290 - Subspecialist Congruency in Coding Blood Pressure Severity at the Initial Hypertension Clinic Visit

Poster Number: 290
Publication Number: 290.1399

Background: In addition to use in billing, ICD-10 codes are employed within learning health systems for population health management, identification of participants for research studies, and data for research. Agreement of ICD-10 code billing with blood pressure (BP) severity by subspecialists seeing youth for hypertension (HTN) is not known, especially around time of diagnosis when patients may be seen at different time points following the recognition of a HTN disorder.

Objective: Determine the congruency of HTN ICD-10 codes with measured abnormal BP at the index subspecialty clinic visit for HTN, and the agreement between nephrology and cardiology.

Design/Methods: Cross-sectional analysis of baseline EHR data from youth referred for HTN disorders in the multisite retrospective SUPERHERO Registry. Inclusion criteria were age < 19 years and initial subspecialty clinic visit for HTN disorders by ICD-10 codes from 1/1/2015–12/31/2022. Exclusion criteria were dialysis, kidney transplantation, or pregnancy by ICD-10 codes. We estimated agreement between HTN or elevated BP by ICD-10 code (ICD-HTN, ICD-elevated) with BP severity measured in clinic (Clinic-elevated, Clinic-HTN), using McNemar’s test for agreement and kappa co-efficient. Clinic-elevated and HTN were categorized using national guidelines. We stratified our population by cardiology vs nephrology clinic visit, used Breslow-Day test for homogeneity for comparison of odds ratios, and calculated sensitivity, specificity, and positive and negative predictive values.

Results: Of 10968 participants, 38% were female and average age was 12.5 years (SD 4.7). Among the 8678 participants with an abnormal clinic-measured BP, there was low-moderate agreement between ICD-HTN and Clinic-HTN with congruency in 39% of cases (Table 1a). Overall, ICD-HTN had 53% specificity and 78% positive predictive value to predict Clinic-HTN and 51% sensitivity and 25% negative predictive value. ICD-HTN in cardiology clinic had higher positive predictive value to predict Clinic-HTN (Table 1b, c).

Conclusion(s): ICD-10 code for HTN captured only some of participants with clinic measured HTN in youth at index visit to a subspecialist for HTN disorders and participants were more likely to receive a diagnosis of HTN for BP in HTN range in cardiology clinic. Our findings could reflect early referral to subspecialty care or variance in available prior BPs at index visit limiting HTN diagnosis by guidelines. Future steps include incorporating medication use and longitudinal data to understand the experience of youth with HTN disorders in pathway of care at tertiary centers for management of BP.