Validation of a PHIS Esophageal Atresia and Tracheoesophageal Fistula Cohort in ICD-10.

ABSTRACT

OBJECTIVE: Validation of a contemporary International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) congenital esophageal atresia/tracheoesophageal atresia (EA/TEF) cohort in the Pediatric Health Information System (PHIS) database. STUDY DESIGN: Database study, validation. SETTING: Tertiary care center. METHODS: Search methods used to validate an ICD-9-CM EA/TEF cohort in PHIS were modified for ICD-10-CM. A retrospectively and prospectively maintained clinical database at a single high-volume EA/TEF center was used for comparison. Patients treated between October 1, 2015 and July 31, 2022 were included. Searches progressively narrowed the cohort by ICD-10-CM diagnosis codes, expansion to include incorrectly coded as 'iatrogenic, age less than 30 days, and use of at least 1 ICD-10-CM procedure code. Results of PHIS data and institution data were compared for accuracy. RESULTS: The most refined search of PHIS and the EA/TEF clinical database yielded 93 and 84 patients, respectively. The sensitivity was 99% and positive predictive value was 94%. A PHIS search using these methods and encompassing 49 children's hospitals yielded an EA/TEF cohort of 2479 patients. CONCLUSION: We present a validated search method in the PHIS database to identify a high-fidelity cohort of EA/TEF patients for multi-institutional study. We have demonstrated that a carefully maintained clinical database may be used to validate cohorts in PHIS. This cohort allows for improved practice variability and outcomes study of EA/TEF patients. Similar methods may be employed to generate other rare disease cohorts in PHIS. LEVEL OF EVIDENCE Level 4.