PEDIATRICS Vol. 110 No. 6 December 2002, pp. 1125-1132
Intersite Differences in Weight Growth Velocity of Extremely Premature Infants
,
,#,||,¶,||,#,¶,
* Departments of Nutrition
Neonatology, Beth Israel Deaconess Medical Center, Boston, Massachusetts
Gerald J. and Dorothy R. Friedman School of Nutrition Science and Policy
|| School of Medicine
¶ Jean Mayer Human Nutrition Research Center on Aging, Tufts University, Boston, Massachusetts
# Department of Maternal and Child Health, Harvard School of Public Health, Boston, Massachusetts
** Biostatistics Research Center, Division of Clinical Care Research
Frances Stern Nutrition Center, New England Medical Center, Boston, Massachusetts
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Objective. To explain differences in weight growth velocity of extremely premature infants among 6 level III neonatal intensive care units (NICUs).
Methods. In 6 NICUs, we studied 564 infants, stratified by gestational age (GA), who were first admissions, survivors, <30 weeks’ GA at birth, and in the NICU at least 16 days. Case mix (eg, birth weight, GA, race, illness severity, prenatal steroids), exposure to medical practices/complications (eg, respiratory support, postnatal steroids, necrotizing enterocolitis, infection), and nutritional intake (kcal/kg/d and protein in g/kg/d) were collected and used to predict weight growth velocity between day 3 and day 28 (or discharge, if transferred early) in multiple linear regression models.
Results. Weight growth velocities varied significantly among the 6 NICUs. Adjustment for case mix and medical factors explained little of this variability, but additional control for calorie and especially protein intake accounted for much of the intersite variability. For the average infant, adjusted growth velocity ranged from 10.4 to 14.3 g/kg/d among the sites studied. The final predictive model, including case mix and medical and nutritional factors, explained 53% of the overall variance in growth velocity. Prolonged (
15 days) exposure to postnatal steroids and greater severity of illness both decreased growth velocity. The model predicted that adding 1 g/kg/d protein to the mean intake for our sample would increase growth by 4.1 g/kg/d.
Conclusions. Variation in nutrition explained much of the difference in growth among the NICUs studied. Mean intake of calories and protein failed to meet recommended levels, and the average growth in only 1 NICU approximated intrauterine growth standards. Increasing nutritional intake into the recommended ranges, in particular of protein, may increase growth of extremely premature infants up to or above intrauterine rates.
Key Words: premature infants • growth • transfer bias • case mix • nutrition • protein • steroids • SNAP
Abbreviations: NICU, neonatal intensive care unit • NEC, necrotizing enterocolitis • GA, gestational age • VLBW, very low birth weight • SNAP, Score for Neonatal Acute Physiology • CI, confidence interval
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
It is difficult to achieve adequate growth in premature infants who are cared for in neonatal intensive care units (NICUs). How well infants grow depends on many factors, including their baseline characteristics (eg, age, size, race, gender, severity of illness, maternal exposures), clinical practices that may facilitate (eg, temperature control, respiratory support, insulin, nutrition) or impede growth (eg, postnatal steroid exposure), and the medical complications (eg, infection, necrotizing enterocolitis [NEC]). Health care professionals strive to provide the combination of NICU practices that optimize health outcomes for each infant.1,2 Advances in medical technology have improved the survival of the youngest, smallest, and sickest infants.3 However, few evidence-based standards of care are available, particularly for complex practices (eg, feeding by special routes), and NICU practices vary widely. Studies that examine the impact of these practices on growth outcomes can provide insight into how best to optimize care for these infants.
Growth is an important health outcome measure in the NICU.4,5 Satisfactory weight gain is associated with shortened lengths of hospital stay,6,7 which in turn may reduce health care costs. Early postnatal growth in weight, head circumference, and length are associated with later growth and cognitive development.8–11 Nevertheless, 97% of infants who are <1500 g at birth are discharged weighing less than the 10th percentile for corrected gestational age (GA).3
Multicenter studies have found that mean growth varies significantly among NICUs.5,12–14 Rubin et al13 reported preliminary findings of significant growth differences in the very low birth weight (VLBW) infants among 7 NICUs. Our growth study examined growth and factors that affect growth in a subset of infants from this "parent" study.
There is no established standard measure of growth for the first 28 days of life, which comprise the neonatal period. Weight loss occurs within the first few days after birth, but weight gain is expected thereafter. Intrauterine weight gain goals15–18 are generally applied once birth weight is regained.19,20 However, these goals do not take into account the degree and duration of initial weight loss, which may also affect ultimate health outcomes. The amount and duration of initial weight loss that are tolerated before corrective interventions are attempted depend on clinical judgment, and this differs depending on the medical practice style of a particular NICU.
The goals of this study were 1) to explain growth differences in extremely premature (<30 weeks’ GA) infants among 6 level III NICUs in New England using a more sensitive and less biased measure of growth than was used in the parent study and 2) to identify NICU practices and complications associated with growth (positive and negative) in extremely premature infants.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Data Sources
The original cohort21 consisted of all VLBW infant admissions to 6 New England NICUs between October 1994 and June 1996. The current analyses are part of a nested multisite cohort study, derived from the original cohort. Institutional Review Board approval was obtained at all sites. (See "Acknowledgments" for a list of the sites and collaborators.)
Sample Selection
The eligibility of infants from the original cohort for our growth study was based on a number of criteria. Only infants admitted within 24 hours of birth to a participating site and who survived for the first 28 days of life (the duration of our study) were eligible for this study. Transfer bias caused by variation in the number and timing of infants who were transferred out of the sites to community NICUs for convalescence was effectively limited in our growth study by restricting eligibility to infants <30 weeks’ GA at birth, thus minimizing the number of infants who were transferred out of the NICUs before 16 days (defined as "early transfers").22 The infants had to spend a minimum of 16 days in the NICU because it provided a reasonable minimum number of days on which to estimate the early growth trajectory of infants in the sample. Additional exclusions to the sample included 13 infants who were transferred to hospitals at which record tracking was not feasible and 4 infants whose length of NICU stay was missing.
The goal for sample selection was to select 600 infants from the eligible infants in the original cohort, 100 from each NICU, stratified into 3 GA categories (
26, 27–28, and 29 weeks’ GA at birth). All infants selected with a 16 day NICU stay were included in the study sample. Infants with a 7- to 15-day NICU stay were labeled "early transfers." For minimizing further the potential impact of early transfer bias, a replacement infant with a 16-day NICU stay was selected, matched on GA category, for each early transfer infant. Previous analyses22 confirmed that these replacements adequately substituted for the early transfers.
Data Collection
This retrospective medical record review collected serial measurements of growth and factors believed to affect growth in extremely premature infants.
Growth
Weights were collected on days of life 0, 3, 7, 10, 14, 21, and 28. When these weights were not available, a weight from the 2 days adjacent to each target day was accepted. When an infant was transferred or discharged from the NICU before 28 days, the weight on the day of transfer (or the day before, as needed) was collected.
Case Mix
Baseline newborn characteristics included GA, birth weight, small for gestational age (<5th percentile weight for age based on Brenner curves23), race, gender, multiple gestation, and Apgar scores. Illness severity was measured using the Score for Neonatal Acute Physiology (SNAP24,25) on days 1 and 3. Maternal characteristics included exposure to prenatal steroids and any prenatal care.
Medical Practices and Complications
Temperature control was defined according to the number of days (in the first 28) the infant spent on a warmer, in an incubator, or in a crib. Respiratory support was defined as the number of days spent on a ventilator or on continuous positive airway pressure or without positive pressure, using either nasal cannula oxygen or room air. Insulin usage (yes/no) was collected on days 0, 3, and 14. Postnatal steroid usage was stratified according to the number of days of exposure. Steroid exposure was categorized into 4 categories (no/low/medium/high) based on 0 days, 1 to 7 days, 8 to 14 days, and 15 or more days of postnatal steroid exposure. Infections were defined as any positive blood culture within the first 28 days of life. The infections were not further classified by type of organism or by whether the clinicians suspected the organism to be a contaminant.26 NEC was defined by Bell’s criteria27 as any documented pneumatosis intestinalis or surgically proven diagnosis within the first 28 days of life. Patent ductus arteriosus was defined as any clinically suspected/proven patent ductus arteriosus that resulted in treatment with medical or surgical intervention within the first 28 days of life.
Nutritional Intake
Intake from parenteral (including intravenous dextrose solutions) and/or enteral nutrition (including feedings by mouth or tube feedings) on days of life 0, 3, 7, 14, 21, and 28 was collected for the full 24-hour period. Raw data on intake (including fluid or formula type, concentration, and volume) were converted to kilocalories (in kcal/kg/d) and protein (in g/kg/d) for total, parenteral, and enteral nutrition based on manufacturers’ information. Nutrition on day 0 (because the intake was so low, eg, an average of 10 kcal/kg/d and 0 g protein/kg/d) and day 28 (because this was the last day of the growth interval) were excluded from analyses because these were unlikely to have an impact on the 28-day growth interval.
Definition of Growth Outcome
The outcome for this study was weight growth velocity (g/kg/d) measured from day of life 3 to 28 (or to transfer date for 40 infants who were transferred before day 28). This estimate of growth reflected the weight losses and gains characteristic of the neonatal period and is expressed as a growth velocity because this is clinically meaningful in the NICU setting.22 Preliminary studies indicated that the best estimate of neonatal growth velocity was based on day 3 and day 28 weights because it takes into account some early weight loss.22 Because each infant had a minimum 16-day stay in the study NICU, growth velocity estimates were based on a minimum of 13 days and a maximum of 25 days. Weight growth velocity was calculated as the weight change (day 3–28) divided by the number of days spent in the NICU after day 3 and divided by birth weight. When an infant was transferred from the NICU before 28 days, the transfer weight was used to calculate the growth velocity (ie, weight change divided by length of stay divided by birth weight).
Statistical Analyses
Infant characteristics and infants’ exposures to specific NICU practices were summarized using means with standard deviations (for continuous variables) or numbers with percentages (for categorical variables). This information was calculated for the whole study sample and for each NICU separately. Analysis of variance and 
2 tests were used to compare characteristics across NICUs.
Missing Data
Temperature control data (ie, time spent on a warmer, in an incubator, or in a crib) were often not available in the medical records, and the consistency of their availability differed among sites. Therefore, these data were eliminated from the analyses. In calculating weight growth velocity, day 0 weight was used in place of a missing day 3 weight for 3 infants (each from a different NICU). Missing values for race (N = 31) were handled by creating an "unknown race" category. For missing enteral nutrition values, the following assumptions were made: 1) full strength when breast milk or formula strength was missing and 2) 20 kcal/oz concentration when breast milk or formula concentration was missing. For parenteral nutrition, dextrose and/or amino acid concentration was missing on 1 of the study days for 14 infants, on 2 of the study days for 5 infants, and on 3 of the study days for 1 infant. Conservative defaults (5% dextrose and 0.5% amino acids) were used for these missing values. The volume for parenteral nutrition was missing for 10 infants, and the default used was no volume. Seven infants had unrealistically high nutritional intake (for the combined total of parenteral and enteral nutritional intake) on 1 of the study days. These measurements were replaced by interpolated values (ie, a feasible amount of nutrition on that study day based on the intake on the adjacent study days).
Regression Analyses
Multiple linear regression models to predict weight growth velocity used the following predictors: NICU sites, case mix (including the patient and maternal characteristics listed above), medical practices and complications, and nutrition. Predictors were added in clusters to test for incremental predictive value while controlling for other predictors in the model as follows: model 1: NICUs only, to test for crude growth velocity differences among sites; model 2: NICUs + case mix, to determine whether case mix factors explained the growth velocity differences among NICUs; model 3: NICUs + case mix + medical practices and complications, to determine whether the medical treatment practices and complications explained the growth velocity differences among NICUs while controlling for the case mix factors; model 4: NICUs + case mix + medical practices and complications + nutritional intake, to test whether nutritional intake explained the growth velocity differences among NICUs while controlling for the case mix and medical predictors. Significance was determined at a 2-sided 0.05 level. Nonsignificant variables were retained in the models because our previous hypothesis was that these might affect growth. The final model conformed to standard regression assumptions of the linearity of predictors and constant variance, normality, and independence of errors.
Interactions
The clinically feasible interactions among the predictors that were prespecified and tested in the context of the final model were: 1) GA and illness severity, 2) GA and steroid exposure, 3) infection and steroid exposure, 4) infection and NEC, and 5) steroid exposure and NEC. Interactions were also tested between site and any significant main effect. Significance for all interactions was determined using a 2-sided test at the .01 level.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Study Sample
On the basis of inclusion criteria, 851 infants from the original cohort were eligible for the growth study. The random selection process resulted in a total of 591 infants, because of a shortfall of 9 eligible cases within the requisite GA categories (7 for 29 weeks’ GA cases, 2 for 26 weeks’ GA cases). Twenty-one early transfer infants were selected. The medical records for 27 cases were unavailable, which resulted in a total sample size of 564 infants including those who were selected as matched replacements for the early transfers.
Infant Characteristics and NICU Practices
Table 1 presents the case mix characteristics and exposures to medical practices and complications for all NICUs combined and each NICU separately. Data on insulin usage were also collected; however, because its use overall was extremely low (3.2%, 2.0%, and 2.1% on days 1, 3, and 14, respectively) in this sample, it was not included in the regression models as a predictor of growth. Table 2 provides the mean total nutritional intake (parenteral + enteral feedings by mouth or tube) for each NICU. As is evident in these tables, case mix and medical and nutritional risk factors differed significantly by NICU. To account for this variation, we used multivariate models for the final analyses.
|
|
Multiple Linear Regression Models
Only 9 infants had missing data for demographics, clinical variables, and/or the growth outcome after the defaults and/or data adjustments (discussed above). These infants were excluded from the regression analyses for a total of 555 cases.
Table 3 presents the average weight growth velocities (in g/kg/d) for a reference infant (as defined in the footnote to the table), based on day 3 to days 16 to 28 growth, for each NICU compared with the reference NICU. NICU F was selected as the reference site because its infants had the slowest mean growth among the NICUs studied. Model 1 provides the crude mean growth velocity for each NICU unadjusted for any of the predictors of growth (case mix, medical practices and complications, and nutrition). Models 2, 3, and 4 provide mean growth velocity, progressively adjusted for the predictors of growth as specified by each model. Table 4 presents the predicted growth velocity estimate, standard error, 95% confidence interval (CI), and P value for each factor in the final regression model (model 4).
|
|
Model 1: NICU Sites Only
The unadjusted weight growth velocities (Table 3) varied significantly among the 6 NICUs (P < .0001). On average, the infants in NICUs A, B, C, and E grew significantly faster than reference NICU F. This unadjusted model explained only 12% of the variation in the growth velocity, however (R2 = 0.12).
Model 2: NICUs + Case Mix
Adjusting for case mix explained a substantial amount of the variation in the growth velocity, as indicated by the increase in the R2 to 0.37 (Table 3). Most of the additional explanatory power of these factors came from the measure of severity of illness (ie, SNAP scores). The ranking for site B dropped relative to the other sites because infants in this NICU had the lowest SNAP scores on average (ie, the healthiest infants, on average). The average infant in NICU B lost his or her advantage of being healthier when controlling for severity of illness. Although case mix factors substantially increased the explained variance of growth itself, they did little to explain the growth differences among the NICUs. Average growth velocity, adjusted for case mix, for infants in NICUs A, B, C, and E remained significantly greater than in site F (P < .0001).
Model 3: NICUs + Case Mix + Medical Practices and Complications
The inclusion of the medical predictors, while adjusting for case mix, provided little additional explanatory value to the model increasing R2 from 0.37 only to 0.42 (Table 3). Of all of the medical practices and complications tested, postnatal steroids had the majority of the impact on growth. The medical predictors also did little to help explain the growth differences among sites. Average growth, adjusted for case mix and medical factors, in NICUs A, B, C, and E remained significantly greater than in NICU F (P < .0001).
Model 4: NICUs + Case Mix + Medical Practices and Complications + Nutritional Intake
With the addition of nutritional intake (Table 3), this predictive model explained 53% of the overall variance in growth velocities and most of the growth differences among sites. Although site differences remain significant (P < .0001) after adjustment for all effects, only NICUs A and C had significantly faster growth than the reference site. Therefore, the variation in the mean calories and protein of infants among the NICUs had the largest impact on explaining the growth differences among the NICUs studied.
In this final model (Table 4), factors significantly associated with a higher growth velocity were lower SNAP on days 1 and 3, black race, singleton pregnancy, prenatal exposure to steroids, less exposure to postnatal steroids, and protein intake (P < .05). On average, infants with higher SNAP scores on day of life 1 (ie, who were sicker) grew more slowly (a mean of 0.1 g/kg/d slower per SNAP point). For example, a moderately ill infant with a SNAP score of 20 points on day 1 would be expected to grow an average of 0.9 g/kg/d slower than a reference infant with a day 1 SNAP score of 11.2 (eg, [20 points - mean SNAP of 11.2 points] x 0.1 g/kg/d). On average (Table 4), black infants grew 1.6 g/kg/d more rapidly than white infants, infants from multiple gestation pregnancies grew 0.9 g/kg/d slower than infants from singleton pregnancies, and infants who were exposed prenatally to steroids grew 1.0 g/kg/d more rapidly than unexposed infants. In contrast, postnatal steroids had a strong negative effect on growth that increased with increasing duration of exposure (ie, 1.0, 1.9, and 3.4 g/kg/d less growth based on 1–7, 8–14, and 15 days of exposure) in comparison with no exposure. The combined predictive contribution to growth velocity of an additional 1 g/kg/d protein to the mean intake for our sample (Table 2; an increase to 1.6, 3.1, 3.5, and 3.5 g protein/kg/d for days 3, 7, 14, and 21, respectively) was estimated to be 4.1 g/kg/d. Our results also suggest an additional 30 kcal/kg/d (ie, an increase in the mean to 74, 104, 124, and 132 kcal/kg/d for days 3, 7, 14, and 21, respectively) might increase growth velocity by approximately 1.4 g/kg/d (95% CI: -1.3–4.2).
The only significant interaction found was between NICU and day 3 SNAP score. Because of the interaction, the relationship between growth velocity and day 3 SNAP score varied by site. Sicker infants tended to grow more slowly in 2 of the NICUs (E and F). In the other 4 NICUs (A, B, C, and D), growth velocity did not vary significantly across the range of illness severities. When tested in the predictive model, this interaction had no significant impact on the growth estimates of the other predictors or the explained variance of the model.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Significant growth differences were previously reported in the average growth of premature infants among these 6 New England NICUs,13 and we confirmed and extended these findings. The current growth study improved on the previous study in several ways. Our sample was defined by a GA rather than a birth weight cutoff to avoid including a disproportionate number of small-for-GA infants28,29 whose growth might not be representative of the entire population of premature infants in the NICU. Only infants who were <30 weeks’ GA at birth were included to focus on infants who were most challenged to grow. Our sampling strategy minimized any bias introduced by case mix differences and transfers out of the NICU.22 We selected a growth measure (day 3 to days 16–28 growth velocity in g/kg/d) that provided a good measure of neonatal growth and minimized the potential bias between sites in growth comparisons.22 Finally, we performed a comprehensive retrospective analysis of clinical and nutritional factors that were expected to explain the inter-NICU differences in growth.
Despite stratified sampling, significant variation among the NICUs remained in the case mix characteristics, as well as medical practices and complications (Table 1) and nutritional intake (Table 2). Model 1 (in Table 3) demonstrated that there were significant differences in average growth velocity among the extremely premature infants in the 6 NICUs in our study. Stage-wise adjustments to control for variation in the predictors of growth helped to explain some but not all of the growth differences among the NICUs studied as the disparity in growth narrowed in the series of regression models (Table 3). Nutrition, in particular protein intake, had the strongest impact of explaining these differences among NICUs, as discussed below.
We also identified several factors that contributed to growth in all sites. The most powerful significant predictors of growth in our models were SNAP score on days 1 and 3, exposure to postnatal steroids, and protein intake. Black race, singleton pregnancies, and exposure to prenatal steroids were also positive and significant but less powerful predictors of growth in predictive modeling.
It is not surprising that a measure of severity of illness (SNAP score) was found to be such a strong predictor of growth. In the clinical setting, healthy infants grow more rapidly than sick infants. The interaction between NICU and day 3 SNAP score is intriguing. Growth was impeded in sicker infants compared with healthier infants in some but not all sites. This differential impact of sickness on growth may represent differences in NICU practices. Aggressive use of parenteral nutrition in the presence of increased nutritional needs during an acute period of illness may help to explain the diminished impact that illness had on growth in 4 of the NICUs studied. In contrast, a conservative approach of restricting nutritional advances until the infant’s medical condition improves may help to explain the negative impact that illness had on infants in the other 2 NICUs. In either case, our results stress the importance of including the SNAP score in the collection of case mix data.
Postnatal corticosteroids have been used in the preterm infant to help improve lung compliance and shorten respiratory support duration,30 but this may be achieved at the expense of optimal growth. The negative effect of these catabolic steroids on growth30–34 was evident in our study when examined as an exposure gradient. The risk-benefit ratio of this medication is controversial.35
Nutritional intake also helped to explain growth differences in our models. Moreover, it had the largest impact on explaining the growth differences among the sites, as mentioned above. Enteral nutrition recommendations for premature infants range from 120 to 165 kcal/kg/d36,37 and 3 to 4 g protein/kg/d,18,38 and the upper end of these ranges are suggested for the smaller, younger infants. The mean calorie and protein intake (in kcal/kg/d and g protein/kg/d) of our study infants differed significantly by site (Table 2). NICU B consistently provided the highest amounts of calories and protein, and these differences often achieved statistical significance. NICU F provided nutrition at the lowest end of the range. It is interesting that the more rapid growth in NICUs A and B must have been attributable in part to "better" nutrition (ie, provision of more kcal/kg/d and g protein/kg/d) compared with the other NICUs, because when nutrition was controlled in model 4, the adjusted growth velocity in both of these NICUs decreased by 1.3 g/kg/d (Table 3). The adjusted growth in NICU C, however, improved when nutrition was controlled for in the model, suggesting that the faster growth of these infants was not attributable to their nutritional intake. Note that the average nutritional intake in this NICU was consistently on the lower end of calories and protein provided among NICUs. Nutrition was the most important explanatory factor accounting for growth differences among sites in our study; however, none of the NICUs provided infants (on average) the recommended levels of nutrition with respect to calories and protein. On the basis of our results, increasing mean protein intake of these infants closer to the recommendations (ie, an additional approximately 1 g protein/kg/d) extrapolates to an increase in growth velocity by >4 g/kg/d. In addition, although calorie intake was not a significant predictor of growth, our data cannot rule out a substantial effect of additional calories (ie, approximately 30 kcal/kg/d) on growth given the wide CI around the growth estimate (Table 4).
There is no established benchmark for neonatal weight growth. Weight growth velocity based on days 3 to 28 (or discharge, if transferred early) purposely uses a baseline measurement after the first few days of weight loss so as not to exclude the period of weight loss that every neonate experiences early in postnatal life. Standards for day 3 to days 16 to 28 growth velocity are, therefore, likely to be below intrauterine growth rates of approximately 15 g/kg/d.15–18 Nevertheless, in our study, the mean growth of an average infant in the NICU with the most rapid growth (NICU A, 14.3 g/kg/d) was very close to intrauterine rates (Table 3). This suggests that improvements in nutritional intake to achieve more closely the recommendations may permit growth at or above intrauterine rates. In fact, such "catch-up" growth may be desirable in this tiny, extremely premature population.
A limitation of this study is that these data were collected in 1994 to 1996. Although our results are consistent with the nutritional intake from another study conducted during a similar time period,20 perhaps recent data would be more favorable. Two more recent studies have suggested that the lower ends of the recommendations (ie, approximately 120 kcal/kg/d and approximately 3 g protein/kg/d) are not enough to support optimal growth in the premature infant.39,40 Suboptimal nutrition, starting early in the neonatal period, contributes to the accumulation of growth deficits early in postnatal life.20,31,39 These early growth deficits may contribute to the small size (for corrected GA) at the time of NICU discharge despite achieving intrauterine growth rates later during their NICU stay.3,41 Studies using calories and protein closer to the upper ranges of the recommendations (eg, 130–150 kcal/kg/d and 3.5–4 g protein/kg/d) need to be conducted in the extremely premature population to explore the growth potential of these tiny infants at this level of nutrition. Another limitation may be generalizability. One must be cautious about applying these findings to all NICUs until they are confirmed by additional research, because the mix of clinical and nutritional practices may not reflect regional practice style. For example, the uniformly low usage of insulin among the NICUs in our study may differ from practices throughout the United States. Although the use of insulin in premature infants may improve growth through increased nutrition, we were unable to test for this as its use was consistently low among our NICUs.
The final predictive model explained 53% of the variation in growth, and infants in only 2 NICUs still grew significantly faster than those in the reference site (Table 3). The slower growth of infants in NICU F may have been because it had the sickest infants and usually fed these infants the lowest amount of calories and protein. NICU A (with the fastest mean growth) had infants who were nearly as sick as NICU F but were treated with low amounts of prenatal steroids, were given high amounts of protein and energy, had fewer boys, and had more black infants than NICU F. For NICU C, ranked second for fastest growth, the conservative use of postnatal steroids may have been the reason that these infants grew faster than those in most other NICUs in our study. The final model left 47% of the growth differences between NICUs as unexplained. There may be unmeasured characteristics, NICU practices, or environmental conditions that explain the residual growth differences in the model among the sites.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Significant differences exist in the mean growth velocities of extremely premature infants among the 6 level III NICUs that we studied. Greater severity of illness, 15 or more days of exposure to postnatal steroids, and lower protein intake (in g protein/kg/d) were the strongest predictors of slower growth. However, variation in average nutritional intake had the largest impact on explaining the growth differences among the NICUs. Improved nutrition, especially with respect to protein, may have a significant and positive effect on growth in extremely premature infants.
|
ACKNOWLEDGMENTS |
|---|
Partial funding for this study was provided by Mead Johnson Nutritionals. Ms Olsen was the Ellen and Ronald Block Scholar at Tufts University, Gerald J. and Dorothy R. Friedman School of Nutrition Science and Policy. The data collection for the parent study cohort was funded by the Agency for Healthcare Research and Quality (RO1 HS07015). This material is based on work partially supported by the US Department of Agriculture, under agreement no. 58-1950-9-001. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the US Department of Agriculture.
NICUs and site co-investigators included Bhavesh Shah (Baystate Medical Center, Springfield, MA), DeWayne Pursley (Joint Program in Neonatology, Beth Israel Deaconess Medical Center and Brigham and Women’s Hospital, Boston, MA), Francis Bednarek and Stuart Weisberger (University of Massachusetts-Memorial Medical Center, Worcester, MA), Ivan Frantz III (New England Medical Center, Boston, MA), and Lewis Rubin (Women and Infants’ Hospital, Providence, RI). We thank the research assistants and other support personnel involved in this study.
|
FOOTNOTES |
|---|
Received for publication Oct 4, 2001; Accepted Jun 18, 2002.
Reprint requests to (I.E.O.) Children’s Hospital of Philadelphia, Division of Gastroenterology and Nutrition (CHOP North, Rm 1561), 34th St and Civic Center Blvd, Philadelphia, PA 19104. Email: olseni{at}email.chop.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
- Wirtschafter DD, Jones KR, Thomas JT. Using health care outcomes to improve patient care in the NICU. Jt Comm J Qual Improv.1994; 20 :57 –65[Medline]
- Horbar JD, Rogowski J, Plsek PE, et al. Collaborative quality improvement for neonatal intensive care. NIC/Q Project Investigators of the Vermont Oxford Network.
Pediatrics.2001; 107
:14
–22
[Abstract/Free Full Text] - Lemons JA, Bauer CR, Oh W, et al. Very low birth weight outcomes of the National Institute of Child Health and Human Development neonatal research network, January 1995 through December 1996. NICHD Neonatal Research Network. Pediatrics.2001; 107(1) . Available at: www.pediatrics.org/cgi/content/full/107/1/e1
- Blackwell MT, Petit K, McAlmon K, Linton P, Richardson DK. Inter-NICU variation in growth, feeding practices and gestational age at discharge (GAdisch) in healthy preterm infants [abstract]. Pediatr Res.1999; 45 :238A
- Clark RH, Bloom BT, Thomas P, Peabody J. Application of the best demonstrated process method to a neonatal growth quality-improvement project. J Perinatol.2000; 20 :487 –488
- LaBarre DJ, Maher MM, Raye JR, Rowe JC. Effect of monitored feeding protocol on VVLBW infants. J Am Diet Assoc.1987; 87 :A-58
- Stave VS, Robbins S, Fletcher AB. A comparison of growth rates of premature infants prior to and after close nutritional monitoring. Clin Proc Child Hosp Nat Med Center.1979; 35 :171 –180
- Lucas A, Morley R, Cole TJ. Randomised trial of early diet in preterm babies and later intelligence quotient.
BMJ.1998; 317
:1481
–1487
[Abstract/Free Full Text] - Ross G, Lipper EG, Auld PA. Physical growth and developmental outcome in very low birth weight premature infants at 3 years of age. J Pediatr.1985; 107 :284 –286[CrossRef][Web of Science][Medline]
- Hack M, Breslau N, Weissman B, Aram D, Klein N, Borawski E. Effect of very low birth weight and subnormal head size on cognitive abilities at school age. N Engl J Med.1991; 325 :231 –237[Abstract]
- Georgieff MK, Mills MM, Zempel CE, Chang PN. Catch-up growth, muscle and fat accretion, and body proportionality of infants one year after newborn intensive care. J Pediatr.1989; 114 :288 –292[CrossRef][Web of Science][Medline]
- Lee JS, Richardson DK, Clark RH. Postnatal growth of infants 30–34 weeks gestational age (GA): a comparison of 63 neonatal intensive care units (NICUs) [abstract]. Pediatr Res.2001; 49 :358A
- Rubin L, Richardson D, Bednareck F, Olsen I, McCormick M. Longitudinal growth in hospitalized very low birth weight infants: identification of patient characteristics and inter-neonatal intensive care unit differences [abstract]. Pediatr Res.1997; 41 :239A
- Clark R, Thomas P, Peabody J. Extrauterine growth retardation remains a serious problem in prematurely born neonates [abstract]. Pediatr Res.2001; 49 :295A
- Lubchenco LO, Hansman C, Dressler M, Boyd E. Intrauterine growth as estimated from liveborn birth-weight data at 24 to 42 weeks of gestation.
Pediatrics.1963; 32
:793
–800
[Abstract/Free Full Text] - Shaffer SG, Quimiro CL, Anderson JV, Hall RT. Postnatal weight changes in low birth weight infants.
Pediatrics.1987; 79
:702
–705
[Abstract/Free Full Text] - Usher R, McLean F. Intrauterine growth of live-born Caucasian infants at sea level: standards obtained from measurements in 7 dimensions of infants born between 25 and 44 weeks of gestation. J Pediatr.1969; 74 :901 –910[CrossRef][Web of Science][Medline]
- Ziegler EE, O’Donnell AM, Nelson SE, Fomon SJ. Body composition of the reference fetus. Growth.1976; 40 :329 –341[Web of Science][Medline]
- Ehrenkranz RA, Younes N, Lemons JA, et al. Longitudinal growth of hospitalized very low birth weight infants.
Pediatrics.1999; 104(2 pt 1)
:280
–289
[Abstract/Free Full Text] - Carlson SJ, Ziegler EE. Nutrient intakes and growth of very low birth weight infants. J Perinatol.1998; 18 :252 –258[Medline]
- Richardson DK, Shah BL, Frantz ID 3rd, Bednarek F, Rubin LP, McCormick MC. Perinatal risk and severity of illness in newborns at 6 neonatal intensive care units.
Am J Public Health.1999; 89
:511
–516
[Abstract/Free Full Text] - Olsen IE, Richardson DK, Schmid CH, Ausman LM, Dwyer JT. Neonatal Growth in Extremely Premature Infants. Medford, MA: Tufts University; 2001
- Brenner WE, Edelman DA, Hendricks CH. A standard of fetal growth for the United States of America. Am J Obstet Gynecol.1976; 126 :555 –564[Web of Science][Medline]
- Richardson DK, Gray JE, McCormick MC, Workman K, Goldmann DA. Score for Neonatal Acute Physiology: a physiologic severity index for neonatal intensive care.
Pediatrics.1993; 91
:617
–623
[Abstract/Free Full Text] - Richardson DK, Phibbs CS, Gray JE, McCormick MC, Workman-Daniels K, Goldmann DA. Birth weight and illness severity: independent predictors of neonatal mortality.
Pediatrics.1993; 91
:969
–975
[Abstract/Free Full Text] - Brodie SB, Sands KE, Gray JE, et al. Occurrence of nosocomial bloodstream infections in six neonatal intensive care units. Pediatr Infect Dis J.2000; 19 :56 –65[CrossRef][Web of Science][Medline]
- Bell MJ, Ternberg JL, Feigin RD, et al. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann Surg.1978; 187 :1 –7[Web of Science][Medline]
- Dammann O, Dammann CE, Allred EN, Veelken N. Fetal growth restriction is not associated with a reduced risk for bilateral spastic cerebral palsy in very-low-birthweight infants. Early Hum Dev.2001; 64 :79 –89[CrossRef][Web of Science][Medline]
- Arnold CC, Kramer MS, Hobbs CA, McLean FH, Usher RH. Very low birth weight: a problematic cohort for epidemiologic studies of very small or immature neonates.
Am J Epidemiol.1991; 134
:604
–613
[Abstract/Free Full Text] - Harkavy KL, Scanlon JW, Chowdhry PK, Grylack LJ. Dexamethasone therapy for chronic lung disease in ventilator- and oxygen-dependent infants: a controlled trial. J Pediatr.1989; 115 :979 –983[CrossRef][Web of Science][Medline]
- Berry MA, Abrahamowicz M, Usher RH. Factors associated with growth of extremely premature infants during initial hospitalization.
Pediatrics.1997; 100
:640
–646
[Abstract/Free Full Text] - Van Goudoever JB, Wattimena JD, Carnielli VP, Sulkers EJ, Degenhart HJ, Sauer PJ. Effect of dexamethasone on protein metabolism in infants with bronchopulmonary dysplasia. J Pediatr.1994; 124 :112 –118[CrossRef][Web of Science][Medline]
- Yeh TF, Torre JA, Rastogi A, Anyebuno MA, Pildes RS. Early postnatal dexamethasone therapy in premature infants with severe respiratory distress syndrome: a double-blind, controlled study. J Pediatr.1990; 117(2 Pt 1) :273 –282[CrossRef][Web of Science][Medline]
- Murphy BP, Inder TE, Huppi PS, et al. Impaired cerebral cortical gray matter growth after treatment with dexamethasone for neonatal chronic lung disease.
Pediatrics.2001; 107
:217
–221
[Abstract/Free Full Text] - Jobe AH. Glucocorticoids in perinatal medicine: misguided rockets? J Pediatr.2000; 137 :1 –3[CrossRef][Web of Science][Medline]
- American Academy of Pediatrics, Committee on Nutrition. Nutritional needs of low-birth-weight infants.
Pediatrics.1985; 75
:976
–986
[Abstract/Free Full Text] - Nutrition and feeding of preterm infants. Committee on Nutrition of the Preterm Infant, European Society of Paediatric Gastroenterology and Nutrition. Acta Paediatr Scand Suppl.1987; 336 :1 –14[Medline]
- Micheli JL, Schutz Y. Protein. In: Tsang RC, Lucas A, Uauy R, Zlotkin S, eds. Nutritional Needs of the Preterm Infant: Scientific Basis and Practical Guidelines. Baltimore, MD: Williams & Wilkins; 1993:29–46
- Embleton NE, Pang N, Cooke RJ. Postnatal malnutrition and growth retardation: an inevitable consequence of current recommendations in preterm infants?
Pediatrics.2001; 107; 270
–273
[Abstract/Free Full Text] - Reis BB, Hall RT, Schanler RJ, et al. Enhanced growth of preterm infants fed a new powdered human milk fortifier: a randomized, controlled trial.
Pediatrics.2000; 106; 581
–588
[Abstract/Free Full Text] - Berry MA, Conrod H, Usher RH. Growth of very premature infants fed intravenous hyperalimentation and calcium-supplemented formula.
Pediatrics.1997; 100
:647
–653
[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2002 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Grover, M. Khashu, A. Mukherjee, and V. Kairamkonda Iatrogenic Malnutrition in Neonatal Intensive Care Units: Urgent Need to Modify Practice JPEN J Parenter Enteral Nutr, March 1, 2008; 32(2): 140 - 144. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. H. Clark, D. H. Chace, A. R. Spitzer, and for the Pediatrix Amino Acid Study Group Effects of Two Different Doses of Amino Acid Supplementation on Growth and Blood Amino Acid Levels in Premature Neonates Admitted to the Neonatal Intensive Care Unit: A Randomized, Controlled Trial Pediatrics, December 1, 2007; 120(6): 1286 - 1296. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Profit, J. A. F. Zupancic, J. B. Gould, and L. A. Petersen Implementing Pay-for-Performance in the Neonatal Intensive Care Unit Pediatrics, May 1, 2007; 119(5): 975 - 982. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Profit, M. C. McCormick, G. J. Escobar, D. K. Richardson, Z. Zheng, K. Coleman-Phox, R. Roberts, and J. A.F. Zupancic Neonatal Intensive Care Unit Census Influences Discharge of Moderately Preterm Infants Pediatrics, February 1, 2007; 119(2): 314 - 319. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E Bertino, A Coscia, M Mombro, L Boni, G Rossetti, C Fabris, E Spada, and S Milani Postnatal weight increase and growth velocity of very low birthweight infants Arch. Dis. Child. Fetal Neonatal Ed., September 1, 2006; 91(5): F349 - F356. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J Profit, J A F Zupancic, M C McCormick, D K Richardson, G J Escobar, J Tucker, W Tarnow-Mordi, and G Parry Moderately premature infants at Kaiser Permanente Medical Care Program in California are discharged home earlier than their peers in Massachusetts and the United Kingdom Arch. Dis. Child. Fetal Neonatal Ed., July 1, 2006; 91(4): F245 - F250. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J W L Puntis Nutritional support in the premature newborn. Postgrad. Med. J., March 1, 2006; 82(965): 192 - 198. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. L. Patel, J. L. Engstrom, P. P. Meier, and R. E. Kimura Accuracy of Methods for Calculating Postnatal Growth Velocity for Extremely Low Birth Weight Infants Pediatrics, December 1, 2005; 116(6): 1466 - 1473. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Pridham, A. Bhattacharya, S. Thoyre, D. Steward, J. Bamberger, J. Wells, C. Green, F. Greer, P. Green-Sotos, and M. O'Brien Exploration of the Contribution of Biobehavioral Variables to the Energy Expenditure of Preterm Infants Biol Res Nurs, January 1, 2005; 6(3): 216 - 229. [Abstract] [PDF]
|
|
|
|
|
|
C. L. Berseth, J. E. V. Aerde, S. Gross, S. I. Stolz, C. L. Harris, and J. W. Hansen Growth, Efficacy, and Safety of Feeding an Iron-Fortified Human Milk Fortifier Pediatrics, December 1, 2004; 114(6): e699 - e706. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
