Usefulness of PREDIC³T Case Type Risk Category in the CRISP Registry

Procedural risk in Congenital Cardiac Catheterization (PREDIC³T) was recently reported as the contemporary procedure-type risk metric by the Congenital Cardiac Catheterization Project on Outcomes (C3PO) registry. The usefulness of this metric has not been evaluated elsewhere. The CRISP registry of Congenital Cardiovascular Interventional Study Consortium (CCISC) data set was analyzed. The study period was 14 years (2009 to 2022). The primary outcome was significant adverse event (SAE). Cases were assigned to the 6 PREDIC³T risk categories. Univariate and multivariable logistic regression models were used to evaluate the association between PREDIC³T and the primary outcome. The model discriminative performance was evaluated by the c-statistic. In a total of 64,419 enrolled cases, PREDIC³T case types were assigned in 59,822 cases (93%). The frequency for PREDIC³T category was 0 = 7,494 (12.5%), 1 = 16,932 (28.3%), 2 = 17,023 (28.5%), 3 = 9,885 (16.5%), 4 = 4,403 (7.4%), and 5 = 4,085 (6.8%). SAE was observed in 2,474 cases (4.1%). The SAE rates for category were 0 = 1.0%, 1 = 2.3%, 2 = 4.0%, 3 = 6.2%, 4 = 8.2%, and 5 = 9.0%. In a multivariable model, PREDIC³T case type risk category (odds ratios for category: 0 = 0.49, 1 = 1.00, 2 = 1.40, 3 = 2.06, 4 = 2.79, and 5 = 3.15; p <0.001) were significantly associated with SAE (c-statistic of 0.707) after adjusting for age, preprocedural inotropic support and systemic illness, low systemic saturation, high pulmonary vascular resistance, and the use of general anesthesia. The PREDIC³T case type risk category was associated with the risk of SAE in the CRISP registry data set and appeared to be a useful procedural risk classification tool.