Changes in Predictive Performance of a Frailty Index with Availability of Clinical Domains. Issue 8 (10th April 2020)

• Record Type: Journal Article
• Authors: Shi, Sandra M.; McCarthy, Ellen P.; Mitchell, Susan; Kim, Dae Hyun
• Is Part Of: Journal of the American Geriatrics Society. Volume 68:Issue 8(2020)
• Page Start: 1771
• Page End: 1777
• Publication Date: 2020-04-10
• Subjects: frailty -- risk prediction -- mortality -- falls -- disability; Geriatrics -- Periodicals; 618.97
• DOI: 10.1111/jgs.16436
• Languages: English
• ISSNs: 0002-8614

Abstract

Abstract : OBJECTIVES: Determine the effects of missing data in frailty identification and risk prediction. DESIGN: Analysis of the National Health in Aging Trends Study. SETTING: Community. PARTICIPANTS: About 6206 older adults. MEASUREMENTS: A 41‐variable frailty index (FI) was constructed with the following domains: comorbidities, activities of daily living (ADLs), instrumental activities of daily living, self‐reported physical limitations, physical performance, and neuropsychiatric tests. We evaluated discrimination after removing single and multiple domains, comparing C‐statistics for predicting 5‐year risk of mortality and 1‐year risks of disability and falls. RESULTS: The full FI yielded a mean of .18 and C‐statistics of .72 (95% confidence interval, .70‐.74) for mortality, .80 (.77‐.82) for disability, and .66 (.64‐.68) for falls. Removal of any single domain shifted the FI distribution, resulting in a mean FI ranging from .13 (removing comorbidities) to .20 (removing ADLs) and frailty prevalence (FI ≥ .25) from 16.0% to 28.7%. Among robust participants models missing ADLs misclassified most often, (19% as pre‐frail). Among pre‐frail and frail participants missing comorbidities misclassified most often(69.2% from pre‐frail to robust, 24% from frail to pre‐frail, and 4.9% from frail to robust). Removal of any single domain minimally changed C‐statistics: mortality, .71‐.73; disability, .79‐.80; and falls, .64‐.66. Removing neuropsychiatric testing and physical performance yielded comparable C‐statistics of .70, .78, and .66 for mortality, ADLs, and falls, respectively. However, removal of three or four domains based on likely availability decreased C‐statistics for mortality (.69, .66), disability (.75, .70), and falls (.64, .63), respectively. CONCLUSION: While FI discrimination is robust to missing information in any single domain, risk prediction is affected by absence of multiple domains. This work informs the application of FI as a clinical and research tool. J Am Geriatr Soc 68:1771‐1777, 2020.