MANUSCRIPT CITATION
Mazhari MYA, Priyadarshi M, Singh P, Chaurasia S, Basu S. Norepinephrine versus Dopamine for Septic Shock in Neonates: A Randomized Controlled Trial. J Pediatr 2025;282:114599. https://doi.org/10.1016/j.jpeds.2025.114599. PMID 40252959.
REVIEWED BY
Gabriel Altit, MDCM, FRCPC, FASE
Neonatologist – Montreal Children’s Hospital,
Associate Professor – Department of Pediatrics – McGill University,
1001 Decarie, Office B05.1513. Montreal, Quebec Canada. H4A 3J1
Principal Investigator of the NeoCardioLab (www.neocardiolab.com).
Email: Gabriel.Altit@mcgill.ca
TYPE OF INVESTIGATION
Treatment
QUESTION
In neonates with fluid-refractory septic shock (P), does norepinephrine (I) compared to dopamine (C) result in a higher proportion of shock reversal at 30 minutes (O)?
METHODS
• Design: This was a double-blind, parallel-group randomized controlled trial (RCT). It was conducted over 18 months from February 2023 to July 2024. The trial was registered (prospective registration with the Clinical Trial Registry of India (2023/02/049357)).
• Allocation: Eligible participants were randomly assigned to receive either norepinephrine (NE) or dopamine (DA) in a 1:1 ratio. The randomization sequence was computer-generated in variable block sizes and stratified by gestational age (<35 weeks and ≥35 weeks). Allocation was kept in sealed, opaque, numbered envelopes, opened by a designated study nurse after enrollment.
• Blinding: The study was double-blind, meaning neither the investigators nor the healthcare personnel (except the study nurse who prepared the infusion) were aware of the assigned treatment. Blinding was maintained as the prepared infusions looked identical and were initiated and titrated at similar volume rates.
• Follow-up period: Outcomes were recorded until hospital discharge. Physiological and acid-base parameters were monitored at baseline, 6 hours, and 24 hours after the diagnosis of shock and initiation of therapy. Shock reversal was assessed at 15 and 30 minutes, and also overall.
• Setting: The study was conducted at a single tertiary care hospital in Northern India.
• Patients: All neonates (term or preterm) admitted to the NICU with fluid-refractory septic shock (FRSS) were included. Term neonates were included up to 28 days of life, and preterm neonates up to 40 weeks postmenstrual age. Inclusion criteria for shock were hypotension (systolic, diastolic, or mean BP ≤ third centile for gestational age) plus any 2 of the following: urine output <0.5 mL/kg/h over preceding 12 hours, capillary filling time ≥4 seconds, serum lactate >5 mMol/L, metabolic acidosis with base excess >−5 mMol/L, or superior vena cava (SVC) flow <40 ml/kg/min. Fluid refractoriness was defined as failure of shock reversal after at least one fluid bolus of 10 ml/kg over 10 minutes. Sepsis diagnosis required a positive blood culture, a positive sepsis screen, or a compatible clinical course judged by the treating team having received antibiotics for at least 5 days. Exclusion criteria included major congenital anomalies, moderate to severe hypoxic ischemic encephalopathy in the first 72 hours, neonates already on cardiovascular medications (inotropes/vasopressors), shock within 24 hours of surgery, and failure to obtain parental consent. • Intervention: Neonates in the NE group received norepinephrine diluted in 5% dextrose solution, initiated at 0.4 mL/h (0.2 µg/kg/min). • Comparison: Neonates in the DA group received dopamine diluted in 5% dextrose solution, initiated at 0.4 mL/h (10 µg/kg/min). • Dosing: Doses were escalated if shock reversal was not achieved at 15 minutes, up to a maximum of 0.6 ml/hour (0.3 µg/kg/min for NE or 15 µg/kg/min for DA). The chosen dosing ranges were considered comparable to common NICU practice and previous studies, although not strictly equivalent. • Outcomes: o Primary outcome: The proportion of neonates with shock reversal at 30 minutes after initiation of vasoactive support. Shock reversal was defined as correction of hypotension (BP > third centile for gestational age) and at least two clinical endpoints: capillary refill time ≤2 seconds, normal pulses with no difference between peripheral and central, and warm extremities.
o Secondary outcomes: Changes in cerebral tissue oxygen saturation (CrSO2), need for additional vasoactive drugs and steroids, time to shock reversal, incidence of hyperglycemia, tachyarrhythmia and sinus tachycardia ≥200/min, need and duration of respiratory support, and changes in acid-base parameters (pH, bicarbonate, base excess, lactate, pCO2) at baseline, 6, and 24 hours. Morbidities (IVH grade III/IV, BPD, NEC stage II/III, ROP requiring treatment) and all-cause mortality were recorded until hospital discharge.
• Analysis and Sample Size: An intention-to-treat analysis was performed. Sample size was calculated based on detecting a 30% greater chance of shock reversal with NE compared to DA (assuming 30% with DA based on a previous study), requiring 80 neonates (40 per group) for 80% power and 5% alpha-error.
• Patient follow-up: 80 out of 138 assessed neonates were eligible and randomized. Three neonates in the NE group and 5 in the DA group required crossover due to shock refractory to at least 3 vasoactive drugs. All enrolled patients were analyzed.
MAIN RESULTS
Eighty neonates were enrolled (41 in NE group, 39 in DA group). Baseline characteristics were comparable between groups. Late onset neonatal sepsis was common (66% NE, 64% DA). 49% (39/80) had culture-positive sepsis, primarily gram-negative organisms (Figure 2).
• Primary outcome: The proportion of neonates achieving shock reversal at 30 minutes was 32% (13/41) in the NE group and 46% (18/39) in the DA group, which was not statistically significantly different (p=0.19).
• Secondary outcomes:
o Requirement for additional vasoactive agents and steroids was similar between groups.
o Overall shock reversal rate and time to shock reversal were also similar.
o Incidence of tachycardia was higher in the DA group (15% vs 2% in NE group, p=0.05).
o Incidence of hyperglycemia was similar overall but higher in the DA group among neonates ≥35 weeks.
o CrSO2 values were similar at 30 minutes and 6 hours but lower in the DA group at 24 hours (70 ± 8 vs 76 ± 7 in NE group, p<0.01).
o Neonates in the DA group were more acidotic at 24 hours (pH 7.19 ± 0.14 vs 7.27 ± 0.10 in NE group, p = 0.04). Base deficit and pCO2 at 24 hours also trended towards worse values in the DA group, but lactate levels were comparable.
o Mortality and other morbidities (IVH grade III/IV, BPD, NEC stage II/III, ROP requiring treatment) were similar between the two groups. The all-cause mortality rate was high overall (59% in NE vs 56% in DA), but similar between groups.
CONCLUSION
Norepinephrine and dopamine demonstrated comparable efficacy as first-line vasoactive agents for neonates with fluid-refractory septic shock based on the primary outcome of shock reversal at 30 minutes. However, the dopamine group experienced a higher incidence of tachycardia, lower pH, and lower cerebral tissue oxygen saturation at 24 hours following treatment initiation. The study is a commendable effort as a double-blind RCT in a difficult-to-study population. It confirms that dopamine and norepinephrine are comparably effective in achieving initial shock reversal in this specific cohort of fluid-refractory septic neonates. The observed adverse effects associated with dopamine, particularly regarding tissue perfusion, warrant further investigation. Larger, multicenter trials are needed to definitively determine the optimal first-line vasoactive agent for neonatal septic shock, ideally incorporating objective measures of tissue perfusion and cardiac function, potentially using adaptive trial designs or deferred consent models given the acute nature of the condition. Such studies should also evaluate long-term neurodevelopmental outcomes.
COMMENTARY
The randomized controlled trial by Mazhari et al. provides valuable comparative data on norepinephrine versus dopamine as first-line agents in neonatal fluid-refractory septic shock1. Despite these insights, several important limitations temper the interpretation of the findings. The study population included both term and preterm infants with a wide range of gestational ages and clinical presentations. The diagnosis of septic shock relied largely on clinical judgment, often in the absence of culture confirmation, with nearly half of participants classified as having “clinical sepsis.” This introduces a risk of misclassification bias and limits internal validity. Standardized sepsis scoring systems or objective definitions for endpoints such as “shock reversal” were not applied, raising concerns about subjectivity, particularly given the reliance on variable clinical signs such as capillary refill or pulse quality.
The primary outcome—shock reversal within 30 minutes—may not be physiologically meaningful in neonates, where hemodynamic responses are dynamic and may evolve over a longer period2. Recruiting infants already in a decompensated state of shock likely contributed to the high mortality rate and may have hindered the ability to assess early treatment effects. The absence of echocardiographic assessment to characterize shock phenotypes is another key limitation. Neonatal septic shock is often hemodynamically complex, involving a mix of vasodilatory and cardiogenic features, sometimes exacerbated by persistent fetal shunts or elevated pulmonary vascular resistance2, 3. Without phenotypic stratification, it is difficult to determine whether either drug was more effective for specific shock subtypes.
Methodologically, the trial used appropriate randomization and blinding strategies, but lacked standardization in critical aspects of drug delivery. Practical variables—such preparation site (bedside vs. pharmacy), carrier fluid use—were not described, which could introduce variability affecting drug efficacy. While the study measured cerebral regional oxygen saturation (CrSO₂), it did not report fractional tissue oxygen extraction (FTOE), which would have provided a more integrative index of oxygen delivery and utilization. Although CrSO₂ at 24 hours was significantly higher in the norepinephrine group (76.0 ± 7.3% vs. 69.5 ± 7.7%, p < 0.01), its clinical significance is unclear. CrSO₂ is influenced not only by perfusion but also by metabolic demand; higher values may reflect reduced cerebral metabolism rather than improved perfusion, potentially signaling adverse neurological effects4, 5.
The study also suffers from the common pitfall of multiple unadjusted comparisons, increasing the risk of type I error. Without statistical correction strategies (e.g., Bonferroni adjustment or false discovery rate control), any significant secondary outcome must be interpreted with caution. Ideally, these findings should have been clearly framed as exploratory. Generalizability is another concern. All patients were enrolled at a single tertiary center in India, where microbial epidemiology and clinical practices differ from those in high-income settings. Pathogens were predominantly gram-negative, including Klebsiella pneumoniae, Acinetobacter spp., and E. coli, with no reported cases of Group B Streptococcus. The feasibility of conducting such a trial in this setting is commendable; however, the consent process (e.g., deferred or opt-out strategies) was not detailed, which is an important ethical consideration in acutely ill neonates. The average gestational age of participants was 33.2 weeks in both treatment arms, indicating that extremely preterm infants (<28 weeks’ gestation) were largely excluded. As such, the findings may not apply to this particularly vulnerable subgroup or to neonates with different underlying sepsis etiologies.
REFERENCES
1. Mazhari MYA, Priyadarshi M, Singh P, Chaurasia S, Basu S. Norepinephrine versus Dopamine for Septic Shock in Neonates: A Randomized Controlled Trial. J Pediatr. 2025;282:114599.
2. Duignan SM, Lakshminrusimha S, Armstrong K, de Boode WP, El-Khuffash A, Franklin O, et al. Neonatal sepsis and cardiovascular dysfunction I: mechanisms and pathophysiology. Pediatr Res. 2024;95(5):1207-1216.
3. Deshpande S, Suryawanshi P, Holkar S, Singh Y, Yengkhom R, Klimek J, et al. Pulmonary hypertension in late onset neonatal sepsis using functional echocardiography: a prospective study. J Ultrasound. 2022;25(2):233-239.
4. Kusaka T, Isobe K, Yasuda S, Koyano K, Nakamura S, Nakamura M, et al. Evaluation of cerebral circulation and oxygen metabolism in infants using near-infrared light. Brain Dev. 2014;36(4):277-283.
5. Kanaprach P, Michel-Macias C, Mazzarello M, Mir M, Rampakakis E, Wutthigate P, et al. Cerebral Saturation and Fractional Tissue Oxygen Extraction Are Associated with Anterior Cerebral Artery Doppler Parameters in Neonates with Congenital Heart Defects. Neonatology. 2025:1-16.
