Frailty Indices Using Claims Data - Examples from a Hospitalized Medicare Population
Author(s)
Michael V. Murphy, BS1, Jessica Duchen, MPH2, Pamela Blumberg, MPH, DrPH1;
1Magnolia Market Access, Bridgewater Township, NJ, USA, 2Magnolia Market Access, Hamden, CT, USA
1Magnolia Market Access, Bridgewater Township, NJ, USA, 2Magnolia Market Access, Hamden, CT, USA
Presentation Documents
OBJECTIVES: Frailty is an indicator of health outcomes in older adults but is difficult to measure in claims data. We describe the frailty of hospitalized Medicare enrollees using two published indices: the Risk Analysis Index (RAI) adapted for ICD-10-CM codes and the Kim Index.
METHODS: Kim and RAI indices were calculated at admission for all hospitalizations in the Medicare 5% SAFs (2017-2022). Kim ranges from 0 to 1 and RAI from 0 to 81; with higher scores indicating increased frailty. Patients were classified into four categories defined by each frailty score for every hospitalization in the timeframe. Exclusions included stays with unknown age or sex, or <12 months of continuous Part A and B enrollment before admission. Descriptive statistics for Charlson Comorbidity index (CCI), length of stay (LOS), discharge status, 30-day readmission, and 30-day mortality for each frailty index category are presented.
RESULTS: Among 2,540,925 hospitalizations, mean Kim score was 0.3 (±0.09), corresponding to Pre-Frail category, and mean RAI score was 32.1 (±10.13), corresponding to Normal. Within Kim, 9.3% classified as Robust (least frail), 44.7% as pre-Frail, 31.6% as Mildly Frail, and 14.4% as Moderate-Severely Frail (most frail). Using RAI score, 28.2% classified as Robust, 39.2% as Normal, 21.5% as Frail, and 11.1% as Very Frail. The correlation coefficient between scores was 0.227; spearman-rank correlation coefficient between categories was 0.219.
Hospitalizations classified as frailest for each score had the highest CCI (Kim:5.9; RAI:5.6), LOS (Kim:7.77; RAI:8.1), in-hospital mortality (Kim:5.3%; RAI:6.9%), and proportion discharged to hospice (Kim:5.9%; RAI:11.4%). Hospitalizations resulting in live discharges of frailest patients had the highest 30-day mortality (Kim:12.0%; RAI:20.9%) and 30-day readmissions (Kim:33.8%; RAI:27.4%).
CONCLUSIONS: While the frailest patients had poorer outcomes with both indices, further analysis is required to determine if frailty can be confidently identified from claims data, or if Kim and RAI are measuring different aspects of frailty.
METHODS: Kim and RAI indices were calculated at admission for all hospitalizations in the Medicare 5% SAFs (2017-2022). Kim ranges from 0 to 1 and RAI from 0 to 81; with higher scores indicating increased frailty. Patients were classified into four categories defined by each frailty score for every hospitalization in the timeframe. Exclusions included stays with unknown age or sex, or <12 months of continuous Part A and B enrollment before admission. Descriptive statistics for Charlson Comorbidity index (CCI), length of stay (LOS), discharge status, 30-day readmission, and 30-day mortality for each frailty index category are presented.
RESULTS: Among 2,540,925 hospitalizations, mean Kim score was 0.3 (±0.09), corresponding to Pre-Frail category, and mean RAI score was 32.1 (±10.13), corresponding to Normal. Within Kim, 9.3% classified as Robust (least frail), 44.7% as pre-Frail, 31.6% as Mildly Frail, and 14.4% as Moderate-Severely Frail (most frail). Using RAI score, 28.2% classified as Robust, 39.2% as Normal, 21.5% as Frail, and 11.1% as Very Frail. The correlation coefficient between scores was 0.227; spearman-rank correlation coefficient between categories was 0.219.
Hospitalizations classified as frailest for each score had the highest CCI (Kim:5.9; RAI:5.6), LOS (Kim:7.77; RAI:8.1), in-hospital mortality (Kim:5.3%; RAI:6.9%), and proportion discharged to hospice (Kim:5.9%; RAI:11.4%). Hospitalizations resulting in live discharges of frailest patients had the highest 30-day mortality (Kim:12.0%; RAI:20.9%) and 30-day readmissions (Kim:33.8%; RAI:27.4%).
CONCLUSIONS: While the frailest patients had poorer outcomes with both indices, further analysis is required to determine if frailty can be confidently identified from claims data, or if Kim and RAI are measuring different aspects of frailty.
Conference/Value in Health Info
2025-05, ISPOR 2025, Montréal, Quebec, CA
Value in Health, Volume 28, Issue S1
Code
MSR14
Topic
Methodological & Statistical Research
Disease
No Additional Disease & Conditions/Specialized Treatment Areas, SDC: Geriatrics