Impact of Social Determinants of Health on Predictive Models for Outcomes After Congenital Heart Surgery

Days Alive Out of Health Care: A Novel Measure of Health Status After Congenital Heart Surgery

BACKGROUND

As pediatric and adult operative mortality improves, longitudinal outcomes beyond survival become increasingly important. "Days alive and out of hospital" is well-studied as a longitudinal, patient-centered outcome in adult populations. However, it has yet to be validated for children, whose health care use and survival differ from that of adults.

OBJECTIVES

This study sought to evaluate days alive and out of health care/hospital as outcomes for neonatal and infant cardiac surgery.

METHODS

Using linked, locally held, clinical registry data, the National Death Index, and Medicaid claims from the CHS-COLOUR (New York Congenital Heart Surgery Collaborative for Longitudinal Outcomes and Utilization of Resources), we describe and compare the distribution of days alive and out of health care/hospital (730 days or date of death minus date of birth, minus the sum of days in health care) over 2 years for New York State Medicaid-enrolled children who underwent initial neonatal or infant cardiac surgery during 2006 to 2018. We assess relative contributions of mortality, inpatient, and outpatient days, test impacts of alternate outpatient day weightings, and assess associations with patient characteristics known to be associated with other, established congenital heart surgical outcomes.

RESULTS

In total, 2,519 Medicaid-enrolled neonates/infants underwent initial cardiac surgery in the first year of life across New York state during 2006 to 2018. Days alive and out of health care (DAoH) (0-730 days) was bimodally distributed, with 10.4% (n = 261) spending <100 days and 21.6% (n = 545) spending >700 DAoH. We found that 5.9% (n = 149) of children died during their birth admission (0 DAoH). Cardiac disease complexity, weight at surgery, noncardiac preoperative comorbidities, and other preoperative risk factors were all significantly associated with the outcomes (P <0.001). When considering only inpatient days, the distribution was more left-skewed, because 40% of health care days in the second year of life was attributable to outpatient days. Our findings were otherwise largely insensitive to health care type weightings.

CONCLUSIONS

In this data set of children with Medicaid coverage, total DAoH is an easily reproducible, patient-centered outcome that should be considered for assessment of longitudinal outcomes for neonates and infants undergoing cardiac surgery. Where outpatient data are not available, days alive and out of hospital might serve as a reasonable alternative.