641 - App Algorithm to Identify Neonates Appropriate for Transfer to a Community Neonatal Intensive Care Unit (cNICU)

Poster Number: 641
Publication Number: 641.456

Regine Fortunov, Baylor College of Medicine, Houston, TX, United States; Felicia Wooten, Texas Children's Hospital, Houston, TX, United States

Background: Clinically improving NICU patients may benefit from care closer to home but identifying patients appropriate for transfer is complex. Transfer of infants to a Community Neonatal Intensive Care Unit (cNICU) also increases high acuity NICU bed availability.

Objective: To increase identification and subsequent transfers of clinically improving patients from the Texas Children’s Hospital (TCH) Main Campus NICUs to TCH affiliate cNICUs through a standardized process.

Design/Methods: In 07/2022, criteria for appropriate transfers were identified (Table 1). Also required was an expected remaining hospital stay that accommodated 5 business days for insurance authorization and a minimum 7-day cNICU stay.

These criteria were imbedded in a Structured Query Language (SQL) query from the TCH Enterprise Data Warehouse within a Qlik Sense app.  The app was linked to the cNICU TCH neonatology call schedule and an ArcGIS Map displaying the distance from the patient’s home to the nearest 3 cNICUs, and the cNICUs capabilities and contact numbers.  Patients were identified each business day.  After parent, medical team and insurance approval, the patient was transferred to the cNICU.

Results: Only 1 inborn patient was transferred to a cNICU in 2021 (Figure 1).  Infants transferred to cNICUs increased from 11 / 965 (1.1%) TCH NICU admissions Pre App (01/22-06/22) to 23 / 983 (2.3%) TCH NICU admissions Post App (07/22-12/22) (p = 0.04). No patients required return transfer to TCH.

During 2022, 21 infants were transferred to the TCH Woodlands NICU, a internal system surgical level 3 NICU. Evaluation is ongoing to determine how the cNICU identification process affected internal system transfers.

Conclusion(s): Standardization of the identification of infants appropriate for transfer resulted in increased patient transfer to cNICUs. Artificial Intelligence model development to identify other important clinical and social factors in successful transfer is planned.