Forecast Vs. Actual: Why Protocol Design Breaks Clinical Supply Plans
By Shivani Tanwar
Clinical supply planning is often described as a forecasting exercise, but anyone who has worked closely with a clinical trial knows that forecasting is only part of the story. A protocol may define patient numbers, visit schedules, treatment durations, randomization ratios, and enrollment targets with impressive precision. Supply plans are developed around those assumptions, manufacturing schedules are created, inventory is positioned, and distribution strategies are established.
Everything appears aligned. Yet, within weeks of study initiation, many clinical supply teams find themselves revisiting assumptions, adjusting forecasts, and responding to demand patterns that look very different from what was expected.
The answer often lies in a simple reality that supply professionals understand well: protocols are designed in a controlled environment, while clinical trials are executed in an unpredictable one. The protocol is translated into forecasting models that feed IRT systems, driving initial depot allocation rules and replenishment parameters that assume stable behavior over time.
The Protocol Is The Starting Point, Not The Outcome
Clinical protocols serve as the foundation for supply planning. However, a protocol is ultimately a model of expected trial behavior. Those assumptions are necessary, but they rarely survive unchanged once trial execution begins. For supply teams, each deviation introduces operational consequences that extend beyond forecasting models and influence inventory positioning, manufacturing schedules, depot stocking strategies, and distribution activities.
Changes in assumptions are often first reflected in forecast variance reports and IRT usage data, which then feed back into replenishment settings and resupply thresholds that drive depot behavior.
Enrollment Forecasts Are Rarely Linear
Enrollment is one of the most influential variables in clinical supply planning. A site expected to recruit 10 patients per month may recruit only three, while another may unexpectedly become a top-performing location. Slower enrollment can result in excess inventory sitting at depots for extended periods, increasing expiry risk and waste. Faster enrollment can create demand spikes that challenge manufacturing schedules and inventory availability.
If enrollment trajectories shift, depot days-of-supply calculations are recalibrated and automated reorder points in IRT systems begin triggering earlier or later than originally configured.
If enrollment accelerates beyond forecast, inventory at regional depots may be consumed faster than expected, requiring supply teams to review safety stock levels, advance manufacturing campaigns, or expedite shipments to maintain continuity of supply. If enrollment slows, inventory may remain in storage longer than planned, increasing expiry risk and forcing reassessment of inventory transfers and future production requirements.
Randomization Creates Hidden Demand Variability
Many supply models assume treatment allocation will closely follow protocol-defined randomization ratios. Over shorter periods, actual patient assignments can deviate significantly from projected ratios. One treatment arm may experience higher than expected demand while another consumes inventory more slowly.
In practice, this can influence depot stocking strategies and resupply triggers. Manufacturing priorities may shift to support future demand, while slower-moving inventory accumulates elsewhere and increases the risk of expiry.
Site Activation Timing Changes Everything
Site activation delays frequently alter demand patterns. Inventory positioned in anticipation of patient enrollment may remain unused for months. Meanwhile, delayed sites may suddenly activate within a short period, creating concentrated demand that was not reflected in the original forecast. Delayed site activations also create misalignment between prepositioned depot stock and actual patient-centered dispensing flows, which can result in stock sitting in the wrong geographic node relative to demand.
From an operational perspective, activation delays often force supply teams to make inventory positioning decisions, redistribute stock between depots, and revise manufacturing schedules. When multiple sites activate simultaneously, demand can increase rapidly, requiring accelerated resupply activities and closer coordination between clinical operations, manufacturing, and logistics teams.
Patient Behavior Is Difficult To Predict
Patients may miss visits, discontinue participation, withdraw consent, or require resupply outside expected timelines. Collectively, these variations can significantly influence inventory consumption patterns across sites and regions.
IRT system dispensing logic often assumes visit adherence according to protocol windows, but early terminations and missed visits introduce nonlinear consumption patterns that are only captured after dispensing data is aggregated across sites.
Missed visits, treatment interruptions, and early discontinuations create variability in inventory consumption patterns. These fluctuations can affect resupply timing, regional inventory levels, and overall forecast accuracy, requiring supply teams to remain agile throughout study execution.
The Cost Of Getting It Wrong
Overestimating demand can lead to unnecessary manufacturing runs, increased storage requirements, additional distribution costs, and higher levels of product waste. Underestimating demand may result in emergency manufacturing, expedited shipments, depot shortages, and potential disruptions to patient treatment.
Buffer stock decisions and manufacturing campaign sizing are often made around forecast bands that expand or contract as enrollment and dispensing data feed back into planning systems.
Building More Resilient Supply Plans
The strongest supply organizations are not necessarily those with the most accurate forecasts. They are often the ones that identify changes early, understand the operational consequences, and adjust manufacturing, inventory, depot strategy, and resupply plans before disruptions occur.
Early signal detection often comes from daily or weekly IRT dispatch reports and deviation alerts in forecasting systems, which trigger rebalancing decisions across depots and production planning cycles.
Conclusion
Forecasts are essential to clinical supply planning, but they should never be confused with certainty. Success is not measured by the ability to predict the future perfectly. It is measured by the ability to adapt when reality differs from the plan. In an environment defined by uncertainty, resilient supply strategies will always outperform perfect forecasts that exist only on paper.
About The Author:
Shivani Tanwar is a healthcare and supply chain professional with a growing interest in clinical supply operations, inventory management, and patient-focused healthcare logistics. With hands-on experience in pharmacy and operational coordination, she is passionate about bridging the gap between planning and real-world execution within healthcare systems. Shivani actively explores topics related to clinical trial supply challenges, forecasting accuracy, process optimization, and operational efficiency. Through her professional insights and industry-focused content, she aims to contribute to meaningful conversations shaping the future of healthcare supply chains.