The role of prognostic score systems in biventricular repair


  • #CH/NEW 01-EP-9
  • Congenital Heart Surgery/Newborn Critical Congenital Cardiac. E-POSTER (ORAL) SESSION
  • E-Poster (oral)

The role of prognostic score systems in biventricular repair

Sergei V. Kalashnikov, David O. Berishvili

A.N.Bakulev NSPCCS, Moscow, Russia

Date, time and location: 2018.05.25 13:30, Exhibition area, 1st Floor. Zone – B

Abstract

Objectives: To investigate sensitivity, specificity and accuracy of the most known prognostic score systems in biventricular repair of different congenital heart diseases combined with left heart structures hypoplasia.

Methods: Retrospective analysis of 56 patients with different congenital heart diseases combined with left heart structures hypoplasia treated in the A.N. Bakulev NMSC within the period from 2003 to 2015. All patients were divided into two groups: 1) patients who successfully underwent biventricular repair (n=48); 2) patients who died after operation (n=8). Median age at hospitalization: 35,9 ± 55,4 days (from 1 day to 7,5 months). Forty patients (71,4%) were operated in the newborn period. Median weight was 3,8 ± 1,1 kg (from 2,17 to 7,6 kg). Mitral valve z-score at hospitalization in general group was-3.10± 0.79 (from -0.84 to -4.70). Aortic valve z-score-3.54± 1.22 (ranged from -0.86 to -7.04). Left ventricular outflow tract z-score was-3.21± 1.04 (from -1.15 to -6.45). End-diastolic left ventricular volume was19.53± 3.98 (from 12 to 35) ml/m2. The following prognostic models were analyzed for each patient: Rhodes score, CHSS-1, Discriminant score, CHSS-2, 2V-score.

Results: Rhodes score: sensitivity = 0,02; specificity=1; accuracy=0,16. CHSS-1: sensitivity=0,55; specificity=0,57; accuracy=0,55. Discriminant score: sensitivity=0,23; specificity=0,87; accuracy=0,32. CHSS-2: sensitivity=0,94; specificity=0,62; accuracy=0,89. 2V-score: sensitivity=0,67; specificity=0,38; accuracy=0,62. Early hospital mortality was 14,3% (8 patients). Median observation period after biventricular repair was 53,3 ± 22,6 months. One year postoperative survival was 83,5% [74,3 - 92,7 in 95% confidence interval]; three years postoperative survival was 79,4% [68,6 – 90,2 in 95% confidence interval] and five years survival was 70,9% [57,4% – 84,4% in 95% confidence interval].

Conclusions: Advantages, limitations and specificity of the most known prognostic score models were revealed in our study. CHSS-2, 2V-score are the most universal and accurate models for all ranges of congenital heart diseases combined with left heart structures hypoplasia.


To top