Developing a Parsimonious Frailty Index for Older, Multimorbid Adults With Heart Failure Using Machine Learning
Section snippets
Methods
The design was a retrospective cohort study of veterans using VHA to assemble the cohort and extract variables. The Supplementary Methods provides further details on the design, methods, and analysis.
The cohort included all patients admitted to any VHA medical center with a principal diagnosis of HF from October 2015 to October 2018. The International Classification of Diseases (ICD) 10th Revision hospital discharge codes were used to identify patients with a principal diagnosis of HF (I09.9,
Results
Our development cohort included 37,431 participants (Supplementary Figure 1). The average age of the participants was 73.4 ± 10.3 years, most were men (98%), and 26,046 were White (69.6%). About 47.6% of population were obese with BMIs >30 and 43.5% had EF ≤40% (Table 1). The mortality rate during the follow-up interval was 63.1% (n = 23,614), with an average time to death of 3.0 ± 2.1 (median 2.6, interquartile range 0.9 to 5.5) years after discharge.
We started with 76 variables (laboratory
Discussion
The present study provides evidence for the predictive validity of a novel pFI for older patients admitted to the hospital with HF. Notably, the pFI is exclusively developed for patients with HF. The pFI is robust because it uses a national sample that integrates baseline outpatient data from the previous 12 months, follows patients after hospitalization for multiple years, and it uses an innovative cascade of machine learning techniques. The pFI only requires 10 variables and can thus improve
Conclusions
A pFI with 10 variables has a similar predictive validity as the VA-FI. The pFI is thus a feasible and practical tool that can be incorporated into routine practice and provides clinically useful insights that are distinct from existing mortality prediction models,38 especially for older adults with multiple co-morbidities. The pFI index applied at the point of care could guide clinical decisions regarding care transitions, urgency of interventions to reduce readmissions, and appropriateness of
Competency in medical knowledge
We propose a pFI with 10 specific deficits derived from full breadth of electronic medical records from outpatient settings using machine learning algorithm. The pFI has better predictive validity than previous approaches, independent of age and EF in predicting mortality and healthcare utilization after hospital discharge for congestive HF.
Translational Outlook
The pFI is a practical clinical decision support tool for day-to-day inpatient use.
Disclosures
The study is funded partially by the Center for Innovations in Quality, Effectiveness and Safety (VA HSRD CIN 13-413), Michael E. DeBakey VA Medical Center, Houston, Texas, USA and CSP 2010A. Dr. Razjouyan is supported by NIH-NHLBI award 1K25HL152006-01. Dr. Horstman is supported by VA HSR&D CDA-2 award IK2-HX003163. Dr. Orkaby is supported by VA CSR&D CDA-2 award IK2-CX001800. Dr. Virani is supported by research grants from The Department of Veterans Affairs Health Services Research and
Acknowledgment
The authors are grateful to the VA Informatics and Computing Infrastructure.
References (38)
The frailty syndrome: definition and natural history
Clin Geriatr Med
(2011)- et al.
Development of a comorbidity index using physician claims data
J Clin Epidemiol
(2000) - et al.
Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases
J Clin Epidemiol
(1992) - et al.
Frailty as a predictor of adverse outcomes in hospitalized older adults: a systematic review and meta-analysis
Ageing Res Rev
(2019) - et al.
Frailty and bleeding in older adults undergoing TAVR or SAVR: insights from the FRAILTY-AVR study
JACC Cardiovasc Interv
(2020) - et al.
Frailty is intertwined with heart failure: mechanisms, prevalence, prognosis, assessment, and management
JACC Heart Fail
(2019) - et al.
Frailty and healthcare utilization among patients with heart failure in the community
JACC Heart Fail
(2013) - et al.
Frailty assessment instruments in heart failure: a systematic review
Eur J Cardiovasc Nurs
(2018) - et al.
2011 update to National Heart Foundation of Australia and Cardiac Society of Australia and New Zealand Guidelines for the prevention, detection and management of chronic heart failure in Australia, 2006
Med J Aust
(2011) - et al.
Management of heart failure
Med J Aust
(2013)
Co-morbidity and potential treatment conflicts in elderly heart failure patients: a retrospective, cross-sectional study of administrative claims data
Drugs Aging
Coexisting frailty with heart failure
Front Physiol
Importance of frailty in patients with cardiovascular disease
Eur Heart J
Diagnostic accuracy of the International Classification of Diseases, Tenth Revision, codes of heart failure in an administrative database
Pharmacoepidemiol Drug Saf
The burden of frailty among U.S. Veterans and its association with mortality, 2002–2012
J Gerontol A Biol Sci Med Sci
Automated extraction of ejection fraction for quality measurement using regular expressions in Unstructured Information Management Architecture (UIMA) for heart failure
J Am Med Inform Assoc
VIReC Factbook Corporate Data Warehouse (CDW) Patient 2.0 Domain
Smoking Status and Factors associated with COVID-19 In-hospital Mortality among US Veterans
Nicotine Tob Res
Cited by (0)
The analysis was partly supported by the use of facilities and resources at the Center for Innovations in Quality, Effectiveness and Safety, CIN 13-413, Michael E. DeBakey VA Medical Center, Houston, Texas and a National Institutes of Health, National Heart, Lung, and Blood Institute, Bethesda, Maryland, K25 funding, number 1K25HL152006-01, to Dr. Razjouyan.