Validation of a Claims-Based Algorithm for Specialist Palliative Care Delivery in Metastatic Cancer

IF 4 2区医学 Q2 CLINICAL NEUROLOGY

Journal of pain and symptom management Pub Date : 2026-03-01 Epub Date: 2025-11-23 DOI:10.1016/j.jpainsymman.2025.11.016

May Hua MD, MS , Zhixin Yang MS , Ling Guo MD, MS , J. Brian Cassel PhD , R. Sean Morrison MD , Guohua Li MD, DrPH

{"title":"Validation of a Claims-Based Algorithm for Specialist Palliative Care Delivery in Metastatic Cancer","authors":"May Hua MD, MS , Zhixin Yang MS , Ling Guo MD, MS , J. Brian Cassel PhD , R. Sean Morrison MD , Guohua Li MD, DrPH","doi":"10.1016/j.jpainsymman.2025.11.016","DOIUrl":null,"url":null,"abstract":"<div><h3>Context</h3><div>The lack of valid methods to identify specialist palliative care (PC) delivery in population-level data impedes comprehensive understanding of its use.</div></div><div><h3>Objective</h3><div>To develop and validate a claims-based algorithm to identify receipt of specialist PC in Medicare beneficiaries with metastatic cancer.</div></div><div><h3>Methods</h3><div>We developed a claims-based algorithm to identify specialist PC, using a physician billing claim from a known PC clinician as the gold standard, retaining candidate variables with a positive predictive value (PPV) >60%. We evaluated algorithm performance and conducted simulation to measure bias resulting from algorithm use when examining the association between specialist PC and outcomes.</div></div><div><h3>Results</h3><div>We identified 1,384,750 claims from 68,121 patients. The prevalence of specialist PC was 3.8% on the claim-level and 26.8% on the patient-level. The provider specialty code for “hospice and palliative care” (PPV 80.4%) and the diagnosis code Z51.5 for “encounter for palliative care\" (PPV 67.5%) were included, where claims were counted as specialist PC if they had either variable. PPV and sensitivity of the algorithm were 68.0% and 83.0% respectively on a claim-level, and 78.4% and 88.3% respectively on a patient-level. Percent bias differed by outcome (hospice 4.2%, hospice enrollment ≥3 days 5.3%, intensive care unit use in the last 30 days of life −1.7%, chemotherapy use in the last 14 days of life −1.3%).</div></div><div><h3>Conclusions</h3><div>A simple algorithm can identify receipt of specialist PC care in Medicare claims for patients with metastatic cancer with reasonable accuracy. Algorithm use results in potentially acceptable amounts of bias, depending on study aims.</div></div>","PeriodicalId":16634,"journal":{"name":"Journal of pain and symptom management","volume":"71 3","pages":"Pages 385-393.e2"},"PeriodicalIF":4.0000,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of pain and symptom management","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0885392425009522","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/11/23 0:00:00","PubModel":"Epub","JCR":"Q2","JCRName":"CLINICAL NEUROLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Context

The lack of valid methods to identify specialist palliative care (PC) delivery in population-level data impedes comprehensive understanding of its use.

Objective

To develop and validate a claims-based algorithm to identify receipt of specialist PC in Medicare beneficiaries with metastatic cancer.

Methods

We developed a claims-based algorithm to identify specialist PC, using a physician billing claim from a known PC clinician as the gold standard, retaining candidate variables with a positive predictive value (PPV) >60%. We evaluated algorithm performance and conducted simulation to measure bias resulting from algorithm use when examining the association between specialist PC and outcomes.

Results

We identified 1,384,750 claims from 68,121 patients. The prevalence of specialist PC was 3.8% on the claim-level and 26.8% on the patient-level. The provider specialty code for “hospice and palliative care” (PPV 80.4%) and the diagnosis code Z51.5 for “encounter for palliative care" (PPV 67.5%) were included, where claims were counted as specialist PC if they had either variable. PPV and sensitivity of the algorithm were 68.0% and 83.0% respectively on a claim-level, and 78.4% and 88.3% respectively on a patient-level. Percent bias differed by outcome (hospice 4.2%, hospice enrollment ≥3 days 5.3%, intensive care unit use in the last 30 days of life −1.7%, chemotherapy use in the last 14 days of life −1.3%).

Conclusions

A simple algorithm can identify receipt of specialist PC care in Medicare claims for patients with metastatic cancer with reasonable accuracy. Algorithm use results in potentially acceptable amounts of bias, depending on study aims.

查看原文本刊更多论文

转移性癌症专科姑息治疗提供的基于索赔的算法验证。

背景：缺乏有效的方法来确定专科姑息治疗（PC）提供人口水平的数据阻碍了其使用的全面理解。目的：开发和验证一种基于索赔的算法，以确定接收转移性癌症的医疗保险受益人的专业PC。方法：我们开发了一种基于索赔的算法来识别专家PC，使用来自已知PC临床医生的医生账单索赔作为金标准，保留阳性预测值（PPV）为60%的候选变量。我们评估了算法的性能，并进行了模拟，以测量在检查专家PC和结果之间的关联时，算法使用造成的偏差。结果：我们从68,121例患者中确定了1,384,750例索赔。专科PC患病率在索赔水平为3.8%，在患者水平为26.8%。包括“临终关怀和姑息治疗”的提供者专业代码（PPV为80.4%）和“姑息治疗就诊”的诊断代码Z51.5 (PPV为67.5%)，如果索赔中有任何一个变量，则将其计算为专家PC。该算法在索赔水平的PPV和灵敏度分别为68.0%和83.0%，在患者水平的PPV和灵敏度分别为78.4%和88.3%。百分比偏差因结局而异（安宁疗护4.2%，安宁疗护登记≥3天5.3%，生命最后30天使用加护病房-1.7%，生命最后14天使用化疗-1.3%）。结论：一种简单的算法可以以合理的准确性识别转移性癌症患者医疗保险索赔中专科PC护理的接收情况。算法的使用会导致潜在的可接受偏差，这取决于研究目的。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Journal of pain and symptom management 医学-临床神经学

CiteScore

8.90

自引率

6.40%

发文量

821

审稿时长

26 days

期刊介绍： The Journal of Pain and Symptom Management is an internationally respected, peer-reviewed journal and serves an interdisciplinary audience of professionals by providing a forum for the publication of the latest clinical research and best practices related to the relief of illness burden among patients afflicted with serious or life-threatening illness.