Quantifying medicine needs: a practical guide to forecasting for health supply chains

Few problems frustrate health workers and patients more than an empty shelf. Stock-outs of essential medicines and supplies are often blamed on late deliveries or weak distribution, yet many begin earlier, with a forecast that did not reflect what facilities actually needed. The opposite failure is just as costly: overstocking leads to expiry, wasted budgets and storage strain. Quantification, the process of estimating what to buy, how much and when, deserves more attention than it usually receives.

Start with the question, not the spreadsheet

Before any numbers are touched, agree what the forecast is for. Is it a national annual procurement plan, a provincial replenishment cycle, or a donor-funded programme with a fixed budget? The purpose determines the time horizon, the level of detail and the people who must be in the room. Quantification works best when programme managers, pharmacists, finance officers and procurement staff share ownership rather than handing the task to a single analyst.

Know your data sources, and their limits

Forecasts usually draw on three kinds of data, each with weaknesses:

• Consumption data from stores and facility registers, which understate demand whenever stock-outs occurred.
• Service data, such as patient visits or enrolments, which show demand but not always what was prescribed.
• Demographic and morbidity estimates, useful for new programmes but only as accurate as their assumptions.

The practical lesson is to triangulate. Where consumption and service data tell different stories, investigate before choosing one. Adjusting consumption for periods without stock is a simple step that is frequently skipped.

Make assumptions visible

Every forecast rests on assumptions about treatment guidelines, coverage targets, growth in demand and wastage. Write them down in plain language and keep them with the forecast. When results later differ from the plan, a documented assumption tells you whether the model was wrong or the world changed. Where policy is uncertain, prepare a low, base and high scenario instead of a single figure, so decision-makers understand the range of risk.

Connect the forecast to budget and procurement timing

A technically sound forecast can still fail if funding is released late or procurement lead times are underestimated. Compare the quantified need with the available budget early, and agree what will be prioritised if the two do not match. Build realistic lead times into the plan, including tender, delivery, quality assurance and distribution, and set reorder points that allow for delay. Many countries also find that splitting large purchases into scheduled deliveries reduces storage pressure and expiry risk.

Review and update regularly

Quantification is not a once-a-year event. Disease patterns, guidelines, prices and facility networks all change. A light quarterly review comparing planned consumption with actual issues and stock levels allows early correction. Simple indicators, such as months of stock on hand and the share of facilities with stock-outs of tracer items, give managers an early warning without heavy reporting burdens.

Practical takeaways

• Define the purpose, horizon and owners of the forecast before collecting data.
• Correct consumption figures for stock-out periods and cross-check against service data.
• Record assumptions openly and use scenarios where policy or funding is uncertain.
• Align quantities with budget availability and realistic procurement lead times.
• Review forecasts at least quarterly and act on stock indicators.

Good quantification is less about sophisticated models than about disciplined habits: shared ownership, honest data, transparent assumptions and regular review. Health systems that build these habits spend public money more wisely and keep essential medicines where patients need them.