Fig. 5

The effect of life cycle pricing on cost-effectiveness analyses. Source: Lakdawalla et al. [51] (graphic borrowed with permission from No Patient Left Behind). Lakdawalla et al. reviewed a set of 20 traditional CEAs published by the US-based Institute for Clinical and Economic Review and accounted for several additional values as well as dynamic pricing [51]. Importantly, they used a “stacked cohort” model to account for patients starting treatment in years post-launch, including patients who start treatment after a drug has gone generic and therefore received extremely inexpensive benefits. Their base case assumption was that drugs would drop in price by 76%, based on the average decline of a historical set of medicines after their loss of exclusivity. However, recognizing that many of the drugs ICER was evaluating were outliers more likely to drop by more than that because they are specialty drugs for orphan disorders (e.g., the cancer drug Tarceva and the multiple sclerosis treatment Tecfidera dropped in price by over 98% after going generic), they included a sensitivity analysis that assumed a 90% price drop after 14 years [51]. Of the 20 drugs, traditional CEA had deemed only 8 to be cost-effective. But the more inclusive generalized CEA approach showed that 17 were cost-effective in the base case, and 18 were cost-effective in the sensitivity analysis. Interestingly, two remained above the cost-effectiveness threshold even in the sensitivity analysis. One was a drug that had failed its clinical trial and was not even approved for the indication that ICER had modeled (in other words, no one is asking society to pay for it). The other was a drug for sickle cell disease that was approved on an accelerated basis based on a biomarker and is still undergoing outcomes studies that would inform its cost-effectiveness. Lakdawalla et al. pointed out that their calculations only accounted for some of the traditionally missing values of these medicines and that adding in other elements, such as caregiver spillover, would likely further improve their cost-effectiveness