Accuracy of venous thromboembolism ICD-10 codes: A systematic review and meta-analysis
Bonnie Liu, Milena Hadzi‐Tosev, Kerolos Eisa, Yang Liu, Kayla J. Lucier, Anchit Garg, Sophie Li, Emily Xu, Siraj Mithoowani, Rick Ikesaka, Nancy M. Heddle, Bram Rochwerg, Shuoyan Ning
Abstract
The identification of venous thromboembolism (VTE) using administrative databases is frequently required for reporting and research. The accuracy of International Classification of Diseases 10th revision (ICD-10) codes for VTE, including deep vein thrombosis (DVT) and pulmonary embolism (PE), remains unclear. We examined the accuracy of ICD-10 codes for identifying VTE in adult and pediatric inpatients and outpatients. For this systematic review and meta-analysis, we searched MEDLINE, EMBASE, Web of Science, CENTRAL, Epistemonikos and McMaster Superfilters from inception to July 25, 2023 for studies evaluating the sensitivity, specificity, positive predictive value (PPV), and/or negative predictive value (NPV) of ICD-10 codes for VTE in any anatomical location. We assessed risk of bias using QUADAS and certainty of evidence using GRADE. We calculated pooled sensitivity and specificity with 95% confidence intervals (CI) using a random-effects model. We included 24 studies in the qualitative synthesis and 7 in the meta-analysis. Pooled sensitivity for any VTE based on ICD-10 codes was 72% (95% CI 60–85%, low certainty); pooled specificity was 82% (95% CI 76–88%, low certainty). The PPV for ICD-10 VTE codes ranged from 0% to 100% (median: 80%) while the NPV ranged from 95.4% to 100% (median: 100%). ICD-10 codes for PE had a higher pooled sensitivity (91%) than for DVT (58%). ICD-10 codes have moderate-to-high sensitivity and specificity for the identification of VTE in electronic databases. The certainty of evidence is low due to inconsistency and risk of bias. Further robust studies validating ICD-10 VTE codes are needed to improve reporting and better understand coding limitations.