| Organizations | Clinton Health Access Initiative (CHAI); mSupply Foundation; University of California, Berkeley |
| Location | Laos |
| Amount | $3,812,005 |
| Approval date | March 2026 |
| Funded by | Unrestricted funds designated for grantmaking |
Summary
This grant will support the implementation of an electronic logistics management information system (eLMIS) tool that provides real-time reporting on health supply stock levels, and fund a randomized controlled trial (RCT) of its use in health clinics in Laos to generate evidence on whether supply chain digitization programs reduce stockouts and increase access to treatment.1
Published: September 2026
The challenge
Many health systems in low- and middle-income countries rely on paper registers to track the stock of health commodities in individual health facilities.2 That information is periodically aggregated at higher levels within a health system, creating an information lag.3 As a result, staff may be less able to accurately predict inventory levels (and reorder commodities in advance), which can lead to either stockouts, making it more difficult for people to access the treatment they need, or overstock, potentially resulting in expired medicines, wasting limited resources.
How this grant could help
This grant will fund CHAI and the mSupply Foundation to implement an electronic logistics management information system (eLMIS), using the Open mSupply tool, in local health clinics across five provinces in Laos, where an eLMIS system is already being implemented at the central, provincial, and district warehouse levels.4 The grant will cover personnel costs, health worker training, technical support, and other costs.5 The grant will also fund personnel, data collection, and other costs for an RCT of the program in 176 health facilities, which includes:6
- tracking stockouts
- gathering data from patients about
- whether they received their preferred product and
- if not, the reasons and where they subsequently sought care.
Our reasoning
Tracking health commodity stock levels across the supply chain using eLMIS software could be a promising way to improve access to treatment by ensuring availability of medicine at points of care, using data to improve information management practices. While many countries have begun to implement eLMIS systems, implementation has not been consistent. For example, in some locations, it has been implemented only for warehouses, not local health facilities, and in other locations it is used to track only a few specific commodities, such as vaccines.7
In addition, there is limited evidence about whether using an eLMIS reduces stockouts or increases access to treatment.8 Most existing research on eLMIS programs focuses only on stockouts,9 but the impact of stockouts on treatment is not straightforward. For example, some health conditions may not require rapid treatment, patients might be referred to other facilities, or patients might seek care in the private sector. An earlier modeling study estimated that reducing stockouts could result in reduced mortality, but there has not been a rigorous evaluation of this hypothesis.10
We believe this grant would fund the first RCT of the use of eLMIS programs in local health clinics and the first to measure the relationship between stockouts and access to treatment. We think the evaluation will generate rigorous evidence on whether supply chain digitization programs reduce stockouts, and whether those stockout reductions lead to increases in patients receiving appropriate treatment.
Both positive and negative results could affect our future grantmaking. If we learn that eLMIS implementation improves patient treatment access, we may consider funding additional implementation. If the study indicates that the program does not reduce stockouts or improve access to treatment, this may mean that barriers other than supply tracking are more important, thus shifting how we prioritize future funding opportunities.
We are recommending this grant primarily for its learning value. Creating cost-effectiveness estimates for value of information grants is more challenging than for direct delivery programs; however, we estimate this grant to be 7 times more cost-effective than our benchmark.
How we could be wrong
The RCT will not help us resolve all of our uncertainties about supporting eLMIS programs. We think it will reduce our uncertainty about two key parameters:
- the effect of eLMIS on stockout rates
- the relationship between stockout reductions and patient treatment access
However, we will continue to have some uncertainty about other factors, such as health outcomes and whether findings from the study, which is being conducted in Laos, apply to settings in Africa, where we expect programs that support eLMIS deployment to be substantially more cost-effective because of higher disease burdens. Because of this, we may eventually want to fund additional trials in other settings.
In addition, we may not be addressing the right bottleneck. The theory of change assumes that the limitations of paper-based stock management are a key barrier to stock availability and treatment. If other factors, such as national stock availability or staff training, are also major constraints, this program may have limited impact.
Sources
This page summarizes our research at the time of approval and was reviewed by the grant recipient.
- 1This grant is a collaboration between CHAI (as the primary grantee and coordinating body), the mSupply Foundation, UC Berkeley, and the Lao Tropical and Public Health Institute (supporting data collection).
- 2"About 25% and 40% of national and sub-national stores respectively still use paper-based stock management systems." WHO/UNICEF, "Effective Vaccine Management (EVM) Global Data Analysis 2009-2020," p. 34.
- 3"In Laos today, data on medicine availability at health centers is collected manually for a limited set of program commodities, such as malaria and vaccines. Health workers count stock by hand at the end of each month, fill out paper forms, and in some cases enter minimal data into DHIS2. This process is time-consuming, fragmented, and not scalable beyond a few programs or products.” CHAI, Responses to GiveWell questions, 2026 (unpublished)
- 4"CHAI began partnering with the government of Lao PDR in 2014, supporting the country's supply chain system reforms. Today, because of the program, all districts across the country have access to real-time stock data for the supplies needed to treat HIV, tuberculosis, malaria, and reproductive health." CHAI, "Lao PDR."
- 5"Personnel . . . mSupply Implementation Training & Supervision . . . mSupply Implementation Hardware/Software." CHAI, mSupply Budget Narrative, 2026.
- 6
- "External evaluator . . . data collection." CHAI, mSupply Budget Narrative, 2026.
- “The project will be based across five provinces in Laos and will evaluate the implementation and outcomes of the Open mSupply system in 176 district health center.” CHAI, "mSupply: Discussion note," 2025. Note that we now expect this to be based in five, not four, provinces.
- "Measured share of products in stock throughout the month (main outcome variable)." UC Berkeley, "Data collection design and power," 2026 (details likely to change from these initial estimates)
- "In addition to health centre outcomes, we could collect patient outcomes with a supplementary data collection activity. . . . During the survey, enumerators would first screen patients on whether they were able to access their preferred health product. . . . For those who answer no, the enumerator will proceed with the short survey described above and invite the patient to share their contact information for a follow-up interview . . . .The goal of the follow-up interview is to collect data on whether they decided to visit an alternative site for care, as well as additional patient outcomes related to the medical care for their condition." UC Berkeley, "Data collection design and power," 2026 (Unpublished; details likely to change from these initial estimates)
- 7"In Laos today, data on medicine availability at health centers is collected manually for a limited set of program commodities, such as malaria and vaccines. Health workers count stock by hand at the end of each month, fill out paper forms, and in some cases enter minimal data into DHIS2. This process is time-consuming, fragmented, and not scalable beyond a few programs or products.” CHAI, Responses to GiveWell questions, 2026 (unpublished)
- 8For instance, stockouts may be caused by other barriers, such as breakdowns in the supply chain or spoilage, rather than poor tracking, and patients may be able to access treatment in other ways when local health facilities are out of a particular commodity.
- 9We are not aware of any RCTs evaluating the impact of facility-level eLMIS deployment on stockout rates, treatment access, or health outcomes. During our literature review, we found Fritz, Herrick, and Gilbert 2021, a modeling study, which draws on three studies of eLMISs. None of the three studies are RCTs, and all three focus on stockouts, not treatment access: Gilbert et al. 2017 ("This retrospective investigates improvements and stabilization of supply chain performance following introduction of the digital information system."), Gilbert et al. 2020 ("Six months of stockout rates pre- and post-introduction, by antigen, were compared via a two-way analysis of variance (ANOVA)."), and Mwencha et al. 2017 ("We used a nonexperimental pre-post study design to compare the previous system with the upgraded management system.").
- 10"This study projects that digitalization of last-mile LMIS would reduce child mortality by improving coverage of lifesaving health commodities." Fritz, Herrick, and Gilbert 2021