OBJECTIVES: The Neonatal Intensive Care Unit (NICU) is a site of care where severe outcomes, high resource use, and workflow complexity are the norm. Obtaining a rapid etiological diagnosis for infants with early-onset rare diseases remains a major challenge and the genetic causes of critical illness are heterogeneous. These diseases often have a severe presentation and unknown prognosis. Emerging research in the field of NICU productivity and workflow is beginning to define (1) healthcare resource use; (2) data elements required for workflow mapping; and (3) how genomic information can improve outcomes and reduce costs. Research is required to estimate the cost of genetic disease in the NICU setting.

METHODS: To estimate clinical and economic outcomes in the NICU, we utilized the 2019 version of the Kids' Inpatient Database (KID), developed by AHRQ. We used Major Diagnostic Category 15 to select all live births, excluded patients with DRG=795 (normal newborn) and included confirmed transfers into the NICU to form the study cohort. Pediatric genetic disease (PGD) was identified using 348 ICD-10-CM codes corresponding to 283 pediatric genetic diseases (reviewed by a medical geneticist), for which the carrier frequency is greater than 1 in 500 in the general population or in an at-risk subpopulation. Cases were then sorted by highest cost and highest mortality.

RESULTS: The top five genetic diseases were identified: Progressive Familial Intrahepatic Cholestasis (Type 2), Corticosterone Methyloxidase Deficiency, Hypophosphatasia, Congenital Amegakaryocytic Thrombocytopenia (MPL), and Combined SAP Deficiency. Compared to controls, defined as NICU patients without genetic diseases, these conditions were associated with significantly higher (5.7-11.4 fold) costs, higher mortality rate (3.1-8.3 fold) and longer hospital stay (2.7-4.4 fold).

CONCLUSIONS: For NICUs that do not routinely incorporate exome or whole genome testing into their workflow, these conditions highlight the need to consider genetic testing when the underlying etiology is unclear.