Agreement of echocardiographic and catheterization-based methods of transpulmonary pressure gradient measurement in dogs

Introduction/Objectives

Doppler echocardiographic estimation of transpulmonary pressure gradient (PG) is widely used to determine severity of pulmonary valve stenosis and indication for transcatheter intervention. The objective of this study was to describe agreement between Doppler echocardiographic methods of transpulmonary PG estimation and direct peak-to-peak pressure gradient at catheterization (PG_cath) in dogs. We hypothesized that with reference to PG_cath, mean echocardiographic PG (PG_echo-mean) would have less bias than peak modal instantaneous echocardiographic PG (PG_echo-peak).

Animals

Client-owned dogs with congenital pulmonary valve stenosis that underwent balloon pulmonary valvuloplasty at one of two veterinary teaching hospitals between June 2012 and May 2022 were included in this study.

Materials and Methods

Cases that underwent transthoracic echocardiography and subsequent balloon pulmonary valvuloplasty separated by 30 or fewer days were retrospectively identified. For each echocardiogram, average PG_echo-mean (mmHg) and average PG_echo-peak (mmHg) were calculated from stored spectral Doppler recordings. Peak right ventricular-to-peak pulmonary artery PG data (mmHg) were obtained from catheterization reports. Bland–Altman analysis was used to assess agreement between echocardiographic and catheterization data.

Results

Data from 209 dogs (n = 215 instances) were evaluated. Proportional bias, greater at higher gradients, was observed for PG_echo-mean versus PG_cath (P<0.001). A constant bias of −38.12 mmHg was observed for PG_echo-peak versus PG_cath (P=0.62). Bias for both echocardiographic variables had wide limits of agreement that increased with PG.

Conclusions

PG_echo-mean and PG_echo-peak underestimated and overestimated PG_cath, respectively, preventing their interchangeability with PG_cath.