Search

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: In Kenya, access to early infant diagnosis and viral load monitoring services for HIV patients on ART is significantly hampered by sample transportation challenges and long turnaround times. Near patient care testing technologies have the potential to obviate such constraints. The Cepheid GeneXpert was launched in 2010 as a TB assay and in 2014 as a potential point of care HIV viral load assay. Whereas it is widely is used for TB in Kenya, its utility for HIV testing has not been evaluated. Objective: To investigate the performance and usability characteristics of the GeneXpert HIV-1 qualitative and quantitative assay. Methods: This was a cross sectional study among 911 HIV Exposed infants and 310 HIV positive adults. Existing machines used for routine TB diagnosis were used in this study. The diagnostic accuracy of the qualitative assay was assessed using Roche CAP/CTM while the quantitative assay was assessed using with Abbott m2000 as the reference assays respectively. Statistical analysis was done using Stata/MP Version 14 for Mac. Concordance values and misclassification were calculated at the clinical cutoff of 1000 cp/ml. Results: The sensitivity, specificity and accuracy of the GeneXpert HIV-1 qualitative assay were 99.23% (95% CI 97.24–99.90%), 98.91% (95% CI 97.76–99.55%) and 99.00% respectively. For the quantitative assay, they were 92.50% (95% CI 79.61–98.43%), 100.00% and 97.00% respectively. All 30 (100%) users reported that the GeneXpert machine was easy to use, workflow was simple and TB diagnosis was not negatively affected. In our hands, the median turn-around time for an individual qualitative and quantitative test was 90 minutes. A total of 58 (4.34%) errors and 28 (2.10%) invalid outcomes were experienced; 44 (3.29%) tests did not run to completion due to power outages. Conclusion: GeneXpert HIV-1 qualitative and quantitative assay is an accurate test for the diagnosis of HIV in infants and for viral load monitoring. At the point of care, the GeneXpert machine's simple work flow, ease of use and short test turnaround time present the potential to improve access to HIV testing and viral load monitoring. To integrate HIV diagnosis into the existing GeneXpert platforms for TB Diagnosis, there is need to scale up the infrastructure and to change the way work is done. [ABSTRACT FROM AUTHOR]