Objective: To increase weight gain in the first 28 days after birth for very low birth weight (VLBW) infants by isolating and sharing meaningful process differences between high- and low-weight-gain centers within a neonatal network.
Design/methods: We identified weight gain as an important target for improvement in 1999 for our national group practice of neonatologists. Site-specific average weight gain during the first 28 days was the primary outcome measure. Our target population was defined as inborn infants who survived and remained in the hospital of birth, whose birth weights were between 401 and 1500 g (VLBW), and who were >22 weeks' estimated gestational age. A team of 6 neonatologists and 1 nurse met, reviewed processes that might influence growth, and developed a structured observation guide for site visits. Weight gain data were obtained from an existing administrative database for the period January 1, 1997, through June 30, 1999. Centers were ranked and divided into upper, middle, and lower thirds. Seven team members visited 1 high- and 1 low-weight-gain center without being informed of the center's performance. Following the site visits, the team isolated 16 meaningful differences between high- and low-weight-gain sites. Meaningful differences were defined as processes observed in all or virtually all (for this project, 6 or 7 of 7 centers) of the high and none or virtually none (for this project, 0 or 1 of 7) of the low centers. The meaningful differences were distributed to our medical directors in August 2000 along with their site-specific weight-gain performance. To document the impact of sharing this material, we compared weight gain in a baseline period of January 1 through December 31, 1999 and a posteducational intervention period of January 1 through September 30, 2001.
Results: Compared with neonates admitted to our national neonatal practice in 1999, neonates admitted in 2001 were similar in birth weight, gestational age at birth, exposure to antenatal steroids, and male gender. Average daily weight gain during the first 28 days increased from 10.4 +/- 6 g for neonates cared for in 1999 to 12.5 +/- 6 g for neonates cared for in 2001. Thirty-nine of 51 (76%) units noted improvements, 4 were unchanged and 8 noted a decrease in average weight gain. Despite similar average lengths of stay, the average discharge weight for neonates sent home increased from 2.15 +/- 0.5 kg for 1999 to 2.29 +/- 0.5 kg for 2001. There were no differences in frequencies of mortality or major morbidities, including severe intraventricular hemorrhage, retinopathy, or necrotizing enterocolitis, between the 2 time periods. An increase in the use of continuous positive airway pressure was noted in the post implementation period.
Conclusions: Variation in common processes can alter clinical outcomes. Although temporal trends in weight gain may be, in part, responsible for this trend, it appears that isolation and implementation of meaningful differences in processes can augment our desire to rapidly improve clinical outcomes.