J+D Forecasting strike the balance between simplicity and complexity when modelling for the liquid tumour market
The Challenge Identify a clear, transparent, and logical way to structure a forecast model to identify the potential for a new therapy in leukaemia. The model would be complex because patients go through different stages of treatment within a single line of treatment (for example, induction chemo followed by consolidation chemo followed by maintenance therapy). …
Identify a clear, transparent, and logical way to structure a forecast model to identify the potential for a new therapy in leukaemia.
The model would be complex because patients go through different stages of treatment within a single line of treatment (for example, induction chemo followed by consolidation chemo followed by maintenance therapy). In addition, not all patients will follow the same treatment pathway in any single line of treatment.
A new forecasting approach was required:
- To instil confidence in the outputs for a go- or no-go decision
- To incorporate changes in the flow of patients and eligibility for therapy
- To handle the complexity of following patients within a line of treatment
- The model must be user friendly and intuitive
- The model must incorporate a robust approach to risk analysis to support the decision
- The approach must support best practice forecasting principles
Ideally the client team wanted to see a cost reduction vs. their current software.
The Approach and Solution
STEP 1: WORKSHOP
- To clarify the challenges in detail
- To meet the team and understand their individual needs
- To identify business and department priorities
- To explore the data available and the specific outputs the team sought
STEP 2: THE SCHEMATIC
- J+D Forecasting created a schematic – a visual structure/layout of the model
- The client reviewed and identified gaps they felt were important to how they and their colleagues would make a final decision to go/not go with the new therapy
- This approach promoted team collaboration and understanding and buy in to the final model.
- The schematic was finalised and signed off as part of J+D’s standard model development process.
STEP 3: SCOPING AND CREATION
Different forecasting methodologies were considered such as a standard epi cross sectional. However, the team settled on a patient flow approach. The patient flow methodology was considered essential as this is the only forecasting approach that allows you to ‘follow’ patients through their treatment journey.
This is a more accurate approach to defining the size of the potential patient pool and therefore, size of potential opportunity for the leukaemia therapy under exploration.
A forecast model structure was created using Onco+, an Excel add-in created by J+D Forecasting. Onco+ is based on a patient flow methodology created to incorporate all the Oncology variables that affect the forecast. For example, stages/regimens/lines of therapy, restrictions, length of therapy, holiday periods, wastage and has a variety of conversion options too.
The therapy, although likely to be approved for late stages, was likely to be approved for earlier stages in the future. Onco+ helped the team evaluate several potential scenarios for up to 3 lines of therapy. It also provided the outputs that the client was looking for (patients, units, packs, gross revenue and net revenue – all by SKU) using a conversion approach that closely matched the reality of the market that was being forecast.
The product is yet to be launched and further exploration is under way after roll-out of Onco+ across multiple departments within the organisation.
Onco+ has provided a balance of the necessary complexity with the appropriate level of simplicity to be usable and transparent across the organisation.
J+D’s approach and the model has given the team greater confidence in the outputs and subsequent decisions they will have to make in taking the product forward.