Gaussian process emulation to improve efficiency of computationally intensive multidisease models: a practical tutorial with adaptable R code.

BACKGROUND: The rapidly growing burden of non-communicable diseases (NCDs) among people living with HIV in sub-Saharan Africa (SSA) has expanded the number of multidisease models predicting future care needs and health system priorities. Usefulness of these models depends on their ability to replicate real-life data and be readily understood and applied by public health decision-makers; yet existing simulation models of HIV comorbidities are computationally expensive and require large numbers of parameters and long run times, which hinders their utility in resource-constrained settings. METHODS: We present a novel, user-friendly emulator that can efficiently approximate complex simulators of long-term HIV and NCD outcomes in Africa. We describe how to implement the emulator via a tutorial based on publicly available data from Kenya. Emulator parameters relating to incidence and prevalence of HIV, hypertension and depression were derived from our own agent-based simulation model and other published literature. Gaussian processes were used to fit the emulator to simulator estimates, assuming presence of noise for design points. Bayesian posterior predictive checks and leave-one-out cross validation confirmed the emulator's descriptive accuracy. RESULTS: In this example, our emulator resulted in a 13-fold (95% Confidence Interval (CI): 8-22) improvement in computing time compared to that of more complex chronic disease simulation models. One emulator run took 3.00 seconds (95% CI: 1.65-5.28) on a 64-bit operating system laptop with 8.00 gigabytes (GB) of Random Access Memory (RAM), compared to > 11 hours for 1000 simulator runs on a high-performance computing cluster with 1500 GBs of RAM. Pareto k estimates were < 0.70 for all emulations, which demonstrates sufficient predictive accuracy of the emulator. CONCLUSIONS: The emulator presented in this tutorial offers a practical and flexible modelling tool that can help inform health policy-making in countries with a generalized HIV epidemic and growing NCD burden. Future emulator applications could be used to forecast the changing burden of HIV, hypertension and depression over an extended (> 10 year) period, estimate longer-term prevalence of other co-occurring conditions (e.g., postpartum depression among women living with HIV), and project the impact of nationally-prioritized interventions such as national health insurance schemes and differentiated care models.

Predictors of Mortality Among Children and Adolescents Living With HIV on Antiretroviral Therapy in Western Kenya.

BACKGROUND: Antiretroviral therapy (ART) has decreased HIV-attributable deaths; however, children and adolescents continue to have high HIV-associated mortality. SETTING: We determined the predictors of death among children and young adolescents living with HIV (CALWH) who died while in care in Western Kenya. METHODS: This retrospective case-control study used electronically abstracted data of 6234 CALWH who received care in Academic Model Providing Access to Healthcare HIV clinics in Western Kenya between January 2002 and November 2022. The cases comprised CALWH who were reported dead by November 2022, while the controls constituted of matched CALWH who were alive and in care. Independent predictors of mortality were determined using univariable and multivariable Cox proportional hazard regression models. Kaplan-Meier analysis ascertained survival. RESULTS: Of the 6234 participants enrolled, slightly more than half were male (51.7%). The mean (SD) age at the start of ART was significantly lower in cases than in controls at 6.01 (4.37) and 6.62 (4.11) ( P < 0.001), respectively. An age of 11 years or older at start of ART (adjusted Hazard Ratio [aHR]: 8.36 [3.60-19.40]), both parents being alive (aHR: 3.06 [1.67-5.60]), underweight (aHR: 1.82 [1.14-2.92]), and World Health Organization stages 3 (aHR: 2.63 [1.12-6.18]) and 4 (aHR: 2.20 [0.94-5.18]) increased mortality; while school attendance (aHR: 0.12 [0.06-0.21]), high CD4 + counts >350 cells/mm 3 (aHR: 0.79 [0.48-1.29]), and low first viral load <1000 copies/mL (aHR: 0.24 [0.14-0.40]) were protective. CONCLUSION: Independent predictors of mortality were age 11 years or older at the start of ART, orphan status, underweight, and advanced HIV disease. Beyond the provision of universal ART, care accorded to CALWH necessitates optimization through tackling individual predictors of mortality.

Are outpatient costs for hypertension and diabetes care affordable? Evidence from Western Kenya.

BACKGROUND: Diabetes and hypertension pose a significant socio-economic burden in developing countries such as Kenya, where financial risk-protection mechanisms remain inadequate. This proves to be a great barrier towards achieving universal health care in such settings unless mechanisms are put in place to ensure greater access and affordability to non-communicable disease (NCD) management services. AIM: This article aims to examine outpatient management services costs for patients with diabetes and hypertension attending public primary healthcare facilities. SETTING: The study was conducted in Busia and Trans-Nzoia counties in Western Kenya in facilities supported by the PIC4C project, between August 2020 and December 2020. METHODS: This cross-sectional survey included 719 adult participants. Structured interviewer-administered questionnaires were used to collect information on healthcare-seeking behaviour and associated costs. The annual direct and indirect costs borne by patients were computed by disease type and level of healthcare facility visited. RESULTS: Patients with both diabetes and hypertension incurred higher annual costs (KES 13 149) compared to those with either diabetes (KES 8408) or hypertension (KES 7458). Patients attending dispensaries and other public healthcare facilities incurred less direct costs compared to those who visited private clinics. Furthermore, a higher proportionate catastrophic healthcare expenditure of 41.83% was noted among uninsured patients. CONCLUSION: Despite this study being conducted in facilities that had an ongoing NCDs care project that increased access to subsidised medication, we still reported a substantially high cost of managing diabetes and hypertension among patients attending primary healthcare facilities in Western Kenya, with a greater burden among those with comorbidities.Contribution: Evidenced by the results that there is enormous financial burden borne by patients with chronic diseases such as hypertension and diabetes; we recommend that universal healthcare coverage that offers comprehensive care for NCDs be urgently rolled out alongside strengthening of lower-level public healthcare systems.

Cross-sectional analysis of factors associated with medication adherence in western Kenya.

OBJECTIVES: Poor medication adherence in low-income and middle-income countries is a major cause of suboptimal hypertension and diabetes control. We aimed to identify key factors associated with medication adherence in western Kenya, with a focus on cost-related and economic wealth factors. SETTING: We conducted a cross-sectional analysis of baseline data of participants enrolled in the Bridging Income Generation with Group Integrated Care study in western Kenya. PARTICIPANTS: All participants were ≥35 years old with either diabetes or hypertension who had been prescribed medications in the past 3 months. PRIMARY AND SECONDARY OUTCOME MEASURES: Baseline data included sociodemographic characteristics, wealth and economic status and medication adherence information. Predictors of medication adherence were separated into the five WHO dimensions of medication adherence: condition-related factors (comorbidities), patient-related factors (psychological factors, alcohol use), therapy-related factors (number of prescription medications), economic-related factors (monthly income, cost of transportation, monthly cost of medications) and health system-related factors (health insurance, time to travel to the health facility). A multivariable analysis, controlling for age and sex, was conducted to determine drivers of suboptimal medication adherence in each overarching category. RESULTS: The analysis included 1496 participants (73.7% women) with a mean age of 60 years (range 35-97). The majority of participants had hypertension (69.2%), 8.8% had diabetes and 22.1% had both hypertension and diabetes. Suboptimal medication adherence was reported by 71.2% of participants. Economic factors were associated with medication adherence. In multivariable analysis that investigated specific subtypes of costs, transportation costs were found to be associated with worse medication adherence. In contrast, we found no evidence of association between monthly medication costs and medication adherence. CONCLUSION: Suboptimal medication adherence is highly prevalent in Kenya, and primary-associated factors include costs, particularly indirect costs of transportation. Addressing all economic factors associated with medication adherence will be important to improve outcomes for non-communicable diseases. TRIAL REGISTRATION NUMBER: NCT02501746.

Examining the responsiveness of the National Health Insurance Fund to people living with hypertension and diabetes in Kenya: a qualitative study.

OBJECTIVES: To assess the responsiveness of the National Health Insurance Fund (NHIF) Supa Cover benefit package to the needs of individuals with diabetes and hypertension in Kenya. DESIGN, SETTING AND PARTICIPANTS: We carried out a qualitative study and collected data using key informant interviews (n=39) and focus group discussions (n=4) in two purposively selected counties in Western Kenya. Study participants were drawn from NHIF officials, county government officials, health facility managers, healthcare workers and individuals with hypertension and diabetes who were enrolled in NHIF. We analysed data using a thematic approach. RESULTS: Study participants reported that the NHIF Supa Cover benefit package expanded access to services for people living with hypertension and diabetes. However, the NHIF members and healthcare workers had inadequate awareness of the NHIF service entitlements. The NHIF benefit package inadequately covered the range of services needed by people living with hypertension and diabetes and the benefits package did not prioritise preventive and promotive services. Sometimes patients were discriminated against by healthcare providers who preferred cash-paying patients, and some NHIF-empanelled health facilities had inadequate structural inputs essential for quality of care. Study participants felt that the NHIF premium for the general scheme was unaffordable, and NHIF members faced additional out-of-pocket costs because of additional payments for services not available or covered. CONCLUSION: Whereas NHIF has reduced financial barriers for hypertension and diabetes patients, to enhance its responsiveness to patient needs, NHIF should implement mechanisms to increase benefit package awareness among members and providers. In addition, preventive and promotive services should be included in NHIF's benefits package and mechanisms to monitor and hold contracted providers accountable should be strengthened.

Evaluating the effectiveness of the National Health Insurance Fund in providing financial protection to households with hypertension and diabetes patients in Kenya.

BACKGROUND: Non-communicable diseases (NCDs) can impose a substantial financial burden to households in the absence of an effective financial risk protection mechanism. The national health insurance fund (NHIF) has included NCD services in its national scheme. We evaluated the effectiveness of NHIF in providing financial risk protection to households with persons living with hypertension and/or diabetes in Kenya. METHODS: We carried out a prospective cohort study, following 888 households with at least one individual living with hypertension and/or diabetes for 12 months. The exposure arm comprised households that are enrolled in the NHIF national scheme, while the control arm comprised households that were not enrolled in the NHIF. Study participants were drawn from two counties in Kenya. We used the incidence of catastrophic health expenditure (CHE) as the outcome of interest. We used coarsened exact matching and a conditional logistic regression model to analyse the odds of CHE among households enrolled in the NHIF compared with unenrolled households. Socioeconomic inequality in CHE was examined using concentration curves and indices. RESULTS: We found strong evidence that NHIF-enrolled households spent a lower share (12.4%) of their household budget on healthcare compared with unenrolled households (23.2%) (p = 0.004). While households that were enrolled in NHIF were less likely to incur CHE, we did not find strong evidence that they are better protected from CHE compared with households without NHIF (OR = 0.67; p = 0.47). The concentration index (CI) for CHE showed a pro-poor distribution (CI: -0.190, p < 0.001). Almost half (46.9%) of households reported active NHIF enrolment at baseline but this reduced to 10.9% after one year, indicating an NHIF attrition rate of 76.7%. The depth of NHIF cover (i.e., the share of out-of-pocket healthcare costs paid by NHIF) among households with active NHIF was 29.6%. CONCLUSION: We did not find strong evidence that the NHIF national scheme is effective in providing financial risk protection to households with individuals living with hypertension and/diabetes in Kenya. This could partly be explained by the low depth of cover of the NHIF national scheme, and the high attrition rate. To enhance NHIF effectiveness, there is a need to revise the NHIF benefit package to include essential hypertension and/diabetes services, review existing provider payment mechanisms to explicitly reimburse these services, and extend the existing insurance subsidy programme to include individuals in the informal labour market.

Understanding the treatment burden of people with chronic conditions in Kenya: A cross-sectional analysis using the Patient Experience with Treatment and Self-Management (PETS) questionnaire.

In Kenya, non-communicable diseases (NCDs) are an increasingly important cause of morbidity and mortality, requiring both better access to health care services and self-care support. Evidence suggests that treatment burdens can negatively affect adherence to treatment and quality of life. In this study, we explored the treatment and self-management burden among people with NCDs in in two counties in Western Kenya. We conducted a cross-sectional survey of people newly diagnosed with diabetes and/or hypertension, using the Patient Experience with Treatment and Self-Management (PETS) instrument. A total of 301 people with diabetes and/or hypertension completed the survey (63% female, mean age = 57 years). They reported the highest treatment burdens in the domains of medical and health care expenses, monitoring health, exhaustion related to self-management, diet and exercise/physical therapy. Treatment burden scores differed by county, age, gender, education, income and number of chronic conditions. Younger respondents (<60 years) reported higher burden for medication side effects (p<0.05), diet (p<0.05), and medical appointments (p = 0.075). Those with no formal education or low income also reported higher burden for diet and for medical expenses. People with health insurance cover reported lower (albeit still comparatively high) burden for medical expenses compared to those without it. Our findings provide important insights for Kenya and similar settings where governments are working to achieve universal health coverage by highlighting the importance of financial protection not only to prevent the economic burden of seeking health care for chronic conditions but also to reduce the associated treatment burden.

Scaling up the primary health integrated care project for chronic conditions in Kenya: study protocol for an implementation research project.

INTRODUCTION: Amid the rising number of people with non-communicable diseases (NCDs), Kenya has invested in strengthening primary care and in efforts to expand existing service delivery platforms to integrate NCD care. One such approach is the AMPATH (Academic Model Providing Access to Healthcare) model in western Kenya, which provides the platform for the Primary Health Integrated Care Project for Chronic Conditions (PIC4C), launched in 2018 to further strengthen primary care services for the prevention and control of hypertension, diabetes, breast and cervical cancer. This study seeks to understand how well PIC4C delivers on its intended aims and to inform and support scale up of the PIC4C model for integrated care for people with NCDs in Kenya. METHODS AND ANALYSIS: The study is guided by a conceptual framework on implementing, sustaining and spreading innovation in health service delivery. We use a multimethod design combining qualitative and quantitative approaches, involving: (1) in-depth interviews with health workers and decision-makers to explore experiences of delivering PIC4C; (2) a cross-sectional survey of patients with diabetes or hypertension and in-depth interviews to understand how well PIC4C meets patients' needs; (3) a cohort study with an interrupted time series analysis to evaluate the degree to which PIC4C leads to health benefits such as improved management of hypertension or diabetes; and (4) a cohort study of households to examine the extent to which the national hospital insurance chronic care package provides financial risk protection to people with hypertension or diabetes within PIC4C. ETHICS AND DISSEMINATION: The study has received approvals from Moi University Institutional Research and Ethics Committee (FAN:0003586) and the London School of Hygiene & Tropical Medicine (17940). Workshops with key stakeholders at local, county, national and international levels will ensure early and wide dissemination of our findings to inform scale up of this model of care. We will also publish findings in peer-reviewed journals.

Network characteristics of a referral system for patients with hypertension in Western Kenya: results from the Strengthening Referral Networks for Management of Hypertension Across the Health System (STRENGTHS) study.

BACKGROUND: Health system approaches to improve hypertension control require an effective referral network. A national referral strategy exists in Kenya; however, a number of barriers to referral completion persist. This paper is a baseline assessment of a hypertension referral network for a cluster-randomized trial to improve hypertension control and reduce cardiovascular disease risk. METHODS: We used sociometric network analysis to understand the relationships between providers within a network of nine geographic clusters in western Kenya, including primary, secondary, and tertiary care facilities. We conducted a survey which asked providers to nominate individuals and facilities to which they refer patients with controlled and uncontrolled hypertension. Degree centrality measures were used to identify providers in prominent positions, while mixed-effect regression models were used to determine provider characteristics related to the likelihood of receiving referrals. We calculated core-periphery correlation scores (CP) for each cluster (ideal CP score = 1.0). RESULTS: We surveyed 152 providers (physicians, nurses, medical officers, and clinical officers), range 10-36 per cluster. Median number of hypertensive patients seen per month was 40 (range 1-600). While 97% of providers reported referring patients up to a more specialized health facility, only 55% reported referring down to lower level facilities. Individuals were more likely to receive a referral if they had higher level of training, worked at a higher level facility, were male, or had more job experience. CP scores for provider networks range from 0.335 to 0.693, while the CP scores for the facility networks range from 0.707 to 0.949. CONCLUSIONS: This analysis highlights several points of weakness in this referral network including cluster variability, poor provider linkages, and the lack of down referrals. Facility networks were stronger than provider networks. These shortcomings represent opportunities to focus interventions to improve referral networks for hypertension. TRIAL REGISTRATION: Trial Registered on ClinicalTrials.gov NCT03543787 , June 1, 2018.

Egocentric social network characteristics and cardiovascular risk among patients with hypertension or diabetes in western Kenya: a cross-sectional analysis from the BIGPIC trial.

OBJECTIVES: Management of cardiovascular disease (CVD) is an urgent challenge in low-income and middle-income countries, and interventions may require appraisal of patients' social networks to guide implementation. The purpose of this study is to determine whether egocentric social network characteristics (SNCs) of patients with chronic disease in western Kenya are associated with overall CVD risk and individual CVD risk factors. DESIGN: Cross-sectional analysis of enrollment data (2017-2018) from the Bridging Income Generation with GrouP Integrated Care trial. Non-overlapping trust-only, health advice-only and multiplex (trust and health advice) egocentric social networks were elicited for each participant, and SNCs representing social cohesion were calculated. SETTING: 24 communities across four counties in western Kenya. PARTICIPANTS: Participants (n=2890) were ≥35 years old with diabetes (fasting glucose ≥7 mmol/L) or hypertension. PRIMARY AND SECONDARY OUTCOMES: We hypothesised that SNCs would be associated with CVD risk status (QRISK3 score). Secondary outcomes were individual CVD risk factors. RESULTS: Among the 2890 participants, 2020 (70%) were women, and mean (SD) age was 60.7 (12.1) years. Forty-four per cent of participants had elevated QRISK3 score (≥10%). No relationship was observed between QRISK3 level and SNCs. In unadjusted comparisons, participants with any individuals in their trust network were more likely to report a good than a poor diet (41% vs 21%). SNCs for the trust and multiplex networks accounted for a substantial fraction of variation in measures of dietary quality and physical activity (statistically significant via likelihood ratio test, adjusted for false discovery rate). CONCLUSION: SNCs indicative of social cohesion appear to be associated with individual behavioural CVD risk factors, although not with overall CVD risk score. Understanding how SNCs of patients with chronic diseases relate to modifiable CVD risk factors could help inform network-based interventions. TRIAL REGISTRATION NUMBER: ClinicalTrials.gov identifier: NCT02501746; https://clinicaltrials.gov/ct2/show/NCT02501746.

Group Medical Visit and Microfinance Intervention for Patients With Diabetes or Hypertension in Kenya.

BACKGROUND: Incorporating social determinants of health into care delivery for chronic diseases is a priority. OBJECTIVES: The goal of this study was to evaluate the impact of group medical visits and/or microfinance on blood pressure reduction. METHODS: The authors conducted a cluster randomized trial with 4 arms and 24 clusters: 1) usual care (UC); 2) usual care plus microfinance (MF); 3) group medical visits (GMVs); and 4) GMV integrated into MF (GMV-MF). The primary outcome was 1-year change in systolic blood pressure (SBP). Mixed-effects intention-to-treat models were used to evaluate the outcomes. RESULTS: A total of 2,890 individuals (69.9% women) were enrolled (708 UC, 709 MF, 740 GMV, and 733 GMV-MF). Average baseline SBP was 157.5 mm Hg. Mean SBP declined -11.4, -14.8, -14.7, and -16.4 mm Hg 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 mm Hg (-8.5 to 0.7), 3.3 mm Hg (-7.8 to 1.2), and 2.3 mm Hg (-7.0 to 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 pairwise comparison hypothesis testing, confidence intervals for GMV-MF were consistent with impacts ranging from substantive benefit to neutral effect relative to UC. Incorporating social determinants of health into care delivery for chronic diseases has potential to improve outcomes. (Bridging Income Generation With Group Integrated Care [BIGPIC]; NCT02501746).