At the national level, forecasting for Maternal, Newborn, and Child Health (MNCH) commodities is a strategic function. The figures generated inform procurement decisions, guide budget discussions, and determine the availability of life-saving products across the country.
When forecasts are inaccurate or disconnected from funding realities, the consequences are felt throughout the health system. Conversely, when forecasts are credible, tested against historical performance, and aligned with resource allocation processes, they strengthen planning, accountability, and service delivery.
The MNCH Forecasting Tool was developed to support this shift.
Strengthening the Technical Foundation of Forecasting
National MNCH forecasting processes typically involve consolidating routine consumption data, reviewing stock status information, validating assumptions, and aligning projections with programmatic targets. While technically sound, these steps can be time-intensive when managed through manual spreadsheets and fragmented datasets.
The MNCH Forecasting Tool provides a structured digital environment for this process. It consolidates standardized data inputs, including automated or semi-automated extraction from systems such as KHIS and applies validation checks – including anomaly detection– to identify inconsistencies. The tool then generates projections based on historical consumption trends and observed variability.
By embedding forecasting logic within a system rather than individual spreadsheets, the tool improves consistency, reproducibility, and confidence in the outputs.
However, strengthening forecasting is not only about generating projections. It is also about evaluating how well previous forecasts performed against actual consumption.
Embedding Forecast Accuracy Assessment
As part of the forecasting exercise, the tool includes an assessment of forecast accuracy against actual consumption.
This functionality enables national teams to compare previously forecasted quantities against realized consumption data. By doing so, the system supports structured reflection on:
- Over-forecasting and under-forecasting patterns
- Commodities with consistently high variance
- Systematic bias in forecasting assumptions
- The reliability of the adopted forecasting methods
Rather than relying on anecdotal impressions, teams can quantitatively assess performance using defined accuracy metrics. This may include measures such as percentage error, absolute deviation, or variance across time periods.
Forecast accuracy results demonstrate that the consumption-based approach, powered by the digital tool and enhanced by machine learning models such as Prophet, achieved markedly higher accuracy than alternative methods. Commodities forecasted using consumption data recorded relatively low Mean Absolute Percentage Errors (MAPE), including Oxytocin at 9.75%, Gentamicin at 38.39%, and Magnesium Sulphate at 45.73%. In contrast, products forecast using non-consumption approaches showed substantially higher errors, including Benzylpenicillin (recent forecast) at 215%, Chlorhexidine (demographic) at 258%, Ferrous Sulphate/Folic Acid (demographic) at 475.39%, and Tetracycline Eye Ointment at 131.60%. This improved accuracy reflects the value of embedding forecasting within a structured digital system that standardizes inputs and performs validation checks to identify and correct data inconsistencies before analysis. By strengthening data quality upstream and applying consistent forecasting logic, the tool produces more reliable and reproducible outputs, ultimately enabling more precise MNCH planning and better alignment of commodity supply with actual need.
Embedding this within the tool institutionalizes learning. Forecasting becomes iterative rather than static. Each cycle informs the next.
This strengthens technical credibility and supports continuous improvement.
Linking Forecasting to Resource Allocation
A technically robust forecast does not automatically translate into full commodity availability. National programs frequently face funding ceilings that are lower than quantified need. In such contexts, prioritization becomes unavoidable.
The integration of the MNCH Forecasting Tool with the SMArT (Strategic Resource Mapping Tool) strengthens this next step. SMArT, is an AI-enabled decision-support tool that optimizes allocation of limited health resources by prioritizing commodities based on consumption data, stock status, and budget constraints.
Building on this integration, early implementation results show how linking forecasts to structured resource prioritization can translate into tangible availability gains. In five focus counties, stock status data from January 2026 indicate that several high-impact MNCH commodities, including Oxytocin (4.6 months of stock), Chlorhexidine gel (4.1 months of stock), and Magnesium Sulphate (3.4 months of stock), were maintained within acceptable stock ranges, reflecting more aligned procurement decisions. At the facility level, the application of the SMArT tool in Trans Nzoia further illustrates this value: despite resource constraints, the tool recommended procuring 5 of 7 MNCH commodities, fully in line with their forecasted quantities, demonstrating deliberate prioritization of essential maternal and newborn health needs. For commodities, the tool optimized allocations within available budgets, recommending 421 units of Magnesium Sulphate against a forecasted need of 30 units, and 239 units of Tetracycline Eye Ointment against a forecasted 39 units, based on an anticipated increase in service delivery demands following the upgrading of the facility to a Level 4 facility. Together, these results demonstrate that integrating forecasting with strategic resource mapping not only strengthens alignment between needs and procurement but also ensures that limited resources are directed toward the commodities that have the greatest impact on MNCH outcomes.
Once projected requirements are generated, SMArT supports comparison of quantified needs against available funding. It enables identification of commodity-level funding gaps and facilitates structured prioritization under constrained budgets. Allocation decisions are documented and guided by transparent criteria rather than informal adjustments.
This linkage creates a coherent decision-making flow: generate reliable forecasts, assess available resources, and apply structured prioritization.
Improving Transparency and Accountability
When forecasting, validation, and prioritization operate in isolation, trade-offs may occur without clear documentation. Assumptions can remain implicit, and adjustments may not be systematically recorded.
By embedding forecast-accuracy assessment and linking outputs to SMArT, the system strengthens transparency at each stage. Funding gaps are visible. The performance of previous methods is measurable. Prioritization decisions are explicit.
The result is not only improved efficiency, but also improved governance.
Operating Within Real-World Constraints
Digital tools do not eliminate structural challenges. Data completeness can vary. Funding shortfalls remain a reality. Disbursement timelines may affect procurement cycles.
However, stronger forecasting systems improve the quality of decisions made within these constraints. They enable national programs to quantify true need, understand the magnitude of funding gaps, and make deliberate, evidence-based trade-offs.
This strengthens stewardship of limited resources.
Toward Institutionalized Digital Forecasting
The long-term value of the MNCH Forecasting Tool lies in institutionalization. Embedding forecasting processes within a standardized digital system reduces reliance on individual expertise and improves continuity across staffing transitions.
When integrated with prioritization platforms such as SMArT, the tool contributes to a more coherent national supply planning architecture. Forecasting, funding analysis, and allocation become connected components of a structured system rather than parallel processes.
Conclusion
National MNCH forecasting is more than a technical requirement. It is a cornerstone of health system planning.
The MNCH Forecasting Tool strengthens the technical integrity of quantification. The SMArT tool strengthens strategic allocation under constrained resources. Together, they support data-driven, transparent, and accountable decision-making for maternal, newborn, and child health commodities.
In resource-constrained environments, improving the quality of planning processes is not procedural. It directly supports continuity of care and the availability of essential health products where they are most needed.
