Disparities in tracheostomies and gastrostomy tubes in infants at the edge of viability.

Abstract

To assess trends of clinical practice of tracheostomy and gastrostomy tube (G-tube) placement over a 25-year period and to assess the association of both procedures with outcomes in extremely low birth weight (ELBW) infants (<1000 g). We reviewed and analyzed data obtained from the National Inpatient Sample dataset from 1993 to 2018. ELBW infants who received a tracheostomy and or G-tubewere included. Regression analysis was performed to assess the association of tracheostomy and/or G-tube placement with mortality after controlling for confounding variables. A total of 620,061 ELBW infants were identified in the weighted sample. Of them, 221,339 were included. G-tube was placed on 4867 (2.2%) infants, tracheostomy was placed on 1788 (0.8%) infants, and both procedures were placed on 1026 (0.46%). Mortality was highest (26.9%) in infants who received tracheostomy only and lowest (5.5%) in those who had G-tube only and (17.8%) in those who had both procedures with adjusted odds ratios of 7.1 (6.3-7.9), 0.7 (0.6-0.8), and 3.7 (3.1-4.4), respectively. The length of stay (LOS) was highest with the combined procedure (219 ± 171), and lowest in the nonintervention group (88 ± 57). The trend of G-tube placement has increased over the years (p < 0.01) but tracheostomy placement has decreased over time. Tracheostomy is associated with increased mortality in ELBW infants and increased LOS in survivors whereas the placement of both tracheostomy and G-tube is independently associated with decreased mortality. Prospective studies are needed to assess the relationship between clinical presentation, the timing of the procedure, and outcomes.