Calcium chloride in neonatal parenteral nutrition: A 15 year experience

Abstract

OBJECTIVE: The objective of this study was to determine if outcomes at our neonatal intensive care units (NICUs) since we began using calcium chloride (CaCl₂) as our preferred calcium additive in order to reduce aluminum (Al) exposure are within expected outcome ranges for NICUs in the U.S. where calcium gluconate in glass vials (CaGlu-Gl) has been the preferred additive.

STUDY DESIGN: A retrospective study of very low birth weight infants born between January 1, 2000 and December 31, 2014. Outcomes in two intensive care units (NICUs) using CaCl₂ were compared to all U.S. NICUs in the Vermont Oxford Network. Primary outcomes were chronic lung disease (CLD), percent requiring supplemental oxygen at 28 days, and mortality excluding early deaths (MEED). The incidence of IV infiltrates of all admissions to the study NICUs in 2013-2014 was compared to the literature.

RESULTS: The incidence of CLD and those requiring oxygen at 28 days were 24.0% vs 28.6% and 46.2% vs 51.8% for the study NICUs compared to all U.S. NICUs, respectively (both p < 0.0001). The MEED was 8.7% vs 10.3% (p < 0.002). All major morbidities were lower at the study NICUs. The incidence of infiltrates was lower than that in the literature.

CONCLUSION: The use of CaCl₂ was not associated with any detectable adverse effects. Calcium chloride appears to be a safe alternative to the use of CaGlu-Gl based upon studies of clinical outcomes.

Keywords

Aluminum calcium chloride calcium gluconate clinical outcomes parenteral nutrition

1 Introduction

In 1997 Bishop et al. reported that neonates who received parenteral nutrition (PN) solutions containing calcium gluconate in glass vials (CaGlu-Gl) compared to PN solutions containing calcium chloride (CaCl₂) had impaired neurological development associated with increased aluminum (Al) exposure from solutions containing CaGlu-Gl [1]. The high Al exposure from PN solutions containing CaGlu-Gl was also associated with reduced bone mass in adolescence [2]. Notably, there were no short-term or long-term adverse effects associated with the use of CaCl₂. Subsequently, CaCl₂ became the preferred calcium additive in Europe in the late 1990’s [1 , 4]. Aluminum is a contaminant introduced into many small and large volume parenteral products during the manufacturing process, including leaching of Al during sterilization of glass containers [5]. After calcium gluconate in plastic vials (CaGlu-Pl) became available in Europe it became the preferred additive due to increased solubility as well as decreased Al contamination [3]. This additive has not been available in the United States leaving CaCl₂ as the only calcium additive approved for use that limits Al intake [1, 6]. In the United States, since CaGlu-Gl and CaCl₂ have been the only calcium additives available for PN, CaGlu-Gl has been predominately used due to concerns related to the compatibility of CaCl₂ and phosphates that might result in adverse clinical outcomes [7, 8]. This has been the case in spite of studies showing that PN solutions containing CaCl₂ have Al concentrations of 4–6 μg/dL compared to concentrations of 23–45 μg/dL for CaGlu-Gl [1, 6].

There have been adverse outcomes associated with possible precipitates of calcium gluconate and phosphates in PN solutions in adults but not in neonates [8 –11]. All the fatal cases received PN solutions that were unfiltered and contained concentrations of calcium and phosphates that were incompatible or borderline compatible based upon visual methods for determining compatibility. Neonatal PN solutions are filtered however. The cause of death associated with calcium phosphate precipitates has been due to pulmonary embolization of precipitates. The smallest pulmonary capillaries in vivo are thought to be from 4–9 microns in diameter. The standard filter for neonatal PN is 0.22 microns. Some in the U.S. have expressed the opinion that CaCl₂ should never be used as the calcium additive in PN even though standard guidelines for compounding and filtration are followed due to a concern that adverse clinical outcomes may still occur due to the greater degree of ionization of calcium chloride compared to calcium gluconate [7].

We began using CaCl₂ as the calcium additive in neonatal PN in January 2000 based upon the randomized controlled trial (RCT) which showed worse neurodevelopmental outcome with the use of CaGlu-Gl [1]. In addition, the FDA has recommended limiting the Al intake due to contamination of PN solutions for preterm infants to <0.19 μmol/kg/day (<5 μg/kg/day) [12] which could only be done in the U.S. by using CaCl₂ as the calcium additive in neonatal PN solutions. We have followed patients prospectively since making this practice change and have not recognized any episodes of pulmonary embolism related to calcium phosphate precipitation in PN solutions. The purpose of this study was to determine if there were any adverse clinical outcomes associated with the use of CaCl₂ in neonatal PN solutions in our NICUs over a 15 year period. Since calcium gluconate has continued to be the preferred calcium additive in the U.S. due to its increased solubility in PN solutions containing phosphates compared to CaCl₂ [13], we compared outcomes at our neonatal intensive care units (NICUs) since we began using CaCl₂ as our preferred calcium additive to the outcomes for NICUs in the U.S. The study was not intended to provide further information on Al content or adverse outcomes related to Al content of PN solutions.

2 Methods

The study was approved by the institutional review boards for Providence Health and Services and Legacy Health Systems. Informed consent was not required. This was a retrospective, multicenter study that included live births and admissions of 401–1500 g birth weight or gestational age of 22 and 0/7 weeks to 29 and 6/7 weeks (VLBW infants) entered into the Vermont Oxford Network (VON) neonatal database. The study included neonatal patients admitted to the Providence St. Vincent NICU and Randall Children’s Hospital at Legacy Emanuel NICU who were born between January 1, 2000 and December 31, 2014. Since the adverse clinical outcomes related to possible calcium and phosphate incompatibility in PN solutions have been pulmonary the primary outcomes were chronic lung disease (CLD) defined as requiring supplemental oxygen at a postmenstrual age of 36 weeks or receiving supplemental oxygen if discharged between 34 and 36 weeks, percent requiring supplemental oxygen at 28 days, and mortality excluding early deaths defined as occurring at <12 hours of age (MEED). Secondary outcomes included a yearly benefit metric (BM) and percent of infants with weight <10th percentile at discharge. The BM is a risk-adjusted composite morbidity score for VLBW infants based upon the incidence of 8 major morbidities: grade 3-4intraventricular hemorrhage, stage 3-4 retinopathy of prematurity, periventricular leukomalacia, any late infection, necrotizing enterocolitis, focal intestinal perforation, CLD, and discharge weight <10th percentile [14]. The score is adjusted for mortality and the following risk factors: gestational age, SGA (birth weight <10th percentile), inborn/outborn, major birth defect, mode of delivery, 1-minute APGAR score <4, multiple birth, and male sex. A higher score denotes fewer morbidities and/or mortality. Weight gain from birth to discharge [15], as well as, other pertinent perinatal and neonatal data related to treatment and outcomes were also recorded. VON criteria were used to define parameters [16].

The study NICUs were compared to all U.S. NICUs for the years 2000 through 2014 by downloading data from the VON Nightingale Internet Reporting System. The aggregated data for all years (2000–2014) from the combined study NICUs was compared to the aggregated data for all years from all NICUs in the VON U.S. database by student t-test for continuous variables and chi-square test for categorical variables. All statistical analyses were performed using statistical software R (version 3.2.0). A P value <0.05 was considered statistically significant.

We also evaluated the incidence of IV infiltrates for all patients admitted to both study NICUs in 2013-2014 by performing a retrospective review of patient records and incident reports.

3 Results

There were 3748 VLBW infants admitted to the study NICUs and 586668 admitted to the VON U.S. NICUs during the study period. Parenteral nutrition was begun on an average of day of life 2 for all VLBW infants and the average number of days of PN was 11 days per infant. There were a total of 3450 infants who survived after 12 hours so that the estimated number of PN orders written during the period of the study would be 37950. The frequency of baseline characteristics and other pertinent clinical data and outcomes of study infants and those in the VON U.S. database are shown in Tables 1 and 2. The incidence of all the major morbidities was lower at the study NICUs compared to all U.S. NICUs. Weight gain velocity was higher at the study NICUs.

Table 1
Descriptive characteristics

Parameter Study N Study U.S. N U.S. P value

GA (WK) 3748 28.2±3.0^a 586 668 28.0^b NS

GA<27 WK 3748 29.3% 586 668 23.0% 0.0001

BW (g) 3748 1066±311^a 586 668 1036^b <0.05

BW<1000 g 3748 41.1% 586 668 32.0% 0.0001

SGA 3742 21.6% 586 473 20.2% 0.033

Inborn 3748 86.8% 586 668 84.8% 0.001

Birth defect 3748 4.9% 586 249 5.0% NS

Apgar^c 3716 30.5% 581 610 28.7% 0.01

Multiple birth 3748 30.4% 586 616 27.3% 0.0001

Males 3748 52.3% 586 450 51.0% NS

C-section 3748 62.3% 586 545 68.5% 0.0001

Prenatal steroids 3745 78.6% 583 525 74.7% 0.0001

Parameter	Study N	Study	U.S. N	U.S.	P value
GA (WK)	3748	28.2±3.0^a	586 668	28.0^b	NS
GA<27 WK	3748	29.3%	586 668	23.0%	0.0001
BW (g)	3748	1066±311^a	586 668	1036^b	<0.05
BW<1000 g	3748	41.1%	586 668	32.0%	0.0001
SGA	3742	21.6%	586 473	20.2%	0.033
Inborn	3748	86.8%	586 668	84.8%	0.001
Birth defect	3748	4.9%	586 249	5.0%	NS
Apgar^c	3716	30.5%	581 610	28.7%	0.01
Multiple birth	3748	30.4%	586 616	27.3%	0.0001
Males	3748	52.3%	586 450	51.0%	NS
C-section	3748	62.3%	586 545	68.5%	0.0001
Prenatal steroids	3745	78.6%	583 525	74.7%	0.0001

Abbreviations: BW, birth weight; GA, gestational age; SGA, birth weight <10th percentile; WK, weeks. ^aMean±standard deviation. ^bMean. ^cApgar <4 at one minute of age.

Table 2

Additional clinical data

Parameter	Study N	Study	U.S. N	U.S.	P value
WT gain	2958	13.4±2.6^a	466 314	12.6^b	<0.05
(g/kg/day)
Length of stay	3150	59.2±29.6^a	491 903	68.8^b	<0.05
(days)
NCPAP	3481	75.0%	559 254	66.7%	0.0001
MV	3481	53.6%	559 389	68.4%	0.0001
Surfactant	3747	50.8%	586 172	63.7%	0.0001
Steroids for	3481	6.2%	557 889	10.3%	0.0001
CLD
PDA	3479	31.8%	558 674	35.8%	0.0001
RDS	3479	62.4%	559 315	74.6%	0.0001
SIVH	3262	5.5%	510 920	8.9%	0.0001
PVL	3277	2.1%	518 300	3.1%	0.002
SROP	2581	6.4%	403 603	8.4%	0.0002
Any late	3375	7.8%	533 784	17.2%	0.0001
infection
NEC	3481	4.4%	559 318	6.2%	0.0001
SIP	3481	2.2%	559 186	2.5%	NS

Abbreviations: CLD, chronic lung disease; MV, treated with mechanical ventilation; NCPAP, treated with nasal continuous positive airway pressure; NEC, necrotizing enterocolitis; PDA, patent ductus arteriosis; PVL, periventricular leukomalacia; RDS, respiratory distress syndrome; SIP, focal intestinal perforation; SIVH, grade 3-4 intra-ventricular hemorrhage; SROP, stage 3-4 retinopathy of prematurity; WT, weight. ^aMean±standard deviation. ^bMean.

3.1 Primary outcomes

Primary outcomes are shown in Table 3 and Figs. 1–3. The percent of infants diagnosed with CLD and those requiring oxygen at 28 days for the total 15 year study period were significantly less for the study NICUs compared to all U.S. NICUs. As can be seen in Figs. 1 and 2, this has been a consistent finding from year to year throughout the study period. The MEED has also been significantly lower at the study NICUs compared to all U.S. NICUs. The MEED has been less than or equal to that for the VON U.S. NICUs for 10 of the 15 study years (Fig. 3).

Table 3
Primary outcomes

Parameter Study N Study U.S. N U.S. P value

CLD 3125 24.0% 486 097 28.6% 0.0001

O₂ at 28 days 2757 46.2% 453 324 51.8% 0.0001

MEED 3450 8.7% 548 997 10.3% 0.002

Parameter	Study N	Study	U.S. N	U.S.	P value
CLD	3125	24.0%	486 097	28.6%	0.0001
O₂ at 28 days	2757	46.2%	453 324	51.8%	0.0001
MEED	3450	8.7%	548 997	10.3%	0.002

Abbreviations: CLD, chronic lung disease; MEED, mortality excluding early deaths; O₂, oxygen.

Fig.1

Percent of infants diagnosed with chronic lung disease (CLD) at the study neonatal intensive care units (solid line) versus those at all U.S. NICUs in the Vermont Oxford Network (dashed line).

Fig.2

Percent of infants requiring oxygen at 28 days of age at the study neonatal intensive care units (solid line) versus those at all U.S. NICUs in the Vermont Oxford Network (dashed line).

Fig.3

Mortality excluding early deaths (MEED) at the study neonatal intensive care units (solid line) versus those at all U.S. NICUs in the Vermont Oxford Network (dashed line).

3.2 Secondary outcomes

The percent of infants with weight <10th percentile at discharge was lower for the study NICUs compared to all U.S. NICUs at 41.1% (1212/2949) versus 52.6% (242 112/460 176), respectively (p = 0.0001). The benefit metric is higher than that for the VON U.S. NICUs in each year of the study period (Fig. 4).

Fig.4

Value of the benefit metric at the study neonatal intensive care units (solid line) versus those at all U.S. NICUs in the Vermont Oxford Network (dashed line).

3.3 Infiltrates

The incidence of IV infiltrates during years 2013-2014 was 1.1% (28 infiltrates/2599 total patients). Only 1 of 9 infiltrates associated with PN at our NICUs required any treatment, which was hyaluronidase infiltration, and all were characterized by only local skin inflammation which resolved.

4 Discussion

This study is the largest study of clinical outcomes associated with the use of CaCl₂ as the preferred calcium additive in neonatal PN. The study found no evidence of adverse outcomes related to the use of CaCl₂ during the study period.

There is evidence of multiple adverse outcomes associated with Al exposure [17]. Adverse neurodevelopmental outcome has been most concerning [1]. The American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) acknowledges that Al contamination of PN solutions is also a risk factor for metabolic bone disease of preterm infants and recommends that “efforts be made to reduce the aluminum content of PN” [18]. Calcium gluconate in glass vials is the additive most responsible for the excessive Al exposure of neonatal patients to Al from PN [1 , 19]. Concerns with regard to the use of CaCl₂ in PN are based upon a single study from 1980 that showed increased ionization of calcium in solutions of water (not in PN solutions) for CaCl₂ (90%) compared to calcium gluconate (70–75%) [20]. In the same paper the authors published compatibility curves for both CaCl₂ and CaGlu in adult PN solutions and, although less calcium can be added for a given concentration of phosphate when using CaCl₂, the authors did not recommend against using CaCl₂ in PN. Recent studies have presented further data on the compatibility of CaCl₂ and phosphates in neonatal PN solutions [21, 22]. Another concern has been a risk of injury due to intravenous infiltrates but the RCT comparing PN solutions containing CaCl₂ to solutions containing CaGlu-Gl found no difference in the incidence or severity of infiltrates [1]. The incidence of IV infiltrates at the study NICUs in the current study was less than that reported in a multicenter study of IV infiltration in NICU patients of 3.8% (28 infiltrates/742 patients) [23]. Finally, although calcium and phosphate concentrations are lower in PN solutions containing CaCl₂ compared to CaGlu, a previous study from one of our NICUs did not find an increase in metabolic bone disease or poor growth in length associated with the use of CaCl₂ compared to reports in the literature from centers that use CaGlu-Gl [24].

Calcium gluconate in plastic vials became available in the U.S. in 2015 and PN solutions compounded with this additive are also low in Al content [3]. Based upon the increased solubility of calcium gluconate and its low Al content, CaGlu-Pl may be the preferred calcium additive in solutions made with mineral phosphates. In times of shortages of this additive, CaCl₂ may be considered as the preferred choice over CaGlu-Gl based upon its low Al content and the lack of evidence of adverse clinical effects associated with the use of CaCl₂ in neonatal PN. Another alternative is to use an organic phosphate such as sodium glycerophosphate (NaGP) as the phosphate additive with either CaGlu-PL or CaCl₂ since this phosphate additive is highly soluble with either calcium gluconate or calcium chloride [25, 26]. At this time NaGP has only been approved for use in the U.S. when the mineral phosphates are unavailable.

There are limitations to this retrospective study. It has been assumed that the majority of NICUs in the VON U.S. database have used CaGlu-Gl as the preferred calcium additive in neonatal PN during the years of the study. This assumption was based upon the recommendations of the American Society for Parenteral and Enteral Nutrition that were in place during the years of this study which stated that CaGlu was the preferred calcium additive for PN [13]. In a small sample of NICUs, not including the two study centers, from all areas of the U.S. for which we have personal knowledge of PN practices only 4 of 22 used CaCl₂ during the years of the study. We have not reported neurodevelopmental outcomes or other long term outcomes but the randomized controlled study found improved neurodevelopmental outcome as well as improved bone mineralization for infants who received CaCl₂ compared to infants who received CaGlu-Gl in neonatal PN [1, 2]. The primary outcomes of CLD, O₂ at 28 days, and MEED are influenced by many factors and it cannot be assumed that the use of CaCl₂ in neonatal PN at the study NICUs during the years of the study has been primarily responsible for positively or negatively affecting clinical outcomes. Although the data in this study suggest that these outcomes were better for infants supported with CaCl₂ it was not practical for us to obtain individual subject data for the 586 668 infants in the U.S. NICU group in order to perform multivariate logistic regression analyses. Our intent was not to prove that these outcomes were better for infants supported with CaCl₂ but to evaluate the safety of using CaCl₂ in neonatal PN solutions. The fact that the primary outcomes have consistently been better for the infants admitted to the study NICUs compared to all VON U.S. NICUs over an extended period of time supports the safety of CaCl₂ which has not been associated with detectable adverse effects in our experience. In addition, the BM, which is a composite measure of morbidity, was risk adjusted for mortality as well as other risk factors that have been shown to effect major outcomes. Strengths of the study are that it includes a large number of subjects and, in addition, the data from the study NICUs was collected contemporaneously with the data from the control group of all U.S. NICUs over a period of 15 years.

In conclusion, our exclusive use of CaCl₂ in neonatal PN was not associated with any apparent increase in adverse outcomes that might be exacerbated by potential insolubility of calcium and phosphate in PN solutions. This finding is consistent with the previously reported RCT that found no adverse effects for preterm infants whose PN contained CaCl₂ compared to those whose PN contained CaGlu-Gl. Provided that standard guidelines for compounding and filtration are followed, CaCl₂ appears to be a safe alternative to CaGlu-Gl in order to reduce Al exposure from PN in neonatal patients when CaGlu-Pl is not available based upon studies of clinical outcomes.

Conflicts of interest

The authors have no conflicts of interest to disclose.

Footnotes

Acknowledgments

Vermont Oxford Network approved the use of its data for this analysis but played no role in the design, conduct, analysis, interpretation, or reporting of this study. The views, conclusions, and opinions expressed are solely those of the authors and do not represent those of the Vermont Oxford Network.

References

Bishop

, Morley

, Day

, Lucas

. Aluminum neurotoxicity in preterm infants receiving intravenous-feeding solutions. N Eng J Med 1997;336(22):1557–61.

Fewtrell

, Bishop

, Edmonds

, Isaacs

, Lucas

. Aluminum exposure from parenteral nutrition in preterm infants: Bone health at 15-year follow-uPediatrics 2009;124:1372–9.

Frey

, Meier

. Polyethylene vials of calcium gluconate reduce aluminum contamination of TPN. Ann Pharmacother 2000;34:811–2.

Pironi

, Zolessi

, Ruggeri

, Paganelli

, Pizzoferrato

, Miglioli

. Bone turnover in short-term and long-term home parenteral nutrition for benign disease. Nutrition 2000;16:272–7.

Bohrer

, do Nascimento

, Binotto

, Becker

, Pomblum

. Contribution of the raw material to the aluminum contamination in parenterals. JPEN J Parenter Enteral Nutr 2002;26:382–8.

Migaki

, Melhart

, Dewar

, Huston

. Calcium chloride and sodium phosphate in neonatal parenteral nutrition containing Trophamine: Precipitation studies and aluminum content. JPEN J Parenter Enteral Nutr 2012;36:470–5.

Driscoll

, Newton

, Bistrian

. Potential hazards of precipitation associated with calcium chloride in parenteral nutrition admixtures: Response to Migaki et al. JPEN J Parenter Enteral Nutr 2012;36:497–8.

Huston

, Migaki

. Issues concerning the safety of calcium chloride in neonatal parenteral nutrition: Response to the letter from Drs Driscoll, Newton, and Bistrian. JPEN J Parenter Enteral Nutr 2012;36:498–500.

Newton

, Driscoll

. Calcium and phosphate compatibility: Revisited again. Am J Health-Syst Pharm 2008;65:73–80.

10.

Hill

, Heldman

, Goo

EDH

, Whippo

, Perkinson

. Fatal microvascular pulmonary emboli from precipitation of a total nutrient admixture solution. JPEN J Parenter Enteral Nutr 1996;20:81–7.

11.

Shay

, Fann

, Jarvis

. Respiratory distress and sudden death associated with receipt of a peripheral parenteral nutrition admixture. Infect Control Hosp Epidemiol 1997;18:814–7.

12.

Food and Drug Administration. Aluminum in large and small volume parenterals used in total parenteral nutrition; amendment; delay in effective date. Fed Reg 2002;67:70691–2.

13.

Mirtallo

, Canada

, Johnson

, et al. Safe practices for parenteral nutrition. JPEN J Parenter Enteral Nutr 2004;28:S39–70.

14.

Kaempf

, Zupancic

JAF

, Wang

, Grunkemeier

. A risk-adjusted, composite outcomes score and resource utilization metrics for very low-birth-weight infants. JAMA Pediatrics 2015;169:459–65.

15.

Patel

, Engstrom

, Meier

, Kimura

. Accuracy of methods for calculating postnatal growth velocity for extremely low birth weight infants. Pediatrics 2005;116:1466–7.

16.

Vermont Oxford Network Database Manual of Operations. Release 19.0. Vermont Oxford Network: Burlington, 2014.

17.

Hernandez-Sanchez

, Tejada-Gonzalez

. Aluminum in parenteral nutrition: A systematic review. Eur J Clin Nutr 2013;67:230–8.

18.

Nehra

, Carlson

, Fallon

, Kalish

, Potemkin

, Gura

, et al. A. S.P.E.N. clinical guidelines: Nutrition support of neonatal patients at risk for metabolic bone disease. JPEN J Parenter Enteral Nutr 2013;37:570–98.

19.

Poole

, Schiff

, Hintz

, Wong

, Mackenzie

, Kerner

. Aluminum content of parenteral nutrition: Measured versus calculated levels. JPGN 2010;50:208–11.

20.

Henry

, Jurgens

Jr , Sturgeon

, Athanikar

, Welco

, Van Leuven

. Compatibility of calcium chloride and calcium gluconate with sodium phosphate in a mixed TPN solution. Am J Hosp Pharm 1980;37:673–4.

21.

Huston

, Christensen

, Karnpracha

, et al. Calcium chloride in neonatal parenteral nutrition: Compatibility studies using laser methodology. PLoS One 2014;5;9(9):e106825. doi: 10.1371/journal.pone.0106828.

22.

Huston

, Christensen

, Alshahrani

, et al. Calcium chloride in neonatal parenteral nutrition solutions with and without added cysteine: Compatibility studies using laser and micro-flow imaging methodology. PLoS One 2015;28;10(8):e0136894.

23.

Wilkins

, Emmerson

AJB

. Extravasation injuries on regional neonatal units. Arch Dis Child Fetal Neontal Ed 2004;89:F274–5.

24.

Huston

, McCulley

, Markell

, Heisel

, Lewallen

. Neonatal parenteral nutrition containing calcium chloride and sodium phosphate: Mineral intake and aluminum content. ICAN Infant Child Adol Nutr 2012;4:137–42.

25.

MacKay

, Anderson

. Physical compatibility of sodium glyerophosphate and calcium gluconate in pediatric parenteral nutrition solutions. JPEN J Parenter Enteral Nutr 2015;39:725–8.

26.

Anderson

, MacKay

. Physical compatibility of calcium chloride and sodium glyerophosphate in pediatric parenteral nutrition solutions. JPEN J Parenter Enteral Nutr 2016;40(8):1166–9.

Calcium chloride in neonatal parenteral nutrition: A 15 year experience

Abstract

Keywords

1 Introduction

2 Methods

3 Results

Table 3 Primary outcomes Parameter Study N Study U.S. N U.S. P value CLD 3125 24.0% 486 097 28.6% 0.0001 O2 at 28 days 2757 46.2% 453 324 51.8% 0.0001 MEED 3450 8.7% 548 997 10.3% 0.002

4 Discussion

Conflicts of interest

Footnotes

Acknowledgments

References

Table 3
Primary outcomes

Parameter Study N Study U.S. N U.S. P value

CLD 3125 24.0% 486 097 28.6% 0.0001

O₂ at 28 days 2757 46.2% 453 324 51.8% 0.0001

MEED 3450 8.7% 548 997 10.3% 0.002