JSI works with clients and countries around the globe to deploy innovative user-centered logistics management information systems (LMIS) and other tools that help people make better decisions and take effective action. We have learned a lot from these experiences, such as: What makes a system implementation successful? Once data is available, how can it be used to make decisions and improve performance? What other HIS solutions and processes are linked to these digital LMIS? We will be sharing stories and lessons learned specifically from Zambia, Tanzania and Ethiopia through our Digital LMIS Innovations Blog Series; stay tuned for exciting episodes coming to a screen near you!
As supply chain professionals working in the global health sector, we have heard countless requests and statements over the years to integrate data from HMIS and LMIS tools. This is partly to streamline reporting channels and to reduce the reporting burden, and partly to attempt to compare information between the two systems. Often this conversation is burdened by a lack of understanding about how the two data sets differ.
Quick Data Systems 101: The data collected in HMIS tools mainly focus on service delivery but usually include indicators related to drug or vaccine supply, such as incidents and duration of stockouts. LMIS data are collected and used for primarily operational supply chain decisions, but are also used to generate performance indicators.
As countries adopt electronic information systems to manage HMIS and LMIS, it is easier to compare service and logistics data, and DHIS2 provides a useful platform for integrating and visualizing these data together.
With funding from the UN Commission on Life-Saving Commodities for Women and Children, JSI, University of Oslo, University of Dar es Salaam, and VillageReach worked with the Ministry of Health, Community Development, Gender, Elderly and Children in Tanzania to develop an integrated dashboard to look at RMNCH and supply data together. This was achieved by adding a data feed from the eLMIS to DHIS2; two systems that had been deployed nationally in 2013–2014.
This ground-breaking work initially focused on technical feasibility of data exchange. And there were certainly hurdles related to that sort of interoperability between the two systems, like how to align data with different reporting cycles, match facility names, and ensure that the database and software were updated to enable data sharing.
But once we got over those hurdles, the effort raised some bigger questions:
- How would decision makers interpret the data and set up decision rules and routine actions based on the data available? For example, how do users determine what is a data quality issue versus a service delivery or a supply problem?
- What data is the most useful to compare? For example, is it meaningful to compare cases of postpartum haemorrhage with vials of oxytocin dispensed if there is no expected ratio to indicate rational use or a supply problem? Perhaps not, but the data might be more useful for supplies and services with a more predictable product-to-service ratio (i.e. contraceptive implants dispensed, implant insertions reported, and tracking implant removals).
The team went through an iterative user requirements process with MOH RMNCH stakeholders to identify metrics that would be useful, creating mock-ups for refinement and feedback and holding workshops to discuss dashboard use and options for visualization. This participatory approach allowed users to define how the dashboard could be most useful to support their routine work and decision making.
While the Tanzania dashboard is an exciting milestone, additional effort is needed to unlock the full potential of this type of integration. It is clear that the first step is to identify and rectify data quality issues to allow comparability. After that, indicators need to be refined, and data interpretation needs to be better defined to allow the creation of business rules—what actions to take—around these indicators, and to support predictive analysis based on the trends in the data. This would allow the dashboard to be used to guide decision making and action—(e.g. where additional supervision is needed or if disease burden is increasing where stock levels should be replenished) to improve health outcomes.
It’s also important to keep the big picture and future of health information systems in mind. While the work in Tanzania demonstrates the feasibility of interoperability between DHIS2 and eLMIS, the country is moving beyond “peer-to-peer” connections between applications. Already, the country has deployed a master facility registry, a human resources information system, financial managements systems, and many different mHealth applications. Hospital management systems are also being used, and there is a strong commitment to deploying electronic health record (EHR) systems.
With this increasing complexity, Tanzania has recently developed a vision for comprehensive interoperability within its health information ecosystem, with a health information mediator serving as the central hub through which all applications will eventually connect. Other countries seeking to link LMIS applications with DHIS2 will also have to grapple with other health information applications, and should see the immediate LMIS-HMIS data exchange as an interim step toward a more holistic and interoperable health information system.
The increasing complexity on the horizon only underscores the contribution this work has made to understand the challenges around data quality and to ensure comprehensive and participatory stakeholder engagement to use data to improve health services and health outcomes.