Tags

Type your tag names separated by a space and hit enter

Prospective Comparison of Preoperative Predictive Performance Between 3 Leading Frailty Instruments.
Anesth Analg. 2020 07; 131(1):263-272.A&A

Abstract

BACKGROUND

Guidelines recommend routine preoperative frailty assessment for older people. However, the degree to which frailty instruments improve predictive accuracy when added to traditional risk factors is poorly described. Our objective was to measure the accuracy gained in predicting outcomes important to older patients when adding the Clinical Frailty Scale (CFS), Fried Phenotype (FP), or Frailty Index (FI) to traditional risk factors.

METHODS

This was an analysis of a multicenter prospective cohort of elective noncardiac surgery patients ≥65 years of age. Each frailty instrument was prospectively collected. The added predictive performance of each frailty instrument beyond the baseline model (age, sex, American Society of Anesthesiologists' score, procedural risk) was estimated using likelihood ratio test, discrimination, calibration, explained variance, and reclassification. Outcomes analyzed included death or new disability, prolonged length of stay (LoS, >75th percentile), and adverse discharge (death or non-home discharge).

RESULTS

We included 645 participants (mean age, 74 [standard deviation, 6]); 72 (11.2%) participants died or experienced a new disability, 164 (25.4%) had prolonged LoS, and 60 (9.2%) had adverse discharge. Compared to the baseline model predicting death or new disability (area under the curve [AUC], 0.67; R, 0.08, good calibration), prolonged LoS (AUC, 0.73; R, 0.18, good calibration), and adverse discharge (AUC, 0.78; R, 0.16, poor calibration), the CFS improved fit per the likelihood ratio test (P < .02 for death or new disability, <.001 for LoS, <.001 for discharge), discrimination (AUC = 0.71 for death or new disability, 0.76 for LoS, 0.82 for discharge), calibration (good for death or new disability, LoS, and discharge), explained variance (R = 0.11 for death or new disability, 0.22 for LoS, 0.25 for discharge), and reclassification (appropriate directional reclassification) for all outcomes. The FP improved discrimination and R for all outcomes, but to a lesser degree than the CFS. The FI improved discrimination for death or new disability and R for all outcomes, but to a lesser degree than the CFS and the FP. These results were consistent in internal validation.

CONCLUSIONS

Frailty instruments provide meaningful increases in accuracy when predicting postoperative outcomes for older people. Compared to the FP and FI, the CFS appears to improve all measures of predictive performance to the greatest extent and across outcomes. Combined with previous research demonstrating that the CFS is easy to use and requires less time than the FP, clinicians should consider its use in preoperative practice.

Authors+Show Affiliations

From the Department of Anesthesiology and Pain Medicine, University of Ottawa, and the Ottawa Hospital, Ottawa, Ontario, Canada. School of Epidemiology and Public Health, University of Ottawa, Ottawa, Ontario, Canada. Ottawa Hospital Research Institute, Ottawa, Ontario, Canada.From the Department of Anesthesiology and Pain Medicine, University of Ottawa, and the Ottawa Hospital, Ottawa, Ontario, Canada.From the Department of Anesthesiology and Pain Medicine, University of Ottawa, and the Ottawa Hospital, Ottawa, Ontario, Canada. Ottawa Hospital Research Institute, Ottawa, Ontario, Canada.School of Epidemiology and Public Health, University of Ottawa, Ottawa, Ontario, Canada. Department of Surgery.From the Department of Anesthesiology and Pain Medicine, University of Ottawa, and the Ottawa Hospital, Ottawa, Ontario, Canada. Ottawa Hospital Research Institute, Ottawa, Ontario, Canada.From the Department of Anesthesiology and Pain Medicine, University of Ottawa, and the Ottawa Hospital, Ottawa, Ontario, Canada. Ottawa Hospital Research Institute, Ottawa, Ontario, Canada.Division of Geriatric Medicine, University of Ottawa, and the Ottawa Hospital, Ottawa, Ontario, Canada.Hôpital Montfort, Ottawa, Ontario, Canada.From the Department of Anesthesiology and Pain Medicine, University of Ottawa, and the Ottawa Hospital, Ottawa, Ontario, Canada.Ottawa Hospital Research Institute, Ottawa, Ontario, Canada. Department of Medicine, University of Ottawa, and the Ottawa Hospital, Ottawa, Ontario, Canada.School of Epidemiology and Public Health, University of Ottawa, Ottawa, Ontario, Canada. Ottawa Hospital Research Institute, Ottawa, Ontario, Canada. Department of Medicine, University of Ottawa, and the Ottawa Hospital, Ottawa, Ontario, Canada.

Pub Type(s)

Comparative Study
Journal Article
Multicenter Study
Research Support, Non-U.S. Gov't

Language

eng

PubMed ID

31569165

Citation

McIsaac, Daniel I., et al. "Prospective Comparison of Preoperative Predictive Performance Between 3 Leading Frailty Instruments." Anesthesia and Analgesia, vol. 131, no. 1, 2020, pp. 263-272.
McIsaac DI, Harris EP, Hladkowicz E, et al. Prospective Comparison of Preoperative Predictive Performance Between 3 Leading Frailty Instruments. Anesth Analg. 2020;131(1):263-272.
McIsaac, D. I., Harris, E. P., Hladkowicz, E., Moloo, H., Lalu, M. M., Bryson, G. L., Huang, A., Joanisse, J., Hamilton, G. M., Forster, A. J., & van Walraven, C. (2020). Prospective Comparison of Preoperative Predictive Performance Between 3 Leading Frailty Instruments. Anesthesia and Analgesia, 131(1), 263-272. https://doi.org/10.1213/ANE.0000000000004475
McIsaac DI, et al. Prospective Comparison of Preoperative Predictive Performance Between 3 Leading Frailty Instruments. Anesth Analg. 2020;131(1):263-272. PubMed PMID: 31569165.
* Article titles in AMA citation format should be in sentence-case
TY - JOUR T1 - Prospective Comparison of Preoperative Predictive Performance Between 3 Leading Frailty Instruments. AU - McIsaac,Daniel I, AU - Harris,Emma P, AU - Hladkowicz,Emily, AU - Moloo,Husein, AU - Lalu,Manoj M, AU - Bryson,Gregory L, AU - Huang,Allen, AU - Joanisse,John, AU - Hamilton,Gavin M, AU - Forster,Alan J, AU - van Walraven,Carl, PY - 2019/10/1/pubmed PY - 2020/8/12/medline PY - 2019/10/1/entrez SP - 263 EP - 272 JF - Anesthesia and analgesia JO - Anesth Analg VL - 131 IS - 1 N2 - BACKGROUND: Guidelines recommend routine preoperative frailty assessment for older people. However, the degree to which frailty instruments improve predictive accuracy when added to traditional risk factors is poorly described. Our objective was to measure the accuracy gained in predicting outcomes important to older patients when adding the Clinical Frailty Scale (CFS), Fried Phenotype (FP), or Frailty Index (FI) to traditional risk factors. METHODS: This was an analysis of a multicenter prospective cohort of elective noncardiac surgery patients ≥65 years of age. Each frailty instrument was prospectively collected. The added predictive performance of each frailty instrument beyond the baseline model (age, sex, American Society of Anesthesiologists' score, procedural risk) was estimated using likelihood ratio test, discrimination, calibration, explained variance, and reclassification. Outcomes analyzed included death or new disability, prolonged length of stay (LoS, >75th percentile), and adverse discharge (death or non-home discharge). RESULTS: We included 645 participants (mean age, 74 [standard deviation, 6]); 72 (11.2%) participants died or experienced a new disability, 164 (25.4%) had prolonged LoS, and 60 (9.2%) had adverse discharge. Compared to the baseline model predicting death or new disability (area under the curve [AUC], 0.67; R, 0.08, good calibration), prolonged LoS (AUC, 0.73; R, 0.18, good calibration), and adverse discharge (AUC, 0.78; R, 0.16, poor calibration), the CFS improved fit per the likelihood ratio test (P < .02 for death or new disability, <.001 for LoS, <.001 for discharge), discrimination (AUC = 0.71 for death or new disability, 0.76 for LoS, 0.82 for discharge), calibration (good for death or new disability, LoS, and discharge), explained variance (R = 0.11 for death or new disability, 0.22 for LoS, 0.25 for discharge), and reclassification (appropriate directional reclassification) for all outcomes. The FP improved discrimination and R for all outcomes, but to a lesser degree than the CFS. The FI improved discrimination for death or new disability and R for all outcomes, but to a lesser degree than the CFS and the FP. These results were consistent in internal validation. CONCLUSIONS: Frailty instruments provide meaningful increases in accuracy when predicting postoperative outcomes for older people. Compared to the FP and FI, the CFS appears to improve all measures of predictive performance to the greatest extent and across outcomes. Combined with previous research demonstrating that the CFS is easy to use and requires less time than the FP, clinicians should consider its use in preoperative practice. SN - 1526-7598 UR - https://www.unboundmedicine.com/medline/citation/31569165/Prospective_Comparison_of_Preoperative_Predictive_Performance_Between_3_Leading_Frailty_Instruments_ L2 - https://doi.org/10.1213/ANE.0000000000004475 DB - PRIME DP - Unbound Medicine ER -