Staff Physician Boston Children's Hospital Revere, Massachusetts, United States
Abstract: Introduction /
Objective: Vocal cord dysfunction (VCD) is a common complication following congenital cardiac due to recurrent laryngeal nerve injury. VCD has been identified as a predictor of feeding dysfunction in neonates with congenital heart disease (CHD) leading to higher resource utilization. Therefore, our institution developed a clinical algorithm to identify patients at risk for feeding dysfunction with mandatory vocal cord screening. Our objective was to compare pre-and post-algorithm implementation clinical metrics to characterize algorithm effectiveness.
Methods: Patients who presented with high-risk post-cardiac surgery diagnoses were screened in accordance with the algorithm (Fig. 1). All screened patients underwent a head and neck exam followed by a fiberoptic exam (FOE) to assess for VCD. Children with VCD and/or symptoms of aspiration would undergo a modified barium swallow (MBS) to assess for the presence of aspiration in accordance with a feeding team specialist. Clinical metrics surveyed include length of stay (LOS), time to FOE, time to enteral feeding initiation, and need for tube feeding at discharge.
Results: We compared 237 children pre-implementation (PI) to 29 children post-implementation (PO) with high-risk post-cardiac surgery diagnoses. Screening improved PO from 22% (n=59) to 90% (n=26; p< 0.01). Three parents/primary team members deferred the vocal cord screening PO. Post-implementation, 50% (n=13) were identified with VCD compared to 14% (n=37; p< 0.01). Post-implementation, 58% (n=15) underwent MBS of which 53% (n=8) were found to have aspiration with feeds. Clinical metrics assessed included: median ICU LOS was 8.7 [3.4, 37.1] vs. 9.5 [6.25,14] days (p=0.10); median hospital LOS was 41.4 [13.8, 75.1] vs. 19 [12.25,30] days (p < 0.01); median time to initial FOE screening was 10 [3, 25] and 6.5 [3.75,10.25] days (p < 0.05); and 38% (n=91) vs. 30% (n=8) children were discharged with enteral access (p < 0.05).
Conclusions: Initial analysis post-implementation of the algorithm suggests improved screening of patients at risk of feeding dysfunction leading to early detection of VCD/aspiration and a significant reduction in LOS in the hospital with lower incidence of discharge with enteral access. Our findings suggest early, universal screening for high-risk children might lead to improved feeding practices and reduce burden on resource utilization.
