Advanced Analytics: COVID-19 Revealed Weaknesses in Existing Systems, Thereby Accelerating Trends Toward Use of Analytical Tools

Related Supply Chain Topics
Related Global Health Areas
April 12, 2021
Lead Paragraph/Summary

This is the first in a series about how GHSC-PSM adapted its programming to the new COVID-19 environment. Check out the secondthird and fourth installments.

Advanced analytics supports decision-making through access to near real-time data and analysis. GHSC-PSM has an advanced analytics team based in the Washington, DC, metro area that supports customized tools for various country programs and for a range of program areas.

Long before COVID-19 and before the availability of sophisticated file sharing and communication technologies, advanced analytics advisors traveled to partner countries to understand business processes to inform the design of tools and obtain data. Until more recently, remote sharing of information and analysis of data was difficult.

The past ten years have brought a significant change in the ability to exchange and analyze data. Internet and cellular connectivity have improved. File-sharing platforms have replaced email attachments. Daily, computer-driven, automatic downloads of data have replaced monthly or quarterly, people-driven, manual transactions.

The speed at which the COVID pandemic spread also accelerated the demand for supply chain data insights in USAID-supported countries. It exposed the weaknesses of existing data systems that couldn't respond to basic questions like, "How much space is in our warehouses now, and how much will there be when large shipments of personal protective equipment arrive?"

Case in point: Zambia. In Zambia, GHSC-PSM designed and deployed four customized advanced analytics tools through a remote engagement model that allows work to continue uninterrupted and without the need for international travel. In addition to replacing time-consuming manual processes, the tools enable more timely action to prevent supply risks. These remote solutions are ideal for supply-chain management both during the COVID-19 pandemic and for years to come. These tools were designed to be automated and easily deployable in other country contexts. The tools include:

Stock Re-distribution Tool

Zambia Stock re-distribution tool

In April 2019, GHSC-PSM implemented the Zambia stock re-distribution tool, a map-based interface for making stock re-distribution decisions remotely from any location. The tool pulls data from the existing LMIS and automatically analyzes it to identify stock transfer opportunities to support timely decision-making at the provincial health offices (PHOs). The tool uses color-coding and other visual cues to recommend possible transfers to mitigate any facility's stockout risk. From May to August 2020, remote use of the tool helped resolve multiple supply risks and prompted the re-distribution of ARVs, essential medicines, viral load reagents and other lab commodities in the Western Province, re-distribution of Depo Provera in Chilanga District and re-distribution of malaria medicines in Luangwa District.

Consumption Anomaly Detection Tool

Consumption anomaly detection tool

Identifying consumption anomalies is vital to determine supply-chain risks, but these anomalies were difficult to pinpoint amidst the large volume of information in Zambia. GHSC-PSM designed and implemented an automated consumption anomaly detection tool in early 2019 by adapting a private-sector Statistical Process Control (SPC) for the supply chain. In 2020, project staff used the consumption anomalies tool monthly (and by remote) to check more than 200,000 records at more than 2,900 facilities to detect and call to the MOH's Pharmacy Unit's attention and chief provincial pharmacists any abnormalities in consumption. This report supports the National Drug Theft Task Force in intelligence gathering, while provincial task forces use the tool to conduct audits and determine supportive supervision needs.

Hub Capacity Tool

GHSC-PSM developed an Excel-based analytics tool in June 2020 to help the Zambian Medical Stores Limited (MSL) make strategic decisions about which commodities to store in a provincial hub. To reduce the number of commodities managed at the central level and move key commodities closer to facilities, MSL used the hub tool in selecting 40 critical public health commodities that could potentially be stored in provincial hubs instead of the central warehouse, reducing the delivery time and improving the reliability of supply. With remote support from GHSC-PSM in June, MSL used the hub capacity tool to determine that the Luanshya Hub (Copperbelt Province) could store only 15 of the potential 40 commodities because of storage constraints—with a two-month stock holding—allowing MSL to refine their distribution strategy for this province. With lessons learned from Luanshya, MSL used the tool to determine which commodities can be stored at all the other hubs.

Stockout Trend Detection Tool

Stockout Trend Detection Tool

Because of international supply interruptions caused by COVID-19, GHSC-PSM expanded routine monitoring of stockout trends from tracer commodities to all 1,000 commodities to prevent supply risks for non-tracer products. The project developed and deployed a stockout trend detection tool to analyze this much larger data set. It simultaneously runs multiple tests on all products to determine whether there is an increasing trend in stockouts. Then, it alerts GHSC-PSM and the project's counterparts in MSL when stockouts of certain commodities are on the rise. 

Removing the Bottleneck of Information

Pre-pandemic, social contact during formal and informal meetings and exchanges had helped mask the silos in which data have been kept, but social isolation created by COVID helped identify and then bridge unnecessary separations.

By revealing risks, COVID has accelerated the trend towards automated, remote data sharing and analysis by making online exchange the only viable option both internationally and within a country's borders. Where governments once resisted providing access to their data, many are now more open to doing so.

Enabled by the continuous, free flow of data, advanced analytics can shift from just high-level strategic analysis – where it has historically operated – to operational and tactical, including order management, distribution, warehousing, commodity consumption trends, expiry management and early warning systems.

Case in point: Haiti. Haiti illustrates the importance of having access to data among various teams. An existing data visibility dashboard already showed stock status in the warehouse based on quality, value and volume. The tool, which uses data from two systems – called SYGDOCC and MACS – helps manage expiries, so data availability timing is critical. A recent improvement has automated data sharing to a daily process, pulling data automatically from MACS, a warehouse management system. The system is not yet able to pull data from SYGDOCC, but the team in Haiti is working toward that goal.

Expertise Shifts from International Consultants

COVID-19 has temporarily grounded the international advisors who would normally lead short-term advanced analytics discussions in USAID-supported countries. Shifting to remove support has tapped into the potential of local staff, who often deferred to the visiting advisors despite their greater knowledge of the local context.

Remote support makes the best use of both advisors and local staff; the advisors focus on their expertise in designing advanced analytics tools, while local staff leads mapping of business processes and identifying needs, leveraging Microsoft Teams and other online tools to facilitate discussions. International advisors now partner with local staff experts to ensure that data dashboards and other tools are designed following the decision processes and data needs of local partners. This new division of labor has elevated local staff in the eyes of their colleagues across partner organizations.

Long-lasting Change

The changes brought by COVID-19 may last far beyond the public health crisis. Advanced analytics has succeeded most where robustness in data exists, allowing partners to see what was previously obscured by too little data or ineffective analysis of vast amounts of data. Presenting what's most important to decision-makers for both COVID-response and existing public health programs has accelerated the appreciation for an open exchange of information. Advanced analytics shows that data access and visibility are at times not sufficient; insightful analysis and presentation of data help enable decisions and actions that create even greater value for the supply chain.