The Agreement Between Transthoracic and Transesophageal Echocardiography in the Assessment of Right Ventricular Diastolic Dysfunction Grades in Adult Patients Undergoing Cardiac Surgery: A Prospective Observational Study

Introduction The importance of right ventricular (RV) diastolic function in cardiac surgery cannot be overstated, as it significantly affects prognosis and long-term outcomes. Conventionally, RV diastolic dysfunction (RVDD) is assessed and graded using criteria from either the American Society of Echocardiography (ASE) or the British Society of Echocardiography (BSE), with measurements done by transthoracic echocardiography (TTE). However, during cardiac surgery, perioperative echocardiographic evaluation is done predominantly by transesophageal echocardiography (TEE). This study aimed to assess the agreement between TTE and TEE in grading RVDD using both ASE and BSE criteria. Methods Key two-dimensional (2D) and Doppler parameters were measured in 81 patients undergoing cardiac surgery by both TTE and TEE after anesthesia induction within 10 minutes of each other, under similar hemodynamic, anesthetic, and ventilatory conditions. RVDD gradings were done separately by TTE and TEE with both ASE and BSE criteria using the measured values of the key parameters by TTE and TEE, respectively. RVDD gradings derived from TTE were compared with those derived from TEE. The tricuspid inflow Doppler and tricuspid annular tissue Doppler parameters were measured in TEE in both mid-esophageal RV inflow-outflow (MERVIO) and deep transgastric RV inflow-outflow (DTGRVIO) views. Gradings were done separately for both views of TEE by using the Doppler values measured in the respective views (TEE-MERVIO and TEE-DTGRVIO). The TTE-derived RVDD grades were compared with those derived by both TEE-MERVIO and TEE-DTGRVIO. Weighted κ values were used to assess observed agreement beyond chance. Inter-rater reliability of the RVDD grades derived by both TTE and TEE (both views) was also checked. Individual 2D and Doppler parameters were compared between TTE and TEE in terms of Bland-Altman limits of agreement. Results As per ASE criteria, disagreement of RVDD by ≥1 grade was seen in 43 (53.1%) patients and by 2 grades in eight (9%) patients when comparing TTE and TEE-MERVIO, yielding a weighted κ of 0.14 (p=0.123). Disagreement by ≥1 grade was observed in 32 (39.5%) patients and by 2 grades in 10 (12.3%) patients when comparing TTE and TEE-DTGRVIO, yielding a weighted κ of 0.3 (p=0.002). Using the BSE Criteria, disagreement of RVDD grades occurred in nine (11.1%) patients when comparing TTE and TEE-MERVIO, yielding an unweighted κ of 0.25 (p=0.295). Disagreement occurred in 12 (14.8%) patients when comparing TTE and TEE-DTGRVIO, yielding an unweighted κ of 0.260 (p=0.187). There was almost perfect agreement between independent raters regarding both TTE- and TEE-derived RVDD grades per the ASE criteria, and substantial to almost perfect agreement per BSE criteria. Bland-Altman analysis of paired data between the TTE- and TEE-measured values of individual 2D and Doppler parameters showed wide limits of agreement. Conclusions This study revealed, at best, only fair agreement between TTE and TEE in grading RVDD. The measured 2D and Doppler echocardiographic parameters showed wide limits of agreement between TTE and TEE. We recommend further research to develop a TEE-based algorithm for grading RVDD, and to evaluate the prognostic effectiveness of perioperative TEE for predicting adverse clinical outcomes associated with RVDD.