New risk assessment model shows differences in care after pediatric CV surgery

With the use of a novel risk-adjustment method, researchers found variation in cardiac ICU performance after pediatric CV surgery, with no associations between volume and case mix-adjusted survival, according to findings presented at Cardiology 2017.

Michael Gaies, MD, MPH, assistant professor of pediatrics at C.S. Mott Children’s Hospital in Ann Arbor, Michigan, and colleagues evaluated quality of cardiac ICU care at hospitals within the Pediatric Cardiac Critical Care Consortium (PC4). They developed a mortality risk-adjustment model to measure the quality of care after CV surgery.

In their mortality risk model, the researchers included risk factors and variables at P < .1 from the first two cardiac ICU postoperative hours that were associated with mortality. Bootstrap resampling (1,000 samples) was used to determine 95% CI and C statistics. A standardized mortality ratio (SMR) was calculated for each hospital; and outliers were considered any SMR with a 95% CI that did not include 1.

Gaies and colleagues analyzed data from 8,543 cardiac ICU admissions across 23 centers. They found the significant risk factors included: age at surgery preterm neonate (OR = 4.62; 95% CI, 2.2-9.8), term neonate (OR = 2.5; 95% CI, 1.3-4.6), any chromosomal abnormality ( OR = 1.58; 95% CI, 1.1-2.3), more than two previous cardiac surgeries (OR = 3.05; 95% CI, 1.7-5.5), any Society of Thoracic Surgeons preoperative risk factor (OR = 2.13; 95% CI, 1.5-3), preoperative mechanical ventilation (OR = 2.49; 95% CI, 1.8-3.5), STS-European Association for Cardio-Thoracic Surgery Congenital Heart Surgery Mortality Category (OR = 1.5; 95% CI, 1.3-1.8), mechanical ventilation at 2 hours after the procedure in the cardiac ICU (OR = 4.57; 95% CI, 1.6-13), maximum vasoactive inotropic score during the first 2 hours after the procedure (OR = 1.02; 95% CI, 1.01-1.03), and on extracorporeal membrane oxygenation during the first postoperative hour (OR = 15.88; 95% CI, 9.8-25.8).

According to the researchers, the model demonstrated good discrimination (C statistic = 0.92) and calibration. Across the PC4 hospitals, the SMR was 0.4 to 1.9, with four hospitals being statistical outliers (two with better-than-expected mortality and two with worse-than-expected mortality).

However, there were no apparent associations with cardiac ICU volume and mortality ratio.

The researchers concluded that the risk-adjustment method was effective for comparative analyses of cardiac ICU quality of care. – by Cassie Homer

Reference:

Gaies M, et al. Abstract 226. Presented at: Cardiology: the 20th Annual Update on Pediatric and Congenital Cardiovascular Disease; Feb. 22-26, 2017; Orlando, Fla.

Disclosure: Gaies reports no relevant financial disclosures.