Search

In 2019, the World Health Organization recognised diabetes as a clinically and pathophysiologically heterogeneous set of related diseases. Little is currently known about the diabetes phenotypes in the population of low- and middle-income countries (LMICs), yet identifying their different risks and aetiology has great potential to guide the development of more effective, tailored prevention and treatment. This study reviewed the scope of diabetes datasets, health information ecosystems, and human resource capacity in four countries to assess whether a diabetes phenotyping algorithm (developed under a companion study) could be successfully applied. The capacity assessment was undertaken with four countries: Trinidad, Malaysia, Kenya, and Rwanda. Diabetes programme staff completed a checklist of available diabetes data variables and then participated in semi-structured interviews about Health Information System (HIS) ecosystem conditions, diabetes programme context, and human resource needs. Descriptive analysis was undertaken. Only Malaysia collected the full set of the required diabetes data for the diabetes algorithm, although all countries did collect the required diabetes complication data. An HIS ecosystem existed in all settings, with variations in data hosting and sharing. All countries had access to HIS or ICT support, and epidemiologists or biostatisticians to support dataset preparation and algorithm application. Malaysia was found to be most ready to apply the phenotyping algorithm. A fundamental impediment in the other settings was the absence of several core diabetes data variables. Additionally, if countries digitise diabetes data collection and centralise diabetes data hosting, this will simplify dataset preparation for algorithm application. These issues reflect common LMIC health systems' weaknesses in relation to diabetes care, and specifically highlight the importance of investment in improving diabetes data, which can guide population-tailored prevention and management approaches. [ABSTRACT FROM AUTHOR]

Broken water pumps continue to impede efforts to deliver clean and economically-viable water to the global poor. The literature has demonstrated that customers’ health benefits and willingness to pay for clean water are best realized when clean water infrastructure performs extremely well (>99% uptime). In this paper, we used sensor data from 42 Afridev-brand handpumps observed for 14 months in western Kenya to demonstrate how sensors and supervised ensemble machine learning could be used to increase total fleet uptime from a best-practices baseline of about 70% to >99%. We accomplish this increase in uptime by forecasting pump failures and identifying existing failures very quickly. Comparing the costs of operating the pump per functional year over a lifetime of 10 years, we estimate that implementing this algorithm would save 7% on the levelized cost of water relative to a sensor-less scheduled maintenance program. Combined with a rigorous system for dispatching maintenance personnel, implementing this algorithm in a real-world program could significantly improve health outcomes and customers’ willingness to pay for water services. [ABSTRACT FROM AUTHOR]

Background: Kenya aims to achieve universal health coverage (UHC) by 2030 and has selected the National Health Insurance Fund (NHIF) as the 'vehicle' to drive the UHC agenda. While there is some progress in moving the country towards UHC, the availability and accessibility to NHIF-contracted facilities may be a barrier to equitable access to care. We estimated the spatial access to NHIF-contracted facilities in Kenya to provide information to advance the UHC agenda in Kenya. Methods: We merged NHIF-contracted facility data to the geocoded inventory of health facilities in Kenya to assign facility geospatial locations. We combined this database with covariates data including road network, elevation, land use, and travel barriers. We estimated the proportion of the population living within 60- and 120-minute travel time to an NHIF-contracted facility at a 1-x1-kilometer spatial resolution nationally and at county levels using the WHO AccessMod tool. Results: We included a total of 3,858 NHIF-contracted facilities. Nationally, 81.4% and 89.6% of the population lived within 60- and 120-minute travel time to an NHIF-contracted facility respectively. At the county level, the proportion of the population living within 1-hour of travel time to an NHIF-contracted facility ranged from as low as 28.1% in Wajir county to 100% in Nyamira and Kisii counties. Overall, only four counties (Kiambu, Kisii, Nairobi and Nyamira) had met the target of having 100% of their population living within 1-hour (60 min) travel time to an NHIF-contracted facility. On average, it takes 209, 210 and 216 min to travel to an NHIF-contracted facility, outpatient and inpatient facilities respectively. At the county level, travel time to an NHIF-contracted facility ranged from 10 min in Vihiga County to 333 min in Garissa. Conclusion: Our study offers evidence of the spatial access estimates to NHIF-contracted facilities in Kenya that can inform contracting decisions by the social health insurer, especially focussing on marginalised counties where more facilities need to be contracted. Besides, this evidence will be crucial as the country gears towards accelerating progress towards achieving UHC using social health insurance as the strategy to drive the UHC agenda in Kenya. [ABSTRACT FROM AUTHOR]

The verbal autopsy (VA) questionnaire is a widely used method for collecting information on cause-specific mortality where the medical certification of deaths in childhood is incomplete. This paper discusses review by physicians and expert algorithms as approaches to ascribing cause of deaths from the VA questionnaire and proposes an alternative, data-derived approach. In this validation study, the relatives of 295 children who had died in hospital were interviewed using a VA questionnaire. The children were assigned causes of death using data-derived algorithms obtained under logistic regression and using expert algorithms. For most causes of death, the data-derived algorithms and expert algorithms yielded similar levels of diagnostic accuracy. However, a data-derived algorithm for malaria gave a sensitivity of 71% (95% Cl: 58-84%), which was significantly higher than the sensitivity of 47% obtained under an expert algorithm. The need for exploring this and other ways in which the VA technique can be improved are discussed. The implications of less-than-perfect sensitivity and specificity are explored using numerical examples. Misclassification bias should be taken into consideration when planning and evaluating epidemiological studies.

Introduction: Testing for recent HIV infection can distinguish recently acquired infection from long-standing infections. Given current interest in the implementation of recent infection testing algorithms (RITA), we report our experiences in implementing a RITA in three pilot studies and highlight important issues to consider when conducting recency testing in routine settings.Methods: We applied a RITA, incorporating a limited antigen (LAg) avidity assay, in different routine HIV service-delivery settings in 2018: antenatal care clinics in Siaya County, Kenya, HIV testing and counselling facilities in Nairobi, Kenya, and female sex workers clinics in Zimbabwe. Discussions were conducted with study coordinators, laboratory leads, and facility-based stakeholders to evaluate experiences and lessons learned in relation to implementing recency testing.Results: In Siaya County 10/426 (2.3%) of women testing HIV positive were classified as recent, compared to 46/530 (8.7%) of women and men in Nairobi and 33/313 (10.5%) of female sex workers in Zimbabwe. Across the study setting, we observed differences in acceptance, transport and storage of dried blood spot (DBS) or venous blood samples. For example, the acceptance rate when testing venous blood was 11% lower than when using DBS. Integrating our study into existing services ensured a quick start of the study and kept the amount of additional resources required low. From a laboratory perspective, the LAg avidity assay was initially difficult to operationalise, but developing a network of laboratories and experts to work together helped to improve this. A challenge that was not overcome was the returning of RITA test results to clients. This was due to delays in laboratory testing, the need for multiple test results to satisfy the RITA, difficulties in aligning clinic visits, and participants opting not to return for test results.Conclusion: We completed three pilot studies using HIV recency testing based on a RITA in Kenya and Zimbabwe. The main lessons we learned were related to sample collection and handling, LAg avidity assay performance, integration into existing services and returning of test results to participants. Our real-world experience could provide helpful guidance to people currently working on the implementation of HIV recency testing in sub-Saharan Africa. [ABSTRACT FROM AUTHOR]

Background: The RTS,S/AS01 vaccine against Plasmodium falciparum malaria infection completed phase III trials in 2014 and demonstrated efficacy against clinical malaria of approximately 36% over 4 years for a 4-dose schedule in children aged 5-17 months. Pilot vaccine implementation has recently begun in 3 African countries. If the pilots demonstrate both a positive health impact and resolve remaining safety concerns, wider roll-out could be recommended from 2021 onwards. Vaccine demand may, however, outstrip initial supply. We sought to identify where vaccine introduction should be prioritised to maximise public health impact under a range of supply constraints using mathematical modelling.Methods and Findings: Using a mathematical model of P. falciparum malaria transmission and RTS,S vaccine impact, we estimated the clinical cases and deaths averted in children aged 0-5 years in sub-Saharan Africa under 2 scenarios for vaccine coverage (100% and realistic) and 2 scenarios for other interventions (current coverage and World Health Organization [WHO] Global Technical Strategy targets). We used a prioritisation algorithm to identify potential allocative efficiency gains from prioritising vaccine allocation among countries or administrative units to maximise cases or deaths averted. If malaria burden at introduction is similar to current levels-assuming realistic vaccine coverage and country-level prioritisation in areas with parasite prevalence >10%-we estimate that 4.3 million malaria cases (95% credible interval [CrI] 2.8-6.8 million) and 22,000 deaths (95% CrI 11,000-35,000) in children younger than 5 years could be averted annually at a dose constraint of 30 million. This decreases to 3.0 million cases (95% CrI 2.0-4.7 million) and 14,000 deaths (95% CrI 7,000-23,000) at a dose constraint of 20 million, and increases to 6.6 million cases (95% CrI 4.2-10.8 million) and 38,000 deaths (95% CrI 18,000-61,000) at a dose constraint of 60 million. At 100% vaccine coverage, these impact estimates increase to 5.2 million cases (95% CrI 3.5-8.2 million) and 27,000 deaths (95% CrI 14,000-43,000), 3.9 million cases (95% CrI 2.7-6.0 million) and 19,000 deaths (95% CrI 10,000-30,000), and 10.0 million cases (95% CrI 6.7-15.7 million) and 51,000 deaths (95% CrI 25,000-82,000), respectively. Under realistic vaccine coverage, if the vaccine is prioritised sub-nationally, 5.3 million cases (95% CrI 3.5-8.2 million) and 24,000 deaths (95% CrI 12,000-38,000) could be averted at a dose constraint of 30 million. Furthermore, sub-national prioritisation would allow introduction in almost double the number of countries compared to national prioritisation (21 versus 11). If vaccine introduction is prioritised in the 3 pilot countries (Ghana, Kenya, and Malawi), health impact would be reduced, but this effect becomes less substantial (change of <5%) if 50 million or more doses are available. We did not account for within-country variation in vaccine coverage, and the optimisation was based on a single outcome measure, therefore this study should be used to understand overall trends rather than guide country-specific allocation.Conclusions: These results suggest that the impact of constraints in vaccine supply on the public health impact of the RTS,S malaria vaccine could be reduced by introducing the vaccine at the sub-national level and prioritising countries with the highest malaria incidence. [ABSTRACT FROM AUTHOR]

Background: In sub-Saharan Africa, women are most likely to receive skilled and adequate childbirth care in hospital settings, yet the use of hospital for childbirth is low and inequitable. The poorest and those living furthest away from a hospital are most affected. But the relative contribution of poverty and travel time is convoluted, since hospitals are often located in wealthier urban places and are scarcer in poorer remote area. This study aims to partition the variability in hospital-based childbirth by poverty and travel time in four sub-Saharan African countries. Methods: We used data from the most recent Demographic and Health Survey in Kenya, Malawi, Nigeria and Tanzania. For each country, geographic coordinates of survey clusters, the master list of hospital locations and a high-resolution map of land surface friction were used to estimate travel time from each DHS cluster to the nearest hospital with a shortest-path algorithm. We quantified and compared the predicted probabilities of hospital-based childbirth resulting from one standard deviation (SD) change around the mean for different model predictors. Results: The mean travel time to the nearest hospital, in minutes, was 27 (Kenya), 31 (Malawi), 25 (Nigeria) and 62 (Tanzania). In Kenya, a change of 1SD in wealth led to a 33.2 percentage points change in the probability of hospital birth, whereas a 1SD change in travel time led to a change of 16.6 percentage points. The marginal effect of 1SD change in wealth was weaker than that of travel time in Malawi (13.1 vs. 34.0 percentage points) and Tanzania (20.4 vs. 33.7 percentage points). In Nigeria, the two were similar (22.3 vs. 24.8 percentage points) but their additive effect was twice stronger (44.6 percentage points) than the separate effects. Random effects from survey clusters also explained substantial variability in hospital-based childbirth in all countries, indicating other unobserved local factors at play. Conclusions: Both poverty and long travel time are important determinants of hospital birth, although they vary in the extent to which they influence whether women give birth in a hospital within and across countries. This suggests that different strategies are needed to effectively enable poor women and women living in remote areas to gain access to skilled and adequate care for childbirth. [ABSTRACT FROM AUTHOR]

Background: The World Health Organization recommends "same-day" initiation of antiretroviral therapy (ART) for HIV patients who are eligible and ready. Identifying efficient, safe, and feasible procedures for determining same-day eligibility and readiness is now a priority. The Simplified Algorithm for Treatment Eligibility (SLATE) study evaluated a clinical algorithm that allows healthcare workers to determine eligibility for same-day treatment and to initiate ART at the patient's first clinic visit.Methods and Findings: SLATE was an individually randomized trial at three outpatient clinics in urban settlements in Johannesburg, South Africa and three hospital clinics in western Kenya. Adult, nonpregnant, HIV-positive, ambulatory patients presenting for any HIV care, including HIV testing, but not yet on ART were enrolled and randomized to the SLATE algorithm arm or standard care. The SLATE algorithm used four screening tools-a symptom self-report, medical history questionnaire, physical examination, and readiness assessment-to ascertain eligibility for same-day initiation or refer for further care. Follow-up was by record review, and analysis was conducted by country. We report primary outcomes of 1) ART initiation ≤28 days and 2) initiation ≤28 days and retention in care ≤8 months of enrollment. From March 7, 2017 to April 17, 2018, we enrolled 600 patients (median [IQR] age 34 [29-40] and CD4 count 286 [128-490]; 63% female) in South Africa and 477 patients in Kenya (median [IQR] age 35 [29-43] and CD4 count 283 [117-541]; 58% female). In the intervention arm, 78% of patients initiated ≤28 days in South Africa, compared to 68% in the standard arm (risk difference [RD] [95% confidence interval (CI)] 10% [3%-17%]); in Kenya, 94% of intervention-arm patients initiated ≤28 days compared to 89% in the standard arm (6% [0.5%-11%]). By 8 months in South Africa, 161/298 (54%) intervention-arm patients had initiated and were retained, compared to 146/302 (48%) in the standard arm (6% [(2% to 14%]). By 8 months in Kenya, the corresponding retention outcomes were identical in both arms (137/240 [57%] of intervention-arm patients and 136/237 [57%] of standard-arm patients). Limitations of the trial included limited geographic representativeness, exclusion of patients too ill to participate, missing viral load data, greater study fidelity to the algorithm than might be achieved in standard care, and secular changes in standard care over the course of the study.Conclusions: In South Africa, the SLATE algorithm increased uptake of ART within 28 days by 10% and showed a numerical increase (6%) in retention at 8 months. In Kenya, the algorithm increased uptake of ART within 28 days by 6% but found no difference in retention at 8 months. Eight-month retention was poor in both arms and both countries. These results suggest that a simple structured algorithm for same-day treatment initiation procedures is feasible and can increase and accelerate ART uptake but that early retention on treatment remains problematic.Trial Registration: Clinicaltrials.gov NCT02891135, registered September 1, 2016. First participant enrolled March 6, 2017 in South Africa and July 13, 2017 in Kenya. [ABSTRACT FROM AUTHOR]

In this study, experiments are conducted to gauge the relative importance of model, initial condition, and driving climate uncertainty for simulations of malaria transmission at a highland plantation in Kericho, Kenya. A genetic algorithm calibrates each of these three factors within their assessed prior uncertainty in turn to see which allows the best fit to a timeseries of confirmed cases. It is shown that for high altitude locations close to the threshold for transmission, the spatial representativeness uncertainty for climate, in particular temperature, dominates the uncertainty due to model parameter settings. Initial condition uncertainty plays little role after the first two years, and is thus important in the early warning system context, but negligible for decadal and climate change investigations. Thus, while reducing uncertainty in the model parameters would improve the quality of the simulations, the uncertainty in the temperature driving data is critical. It is emphasized that this result is a function of the mean climate of the location itself, and it is shown that model uncertainty would be relatively more important at warmer, lower altitude locations. [ABSTRACT FROM AUTHOR]

Background: The 2007 WHO algorithm for diagnosis of smear-negative pulmonary tuberculosis (PTB) including Mycobacterium tuberculosis (MTB) culture was evaluated in a HIV prevalent area of Kenya. Methods: PTB smear-negative adult suspects were included in a prospective diagnostic study (2009-2011). In addition, program data (2008-2009) were retrospectively analysed. At the first consultation, clinical examination, chest X-ray, and sputum culture (Thin-Layer-Agar and Lowenstein-Jensen) were performed. Patients not started on TB treatment were clinically re-assessed after antibiotic course. The algorithm performance was calculated using culture as reference standard. Results: 380 patients were included prospectively and 406 analyzed retrospectively. Culture was positive for MTB in 17.5% (61/348) and 21.8% (72/330) of cases. Sensitivity of the clinical-radiological algorithm was 55.0% and 31.9% in the prospective study and the program data analysis, respectively. Specificity, positive and negative predictive values were 72.9%, 29.7% and 88.6% in the prospective study and 79.8%, 30.7% and 80.8% in the program data analysis. Performing culture increased the number of confirmed TB patients started on treatment by 43.3% in the prospective study and by 44.4% in the program data analysis. Median time to treatment of confirmed TB patients was 6 days in the prospective study and 27 days in the retrospective study. Inter-reader agreement for X-ray interpretation between the study clinician and a radiologist was low (Kappa coefficient = 0.11, 95%CI: 0.09-0.12). In a multivariate logistic analysis, past TB history, number of symptoms and signs at the clinical exam were independently associated with risk of overtreatment. Conclusion: The clinical-radiological algorithm is suboptimal to diagnose smear-negative PTB. Culture increases significantly the proportion of confirmed TB cases started on treatment. Better access to rapid MTB culture and development of new diagnostic tests is necessary. [ABSTRACT FROM AUTHOR]

Objective: To systematically evaluate descriptive measures of spatial access to medical treatment, as part of the millennium development goals to reduce the burden of HIV/AIDS, tuberculosis and malaria.Methods: We obtained high-resolution spatial and epidemiological data on health services, population, transport network, topography, land cover and paediatric fever treatment in four Kenyan districts to develop access and use models for government health services in Kenya. Community survey data were used to model use of government health services by febrile children. A model based on the transport network was then implemented and adjusted for actual use patterns. We compared the predictive accuracy of this refined model to that of Euclidean distance metrics. RESULTS Higher-order facilities were more attractive to patients (54%, 58% and 60% in three scenarios) than lower-order ones. The transport network model, adjusted for competition between facilities, was most accurate and selected as the best-fit model. It estimated that 63% of the population of the study districts were within the 1 h national access benchmark, against 82% estimated by the Euclidean model.Conclusions: Extrapolating the results from the best-fit model in study districts to the national level shows that approximately six million people are currently incorrectly estimated to have access to government health services within 1 h. Simple Euclidean distance assumptions, which underpin needs assessments and against which millennium development goals are evaluated, thus require reconsideration. [ABSTRACT FROM AUTHOR]