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

Effects of Medicare Part D plan switching on prescription drug expenditures: a longitudinal analysis using mixture models

Talisa, Victor (2015) Effects of Medicare Part D plan switching on prescription drug expenditures: a longitudinal analysis using mixture models. Master's Thesis, University of Pittsburgh. (Unpublished)

Submitted Version

Download (1MB)


Background: Medicare beneficiaries enrolled in a Part D drug plan are faced with choosing from a large number of plans with a complex array of attributes. Unnecessarily high spending may lead to cost-related non-adherence (CRN). The extent to which changes in spending are attributable to sponsor renewal of plans offered (compared to termination or consolidation), or beneficiary plan switching (compared to not switching), is not well known.
Research Aims: This study evaluated whether 1) unanticipated plan consolidations or terminations lead to higher gross total and patient out-of-pocket (OOP) costs compared to plan renewals, and 2) plan switching is associated with gross total and patient OOP costs.
Methods: We used Part D data from N=1,187,469 beneficiaries enrolled between 2006 and 2012 who never received income subsidies and who were enrolled continuously through December to January of two consecutive years. Beneficiaries were classified into one of five plan switching groups at each transition period: plan termination—switch (PT-S), plan consolidation—switch (PC-S) or no switch (PC-NS), and plan renewal—switch (PR-S) or no switch (PR-NS). Longitudinal mixed models using an analytical sample of N=17,812 beneficiaries evaluated plan switching group effect on gross total and patient OOP costs, controlling for covariates. Linear mixed models (LMM) and mixture models were compared.
Results: Model diagnostics suggested that the mixture model provided a better fit than LMM. Neither plan termination nor consolidation led to consistently higher gross total spending or patient OOP costs. The lowest gross costs were observed in the PC-S group, with spending in PC-NS, PR-NS and PR-S estimated to be 4.6%, 6.9% and 5.2% higher, respectively. Patient OOP spending was significantly higher among PC-NS and PR-NS compared to PC-S and PR-S. Estimated differences were between 4.8% and 7.5%, equivalent to $24 to $60 higher annual patient OOP spending among individuals who did not switch compared to those that switched.
Conclusions: The public health impact of our results is that increased spending may not be a significant concern for individuals exposed to plan consolidation or termination. However, plan switching may not reduce spending (and consequently risk of CRN) as much as has been shown to be possible in previous studies.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Talisa, Victorvit13@pitt.eduVIT13
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Thesis AdvisorBuchanich, Jeaninejeanine@pitt.eduJEANINE
Committee MemberZhang, Yutingytzhang@pitt.eduYTZHANG
Committee MemberYouk, Adaayouk@pitt.eduAYOUK
Date: 28 September 2015
Date Type: Publication
Defense Date: 20 July 2015
Approval Date: 28 September 2015
Submission Date: 20 July 2015
Access Restriction: 2 year -- Restrict access to University of Pittsburgh for a period of 2 years.
Number of Pages: 78
Institution: University of Pittsburgh
Schools and Programs: School of Public Health > Biostatistics
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: Medicare Part D, Mixture Model, Longitudinal, Healthcare Expenditure Data
Date Deposited: 28 Sep 2015 18:36
Last Modified: 01 Sep 2017 05:15


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item