Maro responds to “raising a high-pressure alarm about pediatric hypertension”

We thank Dr. Weiss1 for his thoughtful engagement with our article2 and particularly how to interpret the findings in the absence of additional measurement of persistent pediatric hypertension (ie, when children with initial findings of clinical hypertension or elevated blood pressure do not experience a subsequent return to normal blood pressure without antihypertensive treatment). Dr. Weiss acknowledges the well-documented underrecognition and underdiagnosis of pediatric hypertension3,-6 despite clinical practice guidelines7 that were streamlined and simplified to improve recognition. Our study2 adds to this literature but finds a clear difference in estimates between children who met the clinical requirements for pediatric hypertension based on blood pressure readings and those who received a coded diagnosis of hypertension in their electronic medical record. Additionally, those with diagnosed hypertension were uniformly less well, suggesting a diagnosis is only conferred at more advanced stages of hypertension. Why is this mismatch so important? Early identification of pediatric hypertension allows earlier intervention strategies for a disease course when much is gained by prevention. While data collected primarily for research may be ideal to design and test new strategies for potential interventions, these types of observational cohorts are few and far between. Real-world data (ie, data collected in routine care, not primarily intended for research) are relatively inexpensive to acquire and can accelerate our understanding of specific clinical areas. However, these data must be fit-for-purpose, and our comparison of clinical and billing data demonstrates meaningful gaps where the billing data do not reflect the clinical phenomenon. Dr. Weiss acknowledges similar findings in pediatric sepsis,8 and other authors have similar findings in adult sepsis9 and obesity.10 Imagine we were to attempt to design an early-stage intervention to reduce pediatric hypertension and we selected a pediatric cohort that met clinical requirements based on blood pressure readings or we selected a pediatric cohort with diagnosis codes. We would be talking about 2 pediatric populations with markedly different severity, but without studies such as ours,2 we might be unaware of the limitations in generalizability. Dr. Weiss suggests that the prevalence we estimated with clinical blood pressure readings is concerningly high and implies misclassification. We acknowledge that our study was likely to yield higher estimates compared with prior work11 based on our inability to limit to ambulatory visits. Moreover, Dr. Weiss is concerned with our treatment of multiple blood pressures recorded on the same day, which follows clinical guidelines that require the average blood pressure to be recorded.7 Dr. Weiss would have preferred that a sensitivity analysis with the lowest reading was included, and we agree that such a sensitivity analysis would have helped dispel concerns. However, clinical guidelines also require 3 distinct measure-days to declare any child hypertensive, guarding against the possibility of incorrect classification from temporary elevations either on the same day or among multiple visits. We do not believe that the magnitude of such potential misclassification changes our core findings: that clinical data yielded higher estimates of pediatric hypertension that rarely translated into administrative diagnoses of hypertension. Real-world data are not without limits. As Dr. Weiss points out, they cannot be used to evaluate measurement issues in pediatric hypertension that arise from using the incorrect blood pressure instrument or using the incorrect limb. However, they allow investigators to understand more than they would otherwise know, giving us opportunities to improve public health. We should use them with eyes wide open for both their strengths and limitations. This work was supported by the Food and Drug Administration through the Department of Health and Human Services (contract number HHSF223201400030I). None to declared.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.