DISCORDANCE BETWEEN ICD-10-CM OBESITY DIAGNOSIS CODES AND EXACT BMI MEASUREMENTS IN ELECTRONIC MEDICAL RECORDS: A NORSTELLALINQ EHR AND CLAIMS VALIDATION STUDY
Author(s)
Spencer Friedman1, ilan behm, MPH2, Eric Mitchell, BS3, Allison Perry, PhD4.
1Norstella, Yardley, PA, USA, 2Norstella, Englewood, CO, USA, 3Norstella, Brooklyn, NY, USA, 4Norstella, New York, NY, USA.
1Norstella, Yardley, PA, USA, 2Norstella, Englewood, CO, USA, 3Norstella, Brooklyn, NY, USA, 4Norstella, New York, NY, USA.
OBJECTIVES: To quantify discordance between ICD-10-CM obesity diagnosis codes and exact EHR-derived BMI measurements and evaluate implications for real-world patient identification and evidence generation.
METHODS: A cross-sectional validation study was conducted using NorstellaLinQ's linked claims and EHR database covering 650+ U.S. health care organizations. Adults (≥18 years) with ≥1 BMI measurement recorded between January 2017 and March 2026 were included. Exact BMI was derived from discrete height and weight fields; maximum recorded BMI was used to capture evidence of obesity across the full observation period. ICD-10-CM obesity coding status was assessed within ±6 months of the maximum BMI measurement. Discordance was defined as coded obesity absent despite BMI ≥30 (undercoding) or an obesity code present with BMI <30 (overcoding). Sensitivity, specificity and PPV were calculated using exact BMI as the reference standard.
RESULTS: Among 18,461,816 patients with ≥1 recorded BMI, 7,445,857 had a BMI ≥30 kg/m² by exact measurement. Against the BMI reference standard, ICD-10 codes demonstrated a sensitivity of 39.5%, specificity of 85.8%, PPV of 65.2%, and NPV of 67.8%. The undercoding rate was 82.0% using condition-specific Z68 codes. In a separate analysis linking BMI-confirmed obese patients to claims records, 25.1% of patients with confirmed BMI ≥30 (1,538,481 of 6,131,913 with available claims linkage) would be missed by ICD-based identification alone. Overcoding occurred in 14.0% of sub-threshold patients, which may partially reflect prior obesity history or treatment-related weight loss rather than true misclassification.
CONCLUSIONS: ICD-10-CM codes demonstrate poor sensitivity for obesity identification, with 82% of BMI-confirmed obese patients lacking a corresponding code and one in four BMI-confirmed patients invisible to ICD-based screening. Reliance on diagnosis codes alone may introduce substantial outcome and cohort misclassification in obesity-related RWE studies, and incorporating exact EHR-derived BMI should be considered when available for HTA evaluations and payer decision-making.
METHODS: A cross-sectional validation study was conducted using NorstellaLinQ's linked claims and EHR database covering 650+ U.S. health care organizations. Adults (≥18 years) with ≥1 BMI measurement recorded between January 2017 and March 2026 were included. Exact BMI was derived from discrete height and weight fields; maximum recorded BMI was used to capture evidence of obesity across the full observation period. ICD-10-CM obesity coding status was assessed within ±6 months of the maximum BMI measurement. Discordance was defined as coded obesity absent despite BMI ≥30 (undercoding) or an obesity code present with BMI <30 (overcoding). Sensitivity, specificity and PPV were calculated using exact BMI as the reference standard.
RESULTS: Among 18,461,816 patients with ≥1 recorded BMI, 7,445,857 had a BMI ≥30 kg/m² by exact measurement. Against the BMI reference standard, ICD-10 codes demonstrated a sensitivity of 39.5%, specificity of 85.8%, PPV of 65.2%, and NPV of 67.8%. The undercoding rate was 82.0% using condition-specific Z68 codes. In a separate analysis linking BMI-confirmed obese patients to claims records, 25.1% of patients with confirmed BMI ≥30 (1,538,481 of 6,131,913 with available claims linkage) would be missed by ICD-based identification alone. Overcoding occurred in 14.0% of sub-threshold patients, which may partially reflect prior obesity history or treatment-related weight loss rather than true misclassification.
CONCLUSIONS: ICD-10-CM codes demonstrate poor sensitivity for obesity identification, with 82% of BMI-confirmed obese patients lacking a corresponding code and one in four BMI-confirmed patients invisible to ICD-based screening. Reliance on diagnosis codes alone may introduce substantial outcome and cohort misclassification in obesity-related RWE studies, and incorporating exact EHR-derived BMI should be considered when available for HTA evaluations and payer decision-making.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
RWD84
Topic
Clinical Outcomes, Economic Evaluation, Real World Data & Information Systems
Topic Subcategory
Data Protection, Integrity, & Quality Assurance, Distributed Data & Research Networks, Health & Insurance Records Systems, Reproducibility & Replicability
Disease
Diabetes/Endocrine/Metabolic Disorders (including obesity), No Additional Disease & Conditions/Specialized Treatment Areas