We use mixed methods to identify drop offs in the hypertension cascade of care, to triangulate findings, and explore potential reasons and solutions for these gaps in care. Our quantitative approach includes analyzing household survey data and HSIP evaluation data to estimate retention across the care cascade. Our qualitative approach includes conducting focus group discussions with patients, providers, and health administrators and using thematic analysis to describe barriers and facilitators to hypertension care. We also conduct a scoping literature review of evidence-based approaches for improving health service delivery for hypertension and other chronic diseases. We integrate findings from this convergent mixed methods study design to recommend potential policy solutions to improve hypertension continuity of care in Tajikistan.
A. Describing the Hypertension Care Cascade
The management of hypertension involves multiple contacts between the provider and service user—screening, diagnosis, treatment, and monitoring—to achieve and maintain BP control, which is essential to reducing complications and preventing mortality. The cascade of care disaggregates these contacts and provides a useful framework for examining patient progression and identifying drop-offs, that is a discontinuity of recommended care by service users along the cascade of care (Figure 1).
We defined critical stages in patient-provider interactions from pre-diagnosis to the initial attainment of BP control and their accompanying measures:
- Diagnosis: The proportion of people with hypertension in the catchment area of interest who are diagnosed at the facility level, where a diagnosis of hypertension follows at least two measurements of BP on two or more health visits with a systolic BP of 140 mm Hg or higher or a diastolic BP of 90 mmHg or higher. See Figure 2 for an overview of clinical guidelines for hypertension management at the PHC level.
- Treatment Initiation: The proportion of people diagnosed with hypertension who initiate treatment. To measure this, we defined the proportion who had been prescribed hypertension medication with a BP result on file.
- Treatment Monitoring: The proportion of people who initiate treatment for hypertension, remain in care, and are followed-up by their service provider. A patient was considered being monitored by the care provider if there was evidence of BP being measured and recorded.
- Blood Pressure Control: The proportion of hypertension patients who have achieved BP control. We defined control as BP < 140/90 mmHg or lower in complex cases, such as diabetes co-morbidity, as specified in Figure 2.
We drew on existing data sources, including routine health information systems and household surveys conducted through the Demographic and Health Survey and a World Bank-funded Impact Evaluation. All quantitative data were available in unlinked anonymized format for analysis.
The World Bank 2018 Household Survey was conducted between April to July 2018 and measured the BP of all consenting adults over 18 years of age in households within the catchment areas of the facilities selected for investments in RBF and CQI in Sogd and Khatlon Oblasts. Eligible households either had a female member with a pregnancy in the past two years, or a family member above the age of 40 years. While these criteria imply that the survey may not have been representative of the adult population, it was the most recent objectively-measured rather than self-reported BP data within the facility catchment areas. In total, this survey provided data on 8,443 adults in Khatlon and 3,701 adults in Sogd Oblast, respectively.
The CQI database aggregated service user data from 198 facilities implementing CQI as part of the ongoing investment project, beginning in 2015. The catchment areas of these facilities included 237,501 individuals above 20 years in Khatlon Oblast and 78,518 individuals above 20 years in Sogd Oblast respectively. This patient-level database contained information on BP measurements, gender, and prescribed medication for hypertension following diagnosis.
The District Health Information System-2 (DHIS-2), the national health information system, provided population sizes for both Oblasts while the 2017 DHS included BP measurements and self-reported prescriptions for 10,718 women aged 15 to 49 years.
To estimate the number of hypertensive individuals in the CQI catchment areas, we multiplied the proportion of non-pregnant surveyed individuals with hypertension from the 2018 World Bank Household Survey (“estimated hypertension prevalence”) and the population in each catchment area. At the Oblast level, the proportion of hypertensive individuals was multiplied with the oblast population in the DHIS-2. Estimates for hypertension diagnosis, monitoring, and blood pressure control were limited to the CQI database and were unavailable for the broader population. The CQI database also documented for the registered hypertension cases the prescribed medicines, a diagnosis of diabetes, and body mass index. BP results within patient records were categorized by level of severity and by timing (ever/last three months).
By triangulating data sources, including household surveys and facility registries, we were able to estimate the proportion of hypertensive individuals who initiated care and were retained along each step of the care cascade. Graphical descriptions of the cascade of care in both Oblasts, by age group, were presented to facilitate assimilation by the target policy and practitioner audience. The quantitative data on retention along the cascade of care illustrated the magnitude of the burden of undiagnosed hypertension and the poor treatment outcomes among diagnosed cases, and motivated discussions on the causes of these gaps.
B. Understanding the Determinants of Retention in Hypertension Care Cascade
To understand the context-specific reasons for drop-offs along the hypertension care cascade, we held focus group discussions (FGDs) with service users, health providers, and health care administrators in Khatlon and Sogd Oblasts. Service users included male and female adult patients from age 18 years and above, and pregnant women with a diagnosis of hypertension. We recruited health providers at the district level and health administrators who were heads of rural health centers (RHCs) or representatives of the Oblast Health Department. In total, we included 208 participants in 18 FGDs (Table 1).
Table 1: Characteristics of focus group discussion participants
Participant Type
|
Oblast
|
CQI site type
|
Number of participants
|
Type of PHC
|
Location
|
Adult males in hypertension care
|
Khatlon
|
Intervention
|
11
|
RHC
|
Balkhi
|
Khatlon
|
Control
|
7
|
RHC
|
Jaihun
|
Sogd
|
Intervention
|
12
|
RHC
|
Jabbor Rasulov
|
Sogd
|
Control
|
11
|
RHC
|
Kanibadam
|
Total
|
|
41
|
|
|
Adult non-pregnant females in hypertension care
|
Khatlon
|
Intervention
|
11
|
RHC
|
Balkhi
|
Khatlon
|
Control
|
12
|
RHC
|
Jaihun
|
Sogd
|
Intervention
|
12
|
RHC
|
Jabbor Rasulov
|
Sogd
|
Control
|
12
|
RHC
|
Kanibadam
|
Total
|
|
47
|
|
|
Pregnant women in hypertension care
|
Khatlon
|
Intervention
|
12
|
DHC
|
Balkhi
|
Khatlon
|
Control
|
12
|
DHC
|
Jaihun
|
Sogd
|
Intervention
|
12
|
DHC
|
Jabbor Rasulov
|
Sogd
|
Control
|
12
|
DHC
|
Kanibadam
|
Total
|
|
48
|
|
|
Health care providers
|
Khatlon
|
Intervention
|
12
|
|
Balkhi
|
Khatlon
|
Control
|
12
|
|
Jaihun
|
Sogd
|
Intervention
|
12
|
|
Jabbor Rasulov
|
Sogd
|
Control
|
12
|
|
Kanibadam
|
Total
|
|
48
|
|
|
Heads of Oblast health departments
|
Khatlon
|
|
12
|
|
Bokhtar
|
Sogd
|
|
12
|
|
Khujand
|
Total
|
|
24
|
|
|
RHC: Rural Health Centre; DHC: District Health Centre
The FGD guides were developed collaboratively by technical experts within the World Bank and Tajikistan. World Bank experts in service delivery, implementation research, and clinical medicine drafted discussion guides with probes to identify the causes of drop-off and retention at each stage in the care cascade, which were reviewed and refined for wording by local experts involved in hypertension care. The discussion guides were also reviewed by local stakeholders involved in the implementation of the ongoing investment project for completeness and by other facilitators of the FGD to ensure that the translated guides were worded to be consistent with the desired meaning.
The final FGD guides were organized by stage in the care cascade and concluded with an invitation to discuss how high BP could be avoided, that is primary prevention measures. See Supplementary Files 1-3 for full guides. For patients with hypertension, the questions in the discussion guide nudged patients to reflect on the positive and negative experiences with diagnosis and treatment, and the potential reasons for patients to discontinue care. Provider and health administrator discussion guides were structured similarly, to describe the current state of hypertension care, including challenges, their causes, and suggestions for improvement. Health administrators were also asked to describe oblast-level factors relating to the care cascade, such as programmatic support for diagnosis, treatment initiation, and follow-up, and to reflect over potential interventions to address causes of drop-offs in the care cascade.
Leveraging established working relationships with RHC staff under the HSIP, the FGDs with patients were carried out in RHC facilities. Patients that met the eligibility criteria on age (18 years or older), hypertension status, and pregnancy status were recruited by RHC family doctors and nurses and invited to participate in a discussion at their RHC. Providers and administrators were contacted directly by HSIP staff and encouraged to participate in a discussion at their Oblast health department or district health center.
Each FGD was led by two to three individuals, some of whom were HSIP staff, one of whom served as a facilitator while the other served as note-taker. The purpose of the FGD was discussed, and participants provided written consent to be involved in the audio-recorded exchange. FGDs were conducted over six days in November 2018, overlapping with supervision visits by HSIP staff as part of the investment project, with each FGD lasting about two hours and including up to 12 participants.
Following the FGDs, detailed summaries were created, drawing on the audio recordings and notes from each session and translated into English for analysis. Two researchers independently conducted a thematic analysis of each summary to identify the barriers and facilitators of retention in the hypertension care cascade. Differences in themes were discussed and reconciled, and the final list of themes was reviewed with a member of the local team that led the FGDs to ensure consistency with initial findings. The final analysis of FGDs described barriers and facilitators by cascade stage and included patient-, provider-, and administrator-level perspectives, which are fully described in a recently-published report [9].
C. Identifying Fit-for-Purpose Solutions to Improve Retention in the Hypertension Care Cascade
We defined fit-for-purpose solutions as interventions that met three criteria: 1) focuses on a barrier to or facilitator of retention in hypertension care identified in the diagnostic process; 2) empirical evidence of effectiveness in the published or grey literature, and 3) perceived by local stakeholders as applicable to PHC in Tajikistan.
Based on the identified barriers to and facilitators of retention in hypertension care, the characteristics of appropriate solutions were defined as illustrated in Table 2 below:
Table 2: Characteristics of main barriers and appropriate solutions for hypertension care
Barrier
|
Solution
|
Wrong understandings of disease and its therapy among people living with the illness
|
Enable understanding of hypertension and its therapy among people with hypertension
|
The high time and monetary costs of seeking care
|
Reduce the time and monetary costs of seeking care
|
Ambiguous and inappropriate clinical guidelines, particularly for remote under-resourced areas
|
Improve the ease of understanding of clinical guidelines and adapting them to ensure their relevance to the context
|
Shortages of human resource for health and equipment for blood pressure monitoring
|
Ensure the supply of trained and equipped health workers
|
A lack of support from peers, family, providers, and the community for initiation and adherence to hypertension care
|
Incorporate support from peers, families, and communities
|
In October 2018, we conducted a review of interventions aimed at addressing the identified barriers in the hypertension care cascade. Using PubMed, we searched for English or English-translated articles published between 2000 and 2018 that evaluated an intervention’s impact on a care cascade-related metric, such as percent referred to care, percent adherent to medication, and change in BP. Search terms are displayed in Figure 3.
We screened the abstracts of resulting articles and retained those that met our inclusion criteria of evaluating a program or intervention’s measurable hypertension care cascade-related outcome. We extracted the following information into a spreadsheet template: the country the intervention took place in, the target population, care cascade focus area, type of program or policy, health service level, health personnel involved, sample size, primary and secondary outcomes, and any available cost information. Whether interventions resulted in statistically significant improvements in outcomes was also indicated as a measure of empirical effectiveness. Studies were also added to our review if conversations with local providers and administrators suggested further research into specific intervention types, including abbreviated provider care guides.
Of the 150 included studies, 14 (9.3 percent) occurred in the World Bank-defined region of East Asia and the Pacific, 22 (14.7 percent) in Europe and Central Asia, 10 (6.7 percent) in Latin America and the Caribbean, 2 (1.3 percent) in the Middle East and North Africa, 67 (45.3 percent) in North America, 10 (6.7 percent) in South Asia, and 21 (14 percent) in Sub-Saharan Africa. A third of the studies (n = 49) focused on community-based interventions across all hypertension cascade stages, while two-thirds (n = 108) of studies focused on primary or secondary care interventions during the treatment compliance stage alone. In some cases, the studies targeted more than one health service level or cascade stage.
Our literature review aimed to scope the range of clinical and non-clinical interventions that addressed the identified barriers to retention in the hypertension care cascade and aligned with the characteristics of appropriate solutions to these barriers. The literature review served as a practical tool in guiding discussions on possible policies or programs for hypertension care in the Tajikistan context. Thus, interventions found to be effective in other contexts were summarized by intervention scope, setting, and personnel, and presented to policymakers at the MoHSP and during FGDs with health administrators, providers, and patients, as illustrated in Figure 4. During this discussion, the initial list of proposed interventions was narrowed to a final set of fit-for-purpose solutions based on stakeholder feedback on the perceived feasibility of implementing the interventions in the Tajikistan context.