Abstract
Primary Subject area
Neonatal-Perinatal Medicine
Background
Therapeutic hypothermia (TH) is the standard treatment for neonatal hypoxic ischemic encephalopathy (HIE) to improve mortality and long-term impairment. Accurate costing algorithms are essential to evaluate cost-effective interventions and identify cost drivers.
Objectives
We aimed to validate the Canadian Neonatal Network (CNN) costing algorithm for HIE infants treated with TH against costs obtained from hospital-based finance software (CPSS) and compare the costs of TH for infants with mild/moderate to those with severe HIE. We aimed to validate the Canadian Neonatal Network (CNN) costing algorithm for HIE infants treated with TH against costs obtained from hospital-based finance software (CPSS) and compare the costs of TH for infants with mild/moderate to those with severe HIE.
Design/Methods
Retrospective cohort study including 98 infants admitted with HIE and receiving TH in a tertiary NICU between 2016 and 2018. Clinical characteristics and CNN costing data were collected from the local CNN database and actual cost were obtained from CPSS. The primary outcome was the difference in total hospital stay cost between CNN algorithm and CPSS. The differences between both algorithms were also identified in 8 different cost centres such as nursing, respiratory, imaging, etc. Costs per patients using both algorithms were compared using Pearson correlation coefficient (r) and paired t-test. Characteristics and costs per infant were compared between infants with mild/moderate HIE and those with severe HIE.
Results
Among the 98 patients with HIE that received TH, 2 (2%) had mild HIE, 75 (77%) had moderate HIE and 21 (21%) had severe HIE on admission. Mortality rate was 10% (10/98) and median length of stay was 12 days [IQR 10-16]. Total mean cost per infant using the CNN algorithm was $32,727 (SD $23,751 and correlated highly to the CPSS mean $28.373(SD $28.989) (r=0.93, p<0.01). There was no significant difference in mean total costs estimated between the algorithms ($1051, 95% CI $-1073, $3174). There was a strong correlation between cost estimates using the CNN algorithm and CPSS in nursing, physician, transfusion and indirect costs (r range 0.94-0.99) (Figure 1). Mean daily costs per infant with mild/moderate HIE ($1579, SD 808) were lower compared to infants with severe HIE ($2069, SD 1518). In both groups, daily costs were higher in the first days of hospitalization and slightly decreased over time (Figure 2).
Conclusion
The CNN algorithm accurately predicts hospital stay costs for infants diagnosed with HIE and received TH in our centre. Severity of encephalopathy and severity of illness are associated with higher hospital costs.
