Pragya Kakani

Pragya Kakani

Assistant Professor, Division of Health Policy and Economics
Department of Population Health Sciences, Weill Cornell Medical College

Dissertation Title: “Essays on Healthcare Provider Markets

 

Chapter 1: National trends in integration between medical practices and pharmacies for self-administered drugs from 2011 to 2019
Physician practices and health systems are increasingly launching dispensaries and retail pharmacies for self-administered drugs. This paper uses a novel dataset of dispensaries and retail pharmacies operated “in-house” by physician practices, health centers, and health systems to describe this trend in Medicare Part D from 2011-2019. There has been substantial growth in the use of in-house pharmacies for high-cost drugs, especially in oncology; the share of oral oncology spending filled in-house increased from 13% to 37% between 2011-2019. There was also outsized growth in other high-cost categories such as anti-virals and immunosuppressants. Within specialty, higher cost drugs were more likely to be filled in-house. Finally, there was important variation across states, plans within states, and among practices in their reliance on in-house pharmacies. These results highlight the growing relevance of integration between physicians and pharmacies in the U.S., especially for high-cost drugs, and the importance of understanding implications for patient care.

Chapter 2: Physician-pharmacy integration and healthcare outcomes: evidence from oral cancer treatments

Vertical integration among healthcare providers is an increasingly common feature of U.S. healthcare, but its welfare implications are theoretically ambiguous. I use novel data to study the effects of growing integration between physician groups and pharmacies dispensing high-cost, self-administered drugs, especially in oncology. More than one-third of spending on oral cancer treatments is now dispensed in this fashion. Using Medicare Fee-for-Service and Part D Claims from 2006-2019, I perform a stacked event study analysis to evaluate the effects of oncology practices launching pharmacies on welfare-relevant healthcare outcomes for patients using high-cost oral cancer treatments. I find point-of-sale prices paid to physician practice-based pharmacies were 1.0% lower than to non-integrated pharmacies, consistent with lower market shares for physician practice-based pharmacies. Pharmacy launch results in 6.2% more patients filling new prescriptions and reduces time-to-fill for new prescriptions subject to prior authorization. However, pharmacy launch does not improve adherence following initiation or result in detectable substitution to higher cost regimens. These results suggest physician-pharmacy integration increases and accelerates treatment initiation for high-cost oral cancer treatments, partly through faster prior authorization, without anti-competitive effects.

Chapter 3: Emergency department use of physicians and nurses (with Amitabh Chandra and Simone Matecna)
The extent to which medical knowledge reduces the use of avoidable health care, such as avoidable emergency department (ED) visits, is unknown, yet central for designing policies to reduce wasteful health care.We compared the ED use patterns of patients with more access to medical knowledge (doctors, nurses and their spouses) to controls based on demographics and comorbidities. We used Medicare Fee-for-Service claims from 2006-2017 linked to occupational and spousal data to identify these groups and coarsened exact matching to define their controls. Our sample included 79,665 physicians, 84,846 physician spouses, 164,423 nurses, and 170,850 nurse spouses. Physicians had 21.4% (95% CI: 20.4%, 22.4%) and nurses had 3.9% (95% CI: 3.8%, 5.1%) fewer ED visits relative to matched controls, due primarily to fewer visits that appear avoidable by virtue of being non-emergent, primary care treatable, or emergent but preventable Physicians and nurses had 26.8% (95% CI: 25.5%, 28.1%) and 6.8% (95% CI: 5.9%, 7.7%) fewer avoidable visits than their controls respectively. Spouses of physicians and nurses had similar ED use as their partners. Lower use of avoidable ED care by patients with more medical knowledge is consistent with the view that significant medical knowledge can reduce some amount of avoidable health care. Nonetheless, 93.2% of avoidable visits are not avoided by nurses, and 73.2% are not avoided by physicians, suggesting limits to the ability of more medical knowledge to reduce waste in health care.

 

Role

Graduation Year

Concentration