That is the finding from Daysal et al. (2025).
We reproduce the results from an RCT showing the life-saving benefits of radiotherapy. We show radiation therapy also has economic returns: Ten years after diagnosis, treatment increases employment by 37 percent and earnings by 45 percent. Previous work has documented a substantial employment drop after a breast cancer diagnosis. Our results imply that radiation therapy can reduce this effect by 70 percent.
In addition to these empirical results, the paper also addresses two key methodological issues in real-world health economic analysis. This first issue is that in the real-world is endogenous assignment to treatment. Patients in worse health tend to receive more intensive medical treatments; thus, patient outcomes who receive advanced treatments may appear worse–not because the intervention doesn’t work–but because the patients are more ill. Secondly, measuring changes in social welfare holistically is problematic as it requires data on not only health but also economic outcomes.
The former issue the authors the authors use changes in clinical guidelines as an instrument for treatment assignment. Specifically, they apply a 2SLS strategy within a difference-in-difference framework. For the latter issue, the authors use integrated administrative data from Denmark (1990-2008). Mortality data comes from the Register of Causes of Death; labor market outcomes come from the Register-Based Labor Force Statistics; income data come from Income Statistics Register; and finally government transfers come from the DREAM dataset, a weekly register of all persons who receive government transfers.
You can read the full paper here.
