Summary
Background Recent research has found that less than half of people identified with an eye problem in Meru county’s screening programme were able to access care, with younger adults being the least likely to receive the care they needed. We aimed to interview and survey members of this ‘left-behind’ group to explore barriers and identify potential solutions using a rapid mixed-methods approach.
Methods First, we conducted interviews to explore perceptions of barriers and potential solutions. Next, we asked a representative sample to rank the suggested solutions by likely impact. Finally, we held a multistakeholder meeting to identify which of the top-ranked interventions offered the best balance of impact, feasibility, cost, and potential risks. We used a deductive matrix and thematic analysis to rapidly analyse the interview data.
Results We conducted 67 interviews. Barriers to access included long queues, conflicting work engagements, and lack of clear information. Proposed solutions focused on reducing queue lengths, providing better counselling and clinic information, holding mop-up clinics, and maintaining adequate stocks & supplies. We conducted ranking surveys with 401 additional people from the left-behind group. All proposed solutions were ranked at moderately-to-highly likely to improve equitable access. Fifteen people attended the multistakeholder meeting, including community representatives. Workshop participants unanimously selected enhanced counselling and SMS reminders as the interventions that offered the best balance of impact, risk, cost, and feasibility. The other proposed solutions were deemed impractical or unaffordable.
Conclusion Rapid mixed-methods and multistakeholder collaboration were used to identify a range of potential service modifications that will be implemented within the ongoing programme. Our approach was centred on the experiences and perceptions of those who face the highest barriers to care.
Evidence before this study Previous research in Kenyan community screening programmes has shown that at least half of those found to have an eye health need will not be able to access care at their local treatment clinic, even if the care is provided free. Work in Meru County has shown that younger adults less are likely than any other sociodemographic group to check-in at their local clinic, but it’s not clear what the specific barriers are for this group. Across the African continent, approximately half of all ambulatory appointments are missed across all specialities, and sociodemographic inequalities are ubiquitous. In pursuit of Universal Health Coverage (UHC) and the Primary Health Care principles of equity and justice, health system managers are increasingly focused on identifying, trying to understand, and then address unequal access to care, however the traditional approach to identifying barriers and solutions has tended to centre around expert opinion rather than engagement with affected groups.
Added value of this study This study builds on previous efforts to introduce routine sociodemographic data collection into the county-wide eye screening programme operating in Meru, Kenya, as well as additional sites in Meru County, Botswana, Nepal, and Uttar Pradesh. Having already identified younger adults as the least likely to receive care in Meru County, this study introduces a novel mixed-methods approach for engaging with members of this left-behind group to rapidly identify barriers and scalable solutions. We used innovative methods to complete interviews and qualitative analysis in under two weeks, followed by a rapid survey to rank the potential solutions that emerged from this work with a representative sample of younger adults who had not been able to access care. Finally, a multistakeholder workshop with strong local and lay representation identified the top-ranked solutions that would be feasible to introduce and test within the ongoing screening programme. In addition to local evidence for action, this study presents an approach that any community-based programme could use to generate robust, non-tokenistic insights from affected communities within a matter of weeks, minimising the research time requirement and number of senior researchers required whilst maintaining rigorous scientific standards.
Implications of all the available evidence Equitably advancing UHC is predicated on identifying and overcoming unique barriers to care, however existing efforts rarely involve consultation or co-creation with affected communities. Building on existing rapid qualitative and mixed-methods methods, we have developed a cutting-edge approach to identify barriers, prioritise solutions, and identify service modifications that are feasible to introduce. We have applied this approach in Meru County, where younger adults – who were the least likely to access care – suggested a bundle of interventions centring on improving the provision of information and SMS reminders. Our research group will use an embedded RCT to implement and test this bundle, in the context of an equity-focused continuous improvement model that we are also implementing in Botswana, India and Nepal to incrementally improve access for all, with a focus on left-behind groups.
Introduction
Improving equitable access to community health services lies at the heart of Universal Health Coverage (UHC) and ‘leaving no one behind’ is the ‘central, transformative promise’ of the Sustainable Development Goals.1–4 WHO’s Thirteenth General Programme of Work states that ‘the main challenge to making progress towards UHC comes from persistent barriers to accessing health services’.
Our research collaborative is developing and testing a novel approach to identify and address inequitable access to care using the ‘IM-SEEN’ approach (‘Improvement studies for evidence-based and equitable innovation’). This involves identifying which groups are being left behind in a given programme; engaging with these groups to understand the unique barriers they face and their ideas for service improvements; and then testing these potential solutions with embedded randomised controlled trials (RCTs).5
We are applying this model in the context of community-based eye screening programmes in Botswana, India, Kenya, and Nepal. Uncorrected visual impairment affects over a billion people worldwide, levying major social and economic costs, despite the availability of highly cost-effective treatments like spectacles and cataract surgery.6 Our first set of findings from a cross-sectional equity analysis of over 4,000 people in Kenya’s Meru county found that only 46% of those found to have an eye need were able to access their free local treatment outreach clinic.7 We found that younger adults, males, and those working in sales, services, or manual jobs were the least likely to receive the care they need. Age was the strongest predictor of poor access, with less than a third of people aged 18-44 years receiving care compared to two thirds of those aged >45 years, even after controlling for severity of eye condition and a wide range of other factors.
Traditionally, ideas for how to improve programmes come from ‘experts’, service providers, or surveys of service users - rather than affected people themselves.8,9 In the context of renewed interest in Primary Health Care10–12 and the insidious persistence of colonialism and epistemic injustice in global health,13–15 increasing attention is being paid to person- and community-centredness. Simply put, advancing equitable access to health services must be done with, rather than to, or on behalf of left-behind groups.9
In this study we aimed to engage with younger adults who had not been able to access eye care in Meru in order to explore their perceptions of how the local services could be modified to improve access. Working within a live programme, we aimed to deliver robust, non-tokenistic, and generalisable findings within a matter of weeks, with a view to testing suggested service modifications with a subsequent embedded RCT.
Methods
Setting
Meru is a county with a population of 1.5 million in central Kenya, 110 miles north of Nairobi. It includes Mount Kenya and Meru National Park. The capital, Meru town, is home to a quarter of a million people. Agriculture is the main source of employment, with khat and tea representing important cash crops. Kenya’s Vision Impact Programme (‘VIP’) has been operating in Meru since July 2022, and has reached over 350,000 people to date, according to internal data. Teams of screeners go house-to-house testing all adults’ vision using a simple smartphone-based app developed by Peek Vision.16 Screeners refer people whose visual acuity falls below 6/12; those who have a red eye or another issue upon basic visual inspection; and anyone who feels they have an eye problem, even if there are no clinical signs and their visual acuity is >6/12. Our research team has been working with screeners to gather sociodemographic data from every person who screened positive and was referred to an outreach clinic for further assessment and treatment between April – July 2023. As stated above, we had previously found that younger adults are the least likely to be checked-in at treatment clinics but we did not know what the main barriers were or what could be done about them.
Research paradigm, theory, and methodology
We used a pragmatist philosophical paradigm17,18 and a phenomenological approach19,20 to explore these young adults’ lived experiences and perceptions of barriers to accessing eye clinics, and potential solutions. We grounded our work in the complementary frameworks developed by Levesque et al and Obrist et al.21,22 Both conceptualise access to care in terms of service and service-user characteristics. This distinction is helpful as our ultimate aim was to identify service modifications that improve accessibility for younger adults. We required mixed methods to answer a multi-layered question: what are the main barriers to accessing eye services in each location and what could be done about them?
Methods overview
This study was conducted in three stages. In Stage 1, we used interviews to generate a long-list of perceived barriers and potential solutions. Then, in Stage 2, to move from subjective experiences to generalisable service modifications, we conducted a telephone survey where a representative sample of younger adults who did not receive care ranked each of the suggested solutions by likely impact. Finally, in Stage 3, these ranked solutions were reviewed by a multistakeholder group who identified a package of interventions to test based on likely impact, feasibility, cost, and risks.
Team composition and reflexivity
This project was part of the broader ‘IM-SEEN’ programme of work that seeks to develop a new, rapid, robust, and responsive approach to continuously improving access to care, starting in the field of community-based eye screening programmes in Botswana, India, Kenya, Nepal. LSHTM-based researchers (LA, AB, MB, JR, DM & MK) working with Kenya’s Ministry of Health eye lead MG, AB and NB from Peek Vision - the screening programme software provider, and SK - the local research lead SK based at KEMRI, had already conducted a collaborative equity assessment of Meru’s VIP programme. LA – a mid-career British clinician, policy advisor, and mixed-methods public health researcher - led the development of the methodological approach to be used in all countries to engage with members of the left behind groups. LA worked closely with SK – a mid-career female Kenyan public health social scientist – to tailor the approach for Meru County, supported by the wider team. LA and SK recruited and trained six local, early-career data collectors (DG, EMM, EM, PK, BN and FG) to conduct the interviews and surveys. We were interested in understanding the barriers and solutions as perceived and described by affected people in their own words. SK and LA facilitated the multistakeholder workshop where findings were interpreted by lay representatives, other members of the left behind group, and local programme managers. This local multistakeholder group collectively made the final decisions on which suggested service modifications to take forward for implementation.
Stage 1: interviews
Recruitment and sample size
Peek Vision – the programme software provider - provided us with a list of every person aged 18-44 years who had not been able to access their clinic appointment in Meru. In random order, we phoned people from this list to invite them to participate in the interviews, and sought recorded verbal informed consent. We tried each person three times before moving on to the next.
We planned to use Guest and colleagues’ approach to determine our sample size based on thematic saturation, using a ‘base’ of 12 interviews followed by runs of two interviews with a 0% new information threshold.23 In other words, we aimed to continue recruiting interviewees until no new themes emerged after two interviews in a row, with a minimum sample size of 14 (‘12+2’ approach).
Interview modality
We wanted to use telephone interviews, based on empirical evidence that they can be completed faster at lower cost than in-person interviews, and with equivalent data richness.24–28 However, we were not entirely convinced that the data would be equivalent. As such, we decided to recruit two separate samples and use both modalities, conducting an embedded mode-comparison study29 that will be reported elsewhere.
Data collection
Three pairs of Kenyan data collectors with at least basic qualitative training and fluency in English, Kiswahili, and the local dialect conducted semi-structured interviews using the topic guide summarised in Box 1 (see Appendix 1 for the full script). For the telephone interviews, calls were made on speakerphone in a private space and recorded using the phone’s inbuilt call recording app. As one data collector conducted the interview, the other noted down the times at which each unique barrier and potential solution was mentioned. After the call, the interview recording was immediately replayed and the data collectors entered verbatim quotes directly from the audio into our analytic matrix. The same process was used for in-person interviews, but with an audio recorder instead of a mobile phone. Our decision to use direct-from-audio transcription was based on findings from a background systematic review that we conducted on rapid qualitative approaches.30 Interviewees did not review their transcribed quotes in the matrix. In-person interviews were conducted in private rooms in four different health facilities where interviewees’ responses could not be overheard. Only the data collectors and the interviewee were present for each interview.
Topic guide
Barriers
In your own words, can you talk me through why we didn’t see you at that clinic?
Probing questions
Are there any other factors that prevented you from attending?
Is there anything else you’d like to share?
Solutions
What would make the biggest difference in addressing these barriers? Probing questions
else would help?
other changes could we make to the programme that would make it easier for you to attend?
there any other specific changes that we could make to the way that the programme or eye clinics run?
mentioned [list their proposed solutions]. Some of these may be beyond our control, but if we managed to [list their proposed programme-related changes], do you think that would be enough?
That’s the end of my questions. Is there anything else you would like to add?
Data analysis
We utilised an abductive analytic approach,19 whereby data collectors initially entered verbatim quotes relating to barriers and solutions into a deductive framework matrix, nesting each quote under one of ten broad a priori themes that had emerged from a literature review that is described in our protocol:31
Costs
Distance and transport
Desire/priority to seek care
Clinical service quality
Facilities
Awareness & communication
Fear
Norms, values, health beliefs
Empowerment, support & capacity
Other (making room for surprising/unexpected themes)
At daily debrief sessions, SK and LA reviewed the matrix with the data collectors and used inductive coding to identify unique barriers and solutions. The decision to use an analytic matrix and collective interpretation was based on the findings of our previous systematic review, which had found these techniques to be rapid and robust.30
Our matrix had one participant per column and one sub-theme per row – with a new row created every time a sub-theme (a unique barrier or solution) was identified. Each sub-theme (e.g. ‘loss of earnings’) was nested under the relevant theme (e.g. ‘costs’) The process of data entry is demonstrated in this short online video (http://tinyurl.com/29asc6nm) and a blank matrix template is available here.
We generated one matrix for the telephone interviews and another for the in-person interviews. This was so that we could compare the themes that emerged from each modality in our embedded study. For our main analysis, presented here, we pooled all barriers and solutions identified using both modalities.
We trained the data collectors over two days and performed fourteen pilot telephone interviews before starting data collection. Videocall debriefing sessions were held at the end of each day.
Additional analysis
Our original equity analysis had also indicated that people with the highest incomes and those who owned a car or truck may have been less likely to attend that those reporting no vehicle ownership and lower incomes. We conducted an additional ten interviews with people who reported earnings in the highest income category to assess whether the barriers they reported differed from those reported by younger adults. We hypothesised that richer people did not access VIP services because they had sought private care after being identified with an eye need during eye screening.
Output and screening
We created a summary list of all of the unique solutions that had emerged from the interviews. Before taking these to a representative sample for ranking, we met with the implementing partner to identify any ideas that would be completely infeasible given the constraints of the programme e.g. providing helicopter transportation. Any interventions that were deemed to be completely infeasible were removed from the list. We asked the director of Peek Vision to independently review these decisions.
Stage 2: telephone survey
Survey instrument
We used the vetted list of solutions to generate a simple telephone-based survey (Appendix 2) where respondents were asked to rank each suggestions from 1-3 on a Likert scale:
It would make a big difference - i.e. if we introduced this change then you or people like you would definitely attend,
It would make a moderate difference - i.e. it would greatly increase the chances, but it would not be enough to guarantee attendance by itself,
It would make a small difference - i.e. it might help a few people, but the impact is likely to be minimal.
The telephone ranking survey was piloted with 26 people.
Sampling and recruitment
We used a 95% confidence interval, a 5% margin of error, and a conservative assumption that the total population size was 1 million people, rendering a minimum sample size of 384. We took the same list of 18–44-year-olds who had not been able to access care, and used random numbers to generate a call order, removing those who had already been included in the qualitative interviews. The same six data collectors tried calling each person three times before moving on to the next.
Data collection and Analysis
Data collectors obtained recorded informed verbal consent, and then read through the survey instrument using an online data entry form. Data collectors entered the respondents’ score for each proposed solution. We calculated the simple mean for each solution, and then ranked solutions by mean score.
Stage 3: multistakeholder workshop
Once we had this ranked list of solutions, we convened an online workshop with representatives from the programme implementer, programme funder, the county and national health ministry teams, and our community advisory board. We facilitated a discussion where each stakeholder shared their perceptions of the likely impact, feasibility, costs, and risks associated with each solution. As external public health and research ‘experts’, LA and SK restricted their contributions to presenting the ranked solution scores, facilitating the discussion, and providing information on the general strength of the international research evidence for each of the proposed solutions. At the end of the workshop, we asked the participants to collectively agree on one or more solution to implement in the VIP programme. Figure 1 provides an overview of our entire approach.
Overview of the sequential mixed-methods approach
This study was approved by KEMRI and LSHTM ethics committees. Those who attended in-person interviews were given a transport reimbursement of KES 500 (USD 3). We used the COREQ checklist to report our study (Appendix 3).
Findings
Interviews
We made 143 phone calls to invite people to participate in in-person and telephone interviews. Three people declined; 29 did not pick up after three calls; and 34 people agreed but either were not home (13 people) or did not arrive at the agreed interview location (21 people) on the day of their in-person interview; six were not eligible as they told us they had actually received care (i.e. they had not been checked-in properly); and four had moved to a different part of the country. In total we conducted 36 telephone interviews and 31 face-to-face interviews over the course of eight days in September 2023. All our participants were aged 18-44 years old and 53.2% were male.
We ended up performing more interviews than were needed to achieve thematic saturation with the 12+2 approach due to the efficiency of our data collectors. At the debrief on day two, they had already completed 24 telephone interviews. Our research leads had not assessed whether saturation had been reached by the time of the call, so – erring on the side of caution - they advised completing a further day of interviews. By the end of day three, 36 telephone interviews had been completed. A detailed retrospective saturation analysis, presented in Appendix 4, concluded that approximately 30 interviews were required to reach thematic saturation (Figure 1). We conducted 31 in-person interviews to enable fair comparison between telephone and in-person interviews for our embedded study.
Accumulation of themes as the interviews progressed
Supplementary Table 1 (Appendix 5) presents the 21 unique barriers that were identified along with all the quotes from both sets of interviews. Direct, indirect, and opportunity costs; long queues; difficulty getting time off work; and insufficient information about opening times and dates emerged as important themes. We also identified several meta-themes; participants were generally able to access the clinic locations but left after seeing long queues of several hundred people. Many felt they could not ‘waste time’ waiting to be seen, given the associated loss of potential earnings.“I choose to go to work to make money rather than spend my days’ time not knowing whether I will get attended to… If I don’t work, I don’t get money. MFK008, in-person
Another important cross-cutting theme was the perceived lack of information about the clinics: where they were, days of operation, opening and closing times, and what services were available. Assumptions around (non-existent) costs and early closures also prevented some people from attending.“I also forgot the exact location where I was to go for the eye check-up and no one followed up to remind me of the place and date.” MT33, telephone
“They did not tell us if we need to come with money or not. Eye treatment, we are usually told to come with money. I assumed I’ll go there and they will ask me for money and I did not have it.” MFK204, in-person
One interviewee also told us that the counselling he had received at the point of referral was inadequate. He wanted more information about what would happen at the clinic and on why attending was important, especially given that he did not even realise he had a problem:“They just told me that I have problems with my eyes and I should visit [town name] dispensary so I did not know what I was going to do there, is it surgery, is it being given medication, is it being tested again? And for me I have always known that my eyes are okay, and on that day they told me that they are not okay. They were very brief and I didn’t know what to expect, so that shock of being told that I have an eye problem which I have never had before is the reason why I did not go.” MFI03, in-person
In terms of novel barriers, one person told us that they left the queue because they were “an introvert” and didn’t like the crowd (MT772, telephone); another felt their eye problem needed emergency treatment and sought care elsewhere (MFK02, in-person). One interviewee specifically named male health seeking behaviour as the main reason he didn’t attend:“As a man it is very hard to prioritize my health as I am manly focused on my family’s wellbeing and It is easy to forget my health needs” (MT250Z, telephone).
Finally, one young man explained that being made to queue alongside women and children was shameful:“There were women on the line. They could have different lines for youths and women for some us to be comfortable because it is shameful to be on the same line with women and children, with worries how they will perceive me as young man. It was a challenge for me to just stand there with women… I had to go back that day without being attended even though right now my eyes have a problem. MT040, telephone
The 25 proposed solutions to improve access centred around reducing the clinic queue lengths so that people could be seen quickly and then get back to work. Ideas included adding more clinics, holding them closer to villages or workplaces, increasing staff punctuality and speed, scheduling fewer people to attend each day, and extending the opening days and hours. The other meta-theme related to the provision of more detailed information around clinic services and opening times. Table 1 presents a summary of all 25 suggested solutions along with illustrative quotes. A full list all solution-related quotes is presented in Appendix 6.
Reviewing feasibility
As planned, we presented the list of all 25 suggestions to senior representatives from the implementing partner. We asked them to identify any suggestions that would be completely infeasible to deliver, given that they are responsible for funding all aspects of the programme. They felt that the programme budget would not permit additional payments for transport reimbursement or attendance incentives. Given that the outreach clinics involve multiple members of staff and large volumes of equipment and supplies, they also felt that it simply wouldn’t be feasible to deliver a door-to-door version of the outreach clinic. These suggestions were removed from the list. The director of Peek Vision agreed with each of these decisions. The remaining 21 suggestions were put to a representative sample of people from the left behind group in a ranking survey.
Survey and selection workshop
It took two days to train the data collectors and pilot the survey, and five days to complete ranking exercise. In total, data collectors called 440 people, of whom 401 completed the survey (response rate = 91%).
All of the suggestions received a mean score between 2.4–2.9, indicating that all of the potential solutions were felt to offer at least a moderate chance of improving attendance. Table 2 presents the scores for individual-level solutions (that could be tested in an individually randomised RCT) and solutions that would be implemented at the clinic level (requiring testing with cluster RCTs).
Our online workshop included representatives from the community advisory board, the screening programme implementing partner organisation, the programme funder, the national eye screening programme office, Kenya Society for the Blind, and the county health department. After our team had presented the survey findings, we facilitated a discussion around each of the proposed solutions in turn.
The majority of the suggested clinic-level interventions required additional human resources or clinic locations. The programme funder and implementing partner recognised the issues around long queue times in some locations, but were very clear that inflation had already taken the programme over budget and there were no spare resources to introduce additional clinics or staff. The top-ranked suggestion related to frustration experienced by people who attended the treatment outreach clinic but were found to have a complex eye problem that required onward referral to the local hospital for specialist spectacles (where care is subsidised but not free). The workshop participants also recognised this problem, but agreed that it was not possible to have advanced ophthalmic care services present at each outreach clinic. The group suggested clarifying the process of tiered referrals during counselling. The workshop participants felt that organising separate queues for different ages and genders would be practically feasible, however imposing separation may cause problems for friends/family/colleagues who attend together. After reviewing all of the suggestions, the workshop participants unanimously agreed to implement and test the following bundle of interventions relating to counselling and the provision of enhanced information in SMS reminders:
– Send a reminder text on the appointment day and the day before
– Clarify what services are available at the outreach clinic and costs linked to these services
– Specify clinic opening and closing time at the point of referral
– Explain why attending clinic is important at the point of referral
Overall, the ethics review process took four months. The interviews took nine days to complete, and the survey took seven days, including training and piloting. We held the multistakeholder workshop one month after the survey had concluded, with the delay driven by scheduling challenges.
The current verbal counselling script and SMS reminder messages that are used in the VIP programme are presented in Box 2. The SMS reminders are currently sent on the day of referral and the day before the clinic appointment. We drafted a new verbal counselling script and SMS reminder message that included the suggested new elements that were agreed in the workshop. We asked the workshop participants to review the new wording via email, as well as two representatives from the left-behind group. Based on this feedback we made three minor changes. The original script and description of these changes is presented in Appendix 7. The enhanced counselling and SMS reminder will be tested in an individual- level RCT in a subsequent study.
Original and new counselling and SMS reminder wording
Usual care counselling, delivered verbally at the point of referral
“I have examined your eyes, and you have a problem. I have referred you in the system and you will receive an SMS with where and when you are supposed to attend treatment. You will come for treatment on <<date>> at <<location>>. The examination will be free and you will be informed of anything else on the material day.
Current SMS reminder, sent on the day of referral and day before the appointment
Dear <<name>>, you were examined and found to have an eye problem. Kindly report on <location> on <<date>> for assessment. For more information contact Meru Referral Hospital.
Enhanced counselling script, based on interview, survey, and workshop feedback
“I have found a problem with your eyes. I am referring you to the outreach treatment clinic that will be held at <<location>> on <<date>> between <<time>> and <<time>>. At the clinic, eye care professionals will perform a specialist assessment and provide any eye drops or medicines that you might need. If you need glasses, the specialists will tell you what kind you need, and what your prescription is. The assessment is completely free. Note that a small proportion of people will be found to have complex eye problems that require onward referral for hospital assessment and special lenses that cost more than standard glasses. However, the vast majority of people have their needs fully met at the outreach triage clinic and do not need hospital referral.
With treatment, you will be able to see more clearly. This will help with your work, seeing faces, and using your phone. It is important that you attend the clinic to protect your vision. The clinic will only be running from <<day>> to <<day>>, so if you don’t manage to attend, you may not be able to get free care again in the future.”
Enhanced SMS reminder, based on interview, survey, and workshop feedback, to be sent on the day of referral, the day before the appointment, and on the day of the appointment
We found that you had an eye problem. Please attend the outreach clinic at <<location>> on <<date>> between <<9AM-2pm>>. The specialist assessment is free
If you are found to have a complex problem, you may be referred to a hospital for further care or specialist glasses, and this may include a fee
However, the vast majority of people who attend the outreach get their eye problem fixed without the need for any further referral
It’s important that you attend to protect your vision, and you might not have a future opportunity to access free care. See you on <<date>>
Sensitivity analysis
We conducted ten additional interviews with high-income individuals aged 18-44 who had not managed to access care, based on previous evidence that this group were potentially facing unique barriers. We triangulated the themes with those from the other interviews. High-income interviewees did not mention loss of wages as a potential barrier, but the cost of treatment and transport were raised, along with conflicting engagements, long queues, and lack of clear information. No unique or discordant barriers were identified. Similarly, the solutions raised were closely aligned with those from the other interviews, including requests for financial subsidies and incentives – suggesting that this group were not necessarily very affluent i.e. our original ‘high-income’ threshold (USD 2,600/year, aligned with the national top income tax threshold) was set too low. The only unique solutions related to a request for more stylish spectacle frames; “Next time come with spectacles of good standard, not for the old people.” (HIG3); the use of Community Health Promoters to remind people to attend; and asking for permission from employers to get time off. Full quotes and codes are presented in Appendix 8.
Discussion
We conducted 67 interviews and 401 solution ranking exercises with people from the sociodemographic group that had previously been found to face the greatest barriers to accessing eye care services in Meru – young adults aged 18-44 years. These people told us that lack of clear information, long queues, and the opportunity costs associated with long queues were the main barriers to receiving care. Whilst lack of clear information and inadequate counselling have emerged from multiple previous studies in diverse settings and populations,32–36 the theme of long queues does not commonly appear in the wider literature or appear as one of the drop-down barrier options that is used in Rapid Assessment of Avoidable Blindness (RAAB) surveys that have been deployed in over 80 countries.37 We postulate that the ubiquity and fundamental intractability of this problem in resource-scarce settings leads to them being perceived as the status quo rather than a specific barrier.
Research from several other African countries suggests that most patients generally wait 1-4 hours to receive ambulatory care,38–43 and even in a well-resourced setting like the UK, 10% of patients are currently waiting more than 12 hours to receive emergency care.44 Younger adults may be more sensitive to long wait times as they are more likely to be in active employment than older adults, and may be more likely to be ‘hustling’ with multiple informal jobs that do not provide protected sick leave, meaning that they experience the greatest potential opportunity costs from waiting in line for several hours.45
Rapidly surveying a representative sample of younger adults enabled us to move from subjective to generalisable themes. For instance, if we had stopped after completing the interviews, we would not know whether the issue of mixing men and women in the queues was a major structural barrier to access for this group, an unusual individual quirk, or something in-between. In the event, we found that scores accorded to all of the 21 potential solutions clustered between 2.5-3.0, indicating that members from the left behind group felt that all of the suggestions were moderately-to-highly likely to improve access to care. We cannot exclude the possibility that the universally high scores are at least partially driven by cultural norms and/or a form social desirability bias.46,47 We took all of the ranked suggestions to a multistakeholder meeting that included representatives from the left behind group, the local community, the programme funder and implementer, and a national eye charity. This group weighed the rankings against considerations of the risks posed by each intervention, their collective estimation of likely impact, the associated costs, and the operational feasibility of implementing each solution. This group reached a consensus agreement to focus on a bundle of solutions aimed at providing timely and clear information about the outreach clinics.
The selected interventions focus on empowering individuals with the information they need. However, a different perspective is that the workshop participants dismissed more highly- ranked service-side solutions in lieu of minor interventions that place responsibility for access on intended service users, in line with the widely debunked ‘information deficit model’.48 Allen has previously argued that efforts to equitably advance UHC and ‘health for all’ should focus on unjust social structures and resist focussing wholly on ‘downstream’ technocratic solutions. In the context of this study, the paucity of good jobs, the absence of adequate social welfare systems and strong labour laws that guarantee paid sick leave are examples of major structural challenges for younger adults, however these were not raised by the interviewees.
The greatest strengths and limitations of our approach pertain to voice and epistemic justice.13 We strove to centre our approach around the perspectives of those with lived experience of not being able to access care. The ‘pose’ and ‘gaze’ of the project15 focus on the credibility and value of this group, as well as locally recruited data collectors’ work in sensemaking through their initial interpretative work as they engaged with the analytic matrix. Findings from this process fed directly into the local programme, with their implications wholly determined by local stakeholders. Although Europeans and Kenyan public health experts were involved in leading the project, the primary interpretation was locally owned and the resulting solutions will be locally implemented for the benefit of local Meru residents.
The programme funder and implementing partner held veto rights as they were ultimately responsible for delivering service modifications and could not be ‘made’ to implement any of the suggested service modifications. However, community, charity, and public health stakeholders reached one accord with the programme leads on the infeasibility of introducing additional staff of clinics given their shared appreciation of how inflation had literally decimated the programme budget. Nevertheless, our approach involved (and necessitated) epistemic tension; even though all of the solutions came from the left-behind group, the final decision on how to modify the programme required consensus between different stakeholder groups with imperfectly overlapping interests and unequal power in the final determination. We feel that including programme managers is a design strength rather than a flaw, as there is no value in presenting programmes with a post-hoc shortlist of unworkable suggestions. Furthermore, the process of bringing them into conversation with community members, eye care advocates, and representatives from the left behind group can create shared understanding and constructive dialogue. Our role as ‘external’ research leads primarily involved developing rapid and robust methods that would surface non-tokenistic and reliable suggestions, as well as convening actors and facilitating the overall process. We aim to work with local partners to embed this approach so that future iterations can be wholly led by local teams.
Finally, our unplanned sampling deviation led us to conduct more interviews than expected. The post-hoc sensitivity analysis presented in Appendix 3 suggests that if we had adherence to our original ‘12+2’ approach, we would have missed four barriers and 12 additional solutions. We agree with Guest and colleagues that current approaches to assessing and reporting saturation in qualitative research are opaque, however our empirical analysis suggests that – for studies like ours – saturation is only reached after three interviews in a row reveal no new themes, after a minimum of 15 initial interviews have already been conducted (the ‘15+3’ approach).23
Conclusions
Interviews and ranking surveys conducted with people who were not able to access eye care services identified 21 barriers and 25 potential solutions, centring around reducing the time it takes to be seen, and providing clearer information about the eye clinics. Multistakeholder workshop participants ended up selecting interventions that did not fundamentally alter how clinics are operated, instead opting for amendments to the way that information is provided. Even so, these service modifications came directly from those who were being left behind, and were rated likely moderately-to-highly likely to improve access. Including service funders and managers in interpretive and deliberative processes means that our findings will result in concrete, locally-led service modifications. Further work will evaluate the impact of these changes using an embedded, pragmatic RCT.
Data Availability
All data produced in the present work are contained in the manuscript.
Appendix 1: Call scripts
Telephone interview consent script and topic guide
Hello, my name is. I am a researcher from KEMRI, working with the Ministriy of Health and the VIP eye screening programme.
Your recently had your eyes screened and were found to need further assessment. Our records indicate that, like many other people, you were unable to attend that appointment.
You are being contacted because you have previously provided consent to be contacted regarding research being conducted for eye care services. I am calling to invite you to participate in a 15-minute interview. Your participation is completely voluntary. This means that you do not have to do it unless you want to.
We want to understand the barriers that prevented you from attending. We are also asking about how we could change the VIP programme to make it easier for people like you to attend appointments.
Before agreeing, here is the background information that you need to know:
We have invited you because, like many other people, you did not attend. We want to hear about the issues that you personally faced that prevented you from attending, and your ideas on how to make things easier. In total we are aiming to interview about 20 people.
Who are we? I work with a group of researchers from the KEMRI and the London School of Hygiene and Tropical Medicine. We are working to improve the VIP eye screening programme. The leaders of the research are Prof Michael Gichangi and Dr Luke Allen.
We will take the responses from all of the interviews and discuss the ideas for improvement with the leaders of the national programme. We hope to use your suggestions to make the programme work better.
We are also conducting a set of face-to-face interviews with other people who did not manage to access care. We want to compare the responses we get from these different approaches.
In this 15-minute interview there are no risks or benefits to you. It is important to note that agreeing or declining to take part does not have any impact on the services you receive.
You can stop the interview at any time.
I will record the interview. Our team will anonymise your data and keep it safe and secure on a password-protected computer in London. When the study is completed, we will write-up our findings and publish them online so that other researchers can use the information to help people in other places.
The KEMRI and London School of Hygiene and Tropical Medicine ethics committees have both approved this study.
You can ask me any questions you like now. I can also give you the email address and phone number of the lead researchers if you’d like to contact them directly [provide the contact details for Sarah Karanja or Luke Allen as required]. If you have any other concerns I can also give you the contact details for the London School of Hygiene and Tropical Medicine Research Governance and Integrity Office.
Do you have any questions?
Are you happy to begin the interview?
[If yes, proceed; if no thank them for their time and end the call]
Opening questions
To start with, during the home visit by the vision impact project team, please describe your experience with the eye screening.
Probes:
How long did the examination take?
Was the test comfortable? If yes, why? If no, why not?
Barrier elicitation questions
You were assessed at home and found to have an eye problem that required further assessment. You were asked to attend the community eye clinic on [date] and reminders were sent to this mobile number.
In your own words, can you talk me through why we didn’t see you at that clinic?
Probing questions
Are there any other factors that prevented you from attending?
Is there anything else you’d like to share?
Of the issues you mentioned, which is the most important?
Solution elicitation questions
The last part of the interview is exploring whether there is anything we could do to address these barriers and make it more likely that you will attend in the future.
So to start, what would make the biggest difference?
Probing questions
What else would help?
What other changes could we make to the programme that would make it easier for you to attend?
Are there any other specific changes that we could make to the way that the programme or eye clinics run?
You mentioned [list their proposed solutions]. Some of these may be beyond our control, but if we managed to [list their proposed programme-related changes], do you think that would be enough?
That’s the end of my questions. Is there anything else you would like to add?
Probing questions
After missing the outreach clinic appointment, did you seek medical care from a different healthcare facility for your eye condition?
If no, why not?
If yes, where did you seek medical care?
Thank you so much for your time.
Call script to invite people to participate in an in-person interview
Hello, my name is. I am a researcher from the Kenya Medical Research Institute (KEMRI), working with the Ministries of Health and Education on the Vision Impact Project eye screening programme.
You recently had your eyes screened at the community outreach programme and was found to need further assessment. Our records indicate that, like many other people, you were unable to attend that appointment.
You are being contacted because you have previously provided consent to be contacted by Ministry of Health partner organisations regarding research being conducted for eye care services. I am calling to invite you to participate in an in-person interview in the next few weeks. Your participation is completely voluntary. This means that you do not have to do it unless you want to.
If you agree, I will arrange to meet you in or near where you live to ask you some questions about the barriers that prevented your child from attending. I also want to ask about how we could change the Vision Impact Project to make it easier for people to attend appointments.
To compensate you for your time you will receive a [KES 500 equivalent] airtime voucher.
If you are interested, I can send you the full study information via WhatsApp or email or talk it through on the phone with you now.
[switch to phone PIL script here as required]
Appendix 2: Quantitative Ranking Survey
Interviewer name
Short-answer text
Participant ID
Short-answer text
Consent
Good morning/afternoon
My name is and I’m calling from the Vision Impact Project eye screening programme. We saw you a few weeks ago and referred you to the local clinic, but we did not see you on your appointed day.
In fact, half of all people who were referred did not attend. We have sought feedback on ways we could improve our service, and I wanted to ask you which of the ideas we have stand the best chance of helping people like you to access care. It should take approximately 15 minutes of your time.
If you are happy to proceed, I need to tell you a bit more about the survey. I will then double- check that you are still happy to proceed.
I will ask you about a set of potential changes that we are thinking about making. I will ask you to rate each one in terms of how likely you think it is to make a difference at helping people access our clinics.
Your responses will help us to shape and improve our services for others, but there are no direct benefits to you for taking part. Thinking about the issues that prevented you from getting care may be distressing to you. If you face any discomfort because of the questions asked, you can skip any question or ask to end the call whenever you choose.
If you don’t want to take part, that’s ok. You can drop out of the survey at any point. Your decision will not affect your health care or your future relations with the Vision Impact Project in any way.
Your anonymised answers will be combined with those from other people and kept safe and secure on password-protected computers in Nairobi and London. None of the data will be used for commercial use. We will publish our findings in a research journal and in a public repository so that other researchers can learn from what we find. You personal information will not be included in our findings and there is no way that you can be identified from any of the reports that we will produce.
If you have any questions, you can ask me now, or I can put you in contact with the study coordinator - Sarah Karanja from Kenya Medical Research Centre. If you have any questions about your rights as a research participant, I can connect you with the Kenya Medical Research Centre Ethics team who approved this survey.
Does that all make sense? Do you have any questions for me? Are you happy for me to start?
The first two potential improvements are aimed at making it easier for people to get to a clinic.
There are three options to choose from:
It would make a small difference - i.e. it might help a few people, but the impact is likely to be minimal
It would make a moderate difference - i.e. it would greatly increase the chances, but it would not be enough by itself to guarantee attendance by itself
It would make a big difference - i.e. if we introduced this change then you or people like you would definitely attend
If we add more clinic locations to reduce the distance people have to travel, how likely is that to make a difference?
1 []
2 []
3 []
Expand the list of referral clinics to include additional public health facilities and private and faith-based hospitals
1 []
2 []
3 []
The next set of improvements are about reminding people about the clinic
Explain why attending clinic is important at the point of referral
1 []
2 []
3 []
Clarify what services are available at the outreach clinic and specify if there are any costs linked to these services
1 []
2 []
3 []
Send a reminder text on the appointment day and the day before
1 []
2 []
3 []
Send a text to individuals who missed the outreach clinic to invite them to attend on another day
1 []
2 []
3 []
Phone people who cannot read to remind them to attend
1 []
2 []
3 []
Use churches and radio broadcasts to remind people to attend
1 []
2 []
3 []
The next set of improvements are about extending clinic opening hours
Keep the clinics open for a greater number of weekdays
1 []
2 []
3 []
Specify clinic opening and closing time at the point of referral
1 []
2 []
3 []
Keep the clinics open until 6pm
1 []
2 []
3 []
Keep the clinics open on the weekend
1 []
2 []
3 []
Hold mop-up clinics a week or so later for those who didn’t manage to attend
1 []
2 []
3 []
Allow people to choose their appointment day
1 []
2 []
3 []
The next set of improvements are about reducing waiting times
Add more staff to reduce the queue waiting time
1 []
2 []
3 []
Refer fewer people for the same day to reduce queue numbers on each day
1 []
2 []
3 []
Give each person a specific appointment slot
1 []
2 []
3 []
Increase the punctuality and speed at which our staff work
1 []
2 []
3 []
Have a separate line for younger people
1 []
2 []
3 []
Have separate lines for women and children
1 []
2 []
3 []
Have a separate line for older people
1 []
2 []
3 []
The final set of improvements are about stocks & supplies
Ensure that clinics are fully stocked so that all medicines and glasses can be provided on- site
1 []
2 []
3 []
Thank you so much for your time, we really appreciate your generosity, your responses will help us to improve the service Is there anything else you would like to feed-back to us?
Long-answer text
Thank you again, goodbye
Appendix 3: COREQ checklist
Appendix 4: Post-hoc saturation analysis
We originally set out to use Guest and colleague’s simple formula for assessing whether thematic saturation had been achieved.[1] They describe saturation as “the point in data collection and analysis when new incoming data produces little or no new information to address the research question”, and identify three components that can be used to establish when this point occurs:
– The base, i.e. the minimum number of interviews to be conducted before calculating whether saturation has been achieved. It is similar to the ‘initial analysis sample’ described by Francis et al.[2] Guest et al suggest base sizes of 4-6, based on a review of the literature.
– The run length, i.e. the number of interviews that we review to assess whether any new themes have emerged, with reference to those that have already emerged in the base. Guest et al suggest conducting runs of 2-3.
– The new information threshold, i.e. the number of new themes identified in each run as a proportion of the themes identified in the base interviews. Guest et al suggest that researchers may want to stipulate that saturation is only achieved when a run identifies no new themes (new information threshold = 0%), or that the new themes identified represent ≤5% of the themes identified in the base interviews.
We set out to use a conservatively large base size of 12 interviews, based on a literature review presented in our protocol, with runs of two, and a 0% new information threshold. We also decided that we would compare the new information (number of themes) identified in the runs against all previous interviews, rather than just those in the base.
As detailed in our findings section, we ended up conducting 36 interviews. By reviewing the number of themes that were identified in each interview, we are able to retrospectively assess the point at which saturation would have been reached using a range of different base, run, and new information threshold permutations.
For our initial analysis (Table 1), we used a base of 12 and compared new information obtained from each run of two interviews against the themes identified from the original base of 12 interviews. We performed individual analyses for barriers and solutions. We highlighted the points at which the new of 0% was met.
Table 1 shows that two interviews in a row were completed with no new barriers identified by interview number 14. By this point, 11 barriers had been identified. In the remaining 22 interviews a further four barriers were identified.
For the solutions, saturation was achieved by interview number 15. By this point 11 unique solutions had been identified. By the end of 36 interviews, a further 12 solutions had been identified.
In summary, had we used Guest and colleagues approach with a conservative base of 12, less conservative runs of two, and a very conservative new information threshold of 0%, we would have stopped after interview number 15, missing out on a third of the unique barriers that emerged across our 36 telephone interviews.
In our next analysis, we counted the total number of barriers and solutions (themes) that emerged from each interview. We ran six additional scenarios, using runs of two, three, and four against the base of all themes identified in the first 12 interviews. We also used runs of two, three, and four against all themes identified in all preceding interviews.
Table 2 shows that using a more conservative base that includes all interview conducted does not materially change the point at which saturation is deemed to have been achieved. However, increasing the run length from three to four was a associated with a much later saturation point; 32 interviews with a 0% new information threshold. Interestingly, the use of a higher new information threshold (<5%) would lead to very similar saturation points of between 14-16 interviews depending on run length.
Finally, we plotted the cumulative themes that emerged over the course of the 36 interviews. Figure 1 illustrates the fact that the majority of unique barriers were identified by interview number seven, followed by a long plateau. Shorter plateaus in the number of solutions arising from interviews 7-11 and 13-16 were followed by incremental gains until levelling off again around interview number 28. The paucity of new information arising from interviews 13-16 drive the saturation decisions in the analyses performed above.
Accumulation of themes as the interviews progressed
Figure 1 suggests that for this specific set of interviews, a base size of 15 and a run length of three might have yielded a more appropriate saturation point, given that new themes continued to accrue until around interview number 28. Table 3 shows that using a conservative base of ≥15 interviews and a new information threshold of 0% leads to saturation by interview number 30, 31, and 32 with respective run lengths of two, three, and four.
Researchers like Coenen et al. have suggested that run lengths of two are adequate,[3] whilst others have recommended runs of three.[2,4] Based on our findings presented in table we decided that we would use a base of ≥15 and runs of three for future studies.
Application to second data set
As we had conducted a second set of 31 in-person interviews, we were able to test whether the ≥15+3 approach would have ended data collection at an appropriate point, i.e. without missing additional themes that arose in further interviews, nor prolonging data collection long after all themes had been identified.
Table 4 shows that using this approach identifies interview number 24 as the point at which saturation is achieved.
Using this cutoff would have led to the inclusion of all solutions, and all but one of the barriers identified across the 31 interviews. Figure 2 plots the saturation point on a line graph.
Accumulation of themes as the interviews progressed
Based on this retrospective analysis of our data, our group has decided to use the 15+3 approach to assess saturation in our future qualitative research projects.
Appendix 7: Original drafts: Intervention counselling script and SMS reminder
Original enhanced counselling script
I have found a problem with your eyes. I am referring you to the outreach treatment clinic that will be held at [location] on [date] between [time] and [time]. At the clinic, eye care professionals will perform a specialist assessment and provide medicines and spectacles as required, all free of charge. Note that a small proportion of people who attend the clinic will be found to have complex eye problems that require onward referral for hospital assessment and specialist glasses. This may incur a cost. However, the vast majority of people have their needs fully met for free at the outreach triage clinic and do not require any further referral.
With treatment, you will be able to see more clearly. This will help with your work, reading, viewing screens, and many other things. It is important that you attend the clinic or your eye problem may get worse. The clinic will only be running from [day] to [day], so if you don’t manage to attend, you may not be able to get free care again in the future.”
Original intervention SMS script
We found that you had an eye problem. Please attend the outreach clinic at <> on <> between 9am-5pm to receive free medicines/spectacles If you are found to have a complex eye problem, you may be referred to a hospital for further care or specialist glasses, and this may include a fee However, the vast majority of people who attend the outreach get their eye problem fixed, for free, without the need for any further referral It’s important that you attend, as your eye problem may get worse, and you might not have a future opportunity to access free care. See you on <
Changes
We made three changes based on feedback:
One of the young men from the left behind group wanted us to change the sentence “It is important that you attend the clinic or your eye problem may get worse”, to make it ‘less scary’. We had included this original sentence based on interview feedback that we needed to better explain the importance of attending. For instance: “They should also be informed the importance because some people may end up blind”. We changed the wording to read; “It’s important that you attend to protect your vision.”
Due to inflation-imposed budget constraints, the clinics are no longer able to provide all glasses and medications for free. At this point in time it is not clear how subsidised the costs will be. We removed wording stating that the treatments will be free. We kept wording stating that the assessment will be free as this still represents a major advantage over attending a high-street provider or hospital clinic.
Representatives from Peek vision advised that “see you on <>” might be misleading, as the day of the appointment might be more than a week away (e.g. next Friday rather than this Friday). We changed this auto-populated section to “see you on <>”.
Acknowledgements
We would like to thank Paul Ringera for reviewing the study protocol and providing invaluable community perspectives and input on this project.
References
