Association of the child opportunity index with in-hospital mortality and persistence of organ dysfunction at one week after onset of Phoenix Sepsis among children admitted to the pediatric intensive care unit with suspected infection

Abstract

The social determinants of health (SDoH) are fundamental factors that contribute to overall health and health-related outcomes. Children living in lower socioeconomic areas have a higher risk of critical illness and worse outcomes compared to children living in more socioeconomically advantaged areas. In this work, we determine whether the Child Opportunity Index (COI 3.0), a multi-dimensional child-specific indicator of neighborhood environment, is associated with in-hospital mortality or persistence of a Phoenix Sepsis Score ≥ 2 at one week following Phoenix Sepsis onset in children admitted to pediatric intensive care units (PICUs) with suspected infection. We performed a retrospective cohort analysis of 63,824 patients with suspected or confirmed infection admission diagnosis in two PICUs in Atlanta, Georgia with a Georgia residential address that could be geocoded and linked to a census tract. The primary outcome was the composite of in-hospital mortality or persistence of a Phoenix Sepsis Score ≥ 2 at one week following Phoenix Sepsis onset. Model performance measures of interest were the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). Models developed with electronic medical record (EMR) data using Egleston (EG) or Scottish Rite (SR) as the training site achieved AUROCs of 0.81-0.84 (95% CI range: 0.8-0.85) and 0.82-0.82 (95% CI range: 0.81-0.83) and AUPRCs of 0.59-0.68 (95% CI range: 0.58-0.69) and 0.62-0.64 (95% CI range: 0.61-0.65) respectively. Despite significant differences in COI 3.0 characteristics and overall in-hospital mortality of children with Phoenix suspected infection between the EG and SR PICUs, the addition of COI 3.0 did not improve the overall model performance metrics. While children admitted to both PICUs were more often from COI 3.0 neighborhoods in the lowest two quintiles, these neighborhood features had less of an impact on the model's predictive performance compared to patient physiologic and biologic features available in the EMR.

Fig 1. Flow diagram of the study cohort. The GEOID is the 11-digit geographical identification number for a given census tract. The COI 3.0 is the third iteration of the Child Opportunity Index.

Fig 2. The distribution of the Phoenix Sepsis Score in children with suspected infection and geocodable Georgia residential addresses and composite primary outcome by pediatric intensive care unit (PICU) site. (A) The distribution of children with suspected infection (n = 14,761) by Phoenix Sepsis Score admitted to either the Children’s Healthcare of Atlanta Egleston (EG) PICU (blue bars) or the Scottish Rite (SR) PICU (orange bars). (B) The distribution of the composite primary outcome (in-hospital mortality or persistence of a Phoenix Sepsis Score of  ≥  2 one week following Phoenix sepsis onset) by PICU site.

Fig 3. SHAP beeswarm plot of the top 20 features for the eXtreme gradient boosting (XGB) models predicting the composite primary outcome of in-hospital mortality or persistence of a Phoenix sepsis score  ≥  2 one week following Phoenix sepsis onset. The SHAP summary plot explaining the relative importance of features for the models trained with Egleston (EG) data (A) with electronic medical record (EMR) data or with (B) EMR and Child Opportunity Index (COI) data and for the models trained with Scottish Rite (SR) data with (C) EMR data or with (D) EMR and COI data.