Skip to Content

Strategies for Efficient Clinical Supply Management and Forecasting

A proactive approach for effective clinical supply management is essential for reducing the risk of supply-related delays, stock-outs, and excess inventory. Utilizing simulations to performing “What If” scenario planning for demand forecasting is a powerful, insightful supply chain management tool, particularly for complex studies with multiple variables. However, forecasts are only as good as the models upon which they are built and the inputs provided.

Strategies for Efficient Clinical Supply Management and Forecasting

Strategies for Efficient Clinical Supply Management and Forecasting

This article explores how to drive excellence within the forecasting process and how to utilize forecasting throughout the clinical study to better plan clinical supply budgets and project timelines and to identify potential supply-related issues before they negatively impact your study.

Key takeaways include:

  • The importance of pre-study forecasting for initial demand planning and smoother site start-up
  • Critical data points that should be included in every forecast
  • How to plan for and manage mid-study changes and determine when reforecasting may be appropriate

Content Summary

Controlling and managing clinical supplies using forecasting
Reforecasting: Planning and handling mid-study changes
Real-time tracking tools to monitor forecast accuracy
Forecasting and optimizing budget spend with supply chain simulation
Forecasting to support alternate supply models


Clinical supply management has many challenges. Clinical trials have many moving parts and clinical supplies need to be prepared well in advance of the actual study start. Supply delays, stock-outs or excess inventory are all undesirable and potentially risky scenarios. In recent years, clinical studies have become more global and complex – involving a multitude of clinical sites and network depots spanning multiple countries and geographic regions, each with a unique set of challenges and requirements. When more countries and regions are added or dropped during a study, clinical supply management activities such as forecasting become important as they can plan for and guide an appropriate response to such changes.

Study protocols themselves are also becoming more complex. Studies that in the past were treated as separate clinical trials are now being combined to accelerate the clinical development process. As an example, dose-escalation and maximum tolerated dose safety studies are being integrated with pharmacokinetics and pharmacodynamics studies to form a single clinical trial. New approaches to study protocol designs such as adaptive trials and specialized basket/bucket studies (used in oncology studies) can increase clinical supply complexity. Variable and unpredictable patient recruitment often has a direct and meaningful impact on clinical supply planning. As drug therapy becomes more personalized with targeted genomic strategies, the patient population available for clinical studies decreases. With limited patient populations, it is critical to provide IMP (investigational medicinal product) to the patient, whenever a patient gets identified. A limited patient population may also mean slower enrollment and adding more sites to achieve the necessary patient enrollment.

Creating an overstock of clinical supplies remains a common tactic to create an inventory buffer. It can mitigate risk due to variations in patient recruitment rates and ensure an adequate supply level is available at all times. This approach is not only costly, often resulting in a large amount of unused or expired inventory which eventually needs to be destroyed, but also creates a new challenge: where to store the excess supplies? Clinical sites have limited storage space, especially if they are conducting multiple studies. Storage fees for holding excess supplies in specialized clinical supply storage facilities, especially for cold or frozen drugs, adds to the study cost. Information driven clinical supply management and forecasting activities can better predict supply needs and optimize both initial and ongoing inventory levels.

If the clinical trial supplies are very limited in quantity, difficult to get and/or expensive, waste must be minimized. Efficient supply management and forecasting can help to reduce clinical trial costs and supply quantities by decreasing the amount of overstock required and establishing optimal quantity and timings for resupply that is in line with the site demand. Some comparator or reference products can be quite costly, limited in supply, and may have a limited shelf life or need long lead times to source. Also, biologics take longer to manufacture and have shorter expiry/re-test dates than conventional pharmaceutical products. Studies that need such products with long lead times will need to take this into account. They may require forecasts for resupply even before the patient’s enrollment.

Proper clinical supply management and forecasting efforts should be able to address most issues that can occur, such as planning for product waste due to temperature excursion during transit or a slight delay in manufacturing and product availability. But even with the best planning, there may be issues that no one could anticipate. Unforeseen changes in the supply plan due to acts of nature, geopolitical unrest, manufacturing issues, safety issues or comparator product shortages or recalls are scenarios that can disrupt even the best-laid plans. For example, the ash cloud generated during the 2010 eruption of Iceland’s Eyjafjallajökull volcano disrupted air traffic over Europe for a month. Engaging clinical supply management and forecasting adds value by evaluating different response strategies and contingency plans. It can identify supply solutions that minimize or even potentially avoid supply chain disruptions.

Controlling and managing clinical supplies using forecasting

A pre-study forecast provides initial demand planning and an overall estimate of supply requirements. A checklist is used to capture basic information and determine what if any missing information is needed to complete the forecast.

Information on checklists should include:

  • Study Start Date – The First-Patient-In (FPI) date, or another key milestone that marks the beginning of a trial.
  • Study Completion Date – The Last-Patient- Out (LPO) date, or another key milestone that marks the end of a trial.
  • Drug Product Information – This includes:
    • Dosage forms & strengths for all drugs used in the study
    • The patient kit composition or dispensing units supplied to a site than to a patient
    • The number of available drug lots and the batch size, or quantity of each kit by type
    • The estimated expiry of the drug (by product type if the expiration date is likely to change)
  • Study Design – The study design includes information regarding:
    • Treatment Groups – The number and description of treatment groups, block size and ratio for randomization; include this information if there is any run period with treatment or without
    • Blocking Type – Include whether the study is central vs. site-based (other types are also possible)
    • Visit Schedule – A list of visits showing when the drug will be given to a patient as well as screening visits and randomization visits. This list should also include the length of time between visits including visit windows. There could also be unplanned visits, so forecasting can determine probability and what happens during those visits
    • Dispensing Schedule – The schedule is by treatment group and shows which kits are given to patients at which visit. This should include the expected change for variable dosing, for example, weight-based dosage or titration-based tolerability
    • Countries, Regions, or Depots – This information shows who is participating in the study and which depots will supply which countries, plus estimated transit time
  • Patient Enrollment – This includes:
    • The number of patients targeted, or the number of patients required to complete the study
    • Randomization cap, or the maximum number of patients randomized before the study is closed to enrollment
    • Randomization or enrollment rates, or the number of new patients per month per site
    • Country start-up dates and/or start-up dates, based on regulatory approval and/ or drug availability
  • Interactive Response Technology (IRT) Values – Forecasting and simulation will assist to determine the optimal IRT settings. Information to be discussed during the User Requirements Specifications (URS) set-up include:
    • Initial Supply Trigger – What triggers the initial supply to be sent to a clinical site
    • Site Start-Up Rates – The estimated rate of site start-up, either expressed as a percentage per month or number of sites per month
    • Resupply Short Window – The number of days in the future to aggregate demand when determining whether future demand will trigger shipment. If demand causes a floor trigger to be reached, the shipment will be sent
    • Resupply Long Window – The number of days of supply to ship to a clinical site when a re-supply is required (this includes resupply short window and is not an addition to it)
    • Do-Not-Dispense-After (DND) – The number of days before an expiration date that a drug may no longer be dispensed to a patient
    • Do-Not-Ship-Offset (DNS) – The number of days before an expiration date that a drug may no longer be shipped to a site or another depot
    • Label Groups – Multiple labels should be listed by country

Since initial forecasting and planning may need to begin before clinical protocol is even finalized all the required information might not be available. If this is the case, discussions between the clinical sponsor and clinical supply management are necessary. Best estimates and calculations based on different scenarios may help arrive at a set of assumptions that can be agreed upon.

Reforecasting: Planning and handling mid-study changes

Calculations and predictions from the initial forecast need to be re-examined throughout a clinical study due to shifting variables, including site changes, patient enrollment rates, and long-term study changes. Accordingly, forecasts should be checked and periodically adjusted based on updated information and actual dosing data.

Changes that could spur reforecasting include:

  • Individual sites not starting as initially planned
  • Changes in the estimated patient enrollment rate
  • Addition or elimination of sites, countries, regions or depots
  • Patients continuing therapy longer or shorter than predicted
  • Delays in study start-up or enrollment
  • Product expiration before the end of the study
  • Additional resupply to sites required due to short dating of available comparator
  • Interim data analysis report for an adaptive trial
  • Revisions in treatment arms

Real-time tracking tools to monitor forecast accuracy

Tools used for forecasting depend on the type, complexity, and size of the study. Small or uncomplicated studies (for example, with a single depot) can be managed with a manual spreadsheet that calculates drug quantities and predicts resupply needs. Quantities may be determined based on a percent overage of a drug dispensed to patients plus an overage for site seeding.

While a simple spreadsheet could work for basic studies, more complex calculations are necessary for complicated studies. Computer-based simulation tools are necessary for forecasting the supply needs of complex studies with multiple variables. These tools can execute many scenarios based on a high-level analysis of variables and probability far more efficiently than undertaking the same calculations manually. They can also simulate real-world situations using the input of various factors likely to occur.

Simulation is a powerful tool which yields a vast amount of information, including:

  • Quantity of medication or ancillaries to plan for the overall study and during any selected period
  • Number, timing, and quantity of packaging campaigns
  • Optimal packaging configurations; the optimal site inventory levels and buffer stocks; the quantity of comparator or reference products needed
  • Number of anticipated shipments
  • Quantification of risk or likelihood of an out-of-stock situation
  • Contingency plans, or “what if ” situations

Simulations can provide many types of data including:

  • IMP (investigational medicinal product) and comparator/reference product quantities, overall and during any period
  • Regional bright stock allocation for demand-led studies (see the section, “Forecasting to support alternate supply models”)
  • Quantification of risk or likelihood of an out-of-stock situation
  • Anticipated number of shipments
  • Optimal site inventory levels and buffer stocks
  • Optimal packaging configurations
  • Number, timing and quantity of packaging campaigns

Two kinds of the simulation are commonly used: simple or single-run and complex multivariate models. A single-run simulation is used to gather data and troubleshoot issues. This type of simulation is generated from a single seed number. While it only represents a single possible instance, it can be useful to drill into a specific issue that may arise, down to a single potential patient supply issue at a clinical site. Multivariate simulations are based on probabilities and simulate subject and supply data over potentially hundreds of simulations across multiple scenarios to arrive upon the optimal clinical supply plan.

Many types of studies can benefit from simulation data. If investigational medicinal products (IMP) or non-investigational medicinal products (NIMPs) are expensive or in limited quantity, simulation data can help with buying decisions and reduce waste. In multiple packaging campaigns, the data can be used to adjust both the timing and quantity of future campaigns and in studies with expiration date concerns, simulation data can be used to determine if a drug will expire before the last patient completes dosing. The data can also be used to determine the required quantities where supply is pooled across studies. Simulation data also help to determine if the correct kits will be available for all sites in studies with dose titration or complex stratifications.

Finally, potential study design changes or adaptive trial designs can draw on simulation data to better understand the likely impact of study changes on clinical supplies. For example, simulation data can show how the clinical supply plan may need to be modified if a treatment arm is dropped or a new country or depot is chosen for the study.

Case study on resupply simulation:
Synopsis: Distribution costs were analyzed based on changing the quantity of drugs provided with each resupply shipment.

Challenge: A clinical sponsor requested IRT resupply outlook windows consisting of a 14-day short window and 21-day long window.

Catalent Solution: A simulation was run to compare the project impact of the sponsor’s requested window of a 21 day to a 42-day long window. The simulation revealed that the longer duration 42-day window would result in a net saving of USD 345,000 in logistics costs.

Source: Catalent Pharma Solutions

Forecasting and optimizing budget spend with supply chain simulation

Simulation of multiple supply and demand scenarios for a study can provide valuable data with regards to decision-making for budget optimization and phasing of budget spend. As referenced earlier, the cost of the investigational or comparator products is not the only supply variables to consider in total budget spend. Supply logistics costs can include expensive shipping containers and temperature monitors, processing fees, courier fees, duties and taxes and supply depots versus direct-to-site shipping costs. Kit packaging and labeling strategies must be balanced against patient dispensing requirements for evaluation against varied enrollment and IRT supply algorithm strategies. Once all known variables have been considered and reasonable assumptions are made, the data obtained from these simulations to weigh total budget costs balanced against supply risks can then be shared and evaluated amongst the study team. The development of scenarios is critical to support informed decision-making for supply strategies and ultimately better control of the budget.

Forecasting to support alternate supply models

In studies using a demand-led approach to clinical supply, forecasting helps identify and plan for variables that are harder to predict. Unlike traditional clinical supply models, a demand-led approach is a late-stage customization model that draws upon regional inventories of bright stock (unlabeled primary packaged IMP/NIMP) versus finished patient kits. Patient kits are not assembled and shipped until “ordered” via the IRT system for a specific clinical site in response to actual patient need.

Not surprisingly, forecasting plays a crucial role in preparing the clinical supply chain to optimize bright stock inventory levels and meet the real-time variability associated with the demand-led approach. Forecasting also provides information on optimal regional bright stock allocation including pooled supplies in demand led studies. By uncoupling secondary packaging (patient kit assembly) from the traditional clinical supply model of pre-assembled kits, the demand-led approach in combination with forecasting can drive greater supply chain efficiency.


All studies can benefit from forecasting regardless of size or complexity. The initial effort involved in gathering required data and making any necessary assumptions is valuable in and of itself. In the end, forecasting efforts should align with the size and complexity of the study and is crucial for very complex studies or those involving high value and/or high scarcity IMP/NIMP or when using a non-traditional supply model, such as demand-led approach.

Complex clinical study supply chain strategy is best informed using advanced/computer-based forecasting models. These simulation models are a tremendous time and effort saver that can run through hundreds of scenarios to identify the optimal solution, which is possible but not practical to do “by hand.”

However, forecasting is not a crystal ball but a tool that can promote more informed decision making and better contingency planning. Clinical supply needs are often dynamic and fluid in today’s complex studies and actuals need to be tracked against the forecast. While the study is ongoing, reforecasting may be needed to adjust the clinical supply plan.

Source: Catalent

Alex Lim is a certified IT Technical Support Architect with over 15 years of experience in designing, implementing, and troubleshooting complex IT systems and networks. He has worked for leading IT companies, such as Microsoft, IBM, and Cisco, providing technical support and solutions to clients across various industries and sectors. Alex has a bachelor’s degree in computer science from the National University of Singapore and a master’s degree in information security from the Massachusetts Institute of Technology. He is also the author of several best-selling books on IT technical support, such as The IT Technical Support Handbook and Troubleshooting IT Systems and Networks. Alex lives in Bandar, Johore, Malaysia with his wife and two chilrdren. You can reach him at [email protected] or follow him on Website | Twitter | Facebook

    Ads Blocker Image Powered by Code Help Pro

    Your Support Matters...

    We run an independent site that is committed to delivering valuable content, but it comes with its challenges. Many of our readers use ad blockers, causing our advertising revenue to decline. Unlike some websites, we have not implemented paywalls to restrict access. Your support can make a significant difference. If you find this website useful and choose to support us, it would greatly secure our future. We appreciate your help. If you are currently using an ad blocker, please consider disabling it for our site. Thank you for your understanding and support.