Abstract
Background:
Incorporating social determinants of health into care delivery for chronic diseases is a priority.
Objectives:
Evaluate the impact of group medical visits and/or microfinance on blood pressure reduction.
Methods:
We conducted a cluster randomized trial with four arms and 24 clusters: 1) usual care (UC); 2) usual care plus microfinance (MF); 3) group medical visits (GMV); and 4) GMV integrated into MF (GMV-MF). The primary outcome was one-year change in systolic blood pressure (SBP). Mixed-effects intention-to-treat models were used to evaluate the outcomes.
Results:
2890 individuals (69.9% women) were enrolled (708 UC, 709 MF, 740 GMV, and 733 GMV-MF). Average baseline SBP was 157.5 mmHg. Mean SBP declined −11.4, −14.8, −14.7, and −16.4 mmHg in UC, MF, GMV, and GMV-MF, respectively. Adjusted estimates and multiplicity-adjusted 98.3% confidence intervals showed that, relative to UC, SBP reduction was 3.9 mmHg (−8.5,0.7), 3.3 mmHg (−7.8, 1.2), and 2.3 mmHg (−7.0, 2.4) greater in GMV-MF, GMV, and MF, respectively. GMV and GMV-MF tended to benefit women, and MF and GMV-MF tended to benefit poorer individuals. Active participation in GMV-MF was associated with greater benefit.
Conclusions:
A strategy combining GMV and MF for individuals with diabetes or hypertension in Kenya led to clinically meaningful SBP reductions associated with cardiovascular benefit. Although the significance threshold was not met in hypothesis testing, confidence intervals for GMV-MF were consistent with impact ranging from substantive benefit to neutral effect. Incorporating social determinants of health into care delivery for chronic diseases has potential to improve outcomes.
Keywords: hypertension, diabetes, social determinants of health, group medical visits, microfinance, Kenya
Condensed Abstract:
Incorporating social determinants of health into care delivery for chronic diseases is an urgent priority. We conducted a cluster randomized trial with 2890 individuals (69.9% women) to evaluate the impact of group medical visits (GMV) and microfinance (MF) on blood pressure reduction. A strategy combining GMV and MF for treating diabetes or hypertension led to clinically meaningful SBP reductions associated with beneficial cardiovascular impact. Although the significance threshold was not met in hypothesis testing, confidence intervals for GMV-MF were consistent with impact ranging from substantive benefit to neutral effect. Incorporating social determinants of health into care delivery for chronic diseases can improve outcomes, especially with active participation and for subgroups of patients.
Introduction
Cardiovascular disease (CVD) is the leading cause of mortality worldwide, with a disproportionate burden in low- and middle-income countries (LMICs) (1). Diabetes and hypertension are major risk factors for CVD (2,3). However, treatment and control rates remain suboptimal worldwide (4). Patients’ financial challenges remain a major barrier to access to care (5), and the economic burden of hypertension and diabetes is staggering (6,7). Out-of-pocket expenditures are high in many LMICs (8), and affordability of care is a significant barrier especially in the setting of inadequate health insurance coverage (9). Patients also confront costs due to transportation (10), distance to health facilities (11), and time lost from work (12). Indeed, economic stability is a major social determinant of health, in addition to social and community context, neighborhood and built environment, education access and quality, and health care access and quality (13).
Thus, innovative management approaches are urgently needed to incorporate social determinants of health into care delivery, in order to simultaneously address socio-economic and health issues (14,15). Two strategies that show promise in LMICs include microfinance (MF) and group medical visits (GMV). MF activities are commonly implemented in low-resource communities to provide access to saving mechanisms and loan opportunities (16). Studies in LMICs have shown clinical, social, and economic benefits of integrating MF and health care in various disease states, although the impact on blood pressure and CVD risk reduction is not well established. Analogously, GMV is defined as a clinical encounter whereby the clinician meets with and provides medical services to a collective group of patients who share similar conditions (17). GMV can improve social cohesion, social support, and trust, and has demonstrated improved outcomes for hypertension and diabetes (17).
When GMV are integrated into MF groups, there is the additional potential to improve health care access and quality, on top of the other social determinants of health. We have previously reported promising preliminary results from a pilot uncontrolled study that integrated GMV and MF, with improvements in systolic blood pressure (SBP) (18). We report here the results of the BIGPIC Study, a cluster randomized controlled trial that aimed to evaluate the impact of MF and GMV on SBP and cardiovascular risk reduction among individuals with diabetes or hypertension (19).
Methods
Setting
Full details of the study design and methods have previously been described (19). Briefly, the Academic Model Providing Access to Healthcare Partnership (AMPATH) was initiated in Kenya in 2001 (20). In partnership with the Government of Kenya, AMPATH’s clinical scope of work includes CVD, hypertension, and diabetes (21,22). AMPATH has developed and implemented a multicomponent facility-based chronic disease management program that has enrolled over 40,000 patients. The multicomponent package includes task redistribution (23), clinical decision support using health information technology (24), consistent and secure medication supply (25), linkage and retention activities (26), community stakeholder engagement (27), and social support. The current study built upon this existing infrastructure in western Kenya. The protocol was approved by the institutional review boards of all participating institutions.
Study Design
After an initial formative phase of qualitative inquiry (28), we used a participatory, iterative design process to develop the GMV and MF interventions (29). We conducted a cluster randomized trial with four arms: 1) AMPATH’s multicomponent facility-based care (usual care, or UC); 2) usual care plus microfinance groups (MF); 3) group medical visits (GMV); and 4) GMV integrated into MF groups (GMV-MF). The unit of randomization was health facility catchment area (cluster), stratified by county, type of health facility, and level of pre-existing MF participation. There were a total of 24 clusters (six per arm). The randomization process was conducted centrally by biostatisticians at Brown University. The participants and research staff could not practicably be blinded to intervention assignments.
Participants
Inclusion criteria for this study included all adults ≥ 35 years of age in the AMPATH chronic disease management program who had diabetes (fasting glucose ≥ 7 mmol/L), impaired fasting glucose (fasting glucose 5.6 – 6.9 mmol/L), or increased risk of developing diabetes (elevated Leicester Risk Assessment score, ≥ 7, scoring details in Appendix) (30). All participants had either diabetes or hypertension (SBP ≥ 140 mmHg or diastolic blood pressure (DBP) ≥ 90 mmHg). Study personnel contacted potential participants via one of the following “recruitment pathways”: newly screened, previously screened but not linked to care, active in care less than six months, previously in care but no clinic visit in past six months, and currently in private care but wishing to transfer to the public sector care system. Exclusion criteria included: acute illness requiring immediate medical attention, terminal illness, and inability to provide informed consent. Individuals who were pregnant or had HIV were also excluded, as they were automatically referred to a higher level of care within the AMPATH care system.
Study Procedures
In the UC arm, participants received AMPATH’s multicomponent chronic disease management care package. The clinical protocol aligned with contemporaneous relevant global and national guidelines (31,32), using drugs contained in the Kenyan national formulary, and included both pharmacotherapy and non-pharmacological interventions (Appendix). There was no charge for the clinical encounter. Standard fees for medications were negotiated between AMPATH and the Kenyan Ministry of Health, and did not vary across the trial arms, so that price differences would not bias study results. At the time of the study, the Kenyan national health insurance plan did not cover outpatient chronic disease medications.
In the MF arm, participants received usual multicomponent clinical care, and were encouraged to create MF groups that met monthly and were organized and supported by AMPATH. AMPATH’s MF program involves the creation of community savings groups, wherein MF group members mobilize and manage their own savings, provide interest-bearing loans to group members, offer a limited form of financial insurance, and contribute to a social fund used for emergency or welfare issues of group members (18). No external funds are provided to the MF groups. All MF group participants received loans from the combined group savings.
In the GMV arm, participants were invited to join a group that met monthly with a community health worker and clinician (either physician or clinical officer) at a location convenient to the community health worker, clinician, and participants. The clinical care package was identical to the multicomponent AMPATH chronic disease management package, but the delivery was in the form of a GMV rather than individual clinician-patient encounter in the health facility. Each GMV began with the measurement of resting electronic blood pressure (BP) (all participants) and blood glucose (diabetes patients), the ascertainment of medication regimen for each participant, and extent of medication adherence. Subsequently, the community health worker facilitated a group discussion about a self-care or health education topic chosen by the group. The community health worker had received training in motivational interviewing, chronic disease self-management techniques, and group-based peer support, and encouraged group members to engage in mutual problem-solving and sharing of ideas. During this time, the clinician reviewed the BP, glucose, and adherence data to determine a clinical management plan for each individual. After the group discussion ended, each participant had a brief individual consultation with the clinician during which their individual management plan was discussed and finalized. Each GMV lasted for approximately two hours.
In the GMV-MF arm, the monthly GMV was integrated into the MF groups, wherein each group meeting consisted of an initial MF portion, followed by the GMV. Thus, participants received their clinical care in a GMV and also participated in MF, as described above.
Individuals were free to partake in any of the group-based experiences (either GMV or MF), without any impact on their ability to access clinical care either through the AMPATH chronic disease management program or a different health care provider.
Outcomes
The primary outcome measure was one-year absolute mean change in SBP, measured by trained study staff using electronic BP machines and standardized procedures (33). Key secondary outcome measures included change in DBP and change in 10-year CVD event risk as measured by the QRISK3 score (34,35). Other secondary outcomes included BP control (SBP < 140 and DBP < 90), change in total and LDL cholesterol, change in the international wealth index (asset-based index of a household’s material well-being used in LMICs) (36), and change in livestock ownership. Pre-specified subgroup analysis included assessing the change in the primary outcome (SBP) in the following groups: sex, age, baseline health insurance status, baseline monthly earnings, baseline wealth category, baseline livestock ownership, baseline land ownership, and recruitment pathway.
Statistical Analysis
Data for each participant were collected at baseline, three and 12 months. We report mean values and standard deviations for continuous measurements, and frequencies and percentages for categorical. We fit linear mixed-effects models to estimate the effect of the intervention on primary and secondary outcomes, using random effects to account for clustering of individuals within health facility catchment area. Our power calculations indicated that, for a target sample size of 2880 and power fixed at 80%, we would be able to detect a mean difference in change in SBP consistent with the contemporaneous published literature of approximately 3 to 16 mmHg (37), over a wide range of plausible intra-cluster coefficient and standard deviation values (further details in supplementary file).
Models for the primary, secondary, and subgroup analyses included as baseline covariates age, sex, recruitment pathway, amount of pre-trial MF activity in the cluster at baseline, type of health facility (dispensary, health center, or subcounty hospital), baseline cluster-specific mean of the outcome (38), and the baseline value of the outcome. We present model estimates and multiplicity-adjusted 98.3% confidence interval (CI) of the primary outcome using an intention-to-treat approach. Hypothesis tests were conducted for the primary outcome only. To preserve an overall type 1 error rate (alpha) of 0.05, the significance threshold for each comparison is 0.017. Secondary outcome CIs were not adjusted for multiple comparisons. CIs reflect clustering by facility.
We conducted a per-protocol analysis using randomization as an instrumental variable (IV) to assess the effect of full participation in each intervention, relative to UC, on SBP change. Participation was defined as the percentage of group sessions (either GMV, MF, or GMV-MF) attended by the participant among those organized by his/her group during the study follow-up period.
All statistical analyses were conducted in R Statistical Software (version 4.0.0) (39). IV analysis used the AER (40) and ivpack (41) routines.
Results
Participants
Of the 3041 individuals who initially satisfied inclusion criteria, 2890 individuals (69.9% women) were enrolled (708 UC, 709 MF, 740 GMV, and 733 GMV-MF). Figure 1 provides the CONSORT diagram, detailing reasons for exclusion, trial arm allocation, and follow-up details at both three and 12 months. Follow-up was successfully achieved at 12 months for 2710 (93.8%) participants, with no difference across trial arms.
Figure 1. Consort Diagram of the BIGPIC study.
Details of participant eligibility, reasons for exclusion, trial arm allocation, and follow-up details at both three and 12 months, are provided. BS = blood sugar; BP = blood pressure; HIV = human immunodeficiency virus; LTFU = lost to follow-up.
Baseline characteristics are shown in Table 1. Average baseline SBP was 157.5 mmHg and DBP was 92.5 mmHg overall, and similar across trial arms. Notably, 31.0% had a QRISK3 score of between 10.0 and 19.9 (elevated risk), and 15.3% had a QRISK3 score of greater than or equal to 20 (high risk). Nearly two-thirds of the study population reported no formal employment, and among those who reported formal employment, over 30% reported earning less than USD $10 per month. Over 75% had an international wealth index of less than 40 (indicator of poverty), 35% owned no livestock, and nearly 20% owned no land. Less than 17% reported having enrolled in the Kenyan governmental national health insurance plan. Additional baseline characteristics are provided in Supplemental Table 1.
Table 1.
Baseline Characteristics of the BIGPIC Study Participants.
| Variable Number of participants | Total 2890 | UC 708 | MF 709 | GMV 740 | GMV-MF 733 |
| Gender | |||||
| Female | 2020 (69.9%) | 522 (73.7%) | 481 (67.8%) | 510 (68.9%) | 507 (69.2%) |
| Male | 870 (30.1%) | 186 (26.3%) | 228 (32.2%) | 230 (31.1%) | 226 (30.8%) |
| Age | |||||
| Mean | 60.7 (12.1) | 59.8 (11.5) | 59.3 (12.9) | 62.1 (12.2) | 61.5 (11.7) |
| Diabetes (Self-Reported or Elevated Blood Glucose) | |||||
| No | 2275 (78.9%) | 518 (73.2%) | 589 (83.1%) | 578 (78.5%) | 590 (81%) |
| Yes | 607 (21.1%) | 190 (26.8%) | 120 (16.9%) | 159 (21.5%) | 138 (19%) |
| Missing | 8 (0.3%) | 0 (0%) | 0 (0%) | 3 (0.4%) | 5 (0.7%) |
| Baseline SBP | |||||
| Mean | 157.5 (19.1) | 155.4 (19.9) | 159.0 (18.8) N | 156.4 (17.5) | 159.1 (20.0) |
| Missing | 3 (0.1%) | 0 (0%) | 0 (0%) | 0 (0%) | 3 (0.4%) |
| Baseline DBP | |||||
| Mean | 92.5 (11.8) | 91.2 (11.9) | 92.3 (11.3) | 92.8 (11.1) | 93.8 (12.6) |
| Missing | 3 (0.1%) | 0 (0%) | 0 (0%) | 0 (0%) | 3 (0.4%) |
| Hypertension | |||||
| Yes (SBP ≥ 140 OR DBP ≥ 90) | 2698 (93.3%) | 637 (90.0%) | 668 (94.2%) | 700 (94.6%) | 693 (94.9%) |
| No (SBP < 140 AND DBP < 90) | 189 (6.5%) | 71 (10.0%) | 41 (5.8%) | 40 (5.4%) | 37 (5.1%) |
| Missing | 3 (0.1%) | 0 (0%) | 0 (0%) | 0 (0%) | 3 (0.4%) |
| Total Cholesterol Class | |||||
| <5.17 mmo/L | 2367 (84.3%) | 568 (82.4%) | 594 (85.2%) | 594 (83.0%) | 611 (86.4%) |
| >=5.17 mmo/L | 442 (15.7%) | 121 (17.6%) | 103 (14.8%) | 122 (17.0%) | 96 (13.96%) |
| Missing | 81 (2.8%) | 19 (2.7%) | 12 (1.7%) | 24 (3.2%) | 26 (3.5%) |
| LDL Cholesterol Class | |||||
| <4.14 mmo/L | 2585 (97.1%) | 624 (96%) | 655 (97.3%) | 668 (97.5%) | 638 (97.4%) |
| >=4.14 mmo/L | 78 (2.9%) | 26 (4%) | 18 (2.7%) | 17 (2.5%) | 17 (2.6%) |
| Missing | 227 (7.9%) | 58 (8.2%) | 36 (5.1%) | 55 (7.4%) | 78 (10.6%) |
| QRISK3 (34)(34)(continuous) | 11.5 (9.7) | 10.8(8.49) | 10.6 (9.4) | 12.7(11.3) | 11.8(9.4) |
| Missing | 84 (2.9%) | 20 (2.8%) | 12(1.7%) | 25 (3.4%) | 27 (3.7%) |
| QRISK3 | |||||
| < 10% | 1542 (55.0%) | 373 (54.2%) | 417 (59.8%) | 376 (52.6%) | 376 (53.3%) |
| 10% – 19.9% | 851 (30.3%) | 236 (34.3%) | 187 (26.8%) | 215 (30.1%) | 213 (30.2%) |
| ≥ 20% | 413 (14.7%) | 79 (11.5%) | 93 (13.3%) | 124 (17.3%) | 117 (16.6%) |
| Missing | 84 (2.9%) | 20 (2.8%) | 12 (1.7%) | 25 (3.4%) | 27 (3.7%) |
| National Hospital Insurance Fund (NHIF) | |||||
| No | 2400 (83.1%) | 576 (81.4%) | 594 (83.8%) | 618 (83.5%) | 612 (83.6%) |
| Yes | 489 (16.9%) | 132 (18.6%) | 115 (16.2%) | 122 (16.5%) | 120 (16.4%) |
| Variable Number of participants | Total 2890 | UC 708 | MF 709 | GMV 740 | GMV-MF 733 |
| Missing | 1 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 1 (0.1%) |
| Monthly Earnings* | |||||
| Not formally employed | 1838 (63.7%) | 483 (68.4%) | 491 (69.3%) | 423 (57.2%) | 441 (60.3%) |
| Less than Kshs 1000 | 315 (10.9%) | 43 (6.1%) | 48 (6.8%) | 120 (16.2%) | 104 (14.2%) |
| Kshs 1000–2999 | 252 (8.7%) | 55 (7.8%) | 48 (6.8%) | 79 (10.7%) | 70 (9.6%) |
| Kshs 3000–4999 | 179 (6.2%) | 42 (5.9%) | 51 (7.2%) | 47 (6.4%) | 39 (5.3%) |
| Kshs >=5000 | 272 (9.4%) | 74 (10.5%) | 67 (9.4%) | 65 (8.8%) | 66 (9%) |
| Don’t Know | 27 (0.9%) | 8 (1.1%) | 4 (0.6%) | 5 (0.7%) | 10 (1.4%) |
| Refused | 3 (0.1%) | 1 (0.1%) | 0 (0%) | 1 (0.1%) | 1 (0.1%) |
| Missing | 4 (0.1%) | 2 (0.3%) | 0 (0%) | 0 (0%) | 2 (0.3%) |
| International Wealth Index class | |||||
| < 15.00 | 583 (20.2%) | 118 (16.7%) | 164 (23.1%) | 154 (20.8%) | 147 (20.1%) |
| 15.01–24.99 | 781 (27%) | 163 (23%) | 200 (28.2%) | 221 (29.9%) | 197 (26.9%) |
| 25.00–39.99 | 820 (28.4%) | 227 (32.1%) | 185 (26.1%) | 193 (26.1%) | 215 (29.4%) |
| ≥ 40 | 704 (24.4%) | 200 (28.2%) | 160 (22.6%) | 172 (23.2%) | 172 (23.5%) |
| Missing | 2 (0.1%) | 0 (0%) | 0 (0%) | 0 (0%) | 2 (0.3%) |
| Livestock | |||||
| No | 1010 (35%) | 266 (37.6%) | 266 (37.6%) | 253 (34.3%) | 225 (30.8%) |
| Yes | 1873 (65%) | 441 (62.4%) | 441 (62.4%) | 485 (65.7%) | 506 (69.2%) |
| Missing | 7 (0.2%) | % (0.1%) | 2 (0.3%) | 2 (0.3%) | 2 (0.3%) |
| Land Ownership in Acres | |||||
| 0 | 527 (19.8%) | 161 (24.8%) | 158 (24.2%) | 113 (16.9%) | 95 (13.8%) |
| 1 | 672 (25.3%) | 180 (27.7%) | 162 (24.8%) | 155 (23.2%) | 175 (25.5%) |
| 2 | 488 (18.4%) | 124 (19.1%) | 119 (18.2%) | 102 (15.2%) | 143 (20.8%) |
| 3 | 310 (11.7%) | 67 (10.3%) | 72 (11%) | 81 (12.1%) | 90 (13.1%) |
| 4 | 169 (6.4%) | 32 (4.9%) | 34 (5.2%) | 48 (7.2%) | 55 (8%) |
| ≥ 5 | 493 (18.5%) | 85 (13.1%) | 109 (16.7%) | 170 (25.4%) | 129 (18.8%) |
| Missing | 231 (8%) | 59 (8.3%) | 55 (7.8%) | 71 (9.6%) | 46 (6.3%) |
Absolute numbers (and percentages) for baseline characteristics across the trial arms·Standard deviation values presented for all variables that report the mean. DBP = diastolic blood pressure; GMV = group medical visits; GMV-MF = group medical visits integrated into microfinance; LDL = low-density lipoprotein; MF = microfinance; SBP = systolic blood pressure; UC = usual care.
1 USD = approximately 100 Kenyan Shillings (Kshs).
Primary Outcome
Estimated average change in SBP (95% CI) was −11.4 (−12.9, −10.0), −14.8 (−16.4, −13.3), −14.7 (−16.2, −13.1), and −16.4 (−18.0, −14.7) mmHg in UC, MF, GMV, and GMV-MF, respectively (Figure 2). Model-based estimates showed that, compared to the UC arm, the mean reduction in SBP was 3.9 mmHg greater in the GMV-MF arm (98.3% CI-8.5 to 0.7, p=0.05), 3.3 mmHg greater in the GMV arm (−7.8 to 1.2, p=0.09), and 2.3 mmHg greater in the MF arm (−7.0 to 2.4, p=0.25) (Central Illustration, Table 2). The estimated intraclass correlation coefficient was 0.02.
Figure 2. Absolute change in systolic blood pressure in each arm of the trial.
Change in systolic blood pressure (mmHg) from baseline to 12 months in each arm of the trial; bars indicate 95% confidence intervals. GMV = group medical visits; GMV-MF = group medical visits integrated into microfinance; MF = microfinance; SBP = systolic blood pressure; UC = usual care.
Central Illustration. Pairwise comparisons of change in systolic blood pressure.
Graphs indicate the point estimates and confidence intervals for primary analysis (adjusted for multiple comparisons), per protocol analysis, and among participants with uncontrolled hypertension. MF = microfinance; GMV = group medical visits; GMV-MF = group medical visits integrated into microfinance; UC = usual care.
Table 2.
Primary and Secondary Outcome Results.
| Outcome | MF vs. UC | GMV vs. UC | GMV-MF vs. UC |
|---|---|---|---|
| Change in Systolic Blood Pressure (mmHg) | −2.3 (−6.3, 1.7) | −3.3 (−7.1, 0.5) | −3.9 (−7.8, 0.0) |
| Change in Systolic Blood Pressure (mmHg) (adjusted for multiple comparisons) | −2.3 (−7.0, 2.4) | −3.3 (−7.8, 1.2) | −3.9 (−8.5, 0.7) |
| Change in Systolic Blood Pressure (mmHg) (per protocol analysis) | −4.2 (−12.3, 4.0) | −5.3 (−10.8, 0.2) | −5.7 (−10.4, −1.0) |
| Change in Systolic Blood Pressure (mmHg) (among individuals with uncontrolled BP at baseline) | −2.2 (−6.4, 2.0) | −3.4 (−7.5, 0.6) | −4.2 (−8.3, 0.0) |
| Change in Diastolic Blood Pressure (mmHg) | −2.4 (−4.1, −0.6) | −1.6 (−3.4, 0.1) | −3.2 (−5.1, −1.5) |
| Change in QRISK3 CVD Risk Score | −0.4 (−1.4, 0.5) | −0.3 (−1.4, 0.7) | −0.9 (−1.9, 0.0) |
| Difference in Blood Pressure Control (%) (mixed effects model) | 11.8 (3.8, 19.9) | 9.5 (1.5, 17.4) | 6.7 (−1.2, 14.8) |
Estimates of treatment effect relative to usual care (UC), at the 12-month follow-up time point, with 95% confidence intervals, for primary and secondary outcomes. Primary outcome results shown before and after adjustment for multiple comparisons. . Difference in change in systolic blood pressure and diastolic blood pressure presented as mmHg, difference in change in QRISK3 CVD risk score presented as numerical unit, and difference in BP control presented as absolute percentage difference. Estimates adjusted for individual and cluster-specific mean of the outcome variable, sex, age, health facility type, baseline microfinance concentration in the cluster, and recruitment pathway. Variables included in the QRISK3 calculation were: systolic blood pressure, age, family history of angina, smoking, total:LDL cholesterol ratio, atrial fibrillation, type II diabetes, type I diabetes, hypertension treatment/medication, chronic renal disease, and rheumatoid arthritis. GMV = group medical visits; GMV-MF = group medical visits integrated into microfinance; MF = microfinance; UC = usual care.
Secondary Outcomes
Analysis of secondary outcomes showed evidence that GMV-MF led to greater decrease in DBP and QRISK3 score compared to UC (Table 2). BP control at 12 months was achieved in 40.2% of participants, and was greater in the MF and GMV arms. We did not find evidence that treatment impacted wealth indices or lipid levels (Supplemental Table 2).
Subgroup Analysis
When restricting the analysis to only those with baseline uncontrolled hypertension, the results were consistent with greater SBP reduction in the GMV-MF arm compared to UC (Table 2). Pre-specified subgroup analysis showed that women experienced greater SBP reductions than men: −3.5 mmHg (−7.6, 0.6), −2.7 mmHg (−6.8, 1.4), and −2.7 mmHg (−6.8, 1.4), in the MF-UC, GMV-UC, and GMV-MF-UC pairwise comparisons, respectively. Younger individuals also tended to benefit more than older individuals. In addition, individuals without health insurance and those with lower wealth and employment indices at baseline experienced greater SBP reductions in the MF and GMV-MF arms (Figure 3, Supplemental Table 3).
Figure 3. Forest plot demonstrating subgroup analyses for the primary outcome (difference in change in systolic blood pressure).
Subgoups include sex, age, monthly earnings, and baseline wealth status. MF = microfinance; GMV = group medical visits; GMV-MF = group medical visits integrated into microfinance; UC = usual care; IWI = international wealth index at baseline. 1 USD = approximately 100 Kenyan Shillings (Kshs).
Per-protocol Analysis
The per-protocol analysis yielded estimates of SBP reductions relative to UC that were greater in magnitude than the intention-to-treat analysis, particularly in the GMV-MF and GMV arms (Table 2). On average during the 12-month follow-up period, the median number (and percentage) of group sessions attended by participants in the different intervention arms was 9 (75%), 7 (67%), and 10 (83%), for the MF, GMV, and GMV-MF arms, respectively. Median attendance at each group session ranged from 10 to 16 individuals across all intervention arms.
Adverse Events
All-cause mortality was 1.1% (35 deaths) in the study population. There were 5, 4, 13, and 13 deaths in the UC, MF, GMV, and GMV-MF arms, respectively. We conducted a Cox proportional hazards regression model adjusting for age, sex, recruitment pathway, and clustering, which did not indicate systematic difference in mortality rates across the trial arms (Supplemental Table 4).
Discussion
In this cluster randomized trial evaluating GMV and MF for individuals with diabetes or hypertension in western Kenya, confidence intervals for GMV-MF were consistent with impact ranging from substantially greater SBP reduction to neutral effect, although the significance threshold was not met in hypothesis testing. When restricted to individuals with uncontrolled hypertension at baseline, the results were consistent with a beneficial impact of GMV-MF. In addition, active participation in the intervention arms was associated with greater improvements in SBP reduction. Our results were consistent with a positive impact on DBP and QRISK3 in the GMV-MF arm relative to UC. BP control was also improved in the MF and GMV arms compared to UC.
The observed improvements in SBP, DBP, BP control, and QRISK3 are clinically meaningful and would yield substantial long-term cardiovascular and mortality benefit (42)(43). We chose change in SBP as our primary outcome, because CVD is the major cause of morbidity and mortality among people with diabetes and pre-diabetes (44,45), and elevated BP is a major contributor to cardiovascular complications among these individuals (46,47). BP improvement among diabetics and pre-diabetics has been demonstrated to be an effective way to reduce CVD risk (48).
GMV may also have benefits that extend beyond hard health outcomes, such as combating isolation, increasing social cohesion, and improving clinician-patient trust (49). We have collected information on changes in social network characteristics, which we will incorporate into future analyses to elucidate whether the network characteristics mediate the relationship between the interventions and the outcomes.
Notably, individuals in the UC arm achieved a substantial reduction in SBP at 12 months. AMPATH’s multicomponent care, as described above, includes a combination of task redistribution, clinical decision support using health information technology, consistent and secure medication supply, linkage and retention activities, community and stakeholder engagement, and social support for patients, many of which are not present in standard public sector health facilities. Thus, it is likely that our “usual care” arm outperformed what would have been expected in similar settings (50).
Our study population was characterized by severe financial poverty; nearly all participants reported no formal employment or daily earnings less than $1.25 per day. However, we did not see evidence of differential change in the wealth indices during the study period, even in the arms with a MF component. It is plausible that a 12-month period of time was not of sufficient duration; successful MF programs have involved implementation for at least 24 months (51). We aim to conduct an extended follow-up of study participants to determine if longer-term involvement with MF-based activities will lead to more substantive improvements in economic and wealth status. In addition, we have conducted a process evaluation of the intervention to assess fidelity to the group meetings, amount of funds saved in the MF groups, and other contextual factors, the analyses of which are currently ongoing and will be the subject of future manuscripts.
Our results support the approach of incorporating social determinants of health into clinical care delivery. In particular, individuals without health insurance and with lower baseline earning/wealth indices tended to benefit more from interventions incorporating MF. Given the impoverishing effects of out-of-pocket expenditures for substantial direct (user fees, medications) and indirect (transportation) costs borne by patients with hypertension and diabetes (52)(6)(8), it is critical to adopt both clinical and socio-economic approaches. The BIGPIC model of combining MF with GMV is one strategy to address this pressing global health issue. Financial risk protection is also critically important for patients with hypertension, diabetes, and other non-communicable diseases. Countries worldwide are expanding universal health coverage in alignment with this objective (53).
We also observed that women tended to benefit more than men in trial arms utilizing GMV, consistent with similar group-based programs focused on maternal and child health (54). We also noted that younger individuals tended to benefit more in both the MF and GMV arms. It is possible that the economic and social network impacts were more important for younger, working-age individuals. These observations and hypotheses merit further investigation in future studies.
We acknowledge the following limitations. First, there was sometimes a delay between date of enrollment of a particular individual into the trial and the complete formation of a group (contingent upon enrollment of a minimum threshold number of individuals), which led to differential exposure to the intervention across participants. Second, it is possible that the duration of follow-up was insufficient to demonstrate a significant benefit. Third, there is the possibility of cross-contamination across the trial arms. However, this scenario is unlikely since the clusters were geographically distinct areas and participants were invited by the research staff to attend the group sessions; in addition, no participants in the UC arm reported having attended a GMV. Finally, our study population might not be fully representative of the general population. However, the economic challenges experienced by our study participants are not dissimilar from a large proportion of the global population.
Conclusion
A strategy combining GMV and MF for treating diabetes or hypertension led to SBP reductions that have been shown to have long-term beneficial impact on cardiovascular health. Although the significance threshold was not met in hypothesis testing, confidence intervals for GMV-MF were consistent with impact ranging from substantive benefit to neutral effect. Greater benefit was observed among those with baseline uncontrolled hypertension and those who actively participated in the group sessions. Incorporating social determinants of health into care delivery for chronic diseases can improve outcomes, especially for subgroups of patients. We anticipate that this approach, and our results, will help to stimulate evaluations of similar strategies for chronic diseases worldwide.
Supplementary Material
Perspectives:
Competency in Systems-Based Practice:
Addressing social determinants of health in the delivery of care for patients with chronic diseases, such as by group medical visits and microfinance, can improve outcomes for certain subgroups.
Translational Outlook:
Further implementation research is needed to expand the integration of effective strategies with chronic disease care and characterize patient subgroups most likely to benefit from these interventions.
Acknowledgments:
The authors would like to thank the BIGPIC research team members for their assistance with all aspects of the project. We would like to thank all study participants for their involvement with this project. We would also like to thank Julia Dickhaus for editorial assistance and assistance with the figures.
Funding:
The BIGPIC trial is supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number R01HL125487. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Abbreviations list:
- AMPATH
Academic Model Providing Access to Healthcare Partnership
- BP
Blood pressure
- CIs
Confidence intervals
- CVD
Cardiovascular disease
- DBP
Diastolic blood pressure
- GMV
Group medical visits
- LMICs
Low- and middle-income countries
- MF
Microfinance
- SBP
Systolic blood pressure
- UC
Usual care
Footnotes
Disclosures:
The authors affirm that there are no conflicts of interest.
Clinical Trial: Clinicaltrials.gov, NCT02501746, registration date: July 17, 2015
Short Tweet: Social determinants of health and clinical care for hypertension and diabetes: group medical visits and microfinance in Kenya improve blood pressure and CVD risk
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Vos T, Lim SS, Abbafati C et al. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 2020;396:1204–1222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Brunner EJ, Shipley MJ, Witte DR, Fuller JH, Marmot MG. Relation between blood glucose and coronary mortality over 33 years in the Whitehall Study. Diabetes care 2006;29:26–31. [DOI] [PubMed] [Google Scholar]
- 3.Lewington S, Clarke R, Qizilbash N, Peto R, Collins R, Collaboration PS. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet 2002;360:1903–13. [DOI] [PubMed] [Google Scholar]
- 4.Chow CK, Teo KK, Rangarajan S et al. Prevalence, awareness, treatment, and control of hypertension in rural and urban communities in high-, middle-, and low-income countries. JAMA 2013;310:959–968. [DOI] [PubMed] [Google Scholar]
- 5.Bank TW. Poverty and Health. 2014.
- 6.Gheorghe A, Griffiths U, Murphy A, Legido-Quigley H, Lamptey P, Perel P. The economic burden of cardiovascular disease and hypertension in low- and middle-income countries: a systematic review. BMC Public Health 2018;18:975. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bommer C, Sagalova V, Heesemann E et al. Global Economic Burden of Diabetes in Adults: Projections From 2015 to 2030. Diabetes care 2018;41:963–970. [DOI] [PubMed] [Google Scholar]
- 8.Bukhman G, Mocumbi AO, Atun R et al. The Lancet NCDI Poverty Commission: bridging a gap in universal health coverage for the poorest billion. Lancet 2020;396:991–1044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Khatib R, McKee M, Shannon H et al. Availability and affordability of cardiovascular disease medicines and their effect on use in high-income, middle-income, and low-income countries: an analysis of the PURE study data. Lancet 2016;387:61–69. [DOI] [PubMed] [Google Scholar]
- 10.Khatib R, Schwalm J-D, Yusuf S et al. Patient and Healthcare Provider Barriers to Hypertension Awareness, Treatment and Follow Up: A Systematic Review and Meta-Analysis of Qualitative and Quantitative Studies. PLoS One 2014;9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Naanyu V, Vedanthan R, Kamano JH et al. Barriers Influencing Linkage to Hypertension Care in Kenya: Qualitative Analysis from the LARK Hypertension Study. J Gen Intern Med 2016;31:304–314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wierzejska E, Giernaś B, Lipiak A, Karasiewicz M, Cofta M, Staszewski R. A global perspective on the costs of hypertension: a systematic review. Arch Med Sci 2020;16:1078–1091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Social Determinants of Health. Department of Health and Human Services.
- 14.Gurewich D, Garg A, Kressin NR. Addressing Social Determinants of Health Within Healthcare Delivery Systems: a Framework to Ground and Inform Health Outcomes. J Gen Intern Med 2020;35:1571–1575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.National Academies of Sciences E, Medicine. Integrating Social Care into the Delivery of Health Care: Moving Upstream to Improve the Nation's Health. Washington, DC: The National Academies Press, 2019. [PubMed] [Google Scholar]
- 16.Lorenzetti LMJ, Leatherman S, Flax VL. Evaluating the effect of integrated microfinance and health interventions: an updated review of the evidence. Health policy and planning 2017;32:732–756. [DOI] [PubMed] [Google Scholar]
- 17.Edelman D, McDuffie JR, Oddone E, Gierisch JM, Williams JW Jr., Nagi A Shared Medical Appointments for Chronic Medical Conditions: A Systematic Review. Washington, DC: Department of Veterans Affairs, 2012. [PubMed] [Google Scholar]
- 18.Pastakia S, Manyara S, Vedanthan R et al. Impact of Bridging Income Generation with GrouP Integrated Care (BIGPIC) on Hypertension and Diabetes in Rural Western Kenya. J Gen Intern Med 2017;32:540–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Vedanthan R, Kamano JH, Lee H et al. Bridging income generation with group integrated care for cardiovascular risk reduction: Rationale and design of the BIGPIC study. Am Heart J 2017;188:175–185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Einterz RM, Kimaiyo S, Mengech HNK et al. Responding to the HIV pandemic: the power of an academic medical partnership. Acad Med 2007;82:812–8. [DOI] [PubMed] [Google Scholar]
- 21.Vedanthan R, Kamano JH, Bloomfield GS, Manji I, Pastakia S, Kimaiyo SN. Engaging the Entire Care Cascade in Western Kenya: A Model to Achieve the Cardiovascular Disease Secondary Prevention Roadmap Goals. Glob Heart 2015;10:313–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Binanay CA, Akwanalo CO, Aruasa W et al. Building Sustainable Capacity for Cardiovascular Care at a Public Hospital in Western Kenya. J Am Coll Cardiol 2015;66:2550–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Vedanthan R, Kumar A, Kamano JH et al. Effect of Nurse-Based Management of Hypertension in Rural Western Kenya. Glob Heart 2020; 15:77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Vedanthan R, Blank E, Tuikong N et al. Usability and feasibility of a tablet-based Decision-Support and Integrated Record-keeping (DESIRE) tool in the nurse management of hypertension in rural western Kenya. Int J Med Inform 2015;84:207–219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Tran DN, Manji I, Njuguna B et al. Solving the problem of access to cardiovascular medicines: revolving fund pharmacy models in rural western Kenya. BMJ Glob Health 2020;5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Vedanthan R, Kamano JH, DeLong AK et al. Community Health Workers Improve Linkage to Hypertension Care in Western Kenya. J Am Coll Cardiol 2019;74:1897–1906. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Akwanalo C, Njuguna B, Mercer T et al. Strategies for Effective Stakeholder Engagement in Strengthening Referral Networks for Management of Hypertension Across Health Systems in Kenya. Glob Heart 2019;14:173–179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Dong R, Leung C, Naert MN et al. Chronic disease stigma, skepticism of the health system, and socio-economic fragility: qualitative assessment of factors impacting receptiveness to group medical visits and microfinance for non-communicable disease care in rural Kenya. PLOS One 2020;Revision under review. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Leung CL, Naert M, Andama B et al. Human-centered design as a guide to intervention planning for non-communicable diseases: the BIGPIC study from Western Kenya. BMC health services research 2020;20:415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Gray LJ, Taub NA, Khunti K et al. The Leicester Risk Assessment score for detecting undiagnosed Type 2 diabetes and impaired glucose regulation for use in a multiethnic UK setting. Diabetic medicine : a journal of the British Diabetic Association 2010;27:887–95. [DOI] [PubMed] [Google Scholar]
- 31.Global Guideline for Type 2 Diabetes. Brussels, Belgium: International Diabetes Federation, 2012. [Google Scholar]
- 32.Chobanian AV, Bakris GL, Black HR et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003;289:2560–72. [DOI] [PubMed] [Google Scholar]
- 33.Riddell MA, Edwards N, Thompson SR et al. GACD Hypertension Research Programme.. Developing consensus measures for global programs: lessons from the Global Alliance for Chronic Diseases Hypertension research program. Globalization and health 2017:13:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hippisley-Cox J, Coupland C, Brindle P. Development and validation of QRISK3 risk prediction algorithms to estimate future risk of cardiovascular disease: prospective cohort study. BMJ. 2017. May 23;357:j2099. doi: 10.1136/bmj.j2099. PMID: 28536104; PMCID: PMC5441081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Schofield P, Crichton N, Chen R. Methods for assessing cardiovascular disease risk in a UK black population. Heart 2012;98:1373–7. [DOI] [PubMed] [Google Scholar]
- 36.Smits J, Steendijk R. The International Wealth Index (IWI). Social Indicators Research 2015;122:65–85. [Google Scholar]
- 37.Mendis S, Johnston SC, Fan W, Oladapo O, Cameron A, Faramawi MF. Cardiovascular risk management and its impact on hypertension control in primary care in low-resource settings: a cluster-randomized trial. Bulletin of the World Health Organization 2010;88:412–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Hooper R, Forbes A, Hemming K, Takeda A, Beresford L. Analysis of cluster randomised trials with an assessment of outcome at baseline. BMJ. 2018. March 20;360:k1121. doi: 10.1136/bmj.k1121. PMID: 29559436. [DOI] [PubMed] [Google Scholar]
- 39.Team RC. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2020. [Google Scholar]
- 40.Kleiber C, Zeileis A. Applied Econometrics with R. New York: Springer-Verlag, 2008. [Google Scholar]
- 41.Jiang Y, Small D. ivpack: Instrumental Variable Estimation. R package version 1.2. 2014. [Google Scholar]
- 42.Ettehad D, Emdin CA, Kiran A et al. Blood pressure lowering for prevention of cardiovascular disease and death: a systematic review and meta-analysis. Lancet 2016;387:957–967. [DOI] [PubMed] [Google Scholar]
- 43.Jaspers NEM, Blaha MJ, Matsushita K et al. Prediction of individualized lifetime benefit from cholesterol lowering, blood pressure lowering, antithrombotic therapy, and smoking cessation in apparently healthy people. Eur Heart J 2020;41:1190–1199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Emerging Risk Factors C, Sarwar N, Gao P et al. Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies. Lancet 2010;375:2215–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Emerging Risk Factors C, Seshasai SR, Kaptoge S et al. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med 2011;364:829–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Adler AI, Stratton IM, Neil HA et al. Association of systolic blood pressure with macrovascular and microvascular complications of type 2 diabetes (UKPDS 36): prospective observational study. BMJ 2000;321:412–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Stamler J, Vaccaro O, Neaton JD, Wentworth D. Diabetes, other risk factors, and 12-yr cardiovascular mortality for men screened in the Multiple Risk Factor Intervention Trial. Diabetes care 1993;16:434–44. [DOI] [PubMed] [Google Scholar]
- 48.Buckley LF, Dixon DL, Wohlford GFt, Wijesinghe DS, Baker WL, Van Tassell BW. Effect of intensive blood pressure control in patients with type 2 diabetes mellitus over 9 years of follow-up: A subgroup analysis of high-risk ACCORDION trial participants. Diabetes Obes Metab 2018;20:1499–1502. [DOI] [PubMed] [Google Scholar]
- 49.Kirsh SR, Aron DC, Johnson KD et al. A realist review of shared medical appointments: How, for whom, and under what circumstances do they work? BMC health services research 2017;17:113–113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Owolabi MO, Gebregziabher M, Akinyemi RO et al. Randomized Trial of an Intervention to Improve Blood Pressure Control in Stroke Survivors. Circ Cardiovasc Qual Outcomes 2019;12:e005904. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Banerjee A, Duflo E, Goldberg N et al. Development economics. A multifaceted program causes lasting progress for the very poor: evidence from six countries. Science 2015;348:1260799. [DOI] [PubMed] [Google Scholar]
- 52.Oyando R, Njoroge M, Nguhiu P et al. Patient costs of hypertension care in public health care facilities in Kenya. 2019;34:e1166–e1178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Wagstaff A, Neelsen S. A comprehensive assessment of universal health coverage in 111 countries: a retrospective observational study. The Lancet Global Health 2020;8:e39–e49. [DOI] [PubMed] [Google Scholar]
- 54.Maldonado LY, Bone J, Scanlon ML et al. Improving maternal, newborn and child health outcomes through a community-based women’s health education program: a cluster randomised controlled trial in western Kenya. BMJ Glob Health 2020;5. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.