Situation
A sponsor had developed a protocol that required patients to taper their dose of prednisone over the course of the trial. Patients could enter the trial on their current prednisone dose, resulting in various prednisone tapering schedules across all patients, spanning several months.
The protocol was designed to include an open label part during which patients’ prednisone dose tapering was initiated. Patients then entered a double-blind part where they were randomized between active and placebo.
Due to the trial design and the prednisone tapering, the trial supplies included 50 different pack types in total. Recruitment was planned to take place in 27 different countries, across 120 sites.
Challenge
- The dispensing plan was very complex and needed to be adapted to each patient’s starting prednisone dose
- Some of the medication was used as both open-label and double-blind kits
- The sponsor was concerned about the amount of medication that would need to be packaged, considering the large number of countries and pack types
- There was a risk that sites would have too many kits to store compared to their refrigerated storage capacity
- The sponsor was very concerned with the potentially large quantity of medication sent to sites that may not be used, resulting in high waste volume across the trial
Solution
- Calyx built an IRT solution that guaranteed the minimum number of kits would be sent to sites while ensuring sites were able to recruit new patients on any starting prednisone dose.
- We defined a dispensing schedule that could automatically adapt to the patient’s initial prednisone dose
- Calyx supply chain optimization experts recommended using buffer stock strategies for new patients only to reduce drug wastage at site
- Calyx IRT inventory management was set up to predict for pack types that the patients would require based on their visit number, tapering level, and the part of the study they were in (open-label vs double-blind)
- We applied fractional prediction (also known as partial prediction) to cover the patient’s needs at each visit as well as any need for a replacement kit
- We worked together with the sponsor to define the likelihood of a replacement kit being required and agreed to predict one replacement kit for every 20 patients
- We also applied prediction capping, to reduce the risk of overstocking sites compared to their storage capacity
Result
- By using advanced inventory management settings in Calyx IRT, we reduced the buffer stock at sites by 60%
- Our prediction method was adapted to each patient’s tapering schedule
- The sponsor avoided failed visits by leveraging fractional prediction for kit replacement
- We helped reduce drug wastage at the site level to the minimum based on the protocol constraints
