As supply chains have globalised and trials become increasingly demanding and far reaching, supply planning can make or break study success. There is therefore a need to develop effective processes that are based on sound principles. In this regard, clinical supply is no different from any other sector where the planning of material supply for end-user consumption has to be coordinated and controlled.
This article aims to build a deeper understanding of the principles that apply, taken from supply chain management (SCM) best practices; then to look at their application in the management of clinical supplies.
The principles of supply planning
Firstly, we need to clarify the terminology that is bandied about in SCM circles.
Demand: there are two kinds of demand, independent and dependent.
- Dependent demand is where there is a known ‘quantity’ relationship between the product or material to be produced, and the component material. For example, if we have a demand for 50 patient kits and there are 30 tablets contained within each kit, the (dependent) demand for tablets will be 1,500 (plus maybe some wastage). These quantity relationships are normally included in a bill of materials (BOM).
- Independent demand is where consumption is based upon factors that cannot be easily predicted, that is, where no known quantity relationship exists. This means that methods for predicting independent demand are often required, typically through forecasting. In the commercial environment, this means sales forecasting for customer markets; in clinical planning, independent demand is typically less volatile than commercial markets, since clinical protocols tend to define the number of subjects on each study, albeit with inherent variation.
Supply is the provision of material required to satisfy demand, in the form of inventory. Inventory is present right the way through the end-to-end supply chain, as things are converted into different, higher-value materials and products. In order to create inventory, we need productive capacity to do the work of conversion. Depending on the various stages required for the conversion, there will be a lead time from placement of an order to receipt.
Masterplanning (or scheduling): since independent demand can be extremely variable and production lead times often relatively long, there needs to be a planning interface that ensures balance between supply and demand. This entails the management of inventory, capacity and lead time to consistently meet demand.
Policy decisions: part of the master planning process is to determine policies that are appropriate for the products and materials being planned. If an end-consumer cannot wait for the product when required, it will be necessary to adopt a make-to-stock (or inventory) policy. In this case, lead time is set to zero, with inventory and capacity being managed to maintain the agreed service level. Where consumers can wait a set period, inventory is set to zero, with capacity and lead time being managed to meet the expected delivery date. Capacity has to be managed in both cases.
Applying the principles in clinical trial supplies
It will be helpful if readers first clear their minds of the IT and other automated support currently available, to focus on the business process itself, which in simple cases can be carried out without IT support.
In the planning process, it is customary always to start at the point closest to the end-consumer (patient) and work upstream. Figure 1 shows a schematic of an end-to-end supply chain feeding clinical trial materials to investigator sites. The assumption here is that the trial sponsor company is managing the entire end-to-end supply chain, since this is the most challenging case.
Independent demand will be driven by the clinical protocols in terms of number of subjects on each study, recruitment rates, site initiation timings, etc. From this, we should be able to generate a forecast profile. Remember that accurate forecasts are often less important than knowledge of the bounds of variation. This is often called the ‘envelope of uncertainty’, and is made up of uncertainty in both quantity and timing. Because of this, masterplanning often requires, along with a base-case forecast, a worst-case (low quantity/timing delay) and best-case (high quantity/early timing). Armed with a set of forecasts, we can set about planning supplies.
Determining policy
Policy will vary depending on the product or material. For simplicity, we will take the case of provisioning the sites with patient packs. If this is a new study, then there will not be any existing inventory, so we must make to order. This means managing capacity and lead time so that the supplies turn up on time. Once initial supplies have been delivered, a decision needs to be made on whether to continue making to order or for inventory at agreed quantity levels. In that case, capacity and inventory levels need to be managed to maintain the inventory parameters.
Creating the master plan or schedule
Some readers here will be familiar with BOM and all the other master files that define formulae, specifications and packaging instructions. It is beyond the scope of this article to cover these in any detail, but in very simple terms the process involves the following steps.
From the independent demand for patient packs, we can produce a production plan. The schedule is based on capacity available at the producer and the lead time required. Figure 2 (above) shows a supply/demand chart. The plan before supply is entered shows the independent demand for week one to week ten. If we didn’t produce any supply of kits, clearly we are in trouble from the outset and are 303 units short by week ten.
The scheduling task is then to establish sufficient capacity, at the correct lead time, to satisfy demand. In this case, 100 kits can be produced in a week and made available at the sites for use. As we enter that supply into the second supply/demand chart, we see how inventory can be maintained at a level where we are never out of stock. In fact, by modelling different supply profiles it is possible to optimise customer service and inventory levels accordingly. Then we can also decide whether to maintain a make-to-inventory or make-to-order policy for the life of the studies.
Having created a master plan for the finished kits, we must also be cognisant of the need to maintain supply of materials in the rest of the supply chain. It is important that the dependent demand for materials is coming through appropriately. In the case of the fully outsourced supply chain as in Figure 1, the entire supply chain should be planned for optimum results.
Prerequisites for ongoing success
To round off the discussion on optimising planning, it may be helpful to take a broader view within the discipline of SCM. An important characteristic of supply chains is that they are extremely cross-functional and cross-company. This introduces all sorts of complexities in addition to the technicalities of material management and planning. Key factors that are prerequisites for success are:
- developing a robust, compliant supply base: keep the supply
- chain as simple as possible with minimum stages
- adopting inclusive planning processes: many are involved in delivering on production plans, including quality, regulatory affairs, CMC teams and external partners
- striving for completely accurate records and master files: records such as Bills of Material, item master files and inventory files can cause havoc with the planning process
- if inaccurate
- producing sound agreements, technical/quality, supply and service level: in the increasingly outsourced/third party industry we now work in, it is vitally important that agreements are as explicit as possible
- building appropriate relationships: not all suppliers are the same
- measuring the vital few: too many metrics can be harmful
- improving constantly: concentrate less on the tools and techniques of improvement and home in on simple methods to reduce complexity and add more value.
