Medic Mobile

Digital tools make it easier to collect data about patients closer to where they live, understand their health needs better, and treat them faster, thereby saving lives. Community Health Workers (CHWs) are increasingly collecting digital data through the course of care delivery. However, despite the widespread championing of digital health tools, the full impact of these technologies has not been realized because CHW-collected data is often considered low quality and unreliable for data-driven decision-making. At a systemic level, mistrust in the quality of the data (a lack of so-called ‘data trust’) limits the potential impact. The primary objective of this research program was to identify inconsistent or problematic data (IoP) occurring across a digitally enabled CHW health system in order to recommend changes in digital health tools and processes that might increase trust in CHW-collected data. In this exploratory study, data were analyzed using a variety of statistical and data science approaches including clustering algorithms, histograms, box plots, and Sankey diagrams. IoP were identified from a suite of 160 tests internally developed to identify IoP at the health platform level. As anticipated, data exhibited issues with accuracy, completeness, and timeliness across digital health forms. While each IoP issue identified could be individually remediated, recommendations provided are centered on a platform-wide (and tool agnostic) approach to data quality in community health.

Rapid diagnostic tests (RDTs) provide point-of-care medical diagnosis without sophisticated laboratory equipment, making them especially useful for community health workers (CHWs). Because the procedure for completing a malaria RDT is error-prone, CHWs are often asked to carry completed RDTs back to their supervisors. Doing so makes RDTs susceptible to deterioration and introduces inefficiencies in the CHWs' workflow. In this work, we propose a smartphone-based RDT capture app, RDTScan, that facilitates the collection of high-quality RDT images to support CHWs in the field. RDTScan does not require an external adapter to control the image capture environment, but instead provides real-time guidance using image processing to obtain the best image possible. During our evaluation study, we found that RDTScan had 98.1% sensitivity and 99.7% specificity against visual inspection of the RDTs. RDTScan helped CHWs capture high-quality RDT images within 18 seconds while enabling a better RDT workflow.

Abstract Background: Surgical male circumcision (MC) safely reduces risk of female-to-male HIV-1 transmission by up to 60%. The average rate of global moderate and severe adverse events (AEs) is 0.8%: 99% of men heal from MC without incident. To reach the 2016 global MC target of 20 million, productivity must double in countries plagued by severe healthcare worker shortages like Zimbabwe. The ZAZIC consortium partners with the Zimbabwe Ministry of Health and Child Care and performed over 120,000 MCs. MC care in Zimbabwe requires in-person, follow-up visits at post-operative days 2,7, and 42. ZAZIC program adverse event rate is 0.4%; therefore, overstretched clinic staff conducted more than 200,000 unnecessary reviews for MC clients without complications. Methods: Through an un-blinded, prospective, randomized, control trial in two high-volume MC facilities, we will compare two groups of adult MC clients with cell phones randomized 1:1 into two groups: 1) routine care (control group N=361) and 2) clients who receive and respond to a daily text with in-person follow-up only if desired or if a complication is suspected (intervention N=361). If an intervention client responds affirmatively to any automated daily text with a suspected AE, an MC nurse will exchange manual, modifiable, scripted texts with the client to determine symptoms and severity, requesting an in-person visit if desired or warranted. Both arms will complete a study-specific, Day 14, in-person, follow-up review for verification of self-reports (intervention) and comparison (control). Data collection includes extraction of routine client MC records, study-specific database reports, and participant usability surveys. Intent-to-treat (ITT) analysis will explore differences between groups to determine if two-way texting (2wT) can safely reduce MC follow-up visits; estimate the cost savings associated with 2wT over routine MC follow-up; and assess the acceptability and feasibility of 2wT for scale-up. Discussion: It is expected that this mobile health intervention will be as safe as routine care while providing distinct advantages in efficiency, costs, and reduced healthcare worker burden. The success of this intervention could lead to adaptation and adoption of this intervention at the national level, increasing efficiency of MC scale up, reducing burdens on providers and patients. Clinical Trial Registration Number NCT03119337

This paper explores the design space of feedback systems that connect care recipients to the community health feedback loop. While related work in this vein has often emphasized gathering feedback for the sake of transparency alone, our study emphasizes opportunities to integrate the collection and use of feedback in ways that may improve the quality or equity of routine health services. We conducted a qualitative study using semi-structured interviews and focus groups with 23 participants in Kenya. Our field study makes current feedback practices visible; and reveals barriers faced by beneficiaries, community health workers, and their supervisors. Our findings identify relevant socio-technical complexities, and we outline concrete opportunities to design feedback systems that support and augment current practices. These contributions to the ICTD literature hold potential to inform the design of feedback systems that engage underserved populations in a systematic and equitable manner.