Study design

This qualitative study included focus group discussions (FGDs) with caregivers and health workers and in-depth interviews (IDIs) with key informants [9, 10]. The qualitative component of the assessment is specifically designed to understand why opportunities for vaccination are being missed and what can be adjusted or done differently to mitigate MOV, in the context of Kenya. However, interventions and recommendations are not drawn from only the qualitative data, but the triangulation of other data sources.

As part of the larger Kenya MOV assessment, the in-country assessment began with a training of field team members, followed by deployment to the field for four days to sites across the country to collect data. Multi-partner brainstorming sessions followed data collection to synthesize all the data (desk review and quantitative and qualitative data) and develop an intervention plan based on the data. The assessment process concluded with a stakeholder meeting at the central government level that aimed to build consensus and advocacy for implementing an endorsed and funded intervention plan to reduce MOV. Only the results of the qualitative component of the assessment are presented in this paper.

Field team recruitment and training

Field team members were staff from the Kenyan Ministry of Health (MoH) and various in-country immunization partners. Each field team consisted of two or three members who were trained in both quantitative and qualitative data collection over the 3 days preceding field deployment, and included at least one member with prior qualitative data collection experience who moderated the FGDs and conducted the IDIs while the other field team members took notes.

The qualitative team, comprised of one lead and two team members, had combined experience of all aspects of a qualitative study including study design, data collection, and analysis. The qualitative team led the qualitative training of the field teams. Training topics included the basics of qualitative research methods (with an emphasis on understanding the overarching research question), the purpose of conducting FGDs and IDIs for the MOV assessment, and notetaking during qualitative data collection. The qualitative team also discussed reflexivity to ensure continuous reflection by the qualitative data collectors on their role in data collection. Following in-class role playing, the MOV field teams reviewed and revised the FGD and IDI guides.

Data collection instruments

The qualitative data collection instruments included semi-structured FGD guides for the caregiver and health worker discussions and IDI guides for key informant interviews. The guides began with easy to answer opening questions related to the topic followed by key questions. The key questions were grouped into three sections: health and vaccination services in the community, attitudes toward vaccination and vaccination compliance, and MOV including reasons behind MOV and suggestions for reducing MOV. The guides ended with closing questions to summarize the discussion or interview. These guides were adapted for the Kenyan context from the generic guides provided in the WHO MOV methodology [10]. They were pre-tested for country-context and ease of understanding during the field staff training [9, 10]. All guides were available in English.

Participant selection

The MOV strategy team purposively selected 10 counties to represent various geographic regions and vaccination performance levels (based on coverage of the third dose of DTP-hepatitis B-Haemophilus influenzae type b or pentavalent vaccine) [24]. Ten counties were selected as this was the maximum number logistically feasible for this assessment given time and budget constraints, but also was thought to provide enough diversity to obtain an understanding of the situation of MOV in Kenya. The counties selected were Bungoma, Kajiado, Kiambu, Kitui, Migori, Mombasa, Nakuru, Taita Taveta, Trans Nzoia, and West Pokot. Within these counties, each MOV field team (one per county) selected one or two health facilities for qualitative data collection. Study participants for the FGDs were selected from similar health facilities as in the quantitative arm, based on size and location; facilities that were sampled for the quantitative arm were excluded from the qualitative arm to avoid contamination. For their respective FGDs, caregivers and health workers were selected from either the same or different health facilities. Because the days available for data collection were limited, ease of logistical access was considered in the final selection of health facilities.

For the caregiver FGDs, MOV field teams approached caregivers with children aged <24 months at the selected health facilities and asked if they were willing to participate in the qualitative arm. Caregivers were asked for their child’s age to assess eligibility. To encompass a wide range of experiences and reduce logistical constraints to participation, MOV field teams did not screen caregivers by reason for visit or if their child had a MOV. To ensure that caregivers were able to express their true opinions, the FGDs were conducted in quiet, comfortable locations beyond earshot of the health workers and other non-participants.

For the health worker FGDs, we purposively sampled health workers from diverse departments of the selected health facilities in order to capture a broad-range of perspectives. As MOV can occur at various types of visits and contact with a health facility, we ensured the inclusion of health workers who are not routinely involved in delivering vaccination services.

Key informants for IDIs were identified based on their perceived influence or insights on immunization services at different levels of the health system. Key informants included those responsible for the NVIP at the county and sub-county levels, as well as health facility directors, office managers, and administrators. Health facility-level participants for IDIs were selected from the same health facilities that were sampled for quantitative or qualitative data collection.

For the qualitative component of the MOV assessment, the overall goal was for 10 field teams to complete 10 FGDs with caregivers, 10 FGDs with health workers, and 40 IDIs with key informants. Based on experiences from previous MOV assessments in other countries, saturation was expected to occur with fewer FGDs and IDIs. However, because the 10 teams simultaneously conducted FGDs and IDIs across the 10 counties, it was impossible to review the data in real time. Each of the 10 field teams aimed to complete one FGD with caregivers, one FGD with health workers, and four IDIs (two with NVIP managers at the county or sub-county level and two with health-facility–level staff members). Qualitative data collection was complete for the team when they had achieved this goal. Individuals that participated in an IDI were excluded from FGD participation, and vice versa. In general, health worker FGD participants had direct roles in patient care while IDI participants had management, supervisory or administrative roles.

Data collection

Fieldwork was conducted over four days in November 2016 (November 4–5 and November 7–8). All FGD caregiver groups were same-sex groups, and field teams prioritized assigning moderators of the same sex. Moderators or interviewers began the discussion or interview by introducing themselves and the notetaker, describing the purpose of the FGD or IDI, establishing rapport with the participants or interviewees and explaining how the information would be used; in addition, they explained issues related to confidentiality and obtained informed consent from the participants. Moderators or interviewers then posed open-ended questions from the guide, starting with simple opening questions; they used probing techniques to elicit depth as needed. Prior to the FGD or IDI, no relationship was established between the moderator, interviewer, or notetaker and the participants or interviewee.

Notetakers captured key issues raised and emotional non-verbal aspects such as facial expressions and gestures in summary notes. Additionally, notetakers recorded verbatim quotes. Notes from the qualitative data collection sessions were typed up, labeled, organized, and stored on a secure website each evening. Because of privacy concerns, no audio recordings were made, and no personal information, such as name, job title, or position (for health workers and key informants), was recorded in the notes.

The FGD and IDI sessions lasted an average of 45 and 30 minutes, respectively. The sessions were conducted in the language the participants were most comfortable with. Interviewers and moderators obtained verbal consent from all participants prior to conducting FGDs or IDIs.

Data analysis

Inductive thematic data analysis was an iterative process with re-readings of the text to discern patterns and identify major and recurring themes derived from the data. Field team members consulted as a team following each FGD or IDI to compare notes and verify verbatim quotes before sending the notes to the qualitative lead. Immediately following fieldwork, the qualitative lead conducted preliminary analyses for the immediate post-fieldwork debriefing. As a part of this process, the qualitative lead discussed major themes with all field team members, explored contradictions between sites, and further investigated unexpected findings. These discussions provided the basis for an initial set of analytic codes. Following the assessment period, the qualitative team continued refining the preliminary list of codes, rooted in the data, and used a system of manual open coding organized using Microsoft Word tables. These codes were then grouped into emerging themes through an iterative and flexible process. The final themes were determined by consensus of the qualitative team.

Ethical approval

The Kenya MoH assessed the MOV assessment protocol and classified it as a program assessment. As such, it was exempt from further review by the Institutional Review Board. The study team obtained verbal consent from all participants prior to the FGD or IDI. The consent was recorded by the facilitator or interviewer on the guide. The verbal consent procedure was approved by the Ministry of Health Kenya.

Participant characteristics and key themes

The MOV field teams conducted 6 FGDs with caregivers, 8 FGDs with health workers, and 35 IDIs with key informants across the 10 counties that were included in the MOV assessment (Table 1). Approximately 50 caregivers and 55 health workers participated in the FGDs; each FGD comprised of 5–13 participants. Thirty-five key informants participated in IDIs. Due to time constraints, not all teams were able to meet their qualitative data collection goal. All the caregiver FGDs were exclusively with female caregivers. Health workers who participated in the health worker FGDs were from various backgrounds and departments at the health facility. Similarly, key informants ranged from those at the county level to those at the health facility level. There were no refusals and no participants dropped out during the discussions/interviews.

Download:

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

Table 1. Qualitative data collected during the Kenya missed opportunities for vaccination assessment, by county, November 2016.

https://doi.org/10.1371/journal.pone.0230783.t001

The key themes that emerged from the thematic analysis included perceptions of healthcare services and vaccination; vaccination checks and integration with other services; health worker staffing shortages; stock-outs of vaccines and vaccination-related materials; and health education. The findings from the IDIs and the caregiver and health worker FGDs are summarized by key themes in the following paragraphs. Quotes included are verbatim quotes recorded during the FGDs or IDIs.

Perceptions of healthcare services and vaccination

In general, participating caregivers expressed positive attitudes toward healthcare workers and health services. Caregivers believed that health workers were generally knowledgeable and efficient: “Nurses are good and they don’t waste time” (Caregiver, Kajiado). Health workers and key informants echoed a similar sentiment with their perceptions of caregiver satisfaction: “We give our best… no complaints from the parents (Others nod in agreement)” (Health worker, Bungoma).

Although caregivers were generally satisfied with healthcare services, many participants noted that there were areas in need of improvement: “I have a feeling that clients may not be very satisfied, arising from long waiting times” (Key informant, Mombasa). Additionally, geographic inaccessibility, restricted clinic hours, and limiting the provision of certain vaccines only to specific days of the week caused some dissatisfaction among caregivers: “We come to [name] dispensary for vaccination on Wednesdays only, because that is when vaccines are available” (Caregiver, Trans Nzoia). Beyond the limited vaccination days, some also reported restricted hours even on vaccination days: “Vaccination services are given in the morning only,” (Key informant, Kiambu) and “Even those that provide services daily, it is at the discretion of those in the unit to accept more clients after 12:00 pm” (Key informant, Mombasa).

Similar to their attitudes toward healthcare services, most caregivers had positive attitudes toward vaccines in general (no specific antigen was described) and to vaccination: “Immunization makes our children get better quickly from diseases that used to kill other children before” (Caregiver, Trans Nzoia). Caregivers were happy to bring their children for vaccination and reported that this was the general attitude among mothers in their communities: “Every mother brings their child for vaccination” (Caregiver, Bungoma).

Despite the generally positive attitudes, some rumors and misconceptions persisted among caregivers. Although many participating caregivers stated that they themselves did not believe the rumors, they had heard of them and knew other caregivers in their communities who did believe them. Caregivers reported feeling hesitant about multiple injections: “…since the introduction of a third injection [Inactivated Polio vaccine (IPV)], mothers fear that there are too many injections and thus do not bring the children on time…” (Caregiver, Bungoma). There were rumors about adverse events related to being given multiple vaccinations: “If you are given vaccines twice, for example, during the routine and during the campaigns, the children can get very sick or die” (Caregiver, Migori). There were also rumors about potential long-lasting, permanent health consequences related to vaccination and the quality of vaccines offered at health facilities: “Health facility gives bad injections—the vaccines themselves are bad, not the injection technique” (Caregiver, Kajiado).

Vaccination checks and integration with other services

Field teams reported a lack of consistency across health facilities on when vaccination checks should be completed and whose responsibility it was. Some health workers and key informants reported checking vaccination status only at vaccination visits, while others would check at inpatient and outpatient visits but not at visits for other services (e.g. eye, dental, nutrition services): “If the child is not here for immunization, they don’t ask, but if the child is here for immunization, they normally ask” (Key informant, Kitui). Some attributed this to the lack of integration of services within a health center, with certain departments siloed: “[Vaccination]…that is the work of the MCH [Maternal and Child Health] staff” (Key informant, Taita Taveta). At other facilities, checking of vaccination status was attributed to personal health worker initiative: “When I’m attending to the children in the OPD [out-patient department], it has always been my culture to check on the MCH booklet and I usually look at their immunization record…” (Key informant, Kitui). Overall, health workers and key informants acknowledged that health staff need to provide a holistic healthcare approach: “When telling clients, don’t just brush off [and] do things shallowly; I think you need to dig deep in terms of owning that patient and finding out everything about that patient, not just why they came to the facility” (Key informant, Kitui).

Health worker staffing shortages

Health workers and key informants discussed the inadequacy of staff resources at health facilities: “Staffing is a constant struggle” (Key informant, Kajiado). This sentiment was also echoed by participating frontline health workers: “Understaffing is really, really, really biting into services” (Health worker, Kitui). Participants explained that the lack of staff resources was not due to a lack of trained staff in Kenya, but rather either due to a lack of resources to hire staff or due to the politics behind staff distribution among health facilities within a county.

Health workers explained how the limited staff resources affected health workers’ ability to provide high quality services to patients: “Quality service is compromised because we don’t have enough health workers in our health facilities, and the sufferer is a very innocent person” (Key informant, Trans Nzoia). When health workers are overwhelmed, they are unable to take the necessary time with each patient and are only able to focus on the presenting complaint: “In facilities with only a single staff, focus is on the sick child. When work is too much, children for vaccination are more likely to be told to ‘come tomorrow because I am busy’” (Key informant, Mombasa). Caregivers were also dissatisfied with long wait times, which are exacerbated by staffing shortages: “In the public facilities, the children are many and thus [there is a] long queue with few health staff to attend to us” (Caregiver, Migori). This may even result in caregivers not visiting health facilities: “At a facility with only one staff conducting all the services, mothers find it hard to bear the long line. Some may not go at all” (Key informant, Bungoma).

Stock-outs of vaccines and vaccination-related supplies

There were also discussions about the lack of vaccines and required vaccination-related supplies. Health workers specifically discussed the stock-outs of bacille Calmette-Guerin (BCG) vaccine: “No children have been vaccinated [with BCG] since November 2015; even some children are now walking having not received BCG” (Health worker, Bungoma). Stock-outs of both vaccines and syringes were reported: “The main challenge is stock-outs of vaccines and SoloShot [syringes] for administering BCG vaccine. These stock-outs, especially for SoloShot [syringes] for BCG, have become worse since devolution” (Key informant, Bungoma). These stock-outs contributed to wide-ranging effects, including health workers’ inability to do their basic jobs: “The health workers are very willing to assist but they are handicapped by lack of other drugs and equipment” (Health worker, Trans Nzoia).

Stock-outs also affected caregivers’ willingness to visit a health facility: “The facilities receive few vials of BCG vaccine with fewer doses than required. This makes the nurse at the facility to inform mothers to come for injections at a specific day in the week… the mothers will bring their children on that day, but I think we can have cases where some don’t come back immediately” (Key informant, Trans Nzoia). Additionally, as caregivers return to their communities and report that certain supplies in the clinic were not available, other caregivers may be reluctant to return for fear of wasting their time and money or of not being able to receive the services they need for their children: “In addition, when there are stock-outs, parents spread the word in the community that there are no commodities (drugs, vaccines, syringes)…so they don’t come…” (Key informant, Bungoma).

Health workers also discussed shortages of other vaccine-related supplies, including mother-and-child health (MCH) booklets, which contain the child’s health history, including vaccinations: “Bring back the mother-and-child booklets! We don’t have these anymore” (Health worker, Kajiado). While substitute notebooks can hold the same information, they do not necessarily have the same structure or sentimental significance as the MCH booklets: “Mothers really enjoyed [the MCH booklets]. It gave [caregivers] a sense of connection and empowerment to have their child’s health history, and it also provided an incentive to bring the card for every health contact” (Health worker, Kajiado).

Health workers and key informants also discussed ongoing problems with the vaccine cold chain. In some facilities, there were refrigerators with adequate storage space but “efficiency is questionable” (Health worker, Kitui). When problems with refrigerators arise, “the technician is not readily available and is expected to come from the county [office]” (Key informant, Nakuru), causing delays and requiring health facilities to improvise in the meantime. Some facilities reported not having refrigerators and therefore needing to further limit vaccination days: “…three facilities schedule vaccination [vaccinate only on a few specified days of each month] as they do not have refrigerators” (Key informant, Bungoma).

Health workers and key informants discussed the various reasons for vaccine stock-outs, including delayed vaccine distribution at the county level, insufficient transport funds for supervision visits and for vaccine pick-up from the subcounty, and problems with the supplier. Ultimately however, many believed that these persistent problems were a result of the national devolution in 2013 in which 47 counties began setting up their own semi-autonomous institutions, impacting the healthcare system in the country. “…Immunization should have remained a national function…it was better organized then…we have more challenges now after devolution…counties do not have money to buy things like syringes because of budgetary constraints….” (Key informant, Bungoma). In addition to budgetary constraints, there are logistic challenges to vaccine distribution and supervision: “The old system [before devolution]…was a better system since now the health facilities pick up the vaccines and the supervision is erratic. There are instances she has taken a motorbike to go supervise the facilities, but some are so far-flung and no funds are available to take a vehicle there” (Key informant, Trans Nzoia).

Health education

Participants in all FGDs and IDIs universally called for increased health education for both caregivers and health workers. Health workers acknowledged that caregivers were generally very desirous and receptive of health education: “They normally ask why, why, why and, then we answer them accordingly” (Key informant, Kitui). Many thought that basic health counseling at or before an appointment was effective, even for those small pockets of cultural and religious resistance and mobile populations in Kenya: “Health education on the importance of vaccination could improve the health-seeking behavior of the mothers” (Caregiver, Kajiado). Caregivers suggested that health facilities should engage the communities and involve their traditional leaders in advocating for vaccination services: “We need more discussions with the traditional leadership so they can hold meetings to inform parents on the importance of immunization and to dispel any traditional beliefs that there is no need to vaccinate” (Caregiver, Trans Nzoia).

Health workers also acknowledged that they both wanted and needed more health education and other on-the-job resources, a sentiment that was independently reinforced by key informants. Better education would help empower health workers to do their jobs better, in addition to reinforcing standard operating procedures and policies such as whether or not to open multi-dose vials: “…[health workers] send mothers away because [they are] reluctant to open a vial of vaccine for one child” (Key informant, Kajiado). Health workers and key informants agreed that health education should also be provided to new and non-NVIP staff members, not just to the seasoned NVIP staff. “There is a big knowledge gap and limited experience amongst healthcare workers offering vaccination services… That zeal and ownership of programmes I used to associate with the NVIP teams at sub-county and health facility level is not there in the new recruits” (Key informant, Taita Taveta). However, opportunities for training were reported to be limited; also, when training was available, the ability to participate was further limited because of understaffing: “For example, today with the immunization workers at the training, we are strapped” (Key informant, Kajiado). Some health workers also raised the issue of the devolution and how it has affected health education: “We are not doing much on health education; since devolution, the county has not adapted the national level plans and activities that were working” (Key informant, Trans Nzoia).

Follow-up actions to reduce MOV and improve coverage and equity

Following field work, all field teams reconvened at the national capital to brainstorm about interventions and to create an intervention plan based on preliminary results from both the quantitative and qualitative data. The multi-partner technical working group on immunization in Kenya has endorsed this plan which aims to address the main identified causes of MOV. To improve health worker knowledge on vaccination across departments, NVIP plans to increase supportive supervision and to create an orientation package specifically targeting non-NVIP staff. Using adult learning strategies, training modules will address vaccination practices and interpersonal communication skills. Improving health worker knowledge, attitudes, and practices across all departments is expected to reduce one of the barriers to timely vaccination of eligible children [43]. The intervention plan also includes an activity to ensure an adequate supply of vaccines and vaccination-related materials by expediting the implementation of a new stock-management module across all counties. Because printing recording tools is the statutory function of counties, NVIP also prioritized efforts to provide counties with electronic copies of the latest versions of all documentation tools (e.g., monitoring charts, summary sheets, tally sheets, MCH booklets) to enable printing of copies. Finally, to clarify immunization-related policy and coordination roles between the national and county governments as a result of the devolution, the MoH plans to disseminate updated NVIP policies and guidelines, along with updated NVIP manuals and standard operating procedures.

Limitations

The FGDs and IDIs were not recorded because of concern that recording would deter participants from speaking honestly and openly. To minimize the possibility that some of the meanings of the transcribed texts were not captured accurately, each FGD and IDI had one or two note takers who were trained to capture verbatim quotes, and teams had field staff with previous qualitative data collection experience, and additional training as moderators. Although note takers were expected to record additional basic information (including total participant size in the FGDs), this information was missing in most of the notes. Additionally, although the target number of participants per FGD was 6–8, some FGDs ended up being larger, which may have impacted group dynamics and discouraged participants from sharing their perspectives. Second, although most FGDs were moderated by someone of the same sex as the participants, this was limited by field staff availability. In a few instances of gender mismatch, cultural sensitivities around gender roles might have affected the ability of participants to express their thoughts openly. However, the likely impact of gender mismatch is minimal because given vaccination is not a sensitive topic in Kenya. Third, all caregivers included in FGDs had received services at a health facility; the perspectives captured are biased toward those who were seeking health services and may not accurately reflect the community’s perspectives. We are confident that this bias had minimal impact on the validity of the findings because the focus of this study was on understanding reasons for MOV in healthcare settings. Nonetheless, these findings are not generalizable to the entire Kenyan community. Finally, these findings identify factors that might contribute to MOV, but these factors may not be causal.