In this qualitative sub study, part of a larger assessment of segmentation counseling in Niger, the majority of health care providers interviewed found the client segmentation process to be of benefit to women seeking family planning. When health care providers use the segmentation tool in their health centers they have more time to discuss family planning with their clients and to ask them pertinent and targeted questions about their family planning needs. The in-depth counseling that accompanies segmentation allows women to better understand the different methods available and increases their confidence that they, and other women in their community, should use family planning. Such in-depth counseling is vital to family planning service provision.17–20 Although the implementation of the tool had some identified bottlenecks, such as stock outs of the segmentation sheet, inadequate training of providers, and translation issues, all providers agreed that the tool greatly improves the quality of family planning services. These findings are timely for the Government of Niger as they consider supporting scale up of the segmentation strategy nationwide. Central to this scale up process, and all scale up processes, is ensuring quality of services and this is where the segmentation tool can be of great use.21–23 If the implementation issues can be systematically addressed during scale up of this approach, the impact on family planning services in Niger could be manifold.
It is clear from the responses that the segmentation trainings, whether the 5-day training or the briefings, gave health providers the basic information needed to implement segmentation-based family planning services and should be scaled up. The quote shown from the provider who was briefed in segmentation shows a detailed understanding of how the segmentation tool is meant to be used, including exactly how the final segment is determined. Other training topics, such as best practices for counseling, family planning technologies and tool translation, were much less commonly reported, especially by providers who were briefed. This indicates the main drawback of only briefing providers on segmentation; there is a lost opportunity for a more in-depth and nuanced understanding about the rationale for segmentation and for deeper engagement on other relevant family planning updates.
One way to support CSIs in their segmentation activities is to assure adequate supervision after training. In our study 69% of the CSIs had received a supervision visit after training and this proportion could be increased to 100%. The responses from those interviewed clearly indicate that the supervision visit was in fact an important opportunity for further coaching and support on segmentation implementation and not just a visit for stock checking and data collection. The example given of a provider who was given assistance in interpreting tool questions during a supervision visit was a typical experience.
The process of implementation, the changes in health service delivery and the impact of segmentation on family planning clients were the most noteworthy findings from our study. The process of implementation analyses showed very few changes to the typical process of providing services to family planning clients, which means that scaling up this strategy will mean just small changes to CSI service delivery (e.g., keeping the segmentation sheet in a client’s health care file folder). We found that all providers segment new clients and some segment returning clients if they have not been segmented previously, and that clients of all types and ages are segmented. However, this information was only qualitatively determined. Future studies of segmentation may want to include quantifiable data on the number and types of women segmented and changes in segmentation rates longitudinally from the time of introduction and onward. This could help to understand the process of implementation more fully than what is available from qualitative data.
When asked about how the segmentation process influenced family planning service delivery, the providers offered rich information about how they felt it changed their work and how they believed it impacted the clients themselves. Almost all providers reported that segmentation increased the amount of time they spent in consultation with a family planning client, allowing them to better explain the different methods and to better understand the client’s background and needs. As noted, this aspect of counseling is an essential component of quality family planning services.17–20 It may be that the process of segmentation, and the questions included in the tool, are less important than the extra time and attention paid to clients as a result of segmentation. This question would need to be explored with more rigorous research but clearly, regardless of the mechanism, we found evidence that the segmentation process results in more individualized family planning services during which women learn much more about family planning methods and why family planning is important than during the pre-segmentation counseling session. We also document provider beliefs that segmentation attracts more family planning clients and that segmentation ultimately decreases family planning rumors. These two findings would need more data and research to be fully verified, but it is likely that family planning demand could grow under segmentation. Notably, due to rapid population growth, the number of women is increasing very quickly and coupled with higher demand, it is possible that a capacity problem may arise in the early years of scaling up segmentation.
As noted above, the process of implementation was not without difficulties. The task of translating the segmentation tool questions into local languages was a main challenge for providers, a process that often started during the training but continued well into the introduction of the tool at the CSIs. When providers started using the questions that had been translated into a local language, often the clients did not understand the translated question and the provider revised the translations yet again. Given the significant challenges with translation, it may be best for implementers to pilot and finalize translated tools in all local languages well before starting the training of providers, including piloting the tools with clients who speak the language.
Another difficulty reported by providers were stock outs. Stock outs of the segmentation sheet were common and resulted in temporary halts of segmentation activities at the CSIs, since the sheets list the questions and are essential to identification of the client’s segment. Preventing stock outs of the segmentation sheets, and less commonly the counseling cards, should be prevented as much as possible during implementation in order to ensure continuation of the program. There may be innovative ways to assist providers in accessing the tool and avoiding stock outs, such as including the tool as part of the Ministry of Health’s vital commodities for health centers. The other important challenge was the need for comprehensive training of providers that has already been noted. Providers also encouraged that more providers are trained at facilities and not just one per facility; this will ensure that all clients are able to be segmented and not just those who come on the day a trained provider is present.
The characteristics of the interviewers may have influenced the results, since they were not gender- or age-matched with interviewees. The interviewees may also have believed that the interviews were part of a performance evaluation and changed their responses to be more positive, although it was made clear during the informed consent process that this was not a consideration. This study was based on in-depth interviews with 16 health care providers. Although the 16 providers varied considerably in gender, age, position at the CSI, and other characteristics, the study could have been improved with a larger number of participants, even though when reviewing the codes we did feel saturation was reached after 6-8 interviews.24 The number of interviews may have especially influenced themes where only a subset of the 16 participants provided responses (e.g., supervision). The inter-rater reliability (IRR) scores in the study showed substantial agreement but there still could have been coding errors occurring that may have influenced the results of this study.16 This is particularly true for the more complex and numerous child codes used for finer content analysis of segmentation themes, where the Kappa score from IRR testing was 0.69 on average. We worked to reduce such errors as much as possible through numerous discussions of coding applications between the master coder and the two other segmentation child code coders. The results of such discussions included changes to the code definitions, recoding of applied codes and additions, reductions or merging of codes. Another limitation in the study is the fact that 46% of the providers, among those who reported the type of training, had received a briefing instead of a full 5-day training in segmentation. Although it is evident that these providers learned how to conduct segmentation “on the job,” the lack of formal training in segmentation may have influenced the depth and sophistication of the responses to our questions about segmentation implementation. However, this may reflect more real-world implementation of the program.