behaviors. However, the strength of this relationship was inconsistent across behaviors and countries (Table 6).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Confidence in relation to communication, skillful decision making, and locus of control related to maternal, newborn, and reproductive health.

https://doi.org/10.1371/journal.pone.0265174.t006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 7. G -ANC effect on health knowledge and uptake of health promoting behaviors.

https://doi.org/10.1371/journal.pone.0265174.t007

Effect of G-ANC experience: Improved communication, support, and care beyond G-ANC meetings.

Women reported communicating with peers and providers and gaining access to additional support outside of G-ANC meetings. Support provided by both peers and providers included advice and emotional support; occasional material support (e.g., food, transport money, pampers); and home visits (e.g., after missing a meeting, having a complication, or delivering). In addition, providers facilitated referrals (particularly helping women find openings for facility delivery during strikes in Kenya) and women experienced prompt respectful clinical care from them outside of meetings. Providers encouraged and enjoyed this additional communication and care.

As one woman explained, “We had unity. When one is in trouble you call a friend, who is a group member” (Woman, KE). Women made a point to note that these were new sources of support, explicitly due to the community formed during G-ANC:

In the group, we were so united, we loved ourselves, we ask after each other, we communicate with each other on phone, we visit each other. In fact, even after delivery we care for each other and the babies. So, they united us like a family even with those we don’t know before. We share problems like a family

(Woman, NG)

Sharing of personal mobile numbers by providers was not explicitly encouraged through the study but spontaneously arose. As one woman described, “this time around you don’t need to beg, they will be the one trying to pass their number across to you!” (Woman, NG). Providers explained that due to their attachment with women in G-ANC, they were anxious to help and wanted to hear updates. They were also confident that women in G-ANC could differentiate danger signs and would call with more serious concerns than women in individual care.

Quantitative data support these qualitative findings. Eighty-five percent of women communicated with each other outside of group meetings, with a similar percentage believing women from their group would reach out to other members if they had a problem or needed help (S3 Table for topics discussed). Four times as many women in G-ANC called or texted their provider over the course of the study compared to those in individual care (Table 8).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 8. Communication with providers and other group members outside of ANC.

https://doi.org/10.1371/journal.pone.0265174.t008

In contrast with their prior expectations or experiences, women also highlighted attentive, respectful care at their ANC facilities outside of G-ANC meetings. They described prompt attendance by staff if they arrived with a problem and encouraging, respectful care during labor and delivery. This improved care outside of routine ANC was repeatedly cited as a reason they believed other women should attend G-ANC. Providers likewise recognized a shift in the relationship women had with the facilities after attending G-ANC:

There is that, what do I say—eeh, continuum of care. Women came to love the health facilities. They bond with the health care providers, even when you are walking outside somebody says hi, so that fear that when you go to the hospital you will be mistreated or you might be slapped while delivering, no one experiences that.

(Provider, KE)

Effect of G-ANC experience: Motivation to continue providing G-ANC.

Providers expressed motivation to continue providing G-ANC as a result of reduced stress and fatigue; belief it better served women and children; and increased personal and professional fulfillment. As one provider replied when asked what she didn’t like about G-ANC:

I like everything about it. (It doesn’t take your time?) No, it doesn’t take my time, and it makes, it brings closer relationship between me and the women. Whenever I went out, like when I go to the market, the women will be seeing you, calling you, this is aunty, aunty that use to teach us in group antenatal. They respect me so much. I feel very happy for that. It gives you a sense of, like you are really helping, as if you’re making a difference. That is why I’m very very strong on doing [G-ANC]. I don’t feel tired when I have group antenatal.

(Provider, NG)

Providers not only appreciated the empowering effect the self-assessments and facilitated discussions had on women, but they also viewed them as helpful in decreasing their own workload. In addition, they reported the structured content of G-ANC reduced their stress and fatigue by simultaneously decreasing the need for repetition and increasing organization and their confidence that all essential topics were being covered:

I am not exhausted, and the women are also not exhausted, and the work is done and at the end of the day I have achieved what I wanted because there is meeting one, there is meeting two, there is meeting three, like that.

(Provider, KE)

About the competence, you don’t collapse because of burn out….you. just go step to step and you handle everything because you have that time…and when you handle this mother, now you be friendly, and now they can feel you and you can feel them

(Provider, KE)

Providers linked their increased personal and professional fulfillment to their ability to provide more and better care while forming strong relationships with women in G-ANC:

I think group ANC has changed my viewing of my work, because in individual ANC you usually see that mother and you do not know that mother…But with group ANC they come for subsequent meetings, you know the mother, even after the mother gives birth you are very happy to know you were attending this mother all through and then she delivered well. She has a healthy baby, so you feel happier with that work—because of you—you have seen this mother all through from ANC until delivery, and the baby is healthy

(Provider, KE).

Future research.

Future research can explicitly test this TOC to provide a better understanding of the proportional contributions of individual aspects of G-ANC to improved experience and outcomes. We encourage others to retain G-ANC components aimed at skill practice, individual action planning, and personal/social accountability in future G-ANC models and to apply more intentional evaluation to clarify their potential value. We also encourage future research to utilize qualitative data in refining quantitative measurements of empowerment as it specifically relates to self-care and health system interactions during pregnancy. Additional research examining differences between G-ANC and individual-ANC in costing, clinical outcomes, long-term effects on provider experience, and broader health system impacts (including equity) will aid policymakers in prioritizing efforts to improve ANC.

Implementation.

We encourage those planning G-ANC implementation to give strong consideration to the system supports necessary for success. G-ANC requires a substantial shift in workflow and new skill development for providers. Lazar et al. found providers repeatedly acknowledged anxiety about facilitation skills, and that confidence in these skills went hand in hand with experiencing their groups with satisfaction. In addition, research has shown that model fidelity (inclusive of faciliatory approach) is associated with improved outcomes [44], warranting development of strong mentoring systems and ongoing fidelity checks.

Application of findings to individual ANC.

While our research is specific to the effects of a group experience, most of the TOC content is not mutually exclusive to individual care. Relationships between provision, experience, and outcomes, may hold across models and are worth exploring. For example, strategies to decrease hierarchy, or perform basic self-assessments (such as the graphical assessment for danger signs) could be applied to individual ANC with relatively minor modifications. “Health talks” commonly given in ANC waiting areas could take lessons learned from G-ANC to effectively integrate more facilitative learning. Likewise, different changes to the provision of individual care could be explored with the objective of creating supportive relationships or partnerships in learning and care.