Treatment persistence refers to the duration of time over which patients are likely to receive treatment with a given regimen. In many oncology settings treatment with a given drug or regimen will persist until that treatment no longer has a positive impact - on prognostics or symptoms (disease continues to progress). However, in some oncology settings patients receive a defined number of cycles of treatment (e.g. haematological malignancies).
An accurate model encapsulating real world market dynamics requires understanding of how both defined number of cycles and treatment to progression will influence a forecast model.
Treatment to progression is often simpler to model as we can use known rates at which patients drop off treatment: often this will be based on progression free survival statistics taken from pre-existing literature in order to determine the proportion of patients still on treatment at any given time.
Cycles of treatment are often slightly more difficult for forecasters to model, and care should be taken to include two things. First a measure of the number of patients who drop out of receiving treatment, during the planned dosing schedule (Often progression free survival). Second a sudden drop in the number of patients receiving treatment at the end of their proposed treatment course. Failure to model either of these events could lead to overestimation of patient numbers and an inaccurate reflection of true market events.
When considering persistency for both methods of dosing a forecaster should always contemplate the intricacies associated with the specific regimen and disease area. For example Azacitidine (one of the recommended treatments for myelodysplastic syndromes) is recommended for patients, regardless of disease progression, in the first 6 cycles of treatment. After the initial 6 months, treatment with the drug should either be continued or discontinued based on response to therapy. These intricacies must be reflected in persistency assumptions. For the duration of the first 6 cycles of treatment, overall survival would be a good indicator of the number of patients likely to be on treatment at any point during that 6 month period. After this initial period progression free survival is likely to be more suitable. The rapid drop in patients after the initial 6 month treatment period must also be factored into assumptions for persistence.
It is important to note that when inputting persistency assumptions you should always reach zero eventually (i.e. all patients discontinue treatment at some point), unless the forecast periodicity (including backlog) is less than the time taken for 100% of patients to discontinue. This zeroing of persistence will prevent unrealistic patient inflation in the model and ensure that patients flow through the model effectively.
Patient persistence is a vital parameter to include when using dynamic patient-based forecasts. It allows more accurate prediction of the:
- Number of patients using a drug at any given time
- Calculation of months of treatment
- Calculation of volume and revenue outputs
It is an essential market driver necessary for modelling and understanding the effects of newly launching regimens on downstream patient flow and degree of overall market disruption.