Post authored by Michael Kohn, Medic’s Product Lead for Data and Analytics. Above, a community health worker in Siaya County, Kenya holds up paper logbook during a review meeting at the health facility.
These days system interoperability is front and center in the ongoing conversation about digital health. With tens of thousands of health workers supported by Community Health Toolkit (CHT) applications, it is important for their data to feed into national health management information systems (HMIS) so that ministry officials can make timely decisions based on accurate data. To that end, the Medic Mobile team has been hard at work designing, prototyping, and building integrations with key platforms like DHIS2, an open-source HMIS platform used in 67 low and middle-income countries.
The Community Health Toolkit + DHIS2
Most health systems have weekly or monthly reporting requirements for community-level activities. In communities without digital health apps, this means health workers must carry around heavy logbooks to manually record all relevant activities. When it is time to submit their data, community health workers summarize what was recorded and share this information with their supervisors, who in turn create paper records of these totals across entire community units or health facilities. This paper record is often passed to yet another individual whose responsibility is to manually key in the data into a health information management system, such as DHIS2.
In communities using digital health apps built with the Community Health Toolkit, we hope to drastically reduce the effort required to fulfill these reporting requirements, while increasing accuracy and timeliness of data submission. So far, we’ve explored several versions of what integrating with DHIS2 to aggregate data already collected on a digital health app might look like.
Manual Export and Import
Early in 2018, we developed a solution to export aggregate data for a community unit in the format required by DHIS2. At the end of the month, a supervisor (for example) reviews the data, downloads a file, and uploads that file to DHIS2 using the “Data Import” function. This solution enables faster and more accurate reporting without altering a health worker’s previous routine; they simply continue to use their app as they normally would. One advantage of this solution is that it is relatively straightforward to configure, quick to deploy, and doesn’t require much technical effort.
Automated Export and Import
Later in the year, we prototyped a solution which automatically sends aggregate data directly to DHIS2 on a fixed schedule each week or month. Community health workers use their digital health app as normal, and at a predefined time, a process runs to aggregate data across the entire community unit/health facility and automatically send that data to DHIS2. This solution utilizes another open-source toolkit, Apache NiFi, to manage the schedule and communication between the two systems.
In this solution, organizations are able to get comfortable with seeing digital health information in parallel with paper-based data, without replacing paper-based reporting entirely. A second dataset can be configured in DHIS2 to record data coming from their community health app, and then comparisons can be made between paper-based reporting and app data, all from within DHIS2.
Fully Integrated HMIS Workflows
In November of last year, five Medic teammates visited two counties in Kenya, shadowing and interviewing dozens of health workers and Ministry of Health teammates about their duties relating to DHIS2 reporting and the various paper tools currently in use. The goal of this trip was to gain a better understanding of how we might support and improve those workflows entirely within CHT apps. The team learned about the available infrastructure, indicator interpretations, meetings conducted throughout the month around DHIS2, and overall end-to-end workflows.
Last month, we completed our initial design for fully integrated workflows to report HMIS data within CHT apps. We are currently iterating and refining the design based on ongoing feedback. Broadly, the workflow will address three main areas of functionality: (1) community health workers will be able to review and sign-off on aggregate monthly summaries; (2) supervisors will be able to review and approve aggregate monthly summaries; and (3) supervisors will be able to submit aggregate data to DHIS2.
We are excited to continue building out our capabilities in this area and expect to start development on the fully supported DHIS2 reporting workflows in the coming months. Please feel free to reach out to me (michael@medicmobile.org) with your questions and experiences about integrations.