Pattern of daily weights among low birth weight neonates in the neonatal intensive care unit: data from a multihospital health-care system

Abstract

The daily measuring and recording of body weight are important parts of neonatal intensive care unit (NICU) patient management. In 1948 Dancis et al.1 published the first widely used grid for charting serial weights of NICU patients. Periodic updates of the Dancis-type postnatal weight grid have been published by Shaffer et al.,2 Wright et al.,3 Lair and Kennedy,4 and Ehrenkranz et al.5 Each of those grids predicts an initial weight loss of 5–20% of birth weight, with the largest percentage losses among the smallest patients. Those grids also predict regaining birth weight after 9–17 days, with greatest number of days required by the smallest patients. Recommendations for early postnatal nutrition and fluid management are ongoing, with several such having been published in the past 10 years.6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 On the basis of these, we hypothesized that NICU patients cared for currently have more rapid weight gain than are displayed on the commonly used postnatal grids.2, 3, 4, 5 To test this we collected, from electronic records, all daily weights of all patients with birth weights 400–2600 g, born during a 30-month period, January 1, 2003–June 30, 2005, cared for in any the NICU's of the Intermountain Health Care (IHC) System. We then compared the patterns of weight change predicted by the present vs previous grids. Data were collected as a deidentified limited data set from archived IHC electronic records. The information collected was limited to the birth weight and all subsequent weights until discharge, death, or transfer to a non-IHC hospital. Data were obtained for patients admitted to the NICU at McKay-Dee Hospital, Ogden, UT, LDS Hospital, Salt Lake City, UT, Primary Children's Medical Center, Salt Lake City, UT, and Utah Valley Regional Medical Center, Provo, UT, with a date of birth from January 1, 2003 through June 30, 2005, and with a birth weight of 400–2600 g. The day of birth was recorded as 'day 0.' The program used for data collection is a modified subsystem of 'clinical workstation'. The 3M Company (Minneapolis, MN) approved the structure and definitions of all data points for use within the program. The data were collected from the electronic medical record, case mix, pharmacy, and laboratory systems. Trained and designated clinical personnel enter additional data. Data are managed and accessed by authorized data analysts. To test our hypothesis, we proposed to make four comparisons between the new grid and three previously published grids.2, 3, 5 Specifically, among various birth weight categories, we sought to compare; (1) the percentage of body weight predicted to be lost at the nadir, (2) the number of days predicted to regain birth weight, (3) the number of days predicted to gain 100 g above birth weight, and (4) the number of days predicted to gain 300 g above birth weight. The IHC Institutional Review Board approved the study. Birth weights and all subsequent daily weights in the hospital were extracted for 1813 consecutive patients, generating over 48 000 individual weight measurements. The racial and ethnic makeup of the 1813 patients was as follows; White 81.7%, Hispanic 13.3%, Asian 1.5%, Pacific Islander 1.3%, Black 1.1%, Native American 0.7%, and Others 0.4%. Figure 1 shows the data for days 0–100. Each of the lines on the figure was drawn from means of all data of neonates within a 200 g birth weight range. For instance, the mean daily weights of patients with birth weights of 400–599 g are shown as the bottom line, originating at 500 g on day 0. Similarly, patients with birth weights of 600–799 g are shown as the line originating at 700 g on day 0, and so forth up through patients with birth weights of 2400–2600 g, shown as the line originating at 2500 g on day 0. When fewer than 20 patients remained in a weight category the line was terminated. For instance, the line depicting daily weights of the 2500 g birth weight group stops on day 23 because most patients had been discharged home by then, with fewer than 20 in that birth weight category remaining in the NICU. No consistent differences in postnatal weight pattern were observed between male and female neonates (Figure 2). Figure 3 is a 'smoothed' version of the postnatal growth grid, created by taking the average of points the day before vs the day after any point not adhering to a smooth line. During the 100 days the lines diverge (Figure 1). Although the lines are approximately 200 g apart on day 0, they spread with increasing age, with those of lowest birth weight showing the greatest divergence (Table 1). The relatively poor weight gain of the smallest birth weight neonates is not apparent in the first 15 days of life (Table 2), but is clearly established by 30 days. Results of the comparative assessments of the present vs previous grids are shown in Tables 3 and 4. As an example, the Shaffer curves2 predict that neonates of 700 g birth weight will lose 14% of their birth weight and will require 15 days to regain birth weight (Table 3). However, the current IHC data predicts that neonates of 700 g will not drop below birth weight. Moreover, the Shaffer curve predicts that neonates of 700 g will require 22 days to gain 100 g above birth weight and will require 38 days to gain 300 g (Table 4). However, the IHC data predicts that they will require only 11 days to gain 100 g and only 25 days to gain 300 g. Five differences were observed in the comparative analysis between the present and previous grids. Specifically, the new grid shows; (1) No predicted weight loss among patients in the categories <900 g birth weight, (2) less initial weight loss among neonates 900–2500 g than predicted on the previous grids, (3) fewer days to regain birth weight in neonates 2000 g than predicted on the previous grids, (4) fewer days to gain 100 and 300 g above birth weight in neonates 1500 g than predicted on the...