Recently, Captario had the opportunity to present to a non-profit organization that uses medicine development as one of its interventions. Knowing that traditional value measurements would be of no interest to this organization, Captario instead created an example drug product model that used QALY. The goal was to use QALY both to measure the efficacy of the intervention and to measure the post-launch impact of the treatment.
Captario defined QALY Improvement (Ph3 QALY True Improvement) as a product of improvements in quality of life (Ph3 QoL Improvements) and life expectancy(Ph3 LE Improvements). Since the model starts before any clinical results, it's impossible to know the efficacy, whereas Captario modeled QoL and LE as log-normal distributions. The Standard Error was then added, providing a range of potential outcomes from the Phase 3 trials (observed effect/improvement). Lastly, comparing the observed effect with a threshold to assess Phase 3 success. Known as effect modeling, it's a bottom-up method to model PTRS rather than static benchmark numbers.
The second part of modeling was the value or impact forecasting. The upper left part of the below figure illustrates the QALY increase for a target region. Having started with the True effect object that was established in the phase 3 success model, Captario then modified this based on the baseline QALY of the target population. However, the QALY improvement might be different if the target population's baseline QALY is different from that of the PH3 trial population. Hence, a delta was calculated for the difference in the QALY baseline, which was then factored into the intervention's true effect. This gives the QALY increase for this particular region. The next step (which is not in the picture, is to multiply the QALY increase by the eligible population to get the total impact.
Captario then ran Monte-Carlo simulations using the model, which provided a wealth of data to analyze. Captario further created new metrics, such as cost per QALY and mean QALY improvement. Since each region had different populations and QALY baselines, it was possible to find areas with more potential to impact positively.
In summary, this exercise led to two conclusions:
1: Using QALY as a measure of efficacy and product value allows getting close to the fundamental value of a drug product. In doing so, it's possible to create models that are comparable in their impact.
2: Captario has only scratched the surface of using QALY, and there are ample opportunities to explore its potential further.
What are your modeling experiences with QALY? Are they good or bad experiences?
Comments