Richard Duszak MD, Presenter: Nothing to Disclose

Nadia Bilal, Abstract Co-Author: Nothing to Disclose

Danny Hughes PhD, Abstract Co-Author: Nothing to Disclose

David Andrew Rosman MD, Abstract Co-Author: Nothing to Disclose

As healthcare payment systems evolve from fee-for-service to value-based episodic bundles, much interest has focused on inpatient payments to hospitals, particularly as they relate to length of stay and readmissions. Physician services, including those of radiologists, have been largely ignored. In anticipation of inpatient bundled physician professional payments, we aimed to develop an evidence-based episode frequency focused strategic framework for identifying encounters where initial bundled payment modeling might be most impactful.

Using a 5% random sample of all approximately 32 million Medicare beneficiaries from 2011, we first used Part A (hospital) claims data to categorize all inpatient admissions using Medicare Severity Diagnosis Related Groups (DRGs). Imaging professional services (defined by Berenson-Eggers Types of Service) rendered by radiologists associated with each episode were then identified separately using Part B (physician services) claims data. Inpatient admission DRGs involving radiologist professional services were frequency ranked by to identify those disproportionately associated with total inpatient encounters.

Of 618,906 identified inpatient admission episodes, imaging professional services were rendered in 430,707 (69.6%), attributable to nearly all (739/742; 99.6%) 742 uniquely identifiable DRGs. Just 4 DRGs, however, accounted for 11.3% of all hospital encounters involving radiologist imaging professional services (470, 871, 292, 194). The top 9 accounted for 20.1%, top 17 for 30.3%, and top 41 for 50.3%. All other 701 combined accounted for 48.5% (1.2% were non-categorizable). The bottom 506 (68.2%) DRG codes together accounted for only 10% of episodes where radiologists rendered imaging professional services.

Efforts to model radiologist inpatient professional services have been hampered by the nearly ubiquitous involvement of imaging in the inpatient setting. A very small number of DRGs, however, account for a disproportionately large share of all admissions involving imaging. These would ideally serve as the basis for initial inpatient imaging bundled payment initiatives.

In the inpatient setting, just 4 DRGs account for 10% of all admissions where radiologists typically render services. Initial bundled payment initiatives should be evidence-based and targeted.

Duszak, R, Bilal, N, Hughes, D, Rosman, D, Bundled Inpatient Radiologist Imaging Professional Services: An Empiric Episode Frequency Focused Approach to Bundled Payment Modeling.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14004054.html