Selected Publications



"Effectiveness and duration of protection provided by the live-attenuated Herpes Zoster vaccine in the Medicare population ages 65 years and older," (with Hector S. Izurieta, Michael Wernecke, Jeffrey A. Kelman, Sarah Wong, Richard Forshee, Douglas R. Pratt, Yun Lu, Qin Sun, Christopher Jankosky, Philip Krause, Chris Worrall, Thomas E. MaCurdy, Rafael Harpaz), Clinical Infection Diseases, 2017 (6): 785-93.

The retrospective cohort study aimed to investigate the effectiveness and durability of herpes zoster vaccination (HZV) in reducing the risk of herpes zoster (HZ) and related complications among aged Medicare beneficiaries. HZV-vaccinated beneficiaries were matched to unvaccinated beneficiaries and HZV-unvaccinated beneficiaries who had received pneumococcal vaccination and HZ-related outcomes in community and hospital settings were analyzed and compared, including ophthalmic zoster and postherpetic neuralgia (PHN). The study found that HZV were effective in preventing herpes zoster, but its effectiveness decreased over time. HZV was also found to be more effective and its effectiveness better preserved over time in preventing PHN and reducing HZ in hospital settings than community HZ.


Skilled Nursing Facilities Payment Models Research Technical Report, (with LLC Acumen), Report prepared for the Centers for Medicare & Medicaid Services, by Acumen, LLC, Burlingame, CA 2017.

This report presents the research conducted by Acumen to develop an alternative to the existing methodology used to pay for Skilled Nursing Facilities (SNF) service under Medicare Part A.


"Anti-Hypertensive Drug Associated Angioedema: Effect Modification by Race/Ethnicity," (with Marsha E. Reichman, Michael Wernecke, David J. Graham, Jiemin Liao, John Yap, Yoganand Chillarige, Mary Ross Southworth, Stephine Keeton, Margie R. Goulding, Katrina Mott, Jeffrey A. Kelman), Pharmacoepidemiology and Drug Safety, 2017.

This study aims to assess the association of ACEIs and ARBs with angioedema in Medicare beneficiaries, focusing on effect modification by race. Results showed that blacks have higher angioedema levels compared to whites in ACEIs, ARBs and beta-blockers. An interaction between black race and ACEI exposure increases the incidence rate for angioedema above that found in white new users of ACEIs. Interaction between black race, ACEI exposure and time of drug exposure results in further effect modification.



"Association Between Changes in CMS Reimbursement Policy and Drug Labels for Erythrocyte-Stimulating Agents With Outcomes for Older Patients Undergoing Hemodialysis Covered by Fee-for-Service Medicare," (with Cunlin Wang, Robert Kane, Mark Levonson, Jeffrey A. Kelman, Michael Wernecke, Joo-Yeon Lee, Steven Kozlowski, Carmen Dekmezian, Zhiwei Zhang, Aliza Thompson, Kimberly Smith, Yu-te Wu, Yuqin Wei, Yoganand Chillarige, Qin Ryan, Chris Worrall, Thomas E. MaCurdy, David J. Graham), JAMA Internal Medicine, 2016 (12): 1818-25.

This study assessed the effects of the 2011 changes to the CMS dialysis reimbursement policy and to the FDA’s drug label for erythropoiesis-stimulating agents (ESAs) on dialysis patients 66 years or older covered by fee-for-service Medicare. The use of ESAs significantly decreased, the rate of blood transfusions moderately increased, the risks of death or major adverse cardiovascular events were unchanged, and the risk of stroke was significantly reduced. A bundled comprehensive payment system for dialysis that includes the costs of ESAs and a revised drug label that recommends more conservative dosing had no adverse effects on older dialysis patients covered by fee-for-service Medicare.


"Stroke, bleeding, and mortality risks in older patients treated with dabigatran or rivaroxaban for nonvalvular atrial fibrillation (AF)," (with David J. Graham, Marsha Reichman, Michael Wernecke, Yahui Hsueh, Rima Izem, Mary Ross Southworth, Yuqin Wei, Jiemin Liao, Margie Goulding, Katrina Mott, Yoganand Chillarige, Thomas MaCurdy, Chris Worrall, Jeffrey A. Kelman), Journal of the American Medical Association, 2016 (11): 1662-71.

This study aimed to evaluate the comparative safety of rivaroxaban versus dabigatran in patients with non-valvular atrial fibrillation who initiated treatment for stroke prevention. The study found that rivaroxaban was associated with significant increases in risk of intracranial hemorrhage and major extracranial bleeding, including major gastrointestinal bleeding compared with dabigatran, along with nonsignificant reduction in risk of thromboembolic stroke.



"Febrile non-haemolytic transfusion reaction occurrence and potential risk factors among the U.S. elderly transfused in the inpatient setting, as recorded in Medicare databases during 2011–2012," (with Forshee RA Menis M, Anderson SA, McKean S, Gondalia R, Warnock R, Johnson C, Mintz PD, Worrall CM, Kelman JA, Izurieta HS), Vox Sanguinis, 2014 (3): 251-61.

This study used Medicare claims data to evaluate risk factors for FNHTR following transfusions of various blood products. Multivariate regression analyses indicate that an individual’s risk for FNHTR increases significantly when transfused with red blood cell- and platelet-containing products, relative to blood plasma. Elevated risk was also linked to clinical conditions including lymphoma, leukemia, a history of recent blood transfusions, and other diseases.


"Babesiosis Occurrence Among the Elderly in the United States, as Recorded in Large Medicare Databases During 2006-2013," (with Mikhail Menis, Richard A. Forshee, Sanjai Kumar, Stephen McKean, Rob Warnock, Hector S. Izurieta, Rahul Gondalia, Chris Johnson, Paul D. Mintz, Mark Walderhaug, Jeffrey Kelman, Steven Anderson), PLoS One, 2015 (10):  e0140332

This study demonstrated the viability of using large claims databases to monitor the spread of vector-borne diseases, in particular babesiosis. During the study period, babesiosis prevalence increased over time, and peaked in states where the tick vector is endemic, as well as during summer months. This study also measured differences in babesiosis occurrence by patient age and sex.



Follow-up Analysis to Evaluate the Impact of Reassignment of Low-Income Subsidy Beneficiaries in Medicare Part D on Health Outcomes, (with LLC Acumen), Report prepared for the Centers for Medicare & Medicaid Services, by Acumen, LLC, Burlingame, CA 2014.

A previous analysis from 2009 was repeated to determine if there are adverse outcomes as a result of the low-income subsidy (LIS) reassignment process. Adverse outcomes included death, inpatient admission, and emergency room (ER) utilization. The 2009 analysis, which is included in the Appendix at the end of this summary, found no differences in these outcomes for LIS beneficiaries who were reassigned and those who were not.


"Cardiovascular and Mortality Risks in Older Medicare Patients Treated with Varenicline or Bupropion for Smoking Cessation," (with David J. Graham, Kunthel By, Stephen McKean, Andrew Mosholder, Cynthia Kornegay, Judith A. Racoosin, Jessica Young, Mark Levenson, Thomas E. MaCurdy, Chris Worrall, and  Jeffrey A. Kelman), Pharmacoepidemiology and Drug Safety, 2014 1205-12.

This study compared cardiovascular and mortality risks in elderly patients treated with varenicline or bupropion for smoking cessation. Cardiovascular and mortality risks were not increased in older patients treated with varenicline compared with bupropion for smoking cessation. A potential increase in the risk of stroke with varenicline could not be excluded.



"Cardiovascular and Mortality Risks in Parkinson’s Disease Patients Treated with Entacapone," (with David J. Graham, James R. Williams, Ya-Hui Hsueh, Katlyn Calia, Mark Levenson, Simone P. Pinheiro, Thomas E. MaCurdy, David Shih, Chris Worrall, and Jeffrey A. Kelman), Movement Disorders 2013 (4): 490-97.

The study used Medicare data to investigate whether entacapone increased cardiovascular and mortality risk compared with the use of a non-levodopa dopamine agonist (DA) or a selective monoamine oxidase type-B inhibitor (MAOBI). Entacapone was not associated with an increased risk of AMI, stroke, or death in elderly patients with Parkinson’s disease.


"Left Bundle Branch Block Predicts Better Survival in Women Than Men Receiving Cardiac Resynchronization Therapy: Long-Term Follow-Up of ~145,000 Patients," (with Zak Loring, Daniel A. Caños, Kimberly Selzman, Naomi D. Herz, Henry Silverman, Thomas E. MaCurdy, Christopher M. Worrall, Jeffrey Kelman, Mary E. Ritchey, Ileana L. Piña and David G. Strauss), JACC: Heart Failure, 2013 (3): 237-44.

This study aimed to test the hypothesis that, in cardiac resynchronization therapy defibrillator (CRT-D) recipients, conventional left bundle branch block (LBBB) diagnosis predicts better survival in women than men. The study found that LBBB diagnosis is associated with greater survival in women than men receiving CRT and this discrepancy is not explained by differences in measured comorbidities. Possible explanations for this difference include that LBBB may have different prognostic significance by gender or that LBBB diagnosis is more often false positive in men compared to women.


"Comparative effectiveness research in practice: the Drug Effectiveness Review Project experience," (with Jordan M. VanLare, Hui-Hsing Wong, Jonathan Gibbs, Rolf Timp, Stephanie Whang, Chris Worrall, Jeffrey Kelman and Patrick H. Conway), Journal of Comparative Effectiveness Research, 2013 (6): 541-50.

This study assessed the effect of the Drug Effectiveness Review Project’s comparative effectiveness research findings on prescribing behavior independently and in conjunction with a Medicaid preferred drug list. The study found that the states that participated in the Drug Effectiveness Review Project and used preferred drug lists have greater utilization of high quality drugs, representing an effective two-pronged strategy to translate comparative effectiveness research into practice.



"Cardiac Resynchronization Therapy Defibrillators: Pre and Postmarket Experience," (with Naomi D. Herz, David G. Strauss, Mary E. Ritchey, Zak Loring, Samantha B. Jacobs, Henry Silverman, Chris Worrall, Jeffrey Kelman, Ileana L. Piña and Daniel A. Caños), Circulation, 2012 (21 Supplement): A14558.

The aim of this study was to determine if baseline characteristics of Medicare beneficiaries receiving CRT-D are similar to those reported in the premarket. Both atrial fibrillation and hypertension were more prevalent in the Medicare cohort than MIRACLE ID while left bundle branch block was less prevalent in Medicare than COMPANION or MIRACLE ID. Overall, this study demonstrated that analyzing claims data identifies vast differences in characteristics between pre- and post-market CRT-D populations.



"Utilizing Medicare Claims Data for Real-Time Drug Safety Evaluations: Is It Feasible?," (with Abraham G. Hartzema, Judith A. Racoosin, Thomas E. MaCurdy, Jonathan M. Gibbs and Jeffrey A. Kelman), Pharmacoepidemiology and Drug Safety, 2011 (7): 684-88.

This study aimed to evaluate the impact of (i) claims information updates and (ii) delay in claim processing (claims delay) on real-time evaluation of health service and drug safety signals using the Medicare database. The study found that claims delay is consistent across time and is minimal, while claims adjudication does not substantially impact the content of clinical information in the Medicare database. It concluded that the Medicare database thus provides consistent information in a prompt enough manner to facilitate medical product safety evaluations in real time.



"Risk of Acute Myocardial Infarction, Stroke, Heart Failure, and Death in Elderly Medicare Patients Treated With Rosiglitazone or Pioglitazone," (with David J. Graham, Rita Ouellet-Hellstrom, Thomas E. MaCurdy, Farzana Ali, Christopher Sholley, Christopher Worrall and Jeffrey A. Kelman), Journal of the American Medical Association, 2010 (4): 411-18.

This study aimed to use survival analysis to determine if the risk of serious cardiovascular harm, defined by acute myocardial infarction, stroke, heart failure, or mortality is increased by rosiglitazone compared with pioglitazone. The study found that patients on rosiglitazone were at an elevated risk of all study outcomes compared to those on pioglitazone