Objectives: We sought to generate a contemporary postnatal weight grid for low birth weight (LBW, <2500 g) neonates using actual data, not mathematically derived idealized weight curves. To do this, we collected electronic data from all neonatal intensive care units (NICU) patients weighing 400-2600 g at birth, during 30 consecutive months (January 2003-June 2005) in all four NICU's within a single health system, Intermountain Health Care (IHC).
Methods: A deidentified limited data set was collected from electronic IHC records. Data were limited to the birth weights and all subsequent daily weights, until discharge, death, or transfer to a non-IHC facility. Lines were drawn connecting the exact daily means of body weight for patients in 200 g birth weight groupings, ranging from 400 to 2600 g. We assessed differences in growth patterns predicted by this grid vs three previously published NICU longitudinal weight grids.
Results: All recorded weights were electronically extracted for 1813 consecutive patients with dates of birth from January 1, 2003 through June 30, 2005. The daily weights of each patient totaled over 48,000 individual data points. The new grid differed from previously published grids, in that the new grid showed; (1) no predicted postnatal weight loss among neonates in categories<900 g birth weight, (2) less initial weight loss among neonates 900-2500 g, (3) fewer days to regain birth weight among neonates<or=2000 g, and (4) fewer days to gain 100 and 300 g above birth weight in neonates<or=1500 g.
Speculation: For charting serial weights of LBW neonates in the NICU the new grid may be a useful standard.