Identifying infants at risk of sudden unexpected death with an automated predictive risk model

Area of Science:

• Pediatrics
• Public Health
• Infant Mortality Research

Background:

• Sudden unexpected infant death (SUID) remains a significant public health concern.
• Existing predictive models for child welfare may offer insights into infant safety.
• Identifying infants at risk for SUID and unsafe sleep is crucial for prevention.

Purpose of the Study:

• To evaluate the Hello Baby predictive risk model (PRM) for its ability to identify infants at high risk of SUID.
• To assess the PRM's utility in identifying infants experiencing non-fatal unsafe sleep events.
• To determine if a PRM developed for foster care risk can be repurposed for infant safety.

Main Methods:

• Retrospective case-control study design.
• Inclusion of SUID and unsafe sleep event cases over 5.5 years in one county.
• Comparison of cases with all births (controls) in the same county.
• Assignment of Hello Baby PRM scores using demographic and clinical data.

Main Results:

• Infants with SUID or unsafe sleep events had significantly higher median PRM scores than controls (17.5 vs. 10, p < 0.001).
• 50% of cases had a PRM score of 17-20, compared to 16% of controls (p < 0.001).
• Demographic and clinical data were similar between cases and controls, except for age in unsafe sleep events.

Conclusions:

• The Hello Baby PRM effectively identifies newborns at elevated risk for SUID and non-fatal unsafe sleep events.
• Early identification enables targeted interventions for high-risk families and modifiable risk factors.
• The model's applicability is contingent on county-level data availability for PRM calculation.