Abstract: PO0992
Identifying Peritoneal Dialysis (PD)-Associated Peritonitis Using Medicare Claims
Session Information
- Peritoneal Dialysis
November 04, 2021 | Location: On-Demand, Virtual Only
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 702 Dialysis: Home Dialysis and Peritoneal Dialysis
Authors
- Young, Eric W., Arbor Research Collaborative for Health, Ann Arbor, Michigan, United States
- Zhao, Junhui, Arbor Research Collaborative for Health, Ann Arbor, Michigan, United States
- Pisoni, Ronald L., Arbor Research Collaborative for Health, Ann Arbor, Michigan, United States
- McCullough, Keith, Arbor Research Collaborative for Health, Ann Arbor, Michigan, United States
- Piraino, Beth M., University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, United States
- Shen, Jenny I., The Lundquist Institute, Torrance, California, United States
- Boudville, Neil, University of Western Australia, Perth, Western Australia, Australia
- Schaubel, Douglas Earl, University of Pennsylvania, Philadelphia, Pennsylvania, United States
- Teitelbaum, Isaac, University of Colorado School of Medicine, Aurora, Colorado, United States
- Perl, Jeffrey, St. Michaels, Toronto, Ontario, Canada
Background
Medicare fee-for-service (FFS) claims offer a population-based approach to PD-associated peritonitis that may offer valuable insights into predictors, trends and preferred practices.
Methods
We used United States Renal Data System (USRDS) standard analysis files for claims (inpatient, outpatient and physician-supplier), eligibility, modality and demographic information. The sample consisted of PD patient-months from 2013 through 2017 characterized by Medicare FFS coverage and paid claims for dialysis or hospital services. We identified ICD-9 and ICD-10 diagnosis codes for peritonitis, including those that do not clearly distinguish peritonitis from catheter infections/inflammation (“catheter codes”). A new peritonitis episode was defined as a peritonitis claim 30+ days from any prior peritonitis claim or 50+ days from the initial peritonitis claim for a prior episode.
Results
The sample included 88,396 adult patients (128,000 observed patient-years), yielding 510,000 peritonitis claims and 75,000 peritonitis episodes. Coding was heterogeneous with no single diagnosis code present on the majority of claims. Peritonitis episodes were inferred from aggregated claims (mean 6.3, median 2). Half of episodes were exclusively outpatient, 7% exclusively inpatient, and 16% exclusively comprised of catheter code claims. The overall peritonitis rate was 0.59 and 0.49 episodes per patient-year with and without inclusion of catheter codes respectively. Peritonitis rates declined by 4%/year from 2013-2017, and varied by age, race (Black > White >Asian), and ESKD vintage.
Conclusion
Coding heterogeneity indicates a lack of standardization and need for clearer coding guidance. We found differences between races, ages, and patient vintages, and declining rates from 2013-2017. These rates are 2-fold higher than reported in US-PDOPPS by Perl et al (AJKD 2020) which is not restricted to Medicare. Claims are an important data source for peritonitis, but more work is needed to validate these rates.
Funding
- Other NIH Support