Outcome prediction in gastroschisis - The gastroschisis prognostic score (GPS) revisited.

ABSTRACT

PURPOSE: The GPS enables risk stratification for gastroschisis and helps discriminate low from high morbidity groups. The purpose of this study was to revalidate GPS's characterization of a high morbidity group and to quantify relationships between the GPS and outcomes. METHODS: With REB approval, complete survivor data from a national gastroschisis registry was collected. GPS bowel injury scoring was revalidated excluding the initial inception/validation cohorts (>2011). Length of stay (LOS), 1st enteral feed days (dFPO), TPN days (dTPN), and aggregate complications (COMP) were compared between low and high morbidity risk groups. Mathematical relationships between outcomes and integer increases in GPS were explored using the entire cohort (2005-present). RESULTS: Median (range) LOS, dPO, and dTPN for the entire cohort (n=849) was 36 (26,62), 13 (9,18), and 27 (20,46) days, respectively. High-risk patients (GPS≥2; n=80) experienced significantly worse outcomes than low risk patients (n=263). Each integer increase in GPS was associated with increases in LOS and dTPN by 16.9 and 12.7days, respectively (p<0.01). COMP rate was also increased in the high-risk cohort (46.3% vs. 22.8%; p<0.01). CONCLUSION: The GPS effectively discriminates low from high morbidity risk groups. Within the high risk group, integer increases in GPS produce quantitatively differentiated outcomes which may guide initial counseling and resource allocation. LEVEL OF EVIDENCE IIb.