How Forecasting and Optimization Work in Clinical Trial Supply Management
Clinical trials have become increasingly costly and complex. Constraints in supply chains, increasingly complex protocols, and higher costs per site mean that optimizing costs at every turn is increasingly important. One important component of cost is in the clinical trial supply chain. More and more, trial sponsors are finding ways to minimize supply chain costs using forecasting up front and optimization throughout their trials.
In this article, we will thoroughly define forecasting and optimization as it relates to the clinical trial supply chain, the current challenges and opportunities, and some of the innovations in this space.
The difference between forecasting and optimization
Often, there is confusion between the terms ‘supply forecasting’ and ‘supply optimization’ in clinical studies. Both forecasting and optimization are conducted in clinical trials to help ensure sites have enough clinical supplies and reduce costs. But forecasting is aimed at providing estimates only, and leaves the burden of complex decisions (overage, production planning, depot supplies, randomization and trial supply management (RTSM) set-up and so on) to the sponsor.
Forecasting in the clinical trial supply chain considers variables such as:
Protocol design
Expected sites opening and patient enrollment
Expected screen fail and dropout rates
Randomization
Stratification
Variable factors such as weight-based dispensing
The visit and dispensing schedule for the trial
Dosages and titrations used in the trials
Finished kit design
Packaging strategy
Labeling strategy
Supply network
Countries involved
Depots, pharmacies, DCT/DTP (decentralized trials, direct-to-patient)
RTSM (IRT) setup
Resupply rules on sites
Thresholds on depots
Current and past enrollment, shipments, inventory, etc.
These and other key factors must be considered to determine how much clinical supply to produce and/or procure and distribute throughout a clinical trial’s execution. From this, DP/DS/API requirements are set to ensure trial success, a controlled service level, and an overall budget. Forecasting most heavily considers enrollment prediction, dispensing schedule, and planned and current inventory to produce demand estimates, inventory consumption, and to inform higher-level decisions such as production planning.
Optimization, on the other hand, considers all of these factors and actuals as new data becomes available to clinical supply managers. By accounting for the intimidatingly complex correlations between all these factors, optimization solutions can help identify the true drug needs and shipment frequencies to reduce uncertainty, reduce supply costs, keep patients safe, and keep the trial going.
Optimization goes beyond forecasting in that it will guide decisions that may prove to be game changers in your studies, either in terms of cost, risk, agility, mitigating uncertainty, facing critical study changes, and many more.
Why are supply forecasting and optimization important for clinical trials?
The need for forecasting and optimization is important to manage clinical trial supply efficiently. Forecasting estimates budgets before the trial happens and gives a general prediction of how much supply will be needed for a trial and when.Optimization helps clinical supply managers to adjust supply parameters as certain variables deviate from initial projections and as new data becomes available.
This means that clinical trial supply forecasting and optimization are all about balancing costs and reducing the amount of drug on site to no more than is needed. On the contrary, you also need to make sure that you don’t have an oversupply at sites or depots, resulting in wasted drug kits and clinical supplies. Ideally, a trial would use only what it needs. But due to delays in the supply chain and not being able to predict the future down to the last dose, trials essentially will always end up with some leftover supply at sites and at depots.
Challenges in forecasting and optimization of clinical trial supply chain
There are as many challenges in forecasting and optimization as there are variables that affect the efficiency of clinical trial supplies. However, understanding the current landscape, we see that these factors are three of the most important challenges to consider when forecasting and optimizing for your studies:
Global logistics are complex
Between global conflicts, supply chain bottlenecks, and other unknowns, supply chains are always accounting for global uncertainty. However, global clinical supply chains also contain a number of known variables that remain a challenge for sponsors who conduct global drug studies. Some of these are differing country regulations around drug supply, currencies, tax structures, as well as unique challenges in logistics in different geographies.
Protocols are getting complex
Clinical trials are becoming complex in their protocols due to a demand to answer more questions in fewer studies and quicker timelines. Dose escalation trials prove tricky to predict supply due to the uncertainty of what dosages patients will be on as trials progress. Changes in dosage can affect packaging/labeling and the allocation of supply across sites. Platform studies may demand not only larger volumes of medication but a greater variety of dosages or formulations to account for differences across disease populations and their subsets. Adaptive trials prove especially complex due to their frequent, unpredictable changes that demand for flexible optimization and logistics. And lastly, pooling drugs across multiple trials, while helpful in efficiently managing resources, demands planning and optimization across a greater number of variables than that of a single trial.
Investigational medical products
Investigational medical products (IMP) are also getting more complex in their composition. This greater complexity calls for greater coordination across different parts of a supply chain. Not only can this mean sourcing of more APIs, but it can mean medicines that require stringent temperature monitoring. A prominent example of this would be the mRNA vaccines used to address the COVID-19 virus needing to be kept at sub zero temperatures to prevent degradation of the drugs themselves. Another example is gene therapies, where due to their sensitive components often require stringent storage and handling requirements.
What are some ways supply forecasting and optimization help clinical trials?
Supply forecasting and optimization play a crucial role in first ensuring that there is enough IMP and other clinical trial supplies for trials. It also seeks to minimize cost and waste of clinical trial supplies. Here are some ways forecasting and optimizing supply chain helps your clinical trials:
Preventing shortages
Accurate forecasting and optimization ensure clinical sites have enough supplies to meet participant needs throughout the trial. By accounting for potential or current disruptions in clinical supply, projected enrollment rates across sites and countries, and estimating demand loads for local and central depots, clinical supply managers can ensure accuracy in their forecasts and prevent shortages at sites.
Budget Planning
Total costs can be calculated in supply forecasting so that drug trial sponsors estimate how much money to allocate to their supply needs for each study. Having this up front cost projection is incredibly handy so that clinical supply managers can balance costs up front. By tweaking different variables such as participating countries, number of sites per country, as well as sourcing and logistics strategies, clinical supply managers get a balanced view of how to plan.
Reducing waste
Wastage of clinical supply can be a serious driver of cost in clinical trials. By seeking to reduce wasted IMP, comparators, and other clinical trial supplies, sponsors not only reduce cost but also benefit from more environmentally sustainable trials.
Foresee risk
Forecasting can help you see the potential risks in your supply chain more clearly. A great forecasting tool with cutting-edge methodologies will give you a bird’s eye view into the data while lending helpful analysis to pinpoint and mitigate potential risks.
Need cutting-edge modeling in an easy-to-use interface?
How can supply optimization help clinical trials?
Supply optimization in trials refines the supply chain process to ensure efficiency, cost-effectiveness, and reliability. Here are some key ways supply optimization helps your clinical trials:
Minimizing overproduction
Optimization reduces your risk of overproducing and storing excess trial supplies by minimizing supply quantities at depots and sites. This obviously leads to lower waste and storage costs.
Ensuring timely delivery
Monitoring and optimizing your clinical trial supply chain can help you ensure that sites and depots receive shipments at the time when they are needed. Timing is critical in clinical trials and can change often, which is why optimization is key throughout your trial. Optimizing timing, like many activities, can ensure minimization of costs and waste.
Dynamic budgeting
Maintaining flexibility in costs is central to what optimization does best. Integrating real-time and dynamic supply chain data points into your decision making will allow you to balance costs more flexibly as your trial moves forward and different variables change.
Monitor and mitigate emergent risk
While forecasting can help you anticipate roadblocks and risks, it won’t be as accurate real-time modeling to pinpoint risk as its occurring. A good optimization tool will allow you to update your reports with the latest data to see oncoming risk more clearly. That way, you can adjust your supply strategies accordingly to mitigate these potential or forthcoming risks.
Adapting to complex protocols
Complex protocols call for solutions like optimization to address complex variables. These might include cycle visit expansions, dose escalation protocols, master protocols that add and subtract cohorts or treatment arms to test different hypotheses, and so on. Powerful optimization tools help you to adapt supply strategies as variables like dosage become clearer in escalation studies, or as platform designs adapt to handle different patient types, and so on.
What are some traditional methods that have been used for clinical supply forecasting and optimization?
Spreadsheets
Notoriously, we have seen spreadsheets as a common way to manage forecasts and optimization. This system is seen by many as a safe, familiar way to plug in data to various functions within a spreadsheet to manage trial supply. But the spreadsheet solution often lacks scalability, is error-prone due to the fallibility of manual data entry, and lacks many advanced features such as stochastic analysis, real-time tracking, and alerts. While many consider it a cost-effective method, it often lacks the decision power of many modern systems, and is overall harder to maintain, standardize, and to account for actuals.
Average-based
Average-based forecasting systems in clinical trial supply chain management involve using historical data to predict future demand for supplies.These systems calculate the average demand over a specific period and use this average to forecast future needs. Averaging can provide clear and specific forecasts, and is a standardized method of forecasting that can be integrated with other systems such as RTSM. However, it often lacks the sophistication needed to make informed, data-driven decisions in complex and dynamic environments, potentially leading to less accurate forecasts.
What are the latest methods and innovations in clinical supply forecasting and optimization?
The global supply chain has seen a boom in technologies and processes in recent decades to better forecast and optimize to better manage everything from costs, lead times, personnel, and beyond. The clinical trial supply chain, while a heavily regulated space, has nevertheless adopted many of these approaches, albeit slower than many other industries. Some of these new approaches and technologies are:
Stochastic models
Supply chain managers are increasingly using stochastic models (also known as Monte Carlo simulations) due to their ability to represent uncertainty and variability more effectively compared to deterministic models. Their extreme flexibility in modeling complex, intertwined, correlated behaviors and characteristics of clinical trials, enables very precise forecasts no matter the complexity of the protocol or the supply chain.
Mathematical Optimization
In the mathematical sense, an optimization problem is set up by defining an objective like cost or waste reduction, variables like RTSM parameters and production quantities, and constraints such as maximum batch size or maximum tolerable risk of missing demand. Specialized computation engines solve for an optimal solution while considering these components. In other words, there is mathematically no way to improve. Finding your optimal production plan falls right into this category.
AI and Machine Learning
One area where trial sponsors are wary to adopt yet see the value in is machine learning (ML) and artificial intelligence (AI). Algorithms in AI/ML can easily detect and replicate patterns and processes. This means that they are increasingly important as life sciences companies continue to train AI models with real supply chain data from their trials.
Takeaways
Clinical trial supply chain is becoming increasingly complex due to the increased complexity of protocols. Forecasting and optimization allow for trial sponsors to minimize supply costs of their investigational medical products, comparators, shipping and inventory, while guaranteeing patients never miss a dose..
Clinical supply forecasting is about planning up front, optimization is about adapting to ongoing changes in your trials.
Forecasting and optimization are becoming increasingly difficult as trial protocols have become more complex, the supply chain continues to prove challenging to predict, and the medicines being tested are of increasing complexity.
Accurate forecasting ensures that clinical sites have enough supply for your trials so that patients are given their assigned drug.
Both forecasting and optimization help manage costs by minimizing overproduction.
While forecasting helps to predict and model initial risk assumptions, optimization allows for monitoring and mitigation of risk as it arises in real time for your trials.
Traditional methods of forecasting and optimization such as spreadsheets and average-based strategies are increasingly less efficient in comparison to newer solutions.
Stochastic modeling as well as advancements in AI/ML are contributing to an increasingly efficient clinical trial supply chain.
About Trialzen
Founded in 2021, Trialzen is a software-as-a-service company focused on providing clinical trial sponsors easy-to-use, scalable supply forecasting and optimization. Their solutions apply the most suitable advanced analytics tools available to model your clinical trials, and to predict and monitor risk with greater control and confidence than traditional industry solutions like spreadsheets.
Trialzen balances its powerful advanced analytics engines with a simple, intuitive user interface that puts trial supply managers in the driver’s seat. No need for a PhD in applied mathematics or many training hours to use. No need for costly consultants or managing middlemen. Simply plug in your data manually or via an RTSM reporting file, run reports based on your needs, get your supply chain under control with in-depth rationalization, and…stay zen!
Note: if you prefer to outsource your clinical supply management, check out our [partnership and certification] program.
