Published online February 1, 2008
PEDIATRICS
Vol. 121
No. 2
February 2008, pp.
e344-e349
(doi:10.1542/peds.2006-3668)
Effect of a Short Course of Prednisolone in Infants With Oxygen-Dependent Bronchopulmonary Dysplasia
Anita Bhandari, MDa,
Craig M. Schramm, MDa,
Claudia Kimble, APRNb,
Mariann Pappagallo, MDb and
Naveed Hussain, MDb
a Division of Pediatric Pulmonology, Connecticut Children's Medical Center, Hartford, Connecticut
b Division of Neonatology, Department of Pediatrics, University of Connecticut School of Medicine, Farmington, Connecticut
 |
ABSTRACT
|
|---|
OBJECTIVE. The purpose of this work was to determine whether oral prednisolone is effective in weaning infants with bronchopulmonary dysplasia, after 36 weeks' postmenstrual age, off supplemental oxygen and to identify factors associated with successful weaning.
METHODS. Data were abstracted from a standardized prospectively collected database at the John Dempsey Hospital NICU. Logistic regression and receiver operating curve analyses were used.
RESULTS. Of 385 infants, 131 (34%) received oral prednisolone and 254 (66%) did not. There was no significant difference in race, gender, birth weight, or gestational age between the groups receiving and not receiving oral prednisolone. Infants in the oral prednisolone group were more likely to have received previous dexamethasone therapy, had longer duration of mechanical ventilation, had longer length of hospital stay, and were more likely to be discharged from the hospital on oxygen. Of those in the oral prednisolone group, 63% responded to treatment. Pulmonary acuity score and PCO2 were the only parameters that remained significant on multiple logistic regression analyses. The oral prednisolone-responsive group had a lower pulmonary acuity score compared with the oral prednisolone-nonresponsive group. A pulmonary acuity score value of 
0.5 had a sensitivity of 20% and specificity of 97.4%, with positive and negative predictive values of 94.1% and 42.1%, respectively. Capillary PCO2 values were significantly lower in the oral prednisolone-responsive group compared with the oral prednisolone-nonresponsive group. In predicting a successful response to oral prednisolone, a capillary PCO2 value of <48.5 mmHg had a sensitivity of 50% and specificity of 89.7%, with positive and negative predictive values of 89.1% and 51.8%, respectively.
CONCLUSIONS. Oral prednisolone therapy is effective in weaning off supplemental oxygen in a postterm infant with oxygen-dependent bronchopulmonary dysplasia who has a pulmonary acuity score of <0.5 and PCO2 of <48.5 mmHg. In addition, if a single course of prednisolone fails, there is no clear benefit of using multiple courses.
Key Words: BPD • prednisolone • oxygen therapy • outcome
Abbreviations: BPD—bronchopulmonary dysplasia • PMA—postmenstrual age • OP—oral prednisolone • No-OP—subjects who did not receive oral prednisolone • OP-R—subjects who responded to oral prednisolone therapy • OP-NR—subjects who did not respond to oral prednisolone therapy • PAS—pulmonary acuity score • FIO2—fraction of inspired oxygen • ROC—receiver operating curve • CI—confidence interval
Bronchopulmonary dysplasia (BPD) is defined as oxygen need at 36 weeks' postmenstrual age (PMA)1 with or without the use of respiratory support and with or without characteristic radiographic changes at 36 weeks' PMA.2 BPD affects 
30% of premature infants with birth weights of <1000 g and is uncommon in infants born at >30 weeks of gestation or >1200 g.3,4 Because inflammation plays a prominent role in the pathogenesis of BPD, intravenous and oral steroids have been used as anti-inflammatory agents to alter the course of lung disease in premature infants. Within this patient population, dexamethasone has been well studied and has been found effective in weaning infants off mechanical ventilation when given moderately early (7–14 days of age) or late (14–28 days of age) in the initial nursery course.5–7 Despite this effect, dexamethasone therapy has not been shown to reduce the total days of hospitalization, duration of supplemental oxygen therapy, or incidence of BPD.5
Because effectiveness of moderately early (7–14 days) and late (14–28 days) dexamethasone therapy in improving pulmonary outcome is well established, at our center, clinical practitioners in both neonatology and pediatric pulmonology have developed the practice of trying a short course of oral steroids (5–10 days) in infants >36 weeks' PMA in an effort to wean infants with oxygen-dependent BPD off of supplemental oxygen. However, the effectiveness of this therapy has not been studied. Thus, the present study was designed to determine whether oral steroid therapy administered at >36 weeks' PMA was effective in weaning infants with BPD off of supplemental oxygen therapy and to identify factors associated with successful weaning. Our hypothesis was that oral prednisolone, when given after 36 weeks' PMA, would help infants with oxygen-dependent BPD wean off supplemental oxygen.
|
METHODS
|
|---|
Data for this retrospective study were abstracted from a standardized, prospectively collected database at the John Dempsey Hospital NICU (Medical Data Systems NIS 3.0, Philadelphia, PA). The following variables were obtained: use of dexamethasone, diuretics, and inhaled corticosteroids; duration of mechanical ventilation and supplemental oxygen use; use of prednisolone after 36 weeks' PMA; and capillary PCO2 at the time of the initiation of prednisolone therapy. Primary outcome of the study was discharge from the NICU without the need for supplemental oxygen. Approval from the institutional review board was obtained to review and publish the data.
Patients
All of the patients with BPD at John Dempsey Hospital NICU from 2000 to 2004 were included in the analysis. Information regarding gestational age, gender, birth weight, race, previous dexamethasone use, length of hospital stay, days on ventilator, and capillary PCO2 within 1 week of starting prednisolone was also gathered. Infants with chromosomal abnormalities and congenital heart disease were excluded from the analyses.
Definition of BPD and Use of Prednisolone
BPD was defined as oxygen dependence at 36 weeks' PMA.1 The subjects were divided into 2 groups, those who received oral prednisolone (OP) and those who did not (No-OP). Among the infants in the OP group, an additional distinction was made among those who responded to OP therapy (OP-R) and those who did not (OP-NR). Response to prednisolone therapy was defined as discharge from the hospital after successfully weaning off supplemental oxygen. Capillary PCO2 values were obtained within a week of starting prednisolone therapy. The dose of prednisolone used was 2 mg/kg per day orally divided twice per day for 5 days, then 1 mg/kg per dose orally daily for 3 days, and then 1 mg/kg per dose every other day for 3 doses. A taper was only done if the patients passed stress oximetry.8,9
Pulmonary Acuity Score
Pulmonary acuity score (PAS) was calculated according to Madan et al10 for all of the patients who received OP using the following equation: PAS = (FIO2) (support score) + medications score, where FIO2 is expressed as the fraction of inspired oxygen if the patient was on a ventilator, continuous positive airway pressure, or hood and as "effective FIO2" if a nasal cannula was used. The support score was 2.5 for ventilator, 1.5 for nasal or endotracheal continuous positive airway pressure, and 1.0 for nasal cannula flow or hood oxygen. Medication score was the sum of 0.20 for systemic steroids, 0.10 for diuretics or inhaled steroids, and 0.05 for methylxanthines or intermittent diuretics.
Outcome Measures
Response to prednisolone therapy was defined as successful weaning off from supplemental oxygen after therapy. Before discontinuation of oxygen therapy, every patient underwent stress oximetry as described by Hussain et al.9 Stress oximetry was performed by a trained respiratory therapist using a Nellcor pulse oximeter (N-20 PA, Nellcor Puritan Bennett, Pleasanton, CA) with an appropriate-sized probe placed on the palm or foot of the infant. Hemoglobin oxygen saturation, heart rate, color, and respiratory effort were measured. The behavioral state of the infant during the test was also noted. Infants were tested in the "quiet-awake" state and then stimulated to achieve an "awake-agitated" or "crying" state. Measurements were recorded both at the existing level of supplemental oxygen and in room air. If infants maintained acceptable oxygen saturation in room air, they were also tested while taking an oral feed. The duration of a complete test was 30 to 45 minutes. The criteria for passing stress oximetry included maintaining oxygen saturation of 
90% in room air while in the stressed state and while feeding, without a significant increase in the work of breathing or color change. Infants were weaned to room air if they passed 2 consecutive stress oximetry tests 1 week apart. The criteria for successful weaning was the ability to remain in room air without desaturations or increased work of breathing for 1 week and to maintain a minimum weight gain of 10 g per day over the period of that week.
Statistical Analysis
Statistical analyses were performed by using Student's t test for continuous variables and the 
2 test for nominal variables. Multiple logistic regression was applied to variables identified as having individual predictive significance (StatView 5.0 [SAS Institute, Inc, Cary, NC]). Receiver operating curve (ROC) analysis was applied to determine best predictive values (Prism 4 [GraphPad Software, San Diego, CA]), and areas under the ROC curves were compared according to the method of Hanley and McNeil.11 A P value of <.05 was used to determine statistical significance.
Sample-Size Calculation
We based our sample-size calculation on the following assumptions: (1) approximately one third of infants with BPD would go home on supplemental oxygen; (2) at least half of the infants ready to go home would have received a late course of OP in an attempt to wean them off oxygen; and (3) to be clinically useful, prednisolone would decrease the need for home O2 therapy by 30%. Under these assumptions, a sample size of 40 O2-dependent patients receiving prednisolone and 40 not receiving prednisolone was needed. If these 80 O2-dependent patients represented one third of the BPD patients getting ready for discharge, 240 records were necessary for review.
|
RESULTS
|
|---|
Demographic data of all of the infants with BPD is as shown in Table 1. The total number of infants studied was 385, of which 131 (34%) received OP and 254 (66%) did not (No-OP). There was no significant difference in race, gender, birth weight, or gestational age between the OP and No-OP groups. Patients in the OP group were more likely to have received previous dexamethasone therapy and had a longer duration of mechanical ventilation and a longer duration of hospital stay. All of the patients had undergone mechanical ventilation. No infants were on mechanical ventilation at the time of OP use. There were 10 deaths in the No-OP group and 3 deaths in the OP group.
All of the infants with BPD who received OP were on supplemental oxygen when it was administered. In contrast, most infants who did not receive prednisolone had already weaned off oxygen by the time they were being readied for discharge. Reasons for not giving prednisolone to the 37 patients in the No-OP group who were discharged on O2 (Table 1) could not be determined from our chart review. Among infants who received OP, approximately two thirds weaned off oxygen and one third did not. The OP-R group was not significantly different in terms of gestational age, birth weight, gender, race, or gestational age at the time of prednisolone use when compared with the OP-NR group (Table 2). With regard to previous therapies, the OP-R group had significantly fewer days on the ventilator when compared with the OP-NR group. All of the patients with BPD who were studied received diuretics, and 98% received albuterol. The majority of OP patients also received inhaled steroids, and a greater number of patients in the OP-NR group received inhaled steroids than in the OP-R group. When multiple regression analyses were applied, there were no significant differences between the groups in terms of previous therapies (Table 3). We also compared OP-R with OP-NR among infants who received single versus multiple courses of OP and found no significant improvement in response to multiple courses (12 of 88 received 2 courses; 2 of 88 received 3 courses; and 1 of 88 received 4 courses, typically 2 weeks apart with a range of 7 days to 4 weeks).
Capillary PCO2 values were significantly lower in the OP-R group as compared with the OP-NR group (P = .0006; Table 2) and remained significant on multiple regression analyses (P = .0028; Table 3). ROC analysis demonstrated moderate significance for PCO2 as a predictor of response to oral steroids with the area under the curve equaling 0.775 (95% confidence interval [CI]: 0.68–0.87) as shown in Fig 1. The optimum predictive cutoff point was found to be a PCO2 <48.5, with a likelihood ratio of 15.6. As a predictor of a patient's likelihood of weaning off supplemental oxygen, a capillary PCO2 value of 48.5 was found to have sensitivity of 50.0% and specificity of 89.7%. The ability of capillary PCO2 of <48.5 to predict the ability to respond to oral steroids (ie, positive predictive value of the test) was 89.1%. Conversely, the ability a capillary PCO2 of 48.5 to predict failure of the oral steroid treatment (ie, negative predictive value) was 51.8%. Concurrent capillary pH values were also obtained and were similar between the 2 groups (7.35 ± 0.03 in OP-NR and 7.37 ± 0.03 in OP-R). To address the metabolic component in the absence of direct bicarbonate data collection, we calculated serum bicarbonate values for all of the patients from their pH and PCO2 data, according to the Henderson-Hasselbach equation. Doing so, we found that the calculated HCO3– was higher in the OP-NR group than in the OP-R group (30.7 ± 0.7 vs 28.1 ± 0.5 mEq/L; P = .0019). However, only 2 patients (both from the OP-R group) had alkalotic pH values (7.45 and 7.46), suggesting that there was little primary metabolic alkalosis in the groups and that the higher HCO3– in the OP-NR patients was in compensation to their primary respiratory acidosis.
We used the PAS to evaluate pulmonary disease acuity at the time the patient received OP. Again, there was a significant difference between the 2 groups, with the OP-NR group having a higher PAS (mean: 0.88) as compared with the OP-R group (mean: 0.63). ROC analysis for PAS yielded an area under the curve equal to 0.714 (95% CI: 0.61–0.82). This area was not statistically different from that for the capillary PCO2 measurements (P = .36). The optimum PAS predictive point was 0.50, with a likelihood ratio of 7.8. As a predictor of our patient's likelihood of weaning off supplemental oxygen, a PAS value of 0.50 was found to have sensitivity of 20.0% and specificity of 97.4%. The positive and negative predictive values of a PAS of 0.50 were 94.1% and 42.1%, respectively.
|
DISCUSSION
|
|---|
The use of oral and intravenous steroids has been studied extensively in infants with BPD. There is little doubt that, when administered early and moderately early, they facilitate weaning from mechanical ventilation and extubation. The use of dexamethasone in preterm very low birth weight infants to prevent BPD has not been recommended because of a lack of long-term pulmonary benefits and clear increases in gastrointestinal and neurologic complications.5–7,12,13 Recent studies have shown shorter days on the ventilator with low-dose dexamethasone when administered after 1 week of life with no increase in gastrointestinal or other adverse effects14 and that the risk of developing neurologic complications may be modified by an infant's risk of chronic lung disease.15 Despite the plethora of literature on the use of dexamethasone in premature infants within a few of weeks of birth, the use of OP in patients with established BPD after 36 weeks' PMA has not been studied. At our institution, pulmonologists and neonatologists frequently use multiple short courses of prednisolone (5–10 days) to continue to wean infants with BPD off oxygen supplementation and mechanical ventilation. This study was, thus, undertaken to evaluate the efficacy of OP in weaning supplemental oxygen in premature infants with BPD after 36 weeks' PMA.
Our demographic data demonstrate that the patients who received prednisolone therapy had more severe lung disease than the ones who did not. Within the OP group, the OP-NR subjects also seemed sicker such that they had a greater number of days on the ventilator, longer length of stay, and higher capillary PCO2 and PAS values when compared with the OP-R group. Patients in the OP-NR group were also more likely to have had >1 course of OP, to have received inhaled steroids and dexamethasone, and to have higher capillary PCO2 values when compared with the OP-R group. A majority of OP patients received inhaled steroids, likely to maximize efforts to wean their oxygen. A greater number of patients in the OP-NR group received inhaled steroids than in OP-R group, possibly reflecting more refractory hypoxemia in the OP-NR patients. In addition, multiple courses of OP did not alter response to therapy. When multiple regression analyses were applied, capillary PCO2 and PAS at the time of the steroid course were the only parameters that were significantly different between the 2 groups of patients.
At our NICU we routinely obtain weekly capillary PCO2 values and stress oximetry in all of the oxygen-dependent infants with BPD. Using ROC analysis, a capillary PCO2 value of <48.5 at the time of prednisolone therapy was found to have a high positive predictive value for weaning off supplemental oxygen (ie, 88%). The negative predictive value of a PCO2 of >48.5 was only fair (50%). Capillary PCO2 is well accepted as a parameter for monitoring lung disease in infants with BPD. Recently, Kovesi et al16 reported that higher capillary PCO2 at the time of discharge from the NICU was associated with a greater likelihood of adverse outcome, including reintubation and rehospitalization, but these investigators were unable to find a reliable cutoff value for capillary PCO2 to predict such an adverse outcome. The higher capillary PCO2 values in OP-NR could not be attributed to diuretic-induced metabolic alkalosis, because capillary pH values were similar in the OP-R and OP-NR groups.
Madan et al10 reported that, when calculated near term through 3 months' corrected age, PAS proved useful in predicting pulmonary morbidity in patients with BPD. In their study, patients who were more likely to have significant pulmonary morbidity up to 3 months of corrected age had higher PAS (median: 0.48) when compared with those who did not (median: 0.38).10 Of interest, the value of 0.48 by Madan et al10 was very similar to our value of 0.50, which predicted responsiveness to OP. In our study, however, PCO2 proved to have a closer correlation with the response to prednisolone therapy than the PAS.
Because we used OP primarily to wean infants off supplemental oxygen, the criteria that are used for determining optimal oxygen saturation are important.17,18 Relatively few studies have been done on such criteria. One such study is from western Australia, which used an "air test" to determine the ability of an infant to maintain saturation at >80% without supplemental oxygen to ensure safe discharge. In this report the readmission rate during the study period was 60% to 64%.19 In our study, patients were discharged from the hospital with or without supplemental oxygen, based on stringent criteria requiring oxygen saturation levels of 85% at various levels of activity. With these criteria, readmission rates at our center have been <15% (unpublished data). The other reason why we use these stringent criteria for discharging patients home on oxygen is the scarcity of home care resources for these fragile infants in our region.
We attempted to create a clinical profile of the group of patients who are likely to respond to OP therapy at 36 weeks' PMA. We speculate that the patients who do not respond to prednisolone have more severe lung disease, as evidenced by higher PCO2 values and higher PAS score in the OP-NR group, and, therefore, could have greater lung inflammation, which results in poor response to prednisolone. Conversely, hypoxemia may be related to other factors causing an increase in the V-Q mismatch, such as a greater reduction in alveolar number or greater alveolar septal fibrosis and fixed airway obstruction. To the best of our knowledge there are no data available on the effect of prednisolone on lung development in the human infant at this age. Data in rats showed no deleterious effects on alveolar development when dexamethasone was given to 3- to 4-week-old rats,20 recognizing the fact that 3- to 4-week rat lung is an almost mature lung and not comparable to the lung of a former premature infant at 36 weeks' PMA.
Although this is the first study to evaluate the response to prednisolone therapy in infants with established BPD, it has certain limitations. This was a retrospective study. Prednisolone was not administered to every infant before being discharged on supplemental oxygen, and we were unable to ascertain from the chart review why it was not given. Although multivariate analyses are able to adjust for known measurable covariates for the primary outcome, significant unknown confounders may have affected these results. In addition, we realize that, although stress oximetry is a tool routinely used at our institution as reported previously, other centers may use different criteria both for purposes of weaning oxygen and discharging infants home on oxygen.8,9
|
CONCLUSIONS
|
|---|
This study provides evidence that OP therapy is effective only in select patients with BPD. The patient most likely to benefit from a short course of OP therapy is an infant with oxygen-dependent BPD who has a capillary PCO2 of <48.5 mmHg or a PAS score of <0.50. The majority of such infants weaned off oxygen after OP therapy. Although a brief course of OP may be considered in patients with higher capillary PCO2 levels or PAS values, there is only a 40% to 50% chance that this therapy will aid in oxygen weaning. In addition, if 1 course of OP fails to help wean off oxygen, there is no clear benefit of using multiple courses unless the PCO2 value has changed significantly in the interim. We do not recommend OP in all patients with oxygen-dependent BPD. It is hoped that these parameters will provide some guidance for the late use of OP in infants with BPD and that independent, prospective analysis will verify the effectiveness of this practice.
|
ACKNOWLEDGMENTS
|
|---|
We thank Vineet Bhandari, MD, DM, for critical review of the article.
|
FOOTNOTES
|
|---|
Accepted Jul 5, 2007.
Address correspondence to Anita Bhandari, MD, Division of Pediatric Pulmonology, Connecticut Children's Medical Center, 282 Washington St, Hartford, CT 06106. E-mail: abhanda{at}ccmckids.org
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES
|
|---|
- Shennan AT, Dunn MS, Ohlsson A, Lennox K, Hoskins EM. Abnormal pulmonary outcomes in premature infants: prediction from oxygen requirement in the neonatal period.
Pediatrics. 1988;82
(4):527
–532[Abstract/Free Full Text]
- Bancalari E, Gerhardt T. Bronchopulmonary dysplasia.
Pediatr Clin North Am. 1986;33
(1):1
–23[ISI][Medline]
- Bhandari A, Bhandari V. Pathogenesis, pathology and pathophysiology of pulmonary sequelae of bronchopulmonary dysplasia in premature infants.
Front Biosci. 2003;8
:e370
–e80[ISI][Medline]
- 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
- Halliday HL, Ehrenkranz RA, Doyle LW. Early postnatal (<96 hours) corticosteroids for preventing chronic lung disease in preterm infants.
Cochrane Database Syst Rev. 2003;
(1):CD001146
- Halliday HL, Ehrenkranz RA, Doyle LW. Moderately early (7–14 days) postnatal corticosteroids for preventing chronic lung disease in preterm infants.
Cochrane Database Syst Rev. 2003;
(1):CD001144
- Halliday HL, Ehrenkranz RA, Doyle LW. Delayed (>3 weeks) postnatal corticosteroids for chronic lung disease in preterm infants.
Cochrane Database Syst Rev. 2003;
(1):CD001145
- Awonuga M, Pappagallo M, Kimble C, Lapin C, Kresch M, Hussain N. "Stress oximetry": a test that predicts supplemental oxygen requirement in infants with bronchopulmonary dysplasia.
Pediatr Res. 2000;47
(4):351A[CrossRef]
- Hussain N, Kimble C, Pappagallo M. ‘Stress oximetry’: a useful test for predicting discontinuation of supplemental oxygen in infants with BPD.
Am J Respir Crit Care Med. 2002;165
(8):A645
- Madan A, Brozanski BS, Cole CH, Oden NL, Cohen G, Phelps DL. A pulmonary score for assessing the severity of neonatal chronic lung disease.
Pediatrics. 2005;115
(4). Available at: www.pediatrics.org/cgi/content/full/115/4/e450
- Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases.
Radiology. 1983;148
(3):839
–843[Abstract]
- American Academy of Pediatrics, Committee of Fetus and Newborn, Canadian Paediatric Society, Fetus and Newborn Committee. Postnatal corticosteroids to treat or prevent chronic lung disease in preterm infants.
Pediatrics. 2002;109
(2):330
–338[Abstract/Free Full Text]
- Thebaud B, Lacaze-Masmonteil T, Watterberg K. Postnatal glucocorticoids in very preterm infants: "the good, the bad, and the ugly"?
Pediatrics. 2001;107
(2):413
–415[Free Full Text]
- Doyle LW, Davis PG, Morley CJ, McPhee A, Carlin JB. Low-dose dexamethasone facilitates extubation among chronically ventilator-dependent infants: a multicenter, international, randomized, controlled trial.
Pediatrics. 2006;117
(1):75
–83[Abstract/Free Full Text]
- Doyle LW, Halliday HL, Ehrenkranz RA, Davis PG, Sinclair JC. Impact of postnatal systemic corticosteroids on mortality and cerebral palsy in preterm infants: effect modification by risk for chronic lung disease.
Pediatrics. 2005;115
(3):655
–661[Abstract/Free Full Text]
- Kovesi T, Abdurahman A, Blayney M. Elevated carbon dioxide tension as a predictor of subsequent adverse events in infants with bronchopulmonary dysplasia.
Lung. 2006;184
(1):7
–13[CrossRef][ISI][Medline]
- Ellsbury DL, Acarregui MJ, McGuinness GA, Eastman DL, Klein JM. Controversy surrounding the use of home oxygen for premature infants with bronchopulmonary dysplasia.
J Perinatol. 2004;24
(1):36
–40[CrossRef][Medline]
- Ellsbury DL, Klein JM. Oxygen-saturation targets in extremely preterm infants.
N Engl J Med. 2003;349(24)
:2362; author reply 2362
- Saletti A, Stick S, Doherty D, Simmer K. Home oxygen therapy after preterm birth in Western Australia.
J Paediatr Child Health. 2004;40(9–10)
:519
–523
- Blanco LN, Frank L. The formation of alveoli in rat lung during the third and fourth postnatal weeks: effect of hyperoxia, dexamethasone, and deferoxamine.
Pediatr Res. 1993;34
(3):334
–340[ISI][Medline]
PEDIATRICS (ISSN 1098-4275). ©2008 by the American Academy of Pediatrics
TOP
ABSTRACT
METHODS
