Article Text

Download PDFPDF

Original research
Qualitative study of pathways to care among adults with diabetes in rural Guatemala
  1. Anita Nandkumar Chary1,2,
  2. Meghna Nandi2,3,
  3. David Flood2,4,
  4. Scott Tschida2,
  5. Katharine Wilcox2,5,
  6. Sophie Kurschner2,6,
  7. Pablo Garcia2,7,
  8. Peter Rohloff2,8
  1. 1Medicine & Emergency Medicine, Baylor College of Medicine, Houston, Texas, USA
  2. 2Center for Research on Indigenous Health, Maya Health Alliance Wuqu' Kawoq, Tecpan, Guatemala
  3. 3Family Medicine, Boston Medical Center, Boston, Massachusetts, USA
  4. 4Medicine, University of Michigan, Ann Arbor, Michigan, USA
  5. 5Family Medicine, University of Illinois Medical Center at Chicago, Chicago, Illinois, USA
  6. 6School of Medicine and Health Sciences, The George Washington University, Washington, District of Columbia, USA
  7. 7Nephrology, Stanford University School of Medicine, Stanford, California, USA
  8. 8Global Health Equity, Brigham and Women's Hospital, Boston, Massachusetts, USA
  1. Correspondence to Dr Anita Nandkumar Chary; anita.chary{at}gmail.com

Abstract

Objective The burden of diabetes mellitus is increasing in low-income and middle-income countries (LMICs). Few studies have explored pathways to care among individuals with diabetes in LMICs. This study evaluates care trajectories among adults with diabetes in rural Guatemala.

Design A qualitative investigation was conducted as part of a population-based study assessing incidence and risk factors for chronic kidney disease in two rural sites in Guatemala. A random sample of 807 individuals had haemoglobin A1c (HbA1c) screening for diabetes in both sites. Based on results from the first 6 months of the population study, semistructured interviews were performed with 29 adults found to have an HbA1c≥6.5% and who reported a previous diagnosis of diabetes. Interviews explored pathways to and experiences of diabetes care. Detailed interview notes were coded using NVivo and used to construct diagrams depicting each participant’s pathway to care and use of distinct healthcare sectors.

Results Participants experienced fragmented care across multiple health sectors (97%), including government, private and non-governmental sectors. The majority of participants sought care with multiple providers for diabetes (90%), at times simultaneously and at times sequentially, and did not have longitudinal continuity of care with a single provider. Many participants experienced financial burden from out-of-pocket costs associated with diabetes care (66%) despite availability of free government sector care. Participants perceived government diabetes care as low-quality due to resource limitations and poor communication with providers, leading some to seek care in other health sectors.

Conclusions This study highlights the fragmented, discontinuous nature of diabetes care in Guatemala across public, private and non-governmental health sectors. Strategies to improve diabetes care access in Guatemala and other LMICs should be multisectorial and occur through strengthened government primary care and innovative private and non-governmental organisation care models.

  • general diabetes
  • anthropology
  • qualitative research
  • primary care
  • public health
  • international health services

Data availability statement

Data are available in a public, open access repository. The interview guide, codebook, coded data, and care maps are available through Dataverse at: https://doi.org/10.7910/DVN/3EO2PD.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • This study demonstrates the utility of care mapping in studying sources of healthcare and care-seeking trajectories for non-communicable diseases.

  • One limitation is data included retrospective self-reported health-seeking behaviours over several years and are subsequently subject to recall and reporting bias.

  • Another limitation of this study is by recruiting participants through a population study evaluating for elevated haemoglobin A1c, individuals with uncontrolled diabetes mellitus are over-represented in our sample.

  • Our study was conducted in majority indigenous populations living in rural areas and may not be generalisable to other areas of Guatemala.

  • This study struggled to recruit equal proportions of men and women.

Introduction

Low-income and middle-income countries (LMICs) face an increasing prevalence of diabetes1 with limited health system resources for disease prevention, screening and management.2–4 Demographic investigations about non-communicable diseases (NCDs) in LMICs have revealed a high degree of unmet need and loss to follow-up in diabetes care.5–9 While valuable, these quantitative survey-based data do not provide insights into sources of care or longitudinal care trajectories of individuals with diabetes in LMICs. This information is particularly important as LMICs face increasing healthcare privatisation and erosion of public health systems.10 11 Individuals with diabetes increasingly use private and non-governmental health resources.3 12–19 Understanding individuals’ diabetes care pathways and decision-making in LMICs can inform health system improvements to address NCDs.

Care mapping is an interview-based process used to reconstruct participants’ pathways to care for a particular illness. The technique was historically developed by the World Health Organization to characterise sources of mental healthcare and referral patterns in LMICs.20 Care mapping is similar to but distinct from patient journey mapping, a quality improvement method that focuses on processes of obtaining care from the patient perspective.21 We coupled care mapping with qualitative interviews to study sources of and healthcare-seeking decisions surrounding diabetes care in rural Guatemala.

Guatemala is a Central American country with an estimated diabetes prevalence of 8.8%.22 Guatemala also has had an over 50% increase in diabetes-related morbidity and mortality over the past decade.23 Guatemala has a mixed health system offering diabetes and NCD care through multiple sectors. The Ministry of Health (MOH) sponsors a network of public hospitals, health centres and posts, which are intended to provide citizens with free care, but face significant resource limitations and medication stockouts leading to out-of-pocket spending.24 The Social Security System provides healthcare to the approximately 18% of the population with formal employment.24 Given limited access to government and social security care, healthcare seekers are increasingly turning to the growing private for-profit and non-governmental healthcare sectors.25 Guatemala also has a rich historical tradition of ethnomedical healing, based largely around herbal remedies, as well as an expanding market of naturopathic products, both of which are used for diabetes care.25 26 Notably, 45% of the Guatemalan population identifies as indigenous Maya.27 Due to longstanding historical inequalities and more recent civil war (1960s–1990s), the Maya population faces structural, linguistic and cultural barriers to biomedical healthcare.25 Indigenous Maya people also experience obstacles to diabetes care, including medication cost, perceived mistreatment, and limited social supports for lifestyle changes.26 28 29

This qualitative study explores pathways to care among individuals with diabetes in rural Guatemala. We aim to offer insights to improve diabetes care in LMICs. Our findings offer a unique contribution to the literature about trajectories of NCD care in LMICs.

Methods

Sampling and recruitment

This qualitative study was a planned substudy conducted as part of a population-based study assessing incidence and risk factors for chronic kidney disease in two rural majority-indigenous Guatemalan sites. The two sites were selected due to their distinct cultural, agricultural and health resource profiles. The first site is located in the highlands in close proximity to a large highway and three cities with major hospitals; it has a majority indigenous population, which cultivates maize, beans and vegetables. The second site is on the southwestern coast, where agricultural production focuses on sugar cane; the population identifies as both indigenous and of mixed descent, and hospital access requires farther travel.

The population study used a mapping technique to randomly sample households in each site, fully detailed elsewhere.30 Briefly, from June 2018 to February 2019, of 1281 eligible individuals recruited from 533 households, 807 non-pregnant adults agreed to participate in the population study. During the initial population study visit, participants received screening blood tests for diabetes as part of an assessment for risk factors for chronic kidney disease. We defined a diagnosis of diabetes as a haemoglobin A1c (HbA1c)≥6.5% or a prior self-reported diagnosis of diabetes. During a follow-up study visit, population study staff informed subjects of their HbA1c result and facilitated referrals to local health centres, as needed. Population study staff then referred subjects with diabetes interested in participating in an interview to qualitative study team members. Inclusion criteria for this qualitative study were: adult (age 18 years+), non-pregnant, HbA1c≥6.5% or a prior self-reported diagnosis of diabetes, Spanish-speaking or Kaqchikel Mayan-speaking, and participation in the population study. Exclusion criteria were: paediatric patients (age<18 years), pregnant, no laboratory or self-reported diagnosis of diabetes, primary speaker of a language other than Spanish or Kaqchikel Mayan, not a participant in the population study.

From November 2018 to June 2019, the qualitative study team approached all 42 participants who enrolled in the population study between June 2018 and February 2019 who had an HbA1c≥6.5% and reported a previous diagnosis of diabetes. The research team aimed to interview approximately 12 participants from each study site, given literature indicating that this sample size typically captures the range of experiences and leads to saturation of qualitative data.31 Participants were approached in sequential order of receipt of laboratory results in both study sites. A total of 29 agreed to the interview, with 13 in the first site and 16 in the second site. Ten individuals declined to participate, and three were travelling at the time of the proposed interview.

Data collection

Trained interviewers conducted semistructured interviews with 29 participants. A single interview was performed with each participant. Interviews explored participants’ experiences of seeking diabetes care and their pathways to diabetes care. Interviewee demographics are detailed in table 1. Interviews occurred approximately 4–6 months after participants’ initial enrolment in the population study. Interviews lasted 20–60 min and occurred in participants’ homes. Depending on participant preference, interviews were conducted in either Spanish or Kaqchikel Maya, with the support of an experienced Kaqchikel interpreter. Based on participants’ preferences, interviews were not digitally recorded. Instead, in cases where one interviewer was present, the interviewer typed detailed notes and directly transcribed short quotes during the interview. In cases where two interviewers were present, the second researcher transcribed responses and quotes in real time. All notes and transcribed quotes were reviewed after the interview for accuracy prior to analysis.

Table 1

Sample demographics

Data analysis

Researchers reviewed notes from the first 12 interviews and developed a codebook using an inductive strategy, in which codes arise from interview data rather than from a priori hypotheses. All interviews were coded by two study team members, and a third team member reviewed coding for discrepancies. Discrepancies in coding were resolved and themes were elucidated by consensus. After the first 20 interviews, no new codes or themes arose, indicating saturation. This was confirmed with formal analysis and review of the other nine interviews. Qualitative analysis was facilitated by the software NVivo Pro, V.11.4.1 for Windows.

The research team collectively reviewed interview notes to generate for each participant a care map, or a digital flowchart depicting each individual’s pathway through sources of diabetes care. Care maps were created to allow visual representation of each individual’s use of distinct healthcare resources outlined in table 2. As described in the care map legend, arrows demarcate sequential and simultaneous use of health resources, and boxes with stars in the lower righthand corner indicate the health sector a participant was using at the time of the interview. A composite care map was generated for each site aggregating individuals’ pathways to care, with each care pathway noted by a different colour. This article features composite care maps separated by site and with individual pathways outlined through four steps of care, for ease of illustration given the visual complexity of care trajectories otherwise. After all care maps were created, two coders independently reviewed each care map to determine the proportion of participants who used each type of health resource and the proportion of participants who used different types of health resources simultaneously.

Table 2

Descriptions of healthcare sectors

Researcher characteristics

In qualitative research, reflexivity refers to consideration of how researchers’ identities and values influence the research process and findings. Our study team included members from both Guatemala and the USA, all of whom have previously studied perceptions of chronic disease in Guatemala. We sought to reduce the impact our identities could have on the research in two ways: (1) interviewers had no involvement in collecting biological samples or disclosing laboratory results; (2) all interviews were conducted in teams with one US researcher and one indigenous researcher or interpreter.

Institutional context and ethics

We follow the Standards for Reporting Qualitative Research guidelines to describe the present investigation.32 The interview guide, codebook, coded data and care maps are available through Dataverse.33

Patient and public involvement statement

Patients and the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.

Results

Four distinct themes emerged regarding pathways to diabetes care and participants’ decision-making: (1) fragmentation in care; (2) the role of pharmacies; (3) perceptions of quality of care; and (4) cost. Pervasive across all themes were participants’ sentiments of challenges in accessing care as well as gaps between participant experience and healthcare provider recommendations.

Fragmented pathways to care

Participants’ distinct pathways to care highlight fragmented diabetes care. The majority of participants initiated diabetes care-seeking due to perceived symptoms of diabetes or acute illness (97%). Only one participant initiated care after a diabetes diagnosis through routine asymptomatic screening at a regular doctor’s visit. The time over which participants reported seeking diabetes care ranged from 3 months to 23 years.

Figure 1, a composite diagram of pathways to care from all participants from the first study site through their fourth step of care, demonstrates that individuals seek diabetes care in multiple health sectors. These include government, private and non-governmental sector biomedical care, as well as ethnomedical therapies and complementary and alternative medicine. Importantly, figure 1 demonstrates that individuals with diabetes do not follow a singular path to care. Rather, the majority of participants in both sites (site 1, 92% and site 2, 100%) accessed care in multiple health sectors. Proportions of participants who sought care in each sector are outlined in table 3. Sectors participants were using at the time of interview are specified in table 4.

Figure 1

Composite care map—site 1. CAM, complementary and alternative medicine.

Table 3

Trends in diabetes care-seeking

Table 4

Sector of diabetes care at time of interview*

Participants’ care pathways also demonstrate discontinuous care. The majority of participants saw multiple providers (90%). As further described below, participants stopped care or changed providers due to cost and lack of economic resources. Even when participants sought care at the same clinical location, such as a government health centre, discontinuity in care resulted from changing healthcare staff and rotating trainees.

The majority of participants sought care with different providers and/or in different health sectors simultaneously, rather than in a linear sequential fashion (79%). For instance, participants used natural treatments or commercialised alternative products concurrently with biomedical medications. They did not perceive the two treatment modalities as mutually exclusive. The care pathway in figure 2 reflects simultaneous care-seeking. Simultaneous care-seeking trends are quantified in table 3.

Figure 2

Participant care map, ID 31. This care map demonstrates simultaneous care-seeking by a woman in her 60s from site 1 over a period of 3.5 years. CAM, complementary and alternative medicine.

Roles of pharmacy

Pharmacies played an important and unique role in diabetes care. The majority of participants (69%) used a pharmacy for consultation or out-of-pocket medication purchasing at some point in their diabetes care trajectory. A significant minority of participants (24%) used the pharmacy to continue purchasing medications even when they were no longer regularly following with a provider. For instance, one participant explained:

Only when I feel very bad do I go to the hospital. Other than that, I just buy my pills. (ID 1112, female in 50s)

Two participants reported that the doctors they saw encouraged participants to stop seeking regular care with them, continue with medications as initially prescribed, and come back only if they had symptoms. For example:

The doctor only told me, ‘When you feel bad, come here,’ but I don’t have another appointment. That [feeling symptoms] is not going to happen if I follow the diet and the medications. I only take the same prescription to the pharmacy for more medication. (ID 1071, female in 40s)

While all participants who used the pharmacy did so to buy medication, 24% also used the pharmacy as a source of medical care, for example, to check their blood sugar when they perceived diabetes symptoms and receive intravenous fluids or medications to manage those symptoms. Figure 3 shows how one participant used the pharmacy as her only source of medical care in this way.

Figure 3

Participant care map, ID 237. A woman in her 50s from site 1 used the pharmacy as a regular source of medical care over the course of 3 years of diabetes care-seeking. CAM, complementary and alternative medicine.

Perceived low-quality public sector care

The majority of participants sought care from a government facility (MOH or Social Security) at some point in their care trajectory (62%). Many participants (59%) sought free care in MOH and Social Security clinics or hospitals as their first or second stop for diabetes care in the healthcare system. However, the MOH lacks resources to consistently provide free medications and laboratory exams and hosts rotating trainees in facilities, which led to perceptions of low quality among 21% of participants:

You can’t trust the health centers or the hospital because they are very neglected… Before, we went to the hospital with confidence. The hospitals served us. Now, they say to you, ‘Look, you have to buy this, you have to do these exams.’ (ID 1273, male in 50s)

In the health center, there are only trainees, no older doctors, therefore they don’t know anything. They don’t have experience. (ID 305, female in 60s)

Participants’ perceptions of low-quality MOH care were also based in experiences of poor communication with government providers (21%). Some participants noted feeling ignored, talked down to or scolded and judged by both MOH and Social Security providers:

The doctor doesn’t ask anything, he doesn’t ask if I feel well or if I have any issues. He only gives me the prescription. He doesn’t ask anything. So, it’s not good in [the social security clinic]. (ID 1307, female in 40s)

I did not want to go to the hospital because I was afraid to hear what they would say. Sometimes I go to the [MOH] health center and they scold you, they scold you and they judge you. (ID 451, female in 40s)

Perceptions of low quality deterred three participants from seeking care in their local MOH health centres or led them to the private sector:

I don’t go to the health center because we think that they don’t give the services we need or have the specialists. So you have to go to the other doctors directly… The health center sends you out, to a hospital, a national hospital or private hospital. (ID 460, male in 50s)

I went to a private doctor, because the health center and the hospital are very neglected. (ID 1273, male in 50s)

Despite inability to afford private sector care, one participant refused to go to his local MOH health centre due to perceived poor-quality care (ID 1266, male in 40s).

Participants occasionally (10%) referred to positive community reputations driving their care-seeking decisions. These reputations were not associated with government facilities but rather with private doctors, clinical facilities in larger cities and specialists:

I felt very desperate, so I left for Xela, because there are good doctors there. (ID 1112, female in 50s)

Cost

Cost significantly influenced participants’ care-seeking. The majority of participants described a financial burden associated with diabetes treatment (66%). Some participants took loans to finance their diabetes care (21%) or relied on family members to pay for treatment costs (21%). Some participants delayed seeking care (14%) or went without medications for periods of time due to lack of economic resources (14%).

Well, there are times that we have to stop taking [medications] because we don’t have the money. But, we have not taken out a loan, because loans don’t go away. It is better to bear it. (ID 1246, female in 50s)

Participants described two major strategies to minimise care costs. First, participants sought free or cheaper sources of care (41%), including through NGOs, medical missions or pharmacies. Second, participants continued to take previously prescribed medications—purchased from the pharmacy, as described above or previously obtained—without regular follow-up care (17%).

I didn’t go to the [community provider] again because of the cost…I didn’t go anymore, because I had the pills. (ID 451, female in 40s)

The care pathway in figure 4 demonstrates how cost, strategies to minimise expenses and perceptions of the government MOH system can impact diabetes care-seeking. This participant initially had Social Security coverage through his employer, but then lost his job and associated coverage. He had no money or insurance to pursue consistent diabetes care. While he desired to see a diabetes care provider, he refused to visit a MOH clinic due to perceived poor-quality care. He instead continued to buy metformin at a pharmacy and intermittently used plant-based treatments. However, on facing significant diabetes complications, including foot infection leading to amputation, he felt forced to pursue urgent care at the public MOH hospital (ID 1266, male in 40s).

Figure 4

Participant care map, ID 1266. A man in his 40s from site 2 sought diabetes care over 12 years. CAM, complementary and alternative medicine.

Cost sometimes trumped participants’ care-seeking preferences. For instance, three participants who expressed dissatisfaction with MOH care continued seeing MOH providers due to care being free. Others sought out perceived high-quality care in the private sector but discontinued it due to associated costs. For instance, one participant hoped to continue with a ‘magnificent’ doctor in a nearby city but could not afford the approximately US$25 fee per appointment (ID 116, male in 50s).

Table 5 provides additional representative quotes for each theme.

Table 5

Representative quotes

Discussion

This qualitative study of diabetes care-seeking in rural Guatemala provides important insights into the fragmented nature of NCD care. It also highlights the roles of cost and perceptions of quality in care-seeking decisions. Together, these findings have major implications for diabetes interventions in Guatemala and other LMICs.

First, care mapping illuminated the fragmented nature of diabetes care, characterised by limited provider continuity, utilisation of pharmacies for clinical care and acute episodic care-seeking for diabetes complications. Participants simultaneously and sequentially used biomedicine, ethnomedicine and commercialised complementary and alternative medicine products for diabetes care. Gaps between biomedical providers’ and participants’ understandings of diabetes management were common, as were participants’ perceptions of poor quality of biomedical care. These features all make participants more vulnerable to diabetes-related morbidity and mortality. Indeed, diabetes care continuity is associated with HbA1c reduction and lower rates of hospitalisation and mortality, as demonstrated in high-income settings.34–38

Other research exploring diabetes care-seeking in Guatemala has also highlighted fragmentation in care pathways. Specifically, qualitative and ethnographic studies in both this study’s highland site and periurban indigenous areas have revealed phenomena of people with diabetes cycling between biomedical and ethnomedical treatments as well as between public and private sector biomedical facilities and pharmacies. Such cycling usually occurs in relation to economic circumstances, acute experiences of diabetes related symptoms and perceptions that private biomedical facilities offer a higher quality of care than public facilities.26 39 In Guatemala, care fragmentation reflects increased healthcare privatisation,25 unequal distribution of healthcare workers and resources in urban and rural areas24 40 and frequent changes in government leadership leading to staff turnover in local health clinics40—challenges faced in other LMICs, as well.10 41 Indeed, investigations of diabetes care in India, and sub-Saharan Africa similarly reveal fragmented trajectories across different health sectors,12 42 43 with initiation of care-seeking following development of symptoms rather than after routine asymptomatic screening.42 44

Strengthening government primary care in LMICs will require longitudinal resource investment and represents a long-term strategy to enhance access to and continuity of diabetes care. Care models that optimise resources and shift tasks to non-physicians are feasible in LMICs, and, whether they occur through public health systems, non-governmental organisations or in multisector partnerships, can improve longitudinal and continuous diabetes care while reducing costs.45 Nurse-led interventions have shown promising results in improving diabetes care access and outcome measures in Guatemala17 and sub-Saharan Africa.19 46 47 Peer-based interventions, in which community non-professionals are trained to provide continued support and promote diabetes self-management, have improved health outcomes in sub-Saharan Africa and Asia19 and have improved care engagement when coupled with mobile health messaging in Honduras and Cambodia.16 48 Community healthcare workers (CHWs) also represent an important resource. In one study in Guatemala, a diabetes programme led by CHWs, who were equipped with a smartphone application to aid in clinical decision-making and supervised by a remote physician, showed improvements in HbA1c.49 Notably, however, a study based in the USA found that individuals with diabetes perceived programmes in which CHWs provided the main intervention as settling for lower-quality care.50 Programme evaluations in LMICs should further explore the social and cultural acceptability of CHWs’ roles in diabetes care.

Second, our study highlights cost to patients as a major barrier to diabetes care, as similarly demonstrated in sub-Saharan Africa and Asia.12 42 44 51 While increasing availability of free public healthcare is an important goal, the expanding private health sector is assuming an increasing role in primary care delivery globally.10 As such, innovative fee-for-service models should be considered as means for improving diabetes care access and continuity in Guatemala and other LMICs. Value-based care, or using outcome and cost data to improve care delivery through performance-associated payment,52 can be leveraged to improve health outcomes while reducing costs for clients. For example, the Clínicas del Azúcar model in Mexico rewards low-income clients for lifestyle modifications and improved health outcomes with reduced membership fees.53 Non-governmental organisations can also play roles in lowering costs through preferential purchasing of generic medications and price negotiations with local suppliers, as described in a study from Guatemala.17

Finally, perceptions of low-quality care related to government facilities’ resource limitations and providers’ communication styles drove participants in our study to change diabetes providers, further affecting care access and continuity. Medication stock-outs and mistreatment by public sector providers have similarly been documented as sources of dissatisfaction with diabetes care elsewhere in Guatemala29 and in sub-Saharan Africa42 and India.12 In global health literature on care quality, these perceptions fall under the domains of effectiveness, equity and person-centredness of care.54 Improvements in effectiveness and equity must occur through investment and resource management changes to make diabetes medications, providers and specialists more readily and consistently available. Person-centredness of diabetes care may be improved through provider education in motivational interviewing, a communication technique intended to elicit an individual’s values and intrinsic desires to change behaviour, which has been associated with improved diabetes outcomes in a variety of settings.55 As studies about diabetes care quality in LMICs have largely focused on poor care alignment of clinical care with recommended evidence-based standards,56 57 patient perceptions of health service quality represent an important direction for future research.

Limitations

Our study faces four limitations. First, pathways to care are based on retrospective self-reported health-seeking behaviours over the course of several years, which are subject to recall and reporting bias. Due to the availability of the qualitative research team and the logistics involved in performing home visits in rural areas, data were collected from each participant in a single interview, rather than longitudinally over repeated interviews. Participants were not always able to provide an exact time course of health-seeking behaviours and may have omitted health resources. Second, by recruiting participants through a population study using elevated HbA1c to evaluate for diabetes as a risk factor for chronic kidney disease, individuals with uncontrolled diabetes mellitus are over-represented in the qualitative study sample. Care trajectories are likely different for individuals who are successfully engaged in treatment with better-controlled disease. Third, our results reflect experiences of a majority indigenous population living in rural areas and may not be generalisable to other areas of Guatemala, particularly given the local strength of ethnomedicine in rural areas as well as unique political–economic, cultural and linguistic barriers to biomedical care faced by indigenous Maya people.25 Fourth, the majority of our participants are women, which is likely related to recruitment through daytime home visits during men’s work hours, and reflects known difficulties of recruiting men into NCD activities in the region.58

Conclusion

Adults with diabetes in Guatemala experienced fragmented diabetes care and faced costs and perceived low-quality care as barriers to diabetes care. Strengthening the national primary healthcare system while innovating in private and non-governmental sectors can help improve diabetes care continuity. This study offers important lessons about improvements needed in diabetes care in Guatemala and other LMICs.

Data availability statement

Data are available in a public, open access repository. The interview guide, codebook, coded data, and care maps are available through Dataverse at: https://doi.org/10.7910/DVN/3EO2PD.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants. This study involves human participants and was approved by the Institutional Review Boards of Partners Healthcare, Boston (2017P002476); the Institute for Nutrition of Central American and Panama, Guatemala (CIE-REV 075/2018); and Wuqu’ Kawoq | Maya Health Alliance, Guatemala (WK-2018-001). Participants gave informed consent to participate in the study before taking part.

Acknowledgments

We thank Wuqu’ Kawoq | Maya Health Alliance staff Eva Leticia Tuiz Ordoñez, Eloin Dhaenens, Glenda Hernandez, Hector Gomez, Irma Yolanda Raquec Teleguario and Alvaro Sir for their assistance in facilitating recruitment of participants from the larger quantitative study, coordinating logistical details around interviews and accompanying qualitative study team members to the interviews. We thank Magda Sotx Muz for providing interpreter services.

References

Footnotes

  • Twitter @anitachary

  • Contributors ANC conceptualised the study design, oversaw data collection and analysis and drafted the manuscript. MN was involved with study design, data collection and analysis and drafted the manuscript. ST KW and SK were involved with study design, data collection and analysis and provided critical revisions of the manuscript. DF and PG were involved with study design, data analysis and critical revision of the manuscript. PR conceptualised the study design, was involved with data analysis and critically revised the manuscript. ANC acts as guarantor for this work.

  • Funding Research was funded by the National Institutes of Health/Fogarty International Center (1R21TW010831-01). MN received support from the Watson Institute for International and Public Affairs Field Research Grant and the Framework in Global Health Program Summer Scholarship at Brown University.

  • Competing interests PR is the principal investigator to receive the National Institutes of Health/Fogarty International Center grant (1R21TW010831-01) that funded this study. ANC receives support from the Houston Veterans Administration Health Services Research and Development Center for Innovations in Quality, Effectiveness and Safety (CIN13-413).

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.