Validity of International Classification of Diseases (ICD)-10 Diagnosis Codes for Identification of Acute Heart Failure Hospitalization and Heart Failure with Reduced Versus Preserved Ejection Fraction in a National Medicare Sample. Issue 2 (23rd January 2023)

• Record Type: Journal Article
• Authors: Bates, Benjamin A.; Akhabue, Ehimare; Nahass, Meghan M.; Mukherjee, Abhigyan; Hiltner, Emily; Rock, Joanna; Wilton, Brandon; Mittal, Garima; Visaria, Aayush; Rua, Melanie; Gandhi, Poonam; Dave, Chintan V.; Setoguchi, Soko
• Is Part Of: Circulation. Volume 16:Issue 2(2023)
• Page Start: e009078
• Publication Date: 2023-01-23
• Subjects: heart failure -- hospitalization -- heart failure, diastolic -- heart failure, systolic -- Medicare -- predictive value of tests -- validation study; Cardiovascular system -- Diseases -- Treatment -- Periodicals; Cardiovascular system -- Diseases -- Research -- Periodicals; Outcome assessment (Medical care) -- Periodicals; Evidence-based medicine -- Periodicals; 616.1007
• DOI: 10.1161/CIRCOUTCOMES.122.009078
• Languages: English
• ISSNs: 1941-7713

Abstract

Abstract : Background: Heart failure (HF) is a leading cause of hospitalization in older adults. Medicare data have been used to assess HF outcomes. However, the validity of ICD-10 diagnosis codes (used since 2015) to identify acute HF hospitalization or distinguish reduced (heart failure with reduced ejection fraction) versus preserved ejection fraction (HFpEF) is unknown in Medicare data. Methods: Using Medicare data (2015–2017), we randomly sampled 200 HF hospitalizations with ICD-10 diagnosis codes for HF in the first/second claim position in a 1:1:2 ratio for systolic HF (I50.2), diastolic HF (I50.3), and other HF (I50.X). The primary gold standards included recorded HF diagnosis by a treating physician for HF hospitalization, ejection fraction (EF)≤50 for heart failure with reduced ejection fraction, and EF>50 for HFpEF. If the quantitative EF was not present, then qualitative descriptions of EF were used for heart failure with reduced ejection fraction/HFpEF gold standards. Multiple secondary gold standards were also tested. Gold standard data were extracted from medical records using standardized forms and adjudicated by cardiology fellows/staff. We calculated positive predictive values with 95% CIs. Results: The 200-chart validation sample included 50 systolic, 50 diastolic, 47 combined dysfunction, and 53 unspecified HF patients. The positive predictive values of acute HF hospitalization was 98% [95% CI, 95–100] for first-position ICD-10 HF diagnosis and 66% [95% CI, 58–74] for first/second-position diagnosis. Quantitative EF was available for ≥80% of patients with systolic, diastolic, or combined dysfunction ICD-10 codes. The positive predictive value of systolic HF codes was 90% [95% CI, 82–98] for EFs≤50% and 72% [95% CI, 60–85] for EFs≤40%. The positive predictive value was 92% [95% CI, 85–100] for HFpEF for EFs>50%. The ICD-10 codes for combined or unspecified HF poorly predicted heart failure with reduced ejection fraction or HFpEF. Conclusions: ICD-10 principal diagnosis identified acute HF hospitalization with a high positive predictive value. Systolic and diastolic ICD-10 diagnoses reliably identified heart failure with reduced ejection fraction and HFpEF when EF 50% was used as the cutoff.