Association of outcomes with organizational characteristics of neonatal intensive care units.

Abstract

Health outcomes may be influenced by organizational and management characteristics. We hypothesized that neonatal intensive care unit managerial practices and organizational processes could affect outcomes of births 500-1499 g, and we assessed this in eight neonatal intensive care units. Multiple-center cohort. Eight acute care neonatal intensive care units in Washington, DC. Infants weighing 500-1499 g, born from October 1, 1994, to February 19, 1997, to a resident of Washington, DC. None. Outcomes included 28-day survival/death, bronchopulmonary dysplasia (BPD), periventricular/intraventricular hemorrhage or periventricular leukomalacia (PIVH/PVL), retinopathy of prematurity (ROP), length of hospital stay, and days of mechanical ventilation. Managerial practices and organizational processes were rated by nurses (n = 218), physicians (n = 73), and respiratory therapists (n = 77). Risk-adjusted logistic and linear mixed models were used to assess the association of outcomes with the caregiver ratings. A lower incidence of PIVH/PVL was associated with better overall scores (p =.0036) and better subscores for leadership (p <.0001), coordination (p =.047), and conflict resolution (p =.020). Better values of the respiratory therapists', nurses', and physicians' scores were associated with lower mortality rates (p =.045) and BPD (p =.0057), PIVH/PVL (p <.0001), and ROP (p =.049), respectively. In multivariate analysis, the joint effects of the professional groups' scores were associated with the incidence of PIVH/PVL (p =.0047); the nurses' scores were associated with lower incidence of PIVH/PVL (p =.0051). The association between the respiratory therapists' scores and lower mortality rate (p =.025) also remained significant in multivariate analysis. We found that the organizational processes and managerial practices as rated by healthcare professionals in neonatal intensive care units were associated with the development of mortality and chronic, severe morbidity. We did not identify specific processes or practices that accounted for our results.