Relationship of Logistics Management Information System Data Quality With Essential Medicine Stockouts in Malawi Using Spearman Rank Correlation Analysis

In the quiet hum of a hospital pharmacy, a simple question can determine whether a patient lives or dies: is the medicine on the shelf? For health systems in low-income countries, the answer often depends on a digital record. These records, known as logistics management information systems, are the nervous system of a supply chain. They track what medicines are needed, what has been ordered, and what has arrived. When these records are accurate, doctors can treat patients. When they are broken or incomplete, the system fails, and essential drugs disappear from the shelves, leaving clinics empty-handed. This failure is not just a clerical error; it is a gap between what the computer says is happening and what is actually happening in the room. Understanding the link between the quality of these digital records and the reality of empty shelves is crucial for fixing health systems that struggle to keep their promises to the people they serve.

In the Blantyre District of Malawi, a researcher set out to measure this exact link. The country relies heavily on a national digital platform called Open-LMIS to manage its medicine supply, yet patients still frequently face shortages of life-saving drugs. The researcher wanted to know if the numbers recorded in this system could predict the actual burden of these shortages. To find out, they looked at six government-run health facilities, ranging from a large central hospital to smaller community clinics. They focused on nine specific essential medicines, tracking their presence over five months in 2023. The goal was to see if the data the system generated—specifically, how often medicines were available and how well orders were filled—matched the reality of how many days patients went without those drugs.

The investigation revealed a stark and clear pattern. At the Madziabango Health Centre, where the digital records showed medicines were available only about 69 percent of the time, the reality was harsh: the clinic suffered from 269 days of stock-outs across the study period. In contrast, the Chilomoni Health Centre, where records showed medicines were available 100 percent of the time, experienced zero days without stock. The data showed that the lower the recorded availability, the higher the number of days the clinic ran out of medicine. This relationship was strong and showed a clear trend, suggesting that the digital record is a reliable mirror of the physical reality. When the system says a facility is running low, it is almost certainly running low, though the small sample size means these results should be interpreted as indicative rather than definitive.

However, the study also uncovered a more complex story about why these shortages happen. The researcher looked at "order fill rates," which measure how much of an order a facility actually receives compared to what it asked for. One might expect that if a facility received more of its ordered medicines, it would have fewer stock-outs. But the data did not show a strong, clear link between receiving more orders and having fewer empty shelves. In fact, some facilities that received very few deliveries still managed to keep some medicines in stock, while others with better delivery rates still faced gaps. This suggests that the problem is not just about how well the local facility tracks its own inventory, but also about a deeper, structural failure in the supply chain itself. The central medical stores sometimes simply do not send the medicines that are ordered, regardless of how well the local clinic reports its needs.

The specific medicines that disappeared most often told a troubling story of their own. Phenobarbitone, a critical drug for treating epilepsy, was missing for 251 days across the facilities studied. Morphine sulphate, essential for pain relief in palliative care, was missing in 60 percent of the observations where it was expected. These are not just numbers; they represent patients going without treatment for seizures or pain. The study found that while the local data quality was a strong predictor of these shortages, the root cause often lay in the broader system failing to deliver the goods. The digital records were accurate enough to show the problem, but the physical supply chain was not robust enough to solve it.

The findings offer a clear path forward for health officials. Because the digital records are so closely tied to the reality of stock-outs, they can be used as an early warning system. If a clinic's recorded availability drops below a certain point, health planners can intervene before the shelves go completely empty. The researcher recommends setting up a simple monitoring dashboard that triggers alerts when availability falls, allowing for proactive redistribution of medicines. At the same time, the lack of a strong link between order fill rates and stock-outs highlights that fixing the local data is not enough. The central supply chain must also be strengthened to ensure that when a clinic orders medicine, it actually arrives. The study concludes that improving the quality of data entry is a low-cost, high-impact way to strengthen the health system, but it must be paired with a commitment to fixing the physical delivery of medicines to ensure that the numbers on the screen translate into health for the people.