Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Variation in diffusion of prescription drugs: mechanisms and cost implications

Metes, Ilinca (2020) Variation in diffusion of prescription drugs: mechanisms and cost implications. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

Submitted Version

Download (1MB) | Preview


In the United States, there is well-documented geographic variation in prescription drug spending and utilization. However, the specific role of physicians, and physician drug adoption behavior on the variation in patient-level prescription drug is still being investigated. This dissertation aims to add to the literature by gaining a better understanding of the role of physician adoption of brand name drugs on geographic variation in patient-level spending, and also utilizing physician peer networks and social network analysis to help elucidate a possible mechanism underlying why some physicians adopt brand name drugs faster than others. The public health relevance of this dissertation rests in improving our understanding of how, and why, new drug adoption drives prescription drug spending in the face of ever rising health care expenditures, and an aging population, that will likely increase demand for chronic disease medications.

Chapter one investigates the association between physician adoption of a moderately novel anti-diabetic drug, sitagliptin, and anti-diabetic drug spending in the Medicare and Medicaid populations in Pennsylvania. We found that anti-diabetic drug spending in both populations were remarkably similar, as were the magnitudes of the associations between sitagliptin adoption and higher, local-level, drug spending. These results highlight how in spite of differences in population characteristics, and the administration of drug benefits in these two distinct programs, physician drug adoption, drug spending, and prescriptions follow similar trends.

Chapter two investigates the association between physician adoption of a highly novel anti-coagulant drug, dabigatran, and both drug and medical-related spending in the Medicare population in Pennsylvania. We found that physician adoption of dabigatran was significantly associated with both higher anti-coagulant drug spending, and higher overall non-drug medical spending in this Medicare cohort. This finding highlights the importance of physician drug adoption behavior, and suggests that areas with higher rates of dabigatran prescribing are not accompanied by cost-offsets in medical-related spending.

Chapter three utilizes social network analysis and instrumental variable modeling to help estimate the fraction of geographic variation in physician drug adoption that can be attributable to peer/social influence. We found that physician drug adoption decisions were significantly influenced by peer adoption behavior across three distinct chronic disease drug-classes (anti-diabetic, anti-coagulant, and anti-hypertensive). Additionally, this study highlights that, consistently across the three drug classes studied, peer influence appears to be explain roughly half of the geographic variation of physician drug adoption.

Taken together, these findings point to the indication that individual characteristics of patients and physicians should be viewed in conjunction with social networks and peer connections when trying to understand variations in behavior, utilization, and spending across the health care system.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Metes, Ilincaidm2@pitt.eduidm2
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairDonohue, Julie M.jdonohue@pitt.edujdonohue
Committee MemberChang, Chung-Chou H.changj@pitt.educhangj
Committee MemberCole, Evan S.evancole@pitt.eduevancole
Committee MemberRoberts, Mark S.mroberts@pitt.edumroberts
Committee MemberDriessen, Juliadriessen@pitt.edudriessen
Date: 29 January 2020
Date Type: Publication
Defense Date: 4 October 2019
Approval Date: 29 January 2020
Submission Date: 20 November 2019
Access Restriction: 2 year -- Restrict access to University of Pittsburgh for a period of 2 years.
Number of Pages: 86
Institution: University of Pittsburgh
Schools and Programs: School of Public Health > Health Policy & Management
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: Prescription Drugs, Physician Behavior, Technology Adoption, Medicare, Medicaid
Date Deposited: 29 Jan 2020 21:31
Last Modified: 01 Jan 2022 06:15


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item