Interhospital variability in health care–associated infections and payments after durable ventricular assist device implant among Medicare beneficiaries
Donald S. Likosky,Guangyu Yang,Min Zhang,Preeti N. Malani,Michael D. Fetters,Raymond J. Strobel,Carol E. Chenoweth,Hechuan Hou,Francis D. Pagani,Michigan Congestive Heart Failure Investigators,Donald S. Likosky,Francis D. Pagani,Ashraf Shaaban Abdel Aziz Abou El Ela,Paul C. Tang,Michael P. Thompson,Keith Aaronson,Supriya Shore,Thomas Cascino,Katherine B. Salciccioli,Min Zhang,Jeffrey S. McCullough,Michelle Hou,Allison M. Janda,Michael R. Mathis,Tessa M.F. Watt,Michael J. Pienta,Alexander Brescia,Austin Airhart,Daniel Liesman,Khalil Nassar
DOI: https://doi.org/10.1016/j.jtcvs.2021.04.074
2021-05-01
Abstract:<h2 class="section-title u-h3 u-margin-l-top u-margin-xs-bottom">STRUCTURED </h2><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Objective</h3><p>To investigate variations across hospitals in infection rates and associated costs, the latter reflected in 90-day Medicare payments. Despite high rates and expenditures of healthcare-associated infections associated with durable ventricular assist device implant, few studies have examined inter-hospital variation and associated costs.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Methods</h3><p>Clinical data on 8,688 patients receiving primary durable ventricular assist devices from July 2008 to July 2017 from The Society of Thoracic Surgeons Interagency Registry for Mechanically Assisted Circulatory Support ("Intermacs") hospitals (n=120) were merged with post-implant 90-day Medicare claims. Terciles of hospital-specific, risk-adjusted infection rates per 100 patient-months were estimated using Intermacs and associated with Medicare payments (among 5,440 Medicare beneficiaries). Primary outcomes included infections within 90 days of implant and Medicare payments.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Results</h3><p>There were 3,982 infections identified among 27.8% (2,417/8,688) of patients developing an infection. The median (25<sup>th</sup>, 75<sup>th</sup> percentile) adjusted incidence of infections (per 100 patient-months) across hospitals was 14.3 (9.3, 19.5) and varied by hospital (range 0.0 – 35.6). Total Medicare payments from implant to 90-days were 9.0% (absolute difference: $13,652) greater in high versus low infection tercile hospitals, p<0.0001. The period between implant to discharge accounted for 73.1% of the difference in payments during the implant to 90-day period across terciles.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Conclusions</h3><p>Healthcare-associated infection rates post-durable ventricular assist device implant varied by hospital and were associated with increased 90-day Medicare expenditures. Interventions targeting preventing infections could improve the value of durable ventricular assist device support from both the societal and hospital perspectives.</p>
surgery,cardiac & cardiovascular systems,respiratory system