Identifying delivery of palliative care in administrative billing claims data: A scoping review.

Abstract

e23128 Background: Palliative care is increasingly recognized as an essential component of whole-person healthcare for people living with serious illnesses. Claims-based data provide a valuable resource for understanding the utilization and delivery of palliative care. However, there is substantial variation in how palliative care services are identified in claims data. We sought to identify the different methods described in the literature. Methods: Following PRISMA guidelines, we developed a search strategy incorporating terms related to (1) “Palliative” and (2) “Medicare,” “claims,” or “billing” across PubMed, Scopus, and Web of Science, published between 2018 and 2025. We conducted an initial screening based on titles and abstracts. Papers meeting the inclusion criteria were further evaluated for eligibility. Studies were excluded if they were not written in English, were not US-based, were limited to abstracts, were not research studies (e.g., book chapters), or assessed palliative care outside of claims data. From the final selection of studies, we extracted key data elements and synthesized findings to identify the different methods used to identify palliative care. Results: Our search retrieved 443 studies. After screening and applying our exclusion criteria, 75 papers were retained for full review. The palliative variable of interest was categorized as palliative care (e.g., encounter for palliative care or specialty code) and palliative intent therapies (e.g., chemotherapy or radiotherapy). 28% (21/75) of papers did not specify the billing codes to measure the palliative variable (n = 7), the billing data files to search these codes (n = 13), or both (n = 1). For papers specifically utilizing Medicare claims (51/75), 33% (17/51) excluded either the billing codes and/or the billing data file information needed to reproduce the results. Conclusions: Our scoping review identified a spectrum of approaches to measuring palliative care services in claims data, with differences in the details specifying the codes used and which billing claims were used to search for these codes (e.g., inpatient, outpatient, carrier Medicare files). Although many papers provided detailed methodological sections when conducting palliative research using claims data, more consistent inclusion of detailed methodology would enhance reproducibility, cross-study comparison and synthesis, and the results of meta-analyses. With respect to Medicare claims data, one suggestion would be to specifically identify the source file information (e.g., “exposure to palliative care (V66.7) ... captured in either inpatient or outpatient Medicare claims data”). By increasing methodological transparency in future research, researchers can enhance the reproducibility of claims-based analyses and the synthesis of this body of knowledge to support establishing the evidence-base for optimal palliative care delivery.