Background
For over a decade, the Bioko Island Malaria Elimination Project (BIMEP) has collaborated with the Ministry of Health and Social Welfare of Equatorial Guinea to procure malaria commodities that are issued to public health facilities (HFs), including prevention materials, diagnostic testing supplies, and antimalarial treatments. Commodities procured by the project are inventoried in a central warehouse by BIMEP staff, who also manage their distributions to public HFs. The BIMEP has developed multiple iterations of Excel-based electronic logistics management information systems (eLMIS) to manage their central warehouse inventory, record stock distributions, and inform malaria commodity procurements.
Rationale
The previous eLMIS developed by BIMEP have supported record keeping efforts; however, forecasting consumption rates, triangulating stock and health services data, and estimating procurement needs were challenging. Such analyses were performed manually and often resulted in delays in reporting to headquarters, which subsequently contributed to stock-outs within the central warehouse, especially during the COVID-19 pandemic when global supply chains were strained. Additionally, these eLMIS were prone to data entry error and data loss, thereby compromising their reliability.
Figure 1: A BIMEP staff member performing routine malaria commodity distributions to public HFs using paper forms, which are subsequently digitized in the DHIS2 Web-browser Data Entry app at the field office.
As budget for malaria-related activities continue to decline, and global supply chains become increasingly strained, the BIMEP determined that an enhanced eLMIS was essential to ensure that an adequate stock of malaria commodities were available to protect vulnerable populations from one of the most frequent illnesses in this small, central African nation.
Already equipped with a robust DHIS2 platform to collect health services data, improve health worker quality of care, and conduct weekly malaria surveillance, the BIMEP saw an opportunity to leverage their existing DHIS2 instance and integrate it with a suite of malaria logistics modules. These modules would empower the team to better forecast procurement requirements, respond to facility needs, and ensure adequate stock and management of malaria commodities at all levels of the public health system.
Objectives
- Integrate World Health Organization (WHO) Malaria Logistics metadata package into existing DHIS2 instance and adapt it to country needs.
- Configure event modules to record stock received into/discarded from central warehouse.
- Configure core logistics indicators for each commodity.
- Configure predictors to estimate current stock-on-hand and estimated months of stock-on-hand at central warehouse given consumption rates to forecast procurement needs.
- Triangulate stock data with health services data to better understand stock performance and losses (stock discarded and stock discrepancies).
Methodology
Leveraging the development of the WHO common logistics data framework, the Aggregate DHIS2 metadata package for Malaria Logistics was integrated into the existing DHIS2 instance and adapted to meet country needs. The BIMEP staff are using the malaria logistics module to track malaria commodity distribution data from the central warehouse to the HF level. The package was adapted to include additional commodities for individual malaria treatment/prevention schemes as well as laboratory supplies for malaria diagnostics. Additional indicators and visualizations were configured to support programmatic efforts.
Figure 2: Malaria stock module adapted from the WHO common logistics data framework metadata package. Additional commodities were configured, and the layout was modified to resemble previous data-collection tools.
A DHIS2 Event program was also configured to model quantities of each malaria commodity received into the central warehouse, along with its expiration date and lot ID. DHIS2 notifications were configured to alert staff when commodities in the central warehouse are nearing expiration date. In addition, DHIS2 predictors were configured to create running totals of both stock received into and stock distributed from the central warehouse in order to generate indicators for estimated stock-on-hand and estimated months of stock-on-hand for each commodity, which are presented in a dashboard for managing stock procurements. Notifications also alert managers to begin procurement procedures when commodity stock-on-hand reaches a critical threshold given predicted consumption rates.
Paper-based forms are carried into the field for collecting data on commodity distributions, which are subsequently entered using the web-based DHIS2 data entry app; however, the transition to using the DHIS2 Android app has begun, and the app is expected to be fully employed by December 2022. The web-based DHIS2 event app is used at the central warehouse to enter stock received upon procurement.
Figure 3: The BIMEP staff members distributing malaria treatments to public HFs. Data is collected on paper forms while simultaneously piloting the use of the DHIS2 Android App.
Results and Discussions
Integrating the WHO malaria logistics metadata package into the existing DHIS2 instance was an easy decision, as the eLMIS design closely resembles the previous workflows, and enables streamlined triangulation of stock and health services data. When brainstorming the design process, guidance from DHIS2 eLMIS team members, @Breno and @George_MC_GUIRE, was instrumental in determining best practices for how to adapt the malaria logistics package to perform more granular modeling. The metadata package was easily augmented to meet the project’s needs for adding malaria commodities and supporting metadata like validation rules, predictors, indicators, and visualizations. Virtual academies offered by DHIS2 and @BAO_Systems helped to build BIMEP staff capacity, and ultimately all modules were able to be configured “in-house”.
Once the modules were integrated into the DHIS2 instance, the BIMEP team was trained on implementing data collection and interpreting results from dashboards. Users with DHIS2 experience from other departments were invited to assist in facilitating virtual trainings, ensuring the fundamental concepts of DHIS2 were well understood. These experienced users have previously benefited from a combination of both virtual and in-person training opportunities offered by DHIS2 and @BAO_Systems, and have received countless hours of on-the-job training. Moreover, these experienced users have also successfully trained dozens of data entry clerks and program managers. Weekly meetings were held to discuss technical challenges, conceptual frameworks, and results, using real program data in DHIS2 dashboards as practical case studies.
Achievements
One of the main benefits of the DHIS2 eLMIS is the automatically forecasted indicators that estimate stock-on hand and months of stock-on hand at the central warehouse. The previous system did not offer features for forecasting stock needs given commodity consumption rates. Consequently, there were multiple occasions when the central warehouse experienced stock-outs due to delays associated with health services and distribution data needing to be manually compiled, and procurement indicators needing to be manually calculated. By the time these calculations had been reported to headquarters for procurement, it was already too late, especially during the pandemic, when the global supply chain was already strained. The indicators and alerts configured in DHIS2 rapidly provided staff with the data they needed for decision making through stock procurement forecasting, which helped to ensure adequate stock-on-hand at the central warehouse and at public HFs.
Figure 4: Malaria commodity inventory dashboard in DHIS2 with indicators generated from predictors to estimate the average quantities distributed to HFs, the current balance in the central warehouse, and the estimated number of months of stock on hand.
An equally important benefit of the eLMIS is that it enables the team to easily triangulate stock distribution data with health services data. These indicator visualizations help the team to better understand stock performance and losses while also providing them with data to inform HF in-charges on how they may improve stock management within their respective HF. Within the first month of implementing the eLMIS in DHIS2, the team immediately noticed data discrepancies between the multiple reporting mechanisms, namely stock distribution reports, weekly malaria reports, and detailed patient registries, all of which are employed in the DHIS2 instance.
Visualizations such as that presented in Figure 5 below, highlight some of the discrepancies occurring in a District Hospital, in which the number of RDTs issued to patients was reported to be greater than the number of tests performed according to digitized patient registries. Even more alarming, the number of RDTs issued to patients was greater than the total number of consultations reported through the weekly malaria surveillance module during that period. Such visualizations, which were previously unattainable within a timely manner, provide the team with easily understandable evidence that they need to support investigations and react quickly. By triangulating stock and health services data, the teams discovered gaps in both internal and external reporting mechanisms, as well as potential sources of stock leakage, which will be discussed in future strategic planning meetings with key stakeholders. Furthermore, the data empowered the BIMEP team to more closely monitor their own internal workflows, thereby increasing accountability and transparency.
Figure 5: The eLMIS was integrated into the existing DHIS2 instance in January 2022, which has been used to captures core malaria indicators since 2018. Visualizations, like the chart above, have enabled the team to better understand discrepancies, like those occurring in a District Hospital in January and March, and provided them with the evidence they needed to investigate. By triangulating stock and health services data, the teams discovered gaps in both internal and external reporting mechanisms, which will be discussed in future strategic planning meetings.
Conclusions
Despite significant achievements, the BIMEP faced a number of challenges as they implemented the eLMIS in DHIS2. Due to the unavailability of certain malaria treatment schemes from pharmaceutical providers, many of the treatments in stock at the central warehouse needed to be repackaged for distribution in order to avoid stock-outs at HFs. Fortunately, the DHIS2 provided flexible means of modeling these changes, and novel indicators enabled the team to continue monitoring their inventory without any substantial complications. As this cost-effective eLMIS was not intended to replace a fit-for-purpose transactional eLMIS, the system is limited in its ability to estimate inventory and track commodity by lot numbers throughout the public health system. Nevertheless, the module created to record stock received into the warehouse has been configured to notify the team when commodities are nearing their expiration date.
Although most of the challenges have been met with solutions made possible by the flexibility of DHIS2, the biggest challenge remains: instilling a data-driven culture within the project to ensure adequate management of malaria commodities. While the tools are available in the eLMIS, the team is still learning how to use the data for decision making. Online academies offered through DHIS2 and @BAO_Systems have strengthened the staff’s technical skills; however, conveying the results generated in the system to the various stakeholders is an equally important process that requires a different set of skills and training. The BIMEP has a wealth of experience working with information systems, and these experiences have taught us that creating a culture of data-use is an iterative process in which progress is made through continuous systems refinement and feedback to all stakeholders.
The BIMEP team is looking at novel ways that they can further leverage their DHIS2 to model other activities and commodities in their inventory while continuously strengthening their efforts in eliminating malaria on Bioko Island.
and glad to read that the DHIS2 Online Academies help in building capacity 