Original Article
155
Severe Bronchopulmonary Dysplasia is Associated with Higher Fluid Intake in Very Low-Birth-Weight Infants: A Retrospective Study Mindy Ming-Huey Guo, MD1 Ching-Hung Chung, PhD2 Hsin-Chun Huang, MD1 Mei-Yung Chung, MD1,3
Hospital, Chang Gung University College of Medicine, Taiwan 2 Department of Dental Laboratory Technology, Min-Hwei College of Health Care Management, Taiwan 3 Department of Respiratory care, Chang Gung University of Science and Technology, Chiayi Campus, Taiwan
Chih-Cheng Chen, MD1
Address for correspondence Mei-Yung Chung, MD, Department of Pediatrics, Kaohsiung Chang Gung Memorial Hospital, No. 123, Ta Pei Road, Niao Sung District, Kaohsiung City, Taiwan (e-mail: [email protected]).
Am J Perinatol 2015;32:155–162.
Abstract
Keywords
► bronchopulmonary dysplasia ► very low-birth-weight infants ► fluid intake ► patent ductus arteriosus ► pulmonary hemorrhage
Objective This study aims to investigate the association between fluid intake in the first 4 days of life and the subsequent severity of bronchopulmonary dysplasia (BPD) in very low-birth-weight infants (VLBWI). Study Design A retrospective chart review of 75 infants with a gestational age of less than 32 weeks and a birth weight of < 1,500 g was performed. Demographic, clinical data, associated maternal risk factors, and amount of fluid received in the first 4 days of life were analyzed. Results Severe BPD was associated with a lower gestational age (27.04 2.073 wks vs. 28.70 1.706 wks, p¼0.001), lower birth weight (981.44 244.54 vs. 1,199.63 165.39 g, p < 0.001), use of surfactant (91.7 vs. 63%, p¼0.002), patent ductus arteriousus (PDA) (70.8 vs. 37%, p¼0.004), pulmonary hemorrhage (14.6 vs. 0%, p¼0.045), and more fluids received from the 2nd to 4th days of life (346.44 42.38 mL/kg vs. 323.91 27.62 mL/kg, p¼0.007). A cut off point of 345 mL/kg of fluids from the 2nd to 4th days of life was selected using receiver operating characteristic curve analysis, and remained a significant risk factor even after multiple logistic regression analysis. Conclusion Our findings demonstrate that VLBWI who received higher fluid intake from the 2nd to 4th days of life are at an increased risk of developing severe BPD.
Bronchopulmonary dysplasia (BPD) is a leading cause of respiratory illness in premature infants, resulting in prolonged hospitalization and oxygen dependency. Premature infants, especially those less than 32 weeks of gestational age, are at the highest risk of developing BPD, due in part to the fact that saccular development of the lung occurs from the 23 to
32 weeks of gestation.1 Not only is the premature lung more susceptible to damage from mechanical and oxidative injury, inadequate fluid clearance may also play a role in the development of BPD.1,2 Optimal fluid intake in the preterm infant is dependent on a myriad of factors, including gestational age, insensible
received September 24, 2013 accepted after revision April 14, 2014 published online June 10, 2014
Copyright © 2015 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662.
DOI http://dx.doi.org/ 10.1055/s-0034-1376393. ISSN 0735-1631.
Downloaded by: University of California, San Francisco. Copyrighted material.
1 Department of Pediatrics, Kaohsiung Chang Gung Memorial
Feng-Shun Chen, MD1
Severe Bronchopulmonary Dysplasia and Fluid Intake
Guo et al.
water losses through a premature epithelial barrier, urine, and stool.3 Very low-birth-weight infants (VLBWI) are more susceptible to fluctuations in iatrogenic fluids because of an inability of the immature kidney to excrete excess water and sodium.4,5 Excessive fluids in the first few days of life may not only impede the natural course of weight loss after birth6 but may also result in an increased risk of patent ductus arteriosus (PDA) and interstitial pulmonary edema. Previous prospective trials have shown that fluid restriction is related to a decreased incidence of PDA and necrotizing enterocolitis, and an inconclusive trend toward improved outcomes in patients with BPD.7 Because the majority of these studies were performed before the advent of widespread surfactant therapy,8,9 it is unclear if these results may be extrapolated to current clinical settings. It is interesting to note that despite recent advances in neonatal care, the incidence of BPD has remained static for the past two decades.10,11 This study was designed with the intention to investigate the association between fluid intake in the first 4 days of life and the subsequent incidence of severe BPD in VLBWI.
Patients and Methods Subjects All infants born at a large tertiary medical center in southern Taiwan within a 2 year period from October 2008 to September 2010 were selected for this study. Inclusion criteria were a gestational age of less than 32 weeks and a birth weight of < 1,500 g. Patients who died before discharge or had incomplete medical records were excluded from this study. BPD was defined as use of supplemental oxygen for more than 28 days of life, and further stratified as mild, moderate, or severe according to the NICHD (National Institute of Child Health and Human Development) 2001 consensus.12 The study population was then split into two groups, those with severe BPD, and those with mild, moderate, or no BPD. The NICHD 2001 consensus criteria was chosen over older definitions of BPD, such as oxygen use at 36 weeks gestational age, due to its better reflection of long-term outcomes, and ability to identify patients with differing degrees of BPD severity.13
throughout the course of hospitalization. Surfactant was administered via endotracheal tube within the first 2 hours of life if there was marked respiratory distress at birth, oxygen supplementation of more than 40% initially, CO2 retention on blood gas analysis, or if characteristic atelectasis, and air bronchograms indicating respiratory distress syndrome were seen on chest X-ray.
Evaluation of Fluid Management As an overall rule, patients were given around 70 to 90 mL/kg/ d on the 1st day of life. Fluid intake, in the form of enteral or parental fluids, was increased gradually on a day to day basis according to clinical condition or physician preference with a maximum of approximately 150 mL/kg/d. More fluids were given if patients were hypotensive, required blood product transfusions, or if they received phototherapy. Daily fluid intake (parental and enteral fluids), urine output, and body weight changes were recorded from the 2nd to the 4th days of life. Only the first 4 days were included in our study as larger differences in amount of fluid intake were more noticeable during this period. Body weight was measured at the same time daily when the infant was fully unclothed on an electronic scale. Two repeat measurements were performed and an average of these measurements was recorded. Both daily fluid intake and body weight changes were represented as a percentage of birth weight. The 1st day of life was excluded due to different times of birth between patients. Cumulative amount of fluid received from the 2nd to 4th days of life was calculated to account for daily fluctuations, and was regarded as a more accurate measure of the total amount of fluids received by each patient. Factors that may result in increased fluid intake were also taken into account, namely, duration of clinically significant hypotension, phototherapy, and amount of blood product transfusions. Symptomatic hypotension was defined as having a mean blood pressure below gestational age that resulted in the need for inotropic agents or fluid boluses in the first 4 days of life. Patients on phototherapy were given 10 to 30% more fluids depending on the intensity of phototherapy. Blood products were given to patients with hemoglobin level of less than 10 mg/dL, or showed symptoms and signs of bleeding tendency.
Statistical Analysis Data Collection A retrospective chart review was performed and demographic data including gender, gestational age, birth weight, parity, and whether the patient was appropriate or small for gestational age was collected. Maternal chart records were reviewed for associated risk factors including the presence of chorioamnionitis, use of antenatal steroids, and history of maternal hypertension or preeclampsia. Patients with PDA were identified by physical examination and confirmed via two-dimensional cardiac Doppler examination within the first 2 weeks of life. Indomethacin prophylaxis was not given on a routine basis during the study period. Pulmonary hemorrhage was determined clinically in patients with persistent fresh bleeding from the endotracheal tube with characteristic changes on chest radiography. Bacterial sepsis was defined by the presence of any positive blood cultures American Journal of Perinatology
Vol. 32
No. 2/2015
SPSS software version 19 (IBM Corp., Armonk, NY) was used for data analysis. Mean and standard deviation was calculated for all continuous variables and compared using student’s ttest. Categorical variables were compared using chi-square tests. A p-value of less than 0.05 was considered statistically significant. ROC curve analysis was performed to determine a cut off value for the cumulative amount of fluids received that was associated with severe BPD. Multivariate logistic regression of all significant variables on univariate analysis was performed by using backward Wald.
Results Demographic Data During our 2 year study period, 108 patients met our inclusion criteria, 8 were excluded due to death before discharge,
Downloaded by: University of California, San Francisco. Copyrighted material.
156
25 were excluded due to incomplete chart records, and the remaining 75 patients were enrolled for analysis. Of the eight patients who died during the study period, four of them died within the first few days of life and could not provide adequate data regarding fluid intake. The other four patients died of the following causes: Candida fungemia with brain involvement, necrotizing enterocolitis with septic shock, pulmonary hypoplasia secondary to oligohydramnios during gestation, and recurrent pneumonia. While these patients survived past the first 4 days of life, causes of death were not related to the development of severe BPD, the main endpoint of our study, and therefore were excluded from analysis. Demographic, fluid balance and data regarding body weight changes of these four patients are included in ►Table 1. Of the 75 patients included in this review, 48 were found to have severe BPD; in comparison, 19 patients were found to have mild BPD, 6 were found to have moderate BPD, and only 2 were found to have no BPD at all according to NIHCD criteria. Those with severe BPD had a lower mean gestational age at birth (27.04 2.073 wks vs. 28.70 1.706 wks, p¼0.001) and less mean birth weight (981.44 244.54 g vs. 1199.63 165.39 g, p < 0.001). Patients with severe BPD were also more likely to be male, appropriate for gestational age, and less likely to be the results of a single birth, although these three factors were not statistically significant (►Table 2).
Prenatal and Postnatal Factors A review of maternal records showed that those with severe BPD had similar rates of maternal chorioamnionitis, antenatal steroid use, and maternal hypertension when compared with those without severe BPD (►Table 2). In terms of postnatal factors, severe BPD was associated with a history of receiving surfactant (91.7 vs. 63%, p¼0.002), presence of PDA (70.8 vs. 37%, p¼0.004), and pulmonary hemorrhage (14.6 vs. 0%, p¼0.045). Apgar score at 1 and 5 minutes after birth, amount of FIO2 used at 24 hours of life, and rate of bacteremia during course of hospitalization were statistically similar between the two groups (►Table 2).
Total Fluid Intake and Body Weight Changes in the First Few Days of Life On average, patients with severe BPD received more fluids from the 2nd to 4th days of life (346.44 42.38 mL/kg vs. 323.91 27.62 mL/kg, p¼0.007). Cumulative net intake was calculated as total cumulative fluid intake minus total cumu-
Guo et al.
lative urine output. Urine output in our study subjects was highly variable (285.22 55.11cc/kg VS 269.39 71.49), with large SDs with no significant difference between those with severe BPD and those without. Net fluid intake (which was calculated as total fluid intake minus urine output) was also higher in patients with severe BPD (61.22 63.06 cc/kg VS 54.51 59.84 cc/kg), but was not statistically significant. However, both patients with and without severe BPD had similar changes in body weight within the first 4 days of life (►Table 3). ROC curve analysis was performed to determine a cut off value for the cumulative amount of fluids received from the 2nd to 4th days of life that was associated with severe BPD. A value of more than 345 mL/kg from the 2nd to 4th days of life was selected due to highest average of sensitivity and specificity ([sensitivity þ specificity]/ 2¼0.6678) (►Fig. 1). Patients who received more than 345 mL/kg of total fluids from the 2nd to 4th days of life were around 6.25 times more likely to develop severe BPD (►Table 4).
Multiple Logistic Regression Factors that were associated with severe BPD on univariate analysis were then subjected to multivariate logistic regression (►Table 5). Both gestational age and a history of pulmonary hemorrhage were not found to be significant enough to be included in the regression model. Two factors were found to be consistently significant after multiple logistic regression: presence of a PDA (p¼0.029), and receiving more than 345 mL/kg of fluids from the 2nd to 4th days of life (p¼0.013) (►Table 5). Due to the retrospective nature of our study, it is possible that increased fluid intake within the first few days of life could be an indicator of disease severity. However, patients who received more than 345 mL/kg of fluids from the 2nd to 4th days of life had a similar duration of hypotension, phototherapy, and similar amounts of blood product transfusions (►Table 6). This suggests that increased fluid intake of more than 345 mL/kg from the 2nd to 4th days of life did not appear to be the result of fluid boluses or increased fluid intake due to hypotension, phototherapy, or blood transfusions.
Discussion In our retrospective study of 75 VLBWI, increased fluid intake within the first 4 days of life was associated with the development of severe BPD, and remained a significant risk
Table 1 Demographic and fluid balance data of patients who died Patient
a
Gestational age (wks)
Birth weight (g)
Cause of death
Net intake (mL/kg)a
Weight changes (% of birth weight)a
1
24
725
Fungemia with Candida albicans
86.89
7.58621
2
29
1,060
Necrotizing enterocolitis
56.60
6.13208
3
26
750
Pneumonia, Stenotrophomonas
54.27
7.33333
4
25
765
Pulmonary hypoplasia
151.76
8.49673
Cumulative amount from the 2nd to 4th days of life. American Journal of Perinatology
Vol. 32
No. 2/2015
157
Downloaded by: University of California, San Francisco. Copyrighted material.
Severe Bronchopulmonary Dysplasia and Fluid Intake
Severe Bronchopulmonary Dysplasia and Fluid Intake
Guo et al.
Table 2 Demographic, prenatal, and postnatal factors Nonsevere BPD (N¼27)
Severe BPD (N¼48)
p-Value
Gestational age (wks)
28.70 1.706a
27.04 2.073a
0.001b
Birth weight (g)
1,199.63 165.39a
981.44 244.54a
0.000b
Male gender
11 (40.7%)
30 (62.5%)
0.069
SGA
5 (18.5%)
5 (10.4%)
0.48
Single
21 (77.8%)
33 (68.8%)
0.171
Twin
6 (22.2%)
9 (18.8%)
Triplet
0 (0%)
6 (12.5%)
8 (29.6%)
14 (29.2%)
0.966
Antenatal steroids
20 (74.1%)
38 (79.2%)
0.613
Maternal hypertension/preeclampsia
6 (22.2%)
4 (8.3%)
0.154
0.110
Demographic data
Parity
Prenatal factors Maternal chorioamnionitis
Postnatal factors Apgar score 1 min 0–3
6 (22.2%)
19 (39.6%)
4–6
12 (44.4%)
22 (45.8%)
7–10
9 (33.3%)
7 (14.6%)
0–3
3 (11.1%)
6 (12.5%)
4–6
5 (18.5%)
15 (31.3%)
7–10
19 (70.4%)
Apgar score 5 min 0.440
27 (56.3%) 0.498 0.192a
0.530
17 (63%)
44 (91.7%)
0.002b
PDA status
10 (37%)
34 (70.8%)
0.004b
Pulmonary hemorrhage
0 (0%)
7 (14.6%)
0.045b
Late-onset sepsis
3 (11.1%)
8 (16.7%)
0.736
FIO2 at 24 h
0.414 0.151
Surfactant use
a
Abbreviations: BPD, bronchopulmonary dysplasia; PDA, patent ductus arteriosus; SGA, small for gestational age. a Data for gestational age, birth weight, and FIO2 at 24 h are presented as mean SD. These were analyzed by student’s t-test and all other factors were analyzed by chi-square test. b Statistically significant difference.
factor even after correcting for birth weight and gestational age. Net fluid intake, which was calculated as total fluid intake minus urine output was also higher in patients with severe but was not statistically significant. This may be due to the large differences in urine output or insensible water losses, both which are highly variable factors. While some previous studies have demonstrated a related decrease in postnatal weight loss14,15; in our study, patients with severe and nonsevere BPD had similar amounts of postnatal body weight loss. The lack of differences in body weight loss between the two groups may have been due to the relatively smaller differences in fluid intake when compared with similar randomized controlled trials in the past.7 A smaller difference between fluid intake between the two study groups, coupled with a smaller sample size may have American Journal of Perinatology
Vol. 32
No. 2/2015
resulted in changes in body weight either not large enough to be detected or were not statistically significant. It has been postulated that extraneous fluid intake in the first few days of life impedes with the physiological contraction of extracellular fluid and diuresis that occurs after birth.16 Overall our study shows that increased fluid intake in the first 4 days of life is associated with the development of severe BPD, and remains an independent risk factor even after factors, such as gestational age and birth weight are taken into consideration. Most similar studies examining the relationship between fluid intake and the development of BPD have been published over 10 years ago.16 In a recent retrospective review of 1,382 extremely low-birth-weight infants by Oh et al published in 2005, BPD was found to be associated with higher fluid intake
Downloaded by: University of California, San Francisco. Copyrighted material.
158
Severe Bronchopulmonary Dysplasia and Fluid Intake
Guo et al.
159
Table 3 Total cumulative fluid intake and body weight loss Nonsevere BPD mean SD (N¼27)
Severe BPD mean SD (N¼48)
p-Value
Total cumulative fluid intake (mL/kg) Day of life 2
94.58 16.26
102 13.81
0.027a
3
204.04 21.53
217.58 21.53
0.020a
4
323.91 27.62
346.44 42.38
0.007a
82.92 33.39
0.860
Total cumulative urine output (mL/kg) 2
84.24 25.37
3
176.73 51.92
188.15 43.93
0.315
4
269.39 71.49
285.22 55.11
0.288
Total cumulative net intake (mL/kg) Day of life 2
10.33 27.69
19.65 33.67
0.225
3
27.29 51.64
29.42 49.77
0.861
4
54.51 59.84
61.22 63.06
0.654
0.911
Cumulative weight change as% of birth weight Day of life 2
1.01 2.12
1.08 3.27
3
3.33 4.04
3.81 6.13
0.718
4
5.69 5.33
6.28 13.85
0.833
Abbreviations: BPD, bronchopulmonary dysplasia; SD, standard deviation. a Statistically significant difference.
and less postnatal weight loss.14 This phenomenon was also seen in a case-control study of 223 infants by Van Marter et al, who demonstrated an association between greater amounts of fluids and net weight gain in the first 4 days of life and the
Fig. 1 ROC curve analysis was performed to determine a cut off value for the cumulative amount of fluids received from the 2nd to 4th days of life that was associated with severe BPD. A value of more than 345 mL/kg from the 2nd to 4th days of life was selected due to highest average of sensitivity and specificity ([sensitivity þ specificity]/ 2¼0.6678). BPD, bronchopulmonary dysplasia; ROC, receiver operating characteristic.
development of BPD and PDA. The authors also found that colloid, but not crystalloid fluids, had a dose-related relationship with higher ventilator settings such as the fraction of inspired oxygen, peak inspiratory pressure, positive endexpiratory pressure, and ventilator rates.15 It is possible that pulmonary edema not only interferes with alveolar gas exchange, but may also affect the function of surfactant, leading to decreased lung compliance necessitating higher ventilator settings and associated barotrauma. Another factor linked to the development of severe BPD in our study included a confirmed PDA. According to the results of several animal studies, the premature lung is ill equipped to handle an increase in precapillary arterial pressure,17 leading to excess fluids or blood flow to be shifted more readily downstream to small capillary vessels.18 It stands to reason that significant left to right shunting through a PDA would increase pulmonary hydrostatic capillary pressures leading to increased pulmonary edema and decreased lung compliance. To date, there has been no conclusive evidence from randomized controlled trials that links the presence of a PDA to the development of severe BPD.19 In addition, our study patients who received surfactant were also more likely to develop severe BPD, although it was not found to be statistically significant after multiple logistic regression. While early rescue surfactant therapy within the first 2 hours of life has been linked to a reduction in the American Journal of Perinatology
Vol. 32
No. 2/2015
Downloaded by: University of California, San Francisco. Copyrighted material.
Day of life
Severe Bronchopulmonary Dysplasia and Fluid Intake
Guo et al.
Table 4 Cut off point 345 (mL/kg) Days 2–4 fluids
Nonsevere BPD (N¼27)
Severe BPD (N¼48)
p-Value 0.001
345 (mL/kg)
4
25
< 345 (mL/kg)
23
23
Abbreviation: BPD, bronchopulmonary dysplasia. Note: Odds ratio: 6.25.
Table 5 Multiple logistic regression analysis p-Value
a
Odds ratio
95% Confidence interval
0.997
0.994–1.000
Gestational age
NSa
Birth weight
0.052
Surfactant use
0.084
3.872
0.833–17.985
Patent ductus arteriosus
0.029
3.908
1.151–13.269
Pulmonary hemorrhage
NSa
Total fluid intake 345 (mL/kg)b
0.013
5.671
1.435–22.416
Not significant enough to be included in the regression model. Total fluid intake345 (mL/kg) in the first 2nd to 4th days of life.
b
severity of BPD later in life,20 patients who received surfactant in our study were likely to be sicker at birth, with marked respiratory distress, characteristic chest radiographic findings or require a FIO2 of more than 40%. In patients with respiratory distress syndrome, the effect of excessive fluids and resulting pulmonary edema may further inhibit surfactant function in an already compromised lung.19,21 Limitations of this study include the retrospective nature of the study design, and a relatively small sample size. Of note, the incidence of moderate to severe BPD in our study was 72%, which is higher than the national Taiwanese average.22 The higher incidence of moderate to severe BPD at our institution might be partially explained by the tertiary nature of our hospital which serves as a referral center to over 30 smaller clinics often in rural areas in southern Taiwan. A portion of our patients may have had inadequate prenatal care, were born emergently, or did not receive antenatal steroids before delivery; this is reflected in the high incidence of low Apgar scores ( 6) at 1 and 5 minutes of life in this study. Although the four
patients who died after the data collection period had ended were not included in our analysis, cause of death was not related to the development of BPD, the main endpoint of this study. Further data regarding the fluid balance of these four patients shows large discrepancies in net fluid intake ranging from 54.27 to 151.76 mL/kg (►Table 1). This seems to suggest the amount of net fluid intake was not related to mortality. No data on electrolytes, serum creatinine or blood urea nitrogen was collected for this study; data that would also reflect fluid and electrolyte balance in the first few days of life. In conclusion, we present a retrospective study that shows an association between increased fluid intake in the first 4 days of life and the development of severe BPD. Higher fluid intake remained an independent risk even after accounting for differences in gestational age, birth weight, and urine output. Moderate fluid restriction may prove to be beneficial in the prevention of severe BPD in VLBWI. More randomized prospective trials are needed to elucidate the effect of fluids on the development of BPD in the future.
Table 6 Causes of increased fluid intake from days 2 to 4 TFI < 345 (mL/kg) mean SD (N¼ 46)
TFI 345 (mL/kg) mean SD (N¼29)
p-Value
Days of hypotension
1.478 1.709
1.931 1.751
0.272
Days of phototherapy
2.000 1.054
2.410 0.867
0.081
Total amount of transfusions (mL/kg)
15.280 20.24
21.252 24.162
0.252
Packed RBC (mL/kg)
3.271 7.154
3.957 8.652
0.710
Fresh frozen plasma (mL/kg)
11.569 15.419
15.288 16.437
0.325
Platelets (mL/kg)
0.440 2.100
2.010 6.715
0.232
Abbreviations: RBC, red blood cell; SD, standard deviation; TFI, total fluid intake. American Journal of Perinatology
Vol. 32
No. 2/2015
Downloaded by: University of California, San Francisco. Copyrighted material.
160
Severe Bronchopulmonary Dysplasia and Fluid Intake
13 Ehrenkranz RA, Walsh MC, Vohr BR, et al; National Institutes of
1 Langston C, Kida K, Reed M, Thurlbeck WM. Human lung growth in
2 3 4 5
6
7
8
9
10
11
12
late gestation and in the neonate. Am Rev Respir Dis 1984;129(4): 607–613 Baraldi E, Filippone M. Chronic lung disease after premature birth. N Engl J Med 2007;357(19):1946–1955 Oh W. Fluid and electrolyte management of very low birth weight infants. Pediatr Neonatol 2012;53(6):329–333 Arant BS Jr. Developmental patterns of renal functional maturation compared in the human neonate. J Pediatr 1978;92(5):705–712 Ross B, Cowett RM, Oh W. Renal functions of low birth weight infants during the first two months of life. Pediatr Res 1977; 11(11):1162–1164 Stonestreet BS, Bell EF, Warburton D, Oh W. Renal response in lowbirth-weight neonates. Results of prolonged intake of two different amounts of fluid and sodium. Am J Dis Child 1983;137(3):215–219 Bell EF, Acarregui MJ. Restricted versus liberal water intake for preventing morbidity and mortality in preterm infants. Cochrane Database Syst Rev 2008; ((1):CD000503 Bell EF, Warburton D, Stonestreet BS, Oh W. Effect of fluid administration on the development of symptomatic patent ductus arteriosus and congestive heart failure in premature infants. N Engl J Med 1980;302(11):598–604 Lorenz JM, Kleinman LI, Kotagal UR, Reller MD. Water balance in very low-birth-weight infants: relationship to water and sodium intake and effect on outcome. J Pediatr 1982;101(3):423–432 Smith VC, Zupancic JA, McCormick MC, et al. Trends in severe bronchopulmonary dysplasia rates between 1994 and 2002. J Pediatr 2005;146(4):469–473 Shah PS, Sankaran K, Aziz K, et al; Canadian Neonatal Network. Outcomes of preterm infants
155
Severe Bronchopulmonary Dysplasia is Associated with Higher Fluid Intake in Very Low-Birth-Weight Infants: A Retrospective Study Mindy Ming-Huey Guo, MD1 Ching-Hung Chung, PhD2 Hsin-Chun Huang, MD1 Mei-Yung Chung, MD1,3
Hospital, Chang Gung University College of Medicine, Taiwan 2 Department of Dental Laboratory Technology, Min-Hwei College of Health Care Management, Taiwan 3 Department of Respiratory care, Chang Gung University of Science and Technology, Chiayi Campus, Taiwan
Chih-Cheng Chen, MD1
Address for correspondence Mei-Yung Chung, MD, Department of Pediatrics, Kaohsiung Chang Gung Memorial Hospital, No. 123, Ta Pei Road, Niao Sung District, Kaohsiung City, Taiwan (e-mail: [email protected]).
Am J Perinatol 2015;32:155–162.
Abstract
Keywords
► bronchopulmonary dysplasia ► very low-birth-weight infants ► fluid intake ► patent ductus arteriosus ► pulmonary hemorrhage
Objective This study aims to investigate the association between fluid intake in the first 4 days of life and the subsequent severity of bronchopulmonary dysplasia (BPD) in very low-birth-weight infants (VLBWI). Study Design A retrospective chart review of 75 infants with a gestational age of less than 32 weeks and a birth weight of < 1,500 g was performed. Demographic, clinical data, associated maternal risk factors, and amount of fluid received in the first 4 days of life were analyzed. Results Severe BPD was associated with a lower gestational age (27.04 2.073 wks vs. 28.70 1.706 wks, p¼0.001), lower birth weight (981.44 244.54 vs. 1,199.63 165.39 g, p < 0.001), use of surfactant (91.7 vs. 63%, p¼0.002), patent ductus arteriousus (PDA) (70.8 vs. 37%, p¼0.004), pulmonary hemorrhage (14.6 vs. 0%, p¼0.045), and more fluids received from the 2nd to 4th days of life (346.44 42.38 mL/kg vs. 323.91 27.62 mL/kg, p¼0.007). A cut off point of 345 mL/kg of fluids from the 2nd to 4th days of life was selected using receiver operating characteristic curve analysis, and remained a significant risk factor even after multiple logistic regression analysis. Conclusion Our findings demonstrate that VLBWI who received higher fluid intake from the 2nd to 4th days of life are at an increased risk of developing severe BPD.
Bronchopulmonary dysplasia (BPD) is a leading cause of respiratory illness in premature infants, resulting in prolonged hospitalization and oxygen dependency. Premature infants, especially those less than 32 weeks of gestational age, are at the highest risk of developing BPD, due in part to the fact that saccular development of the lung occurs from the 23 to
32 weeks of gestation.1 Not only is the premature lung more susceptible to damage from mechanical and oxidative injury, inadequate fluid clearance may also play a role in the development of BPD.1,2 Optimal fluid intake in the preterm infant is dependent on a myriad of factors, including gestational age, insensible
received September 24, 2013 accepted after revision April 14, 2014 published online June 10, 2014
Copyright © 2015 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662.
DOI http://dx.doi.org/ 10.1055/s-0034-1376393. ISSN 0735-1631.
Downloaded by: University of California, San Francisco. Copyrighted material.
1 Department of Pediatrics, Kaohsiung Chang Gung Memorial
Feng-Shun Chen, MD1
Severe Bronchopulmonary Dysplasia and Fluid Intake
Guo et al.
water losses through a premature epithelial barrier, urine, and stool.3 Very low-birth-weight infants (VLBWI) are more susceptible to fluctuations in iatrogenic fluids because of an inability of the immature kidney to excrete excess water and sodium.4,5 Excessive fluids in the first few days of life may not only impede the natural course of weight loss after birth6 but may also result in an increased risk of patent ductus arteriosus (PDA) and interstitial pulmonary edema. Previous prospective trials have shown that fluid restriction is related to a decreased incidence of PDA and necrotizing enterocolitis, and an inconclusive trend toward improved outcomes in patients with BPD.7 Because the majority of these studies were performed before the advent of widespread surfactant therapy,8,9 it is unclear if these results may be extrapolated to current clinical settings. It is interesting to note that despite recent advances in neonatal care, the incidence of BPD has remained static for the past two decades.10,11 This study was designed with the intention to investigate the association between fluid intake in the first 4 days of life and the subsequent incidence of severe BPD in VLBWI.
Patients and Methods Subjects All infants born at a large tertiary medical center in southern Taiwan within a 2 year period from October 2008 to September 2010 were selected for this study. Inclusion criteria were a gestational age of less than 32 weeks and a birth weight of < 1,500 g. Patients who died before discharge or had incomplete medical records were excluded from this study. BPD was defined as use of supplemental oxygen for more than 28 days of life, and further stratified as mild, moderate, or severe according to the NICHD (National Institute of Child Health and Human Development) 2001 consensus.12 The study population was then split into two groups, those with severe BPD, and those with mild, moderate, or no BPD. The NICHD 2001 consensus criteria was chosen over older definitions of BPD, such as oxygen use at 36 weeks gestational age, due to its better reflection of long-term outcomes, and ability to identify patients with differing degrees of BPD severity.13
throughout the course of hospitalization. Surfactant was administered via endotracheal tube within the first 2 hours of life if there was marked respiratory distress at birth, oxygen supplementation of more than 40% initially, CO2 retention on blood gas analysis, or if characteristic atelectasis, and air bronchograms indicating respiratory distress syndrome were seen on chest X-ray.
Evaluation of Fluid Management As an overall rule, patients were given around 70 to 90 mL/kg/ d on the 1st day of life. Fluid intake, in the form of enteral or parental fluids, was increased gradually on a day to day basis according to clinical condition or physician preference with a maximum of approximately 150 mL/kg/d. More fluids were given if patients were hypotensive, required blood product transfusions, or if they received phototherapy. Daily fluid intake (parental and enteral fluids), urine output, and body weight changes were recorded from the 2nd to the 4th days of life. Only the first 4 days were included in our study as larger differences in amount of fluid intake were more noticeable during this period. Body weight was measured at the same time daily when the infant was fully unclothed on an electronic scale. Two repeat measurements were performed and an average of these measurements was recorded. Both daily fluid intake and body weight changes were represented as a percentage of birth weight. The 1st day of life was excluded due to different times of birth between patients. Cumulative amount of fluid received from the 2nd to 4th days of life was calculated to account for daily fluctuations, and was regarded as a more accurate measure of the total amount of fluids received by each patient. Factors that may result in increased fluid intake were also taken into account, namely, duration of clinically significant hypotension, phototherapy, and amount of blood product transfusions. Symptomatic hypotension was defined as having a mean blood pressure below gestational age that resulted in the need for inotropic agents or fluid boluses in the first 4 days of life. Patients on phototherapy were given 10 to 30% more fluids depending on the intensity of phototherapy. Blood products were given to patients with hemoglobin level of less than 10 mg/dL, or showed symptoms and signs of bleeding tendency.
Statistical Analysis Data Collection A retrospective chart review was performed and demographic data including gender, gestational age, birth weight, parity, and whether the patient was appropriate or small for gestational age was collected. Maternal chart records were reviewed for associated risk factors including the presence of chorioamnionitis, use of antenatal steroids, and history of maternal hypertension or preeclampsia. Patients with PDA were identified by physical examination and confirmed via two-dimensional cardiac Doppler examination within the first 2 weeks of life. Indomethacin prophylaxis was not given on a routine basis during the study period. Pulmonary hemorrhage was determined clinically in patients with persistent fresh bleeding from the endotracheal tube with characteristic changes on chest radiography. Bacterial sepsis was defined by the presence of any positive blood cultures American Journal of Perinatology
Vol. 32
No. 2/2015
SPSS software version 19 (IBM Corp., Armonk, NY) was used for data analysis. Mean and standard deviation was calculated for all continuous variables and compared using student’s ttest. Categorical variables were compared using chi-square tests. A p-value of less than 0.05 was considered statistically significant. ROC curve analysis was performed to determine a cut off value for the cumulative amount of fluids received that was associated with severe BPD. Multivariate logistic regression of all significant variables on univariate analysis was performed by using backward Wald.
Results Demographic Data During our 2 year study period, 108 patients met our inclusion criteria, 8 were excluded due to death before discharge,
Downloaded by: University of California, San Francisco. Copyrighted material.
156
25 were excluded due to incomplete chart records, and the remaining 75 patients were enrolled for analysis. Of the eight patients who died during the study period, four of them died within the first few days of life and could not provide adequate data regarding fluid intake. The other four patients died of the following causes: Candida fungemia with brain involvement, necrotizing enterocolitis with septic shock, pulmonary hypoplasia secondary to oligohydramnios during gestation, and recurrent pneumonia. While these patients survived past the first 4 days of life, causes of death were not related to the development of severe BPD, the main endpoint of our study, and therefore were excluded from analysis. Demographic, fluid balance and data regarding body weight changes of these four patients are included in ►Table 1. Of the 75 patients included in this review, 48 were found to have severe BPD; in comparison, 19 patients were found to have mild BPD, 6 were found to have moderate BPD, and only 2 were found to have no BPD at all according to NIHCD criteria. Those with severe BPD had a lower mean gestational age at birth (27.04 2.073 wks vs. 28.70 1.706 wks, p¼0.001) and less mean birth weight (981.44 244.54 g vs. 1199.63 165.39 g, p < 0.001). Patients with severe BPD were also more likely to be male, appropriate for gestational age, and less likely to be the results of a single birth, although these three factors were not statistically significant (►Table 2).
Prenatal and Postnatal Factors A review of maternal records showed that those with severe BPD had similar rates of maternal chorioamnionitis, antenatal steroid use, and maternal hypertension when compared with those without severe BPD (►Table 2). In terms of postnatal factors, severe BPD was associated with a history of receiving surfactant (91.7 vs. 63%, p¼0.002), presence of PDA (70.8 vs. 37%, p¼0.004), and pulmonary hemorrhage (14.6 vs. 0%, p¼0.045). Apgar score at 1 and 5 minutes after birth, amount of FIO2 used at 24 hours of life, and rate of bacteremia during course of hospitalization were statistically similar between the two groups (►Table 2).
Total Fluid Intake and Body Weight Changes in the First Few Days of Life On average, patients with severe BPD received more fluids from the 2nd to 4th days of life (346.44 42.38 mL/kg vs. 323.91 27.62 mL/kg, p¼0.007). Cumulative net intake was calculated as total cumulative fluid intake minus total cumu-
Guo et al.
lative urine output. Urine output in our study subjects was highly variable (285.22 55.11cc/kg VS 269.39 71.49), with large SDs with no significant difference between those with severe BPD and those without. Net fluid intake (which was calculated as total fluid intake minus urine output) was also higher in patients with severe BPD (61.22 63.06 cc/kg VS 54.51 59.84 cc/kg), but was not statistically significant. However, both patients with and without severe BPD had similar changes in body weight within the first 4 days of life (►Table 3). ROC curve analysis was performed to determine a cut off value for the cumulative amount of fluids received from the 2nd to 4th days of life that was associated with severe BPD. A value of more than 345 mL/kg from the 2nd to 4th days of life was selected due to highest average of sensitivity and specificity ([sensitivity þ specificity]/ 2¼0.6678) (►Fig. 1). Patients who received more than 345 mL/kg of total fluids from the 2nd to 4th days of life were around 6.25 times more likely to develop severe BPD (►Table 4).
Multiple Logistic Regression Factors that were associated with severe BPD on univariate analysis were then subjected to multivariate logistic regression (►Table 5). Both gestational age and a history of pulmonary hemorrhage were not found to be significant enough to be included in the regression model. Two factors were found to be consistently significant after multiple logistic regression: presence of a PDA (p¼0.029), and receiving more than 345 mL/kg of fluids from the 2nd to 4th days of life (p¼0.013) (►Table 5). Due to the retrospective nature of our study, it is possible that increased fluid intake within the first few days of life could be an indicator of disease severity. However, patients who received more than 345 mL/kg of fluids from the 2nd to 4th days of life had a similar duration of hypotension, phototherapy, and similar amounts of blood product transfusions (►Table 6). This suggests that increased fluid intake of more than 345 mL/kg from the 2nd to 4th days of life did not appear to be the result of fluid boluses or increased fluid intake due to hypotension, phototherapy, or blood transfusions.
Discussion In our retrospective study of 75 VLBWI, increased fluid intake within the first 4 days of life was associated with the development of severe BPD, and remained a significant risk
Table 1 Demographic and fluid balance data of patients who died Patient
a
Gestational age (wks)
Birth weight (g)
Cause of death
Net intake (mL/kg)a
Weight changes (% of birth weight)a
1
24
725
Fungemia with Candida albicans
86.89
7.58621
2
29
1,060
Necrotizing enterocolitis
56.60
6.13208
3
26
750
Pneumonia, Stenotrophomonas
54.27
7.33333
4
25
765
Pulmonary hypoplasia
151.76
8.49673
Cumulative amount from the 2nd to 4th days of life. American Journal of Perinatology
Vol. 32
No. 2/2015
157
Downloaded by: University of California, San Francisco. Copyrighted material.
Severe Bronchopulmonary Dysplasia and Fluid Intake
Severe Bronchopulmonary Dysplasia and Fluid Intake
Guo et al.
Table 2 Demographic, prenatal, and postnatal factors Nonsevere BPD (N¼27)
Severe BPD (N¼48)
p-Value
Gestational age (wks)
28.70 1.706a
27.04 2.073a
0.001b
Birth weight (g)
1,199.63 165.39a
981.44 244.54a
0.000b
Male gender
11 (40.7%)
30 (62.5%)
0.069
SGA
5 (18.5%)
5 (10.4%)
0.48
Single
21 (77.8%)
33 (68.8%)
0.171
Twin
6 (22.2%)
9 (18.8%)
Triplet
0 (0%)
6 (12.5%)
8 (29.6%)
14 (29.2%)
0.966
Antenatal steroids
20 (74.1%)
38 (79.2%)
0.613
Maternal hypertension/preeclampsia
6 (22.2%)
4 (8.3%)
0.154
0.110
Demographic data
Parity
Prenatal factors Maternal chorioamnionitis
Postnatal factors Apgar score 1 min 0–3
6 (22.2%)
19 (39.6%)
4–6
12 (44.4%)
22 (45.8%)
7–10
9 (33.3%)
7 (14.6%)
0–3
3 (11.1%)
6 (12.5%)
4–6
5 (18.5%)
15 (31.3%)
7–10
19 (70.4%)
Apgar score 5 min 0.440
27 (56.3%) 0.498 0.192a
0.530
17 (63%)
44 (91.7%)
0.002b
PDA status
10 (37%)
34 (70.8%)
0.004b
Pulmonary hemorrhage
0 (0%)
7 (14.6%)
0.045b
Late-onset sepsis
3 (11.1%)
8 (16.7%)
0.736
FIO2 at 24 h
0.414 0.151
Surfactant use
a
Abbreviations: BPD, bronchopulmonary dysplasia; PDA, patent ductus arteriosus; SGA, small for gestational age. a Data for gestational age, birth weight, and FIO2 at 24 h are presented as mean SD. These were analyzed by student’s t-test and all other factors were analyzed by chi-square test. b Statistically significant difference.
factor even after correcting for birth weight and gestational age. Net fluid intake, which was calculated as total fluid intake minus urine output was also higher in patients with severe but was not statistically significant. This may be due to the large differences in urine output or insensible water losses, both which are highly variable factors. While some previous studies have demonstrated a related decrease in postnatal weight loss14,15; in our study, patients with severe and nonsevere BPD had similar amounts of postnatal body weight loss. The lack of differences in body weight loss between the two groups may have been due to the relatively smaller differences in fluid intake when compared with similar randomized controlled trials in the past.7 A smaller difference between fluid intake between the two study groups, coupled with a smaller sample size may have American Journal of Perinatology
Vol. 32
No. 2/2015
resulted in changes in body weight either not large enough to be detected or were not statistically significant. It has been postulated that extraneous fluid intake in the first few days of life impedes with the physiological contraction of extracellular fluid and diuresis that occurs after birth.16 Overall our study shows that increased fluid intake in the first 4 days of life is associated with the development of severe BPD, and remains an independent risk factor even after factors, such as gestational age and birth weight are taken into consideration. Most similar studies examining the relationship between fluid intake and the development of BPD have been published over 10 years ago.16 In a recent retrospective review of 1,382 extremely low-birth-weight infants by Oh et al published in 2005, BPD was found to be associated with higher fluid intake
Downloaded by: University of California, San Francisco. Copyrighted material.
158
Severe Bronchopulmonary Dysplasia and Fluid Intake
Guo et al.
159
Table 3 Total cumulative fluid intake and body weight loss Nonsevere BPD mean SD (N¼27)
Severe BPD mean SD (N¼48)
p-Value
Total cumulative fluid intake (mL/kg) Day of life 2
94.58 16.26
102 13.81
0.027a
3
204.04 21.53
217.58 21.53
0.020a
4
323.91 27.62
346.44 42.38
0.007a
82.92 33.39
0.860
Total cumulative urine output (mL/kg) 2
84.24 25.37
3
176.73 51.92
188.15 43.93
0.315
4
269.39 71.49
285.22 55.11
0.288
Total cumulative net intake (mL/kg) Day of life 2
10.33 27.69
19.65 33.67
0.225
3
27.29 51.64
29.42 49.77
0.861
4
54.51 59.84
61.22 63.06
0.654
0.911
Cumulative weight change as% of birth weight Day of life 2
1.01 2.12
1.08 3.27
3
3.33 4.04
3.81 6.13
0.718
4
5.69 5.33
6.28 13.85
0.833
Abbreviations: BPD, bronchopulmonary dysplasia; SD, standard deviation. a Statistically significant difference.
and less postnatal weight loss.14 This phenomenon was also seen in a case-control study of 223 infants by Van Marter et al, who demonstrated an association between greater amounts of fluids and net weight gain in the first 4 days of life and the
Fig. 1 ROC curve analysis was performed to determine a cut off value for the cumulative amount of fluids received from the 2nd to 4th days of life that was associated with severe BPD. A value of more than 345 mL/kg from the 2nd to 4th days of life was selected due to highest average of sensitivity and specificity ([sensitivity þ specificity]/ 2¼0.6678). BPD, bronchopulmonary dysplasia; ROC, receiver operating characteristic.
development of BPD and PDA. The authors also found that colloid, but not crystalloid fluids, had a dose-related relationship with higher ventilator settings such as the fraction of inspired oxygen, peak inspiratory pressure, positive endexpiratory pressure, and ventilator rates.15 It is possible that pulmonary edema not only interferes with alveolar gas exchange, but may also affect the function of surfactant, leading to decreased lung compliance necessitating higher ventilator settings and associated barotrauma. Another factor linked to the development of severe BPD in our study included a confirmed PDA. According to the results of several animal studies, the premature lung is ill equipped to handle an increase in precapillary arterial pressure,17 leading to excess fluids or blood flow to be shifted more readily downstream to small capillary vessels.18 It stands to reason that significant left to right shunting through a PDA would increase pulmonary hydrostatic capillary pressures leading to increased pulmonary edema and decreased lung compliance. To date, there has been no conclusive evidence from randomized controlled trials that links the presence of a PDA to the development of severe BPD.19 In addition, our study patients who received surfactant were also more likely to develop severe BPD, although it was not found to be statistically significant after multiple logistic regression. While early rescue surfactant therapy within the first 2 hours of life has been linked to a reduction in the American Journal of Perinatology
Vol. 32
No. 2/2015
Downloaded by: University of California, San Francisco. Copyrighted material.
Day of life
Severe Bronchopulmonary Dysplasia and Fluid Intake
Guo et al.
Table 4 Cut off point 345 (mL/kg) Days 2–4 fluids
Nonsevere BPD (N¼27)
Severe BPD (N¼48)
p-Value 0.001
345 (mL/kg)
4
25
< 345 (mL/kg)
23
23
Abbreviation: BPD, bronchopulmonary dysplasia. Note: Odds ratio: 6.25.
Table 5 Multiple logistic regression analysis p-Value
a
Odds ratio
95% Confidence interval
0.997
0.994–1.000
Gestational age
NSa
Birth weight
0.052
Surfactant use
0.084
3.872
0.833–17.985
Patent ductus arteriosus
0.029
3.908
1.151–13.269
Pulmonary hemorrhage
NSa
Total fluid intake 345 (mL/kg)b
0.013
5.671
1.435–22.416
Not significant enough to be included in the regression model. Total fluid intake345 (mL/kg) in the first 2nd to 4th days of life.
b
severity of BPD later in life,20 patients who received surfactant in our study were likely to be sicker at birth, with marked respiratory distress, characteristic chest radiographic findings or require a FIO2 of more than 40%. In patients with respiratory distress syndrome, the effect of excessive fluids and resulting pulmonary edema may further inhibit surfactant function in an already compromised lung.19,21 Limitations of this study include the retrospective nature of the study design, and a relatively small sample size. Of note, the incidence of moderate to severe BPD in our study was 72%, which is higher than the national Taiwanese average.22 The higher incidence of moderate to severe BPD at our institution might be partially explained by the tertiary nature of our hospital which serves as a referral center to over 30 smaller clinics often in rural areas in southern Taiwan. A portion of our patients may have had inadequate prenatal care, were born emergently, or did not receive antenatal steroids before delivery; this is reflected in the high incidence of low Apgar scores ( 6) at 1 and 5 minutes of life in this study. Although the four
patients who died after the data collection period had ended were not included in our analysis, cause of death was not related to the development of BPD, the main endpoint of this study. Further data regarding the fluid balance of these four patients shows large discrepancies in net fluid intake ranging from 54.27 to 151.76 mL/kg (►Table 1). This seems to suggest the amount of net fluid intake was not related to mortality. No data on electrolytes, serum creatinine or blood urea nitrogen was collected for this study; data that would also reflect fluid and electrolyte balance in the first few days of life. In conclusion, we present a retrospective study that shows an association between increased fluid intake in the first 4 days of life and the development of severe BPD. Higher fluid intake remained an independent risk even after accounting for differences in gestational age, birth weight, and urine output. Moderate fluid restriction may prove to be beneficial in the prevention of severe BPD in VLBWI. More randomized prospective trials are needed to elucidate the effect of fluids on the development of BPD in the future.
Table 6 Causes of increased fluid intake from days 2 to 4 TFI < 345 (mL/kg) mean SD (N¼ 46)
TFI 345 (mL/kg) mean SD (N¼29)
p-Value
Days of hypotension
1.478 1.709
1.931 1.751
0.272
Days of phototherapy
2.000 1.054
2.410 0.867
0.081
Total amount of transfusions (mL/kg)
15.280 20.24
21.252 24.162
0.252
Packed RBC (mL/kg)
3.271 7.154
3.957 8.652
0.710
Fresh frozen plasma (mL/kg)
11.569 15.419
15.288 16.437
0.325
Platelets (mL/kg)
0.440 2.100
2.010 6.715
0.232
Abbreviations: RBC, red blood cell; SD, standard deviation; TFI, total fluid intake. American Journal of Perinatology
Vol. 32
No. 2/2015
Downloaded by: University of California, San Francisco. Copyrighted material.
160
Severe Bronchopulmonary Dysplasia and Fluid Intake
13 Ehrenkranz RA, Walsh MC, Vohr BR, et al; National Institutes of
1 Langston C, Kida K, Reed M, Thurlbeck WM. Human lung growth in
2 3 4 5
6
7
8
9
10
11
12
late gestation and in the neonate. Am Rev Respir Dis 1984;129(4): 607–613 Baraldi E, Filippone M. Chronic lung disease after premature birth. N Engl J Med 2007;357(19):1946–1955 Oh W. Fluid and electrolyte management of very low birth weight infants. Pediatr Neonatol 2012;53(6):329–333 Arant BS Jr. Developmental patterns of renal functional maturation compared in the human neonate. J Pediatr 1978;92(5):705–712 Ross B, Cowett RM, Oh W. Renal functions of low birth weight infants during the first two months of life. Pediatr Res 1977; 11(11):1162–1164 Stonestreet BS, Bell EF, Warburton D, Oh W. Renal response in lowbirth-weight neonates. Results of prolonged intake of two different amounts of fluid and sodium. Am J Dis Child 1983;137(3):215–219 Bell EF, Acarregui MJ. Restricted versus liberal water intake for preventing morbidity and mortality in preterm infants. Cochrane Database Syst Rev 2008; ((1):CD000503 Bell EF, Warburton D, Stonestreet BS, Oh W. Effect of fluid administration on the development of symptomatic patent ductus arteriosus and congestive heart failure in premature infants. N Engl J Med 1980;302(11):598–604 Lorenz JM, Kleinman LI, Kotagal UR, Reller MD. Water balance in very low-birth-weight infants: relationship to water and sodium intake and effect on outcome. J Pediatr 1982;101(3):423–432 Smith VC, Zupancic JA, McCormick MC, et al. Trends in severe bronchopulmonary dysplasia rates between 1994 and 2002. J Pediatr 2005;146(4):469–473 Shah PS, Sankaran K, Aziz K, et al; Canadian Neonatal Network. Outcomes of preterm infants
