Aaron P Mitchell, Aaron N Winn, Patrick Augello, Grace Gallagher, Akriti Mishra Meza, Abdullah Abdelaziz, Gabrielle Guzman, Nina Galanter, Nirjhar Chakraborty, Maria Klimchuk, Susan Alsamarai, S Vincent Rajkumar, Peter B Bach, Mithat Gönen, Stacie B Dusetzina
Industry payments were associated with increased use of the promoted cancer treatment, regardless of whether that treatment was guideline-preferred or nonpreferred for a given patient.
PURPOSE: Payments from the drug industry influence physicians' prescribing decisions. Whether industry payments affect the likelihood that patients receive the guideline-preferred treatment for their cancer is unknown.
METHODS: We used fee-for-service Medicare claims to identify patients with an incident cancer diagnosis from 2014 to 2020. We used the National Comprehensive Cancer Network (NCCN) guidelines to identify cancers with variation in clinical benefit among recommended treatment options and to identify the preferred treatment as of each patient's diagnosis date. The primary exposure was drug industry payments received by the patient's medical oncologist during the year before diagnosis, categorized as no payment, payment for preferred treatment only, payment for nonpreferred treatment only, or payment for both. The primary outcome was receipt of a preferred treatment.
RESULTS: There were 14 cancer scenarios comprising 15,607 patients; 60.2% were managed by oncologists who received no payment, 21.9% by oncologists who received payment for guideline-preferred treatment, 30.6% by oncologists who received payment for nonpreferred treatment, and 12.7% by oncologists who received both payment types. Among patients whose oncologist received no payments, 41.6% received payments for preferred treatment versus 51.5% of those whose oncologist received payments for guideline-preferred treatment only, 34.7% of those whose oncologist received payments for nonpreferred treatment only, and 42.9% of those whose oncologist received payments for both payment types. After adjustment, payment for guideline-preferred treatment was associated with a greater likelihood of receiving preferred treatment (risk ratio [RR] = 1.06 [95% CI, 1.01 to 1.13]) and payment for nonpreferred treatment was associated with a lower likelihood (RR, 0.93 [95% CI, 0.89 to 0.98]). Association magnitude increased as payment dollar value increased.
CONCLUSION: Industry payments were associated with increased use of the promoted cancer treatment, regardless of whether that treatment was guideline-preferred or nonpreferred for a given patient.