Objective. To characterize use of volume infusion (VI) for infants who are ≥34 weeks' gestational age and receive intensive cardiopulmonary resuscitation (CPR; defined as >1 minute of positive-pressure ventilation and chest compressions, with or without the administration of medications) in the delivery room and are admitted to the NICU.
Methods. A retrospective review of a resuscitation registry between January 1999 and June 2001 was conducted.
Results. Of 37 972 infants, 23 received CPR, including 13 with VI. Ten of 13 received VI for persistent bradycardia despite CPR, and only 3 of 13 received VI for suspicion of hypovolemia with poor perfusion. More VI versus no VI infants had Apgar scores ≤2 at 5 and 10 minutes. VI versus no VI infants had lower cord arterial pH, had higher arterial partial pressure of carbon dioxide, had larger base deficit, required longer CPR, and required more epinephrine. On admission to the NICU, VI versus no VI infants had lower blood pressure and larger base deficit over the first 2 hours but did not differ in arterial pH, arterial partial pressure of carbon dioxide, heart rate, mortality, or use of additional VI or buffer.
Conclusions. VI is rarely given for overt hypovolemia and more often for asphyxiated infants who are slow to respond to intensive CPR. Persistent postnatal hypotension in VI infants suggests that other factors, eg, myocardial dysfunction, may be important contributors to lack of response to CPR.
The use of medications during neonatal resuscitation is an uncommon event. Data from a large metropolitan delivery service (>15 000 deliveries per year) indicate that medications are required in only 0.12% of all live-born deliveries.1 The infrequent use of medications during resuscitation of newborns in the delivery room (DR) has impeded the design and completion of rigorous investigations to determine the most effective medications and/or dosing regimens. No studies have examined the efficacy of volume infusions (VIs) during DR resuscitation. Indeed, the only studies related to the use of VI in newborns have been conducted in the NICU in the postnatal period in preterm infants. Current neonatal resuscitation guidelines recommend that if an infant is not responding to intensive resuscitation and/or there is evidence of blood loss, then volume expansion may be indicated.2 The guidelines emphasize that in some cases, there will be signs of shock with no evidence of blood loss as may occur with fetal-maternal hemorrhage. Such infants may appear pale, with weak pulses and delayed capillary refill, and have a persistently high or low heart rate despite effective ventilation, chest compressions, and epinephrine. More commonly, normovolemic, asphyxiated infants may have the same appearance.3 Thus, distinguishing normovolemic infants with asphyxial shock from those who are in hypovolemic shock is often problematic in the DR. On the basis of the current recommendations, any infant who does not respond to what otherwise seems to be effective resuscitation is likely to receive volume expansion in the DR, pending diagnostic studies that delineate the cause and suggest more specific therapies.4 The use of volume expansion is an important consideration because it could be detrimental to an asphyxiated infant with decreased myocardial function and low cardiac output. Given these considerations and the paucity of data available regarding the use of VI, we undertook this retrospective study with the following objectives: (1) to determine the frequency of VIs during intensive resuscitation, (2) to delineate the indications for its administration, and (3) to determine the immediate hemodynamic effects on admission to the NICU in infants who are ≥34 weeks' gestational age (GA) and receive cardiopulmonary resuscitation (CPR; positive-pressure ventilation and chest compressions, with or without the administration of medications) for >1 minute in the DR.
Parkland Memorial Hospital in Dallas, Texas, is a large county facility with >15 000 deliveries and ∼1100 admissions to the NICU per year. High-risk deliveries and those with unanticipated difficulties in neonatal transition are attended by a resuscitation team. The resuscitation team is composed of a senior pediatric resident or neonatal nurse practitioner, a neonatal nurse with special training in resuscitation and a neonatal respiratory therapist. A neonatal fellow and/or attending physician is also present for selected high-risk patients and cases that require CPR. For infants who require active resuscitation in the DR (eg, bag/mask ventilation [BMV] and beyond), an obstetric circulating nurse records interventions and heart rate every 30 seconds on a standardized form. The resuscitation team is trained to communicate out loud with one another so that all involved (including the obstetric nurse recording the resuscitation interventions) can hear the assessments, responses, and next steps to be taken. If any clarifications are needed, the obstetric nurse will obtain the information before transport of the infant to the NICU. The resuscitation nurse and doctor must sign off that they are in agreement with the resuscitation record before leaving the DR. All infants who require BMV for >2 to 3 minutes as well as all who receive CPR and/or VI in the DR are triaged to the NICU.
In 1990, a resuscitation registry was developed to collect prospectively information regarding DR stabilization and resuscitation for all deliveries attended by the resuscitation team if the infant was triaged to the NICU. Information collected in the registry included date of birth, race, gender, birth weight, obstetric GA, pediatric GA, maternal complications, mode of delivery, minute of life that resuscitation team arrived, all resuscitation interventions in the DR, Apgar scores, cord gas parameters (pH, arterial partial pressure of carbon dioxide [Pco2], arterial partial pressure of oxygen, and base excess), DR temperature, reason for admission to the NICU, mode of ventilatory support required at admission, admission temperature, respiratory rate, heart rate, mean arterial blood pressure (MAP), oxygen requirement, O2 saturations, arterial blood gas parameters (pH, Paco2, arterial oxygen pressure, and base excess), dextrose stick, and hematocrit (HCT). All parameters were recorded by the resuscitation nurse within the first 4 hours after birth. The neonatal resuscitation registry complies with Health Insurance Portability and Accountability Act guidelines and has been approved for use by the institutional review board of the University of Texas, Southwestern Medical Center of Dallas.
Intensive CPR was defined as the need for >1 minute of positive-pressure ventilation and chest compressions, with or without the administration of medications in the DR. Identification of all patients who required intensive CPR in the DR was determined by searching the resuscitation registry. All infants who were ≥34 weeks' GA and received intensive CPR and were admitted to the NICU between January 1999 and June 2001 underwent a retrospective chart and resuscitation registry review. The following data were retrieved from each case: maternal obstetric history including history of potential fetal blood loss, birth weight, GA (a pediatric assessment as determined by the Dubowitz examination5), and indications for and sequence of resuscitation interventions (need and timing for BMV, intubation, CPR, and medications). Duration of CPR was defined as time to cessation of cardiac compressions. Additional data retrieved from each case included 1-, 5-, and 10-minute Apgar scores, cord arterial blood gas measurements, MAP, heart rate, and base deficit over the first 3 hours after admission to the NICU; initial arterial blood gas measurements, initial ventilator settings, and use of volume expansion and sodium bicarbonate in the NICU; admission HCT; neonatal mortality (defined as death before discharge home from the NICU); and placental pathology (if available).
Hypotension was defined as a MAP of <35 mm Hg.6 Volume loss was defined on the basis of a clinical history of risk for blood loss, eg, cord accident coupled with HCT <35% on arrival to the NICU. All infants who are admitted to the NICU have a peripheral spun HCT measured at admission using a HemataStat-II machine (Separation Technologies, Inc, Altamonte Springs, FL). It is clinical practice to request a Kleihauer-Betke test on the mothers of infants with unexplained low admission HCT. All VIs were of normal saline.
Statistical methods used to compare infants who did not receive VI in the DR versus those who did included 2-tailed t test with Bonferroni correction for repeated measures, Mann-Whitney rank sum test for nonparametric data, and χ2 analysis where appropriate. P < .05 was considered statistically significant. All data are expressed as the mean ± SD unless otherwise stated.
During the 30-month study period, there were 37 977 births of infants who were ≥34 weeks' GA. Five asystolic infants received intensive CPR in the DR and did not respond to be triaged to the NICU. Two of these infants were classified as stillborn, and 3 were live-born infants with initial bradycardia and subsequent asystole that did not respond to intensive resuscitation. Four infants (the 3 live-born and 1 stillborn) had documentation of VI in the DR during CPR. Because of incomplete records, it is unknown whether the other stillborn infant received VI in the DR, but it is possible that he did under the indication of poor response to resuscitation. Infants who died in the DR are not included in the data for the remainder of this report. A total of 36 516 infants who were triaged to the regular newborn nursery presumably did not require more than brief BMV if any assistance was needed; however, infants who were triaged to the regular newborn nursery are not included in the resuscitation registry. A total of 1456 infants who were >34 weeks’ GA were triaged to the NICU (Table 1). Twenty-three (2%) infants fulfilled the criteria for intensive CPR and survived to be admitted to the NICU (Table 2). All of the 23 infants were singleton deliveries. Thirteen (57%) of the 23 infants received VI during the DR resuscitation. The clinical indication for the initial use of VI was an inadequate heart rate despite CPR and epinephrine for 10 (77%) of 13 infants and poor perfusion after intensive CPR coupled with a clinical history suspicious for volume loss for 3 (23%) of 13 infants. The volume of normal saline infused for the 13 infants was 21 ± 14 mL/kg. During the study period, no infants who were ≥34 weeks' GA received volume expansion in the DR outside the context of CPR. Perinatal factors associated with infants who received VI versus no VI during intensive resuscitation are shown in Table 3. Half of the entire cohort had a history of meconium-stained amniotic fluid. Infants who received VI had a higher incidence of decelerations noted on antepartum fetal heart rate monitoring (P = .003) and were more likely to have a history supportive of potential blood loss (P = .046).
Infants who received VI in the DR were significantly more asphyxiated than infants who did not receive VIs (Table 4). Although both groups were of similar gestation with comparable low 1-minute Apgar scores, significantly more infants in the VI group continued to have Apgar scores ≤2 at both 5 (P < .001) and 10 (P = .02) minutes of age. Median Apgar scores for no VI versus VI infants were 1 versus 0 at 1 minute, 3 versus 0 at 6 minutes, and 5 versus 3 at 10 minutes. Infants who received VI versus no VI in the DR had greater derangements of placental gas exchange with lower arterial cord gas pH (6.83 ± 0.19 [n = 10] vs 7.10 ± 0.27 [n = 9]; P = .024]), higher arterial Pco2 (100 ± 27 vs 69 ± 28 mm Hg; P = .024), and larger base deficits (22.9 ± 7.3 vs 12.6 ± 12 mEq/L; P = .037). The duration of CPR was more than twice as long in the VI versus no VI group (9 ± 4 vs 4 ± 1 minute; P < .001). All 13 infants in the VI group required epinephrine before Vis, whereas only 3 of 10 infants in the no VI group received epinephrine (P < .001). Although all pediatric providers are trained to follow Neonatal Resuscitation Program guidelines, 5 of 7 of the no VI infants who should have received epinephrine did not. Administration was delayed by lack of a route for delivery (either difficulty in securing an airway or the presence of an inexperienced provider, ie, the need for the resuscitation team was not anticipated before the delivery). All 5 infants subsequently responded without the medication. In addition, infants who received VI versus no VI in the DR were more likely to receive bicarbonate during the resuscitation (P = .046).
Clinical Characteristics on Admission to the NICU
All infants in both groups remained intubated on arrival to the NICU. There was no difference in admission ventilator settings for VI versus no VI infants (peak inspiratory pressure was 25 ± 7 vs 29 ± 4 cm H2O with ventilator rates of 42 ± 12 vs 36 ± 8 breaths per minute). More infants who received VI in the DR were hypotensive on arrival to the NICU (9 of 13) compared with no VI infants (1 of 10; P = .01). VI infants had significantly lower blood pressure (32 ± 13 vs 49 ± 12 mm Hg; P = .004; Table 5) compared with no VI infants on arrival to the NICU. The lower blood pressure in the VI group persisted over the next 2 hours (Fig 1A). Six of these infants received additional VI; however, only 1 infant was given pressor support. On arrival to the NICU, heart rate (Fig 1B), arterial pH, and Pco2 did not differ between the 2 groups; however, the VI versus no VI group had more pronounced metabolic acidosis as evidenced by a significantly larger base deficit (24.4 ± 6.7 vs 16.5 ± 8.8 mEq/L; P = .023) that persisted over the next 2 hours (Fig 1C). In the NICU, the 2 groups did not differ in the proportion who received additional VIs or sodium bicarbonate for stabilization. Mortality after admission to the NICU did not differ between the 2 groups.
Infants who received VI in the DR had lower admission HCT (41 ± 13%; [range: 15–58%] vs 54 ± 8% [range: 46–69%; P = .017), but only 3 (23%) VI infants met the definition of volume loss (a clinical history of risk for blood loss [cord accident (n = 1), fetomaternal transfusion (n = 1), and abruption (n = 1]) coupled with HCT <35% on arrival to the NICU. The 1 infant with fetomaternal transfusion was suspected to be anemic because of profound pallor and shock at delivery. In this case, the admission HCT was nondetectable and thus was recorded as 15% because the lower limits of reliable detection for HCT is 16% on the machine used for measurement. A Kleihauer-Betke test on the mother confirmed massive fetomaternal transfusion. Two infants in the no VI group had elevated HCTs of 65% and 69%, respectively.
Placental pathology was available for 85% of VI infants and for 50% of the no VI patients (Table 6). Significant abnormalities (acute chorioamnionitis, acute funisitis, chronic villitis, and fetal thrombotic vasculopathy) were common in both.
The data in this retrospective report indicate that when intensive CPR is required in the DR, VI as part of treatment was common and was administered to 57% of such infants. Although in the minority, the mothers of infants who received VI were more likely to present with a history suggestive of an increased risk for fetal blood loss compared with infants who received no VI. Infants who received VI compared with no VI infants exhibited more fetal heart rate abnormalities during labor, more severe fetal and neonatal acidemia, lower Apgar scores at 5 and 10 minutes, and lower blood pressures on admission to the NICU and during the first 2 postnatal hours.
Intensive CPR in the DR is an extremely rare event and thus one that is very difficult to study prospectively in humans. Before this retrospective report, there has been no published information examining the use or efficacy of VIs during DR resuscitation. The limitations of our study include its retrospective nature and the possibility of type 2 errors given the rarity of intensive CPR in the DR. Moreover, we had limited data for infants who were given intensive CPR in the DR and never had return of spontaneous circulation. In addition, we focused on the near-term and term infant rather than on premature infants. The outcomes examined are short term and do not elucidate long-term neurodevelopmental outcome after intensive CPR. It is possible that our definition of volume loss may be falsely limiting, as it required both a clinical history consistent with blood loss and a HCT <35% on arrival in the NICU. Such a restrictive definition may have excluded some infants who experienced acute perinatal hemorrhage or lack of placental transfusion and had not yet experienced fluid shifts in the equilibration process to lower the HCT. A particular strength of the study is that the sequence of DR interventions and responses was recorded during the resuscitation by a Neonatal Resuscitation Program trained observer (obstetric nurse) who was not on the resuscitation team and prospectively entered in a resuscitation registry.
Although the use of VI was common during intensive CPR, the overall need for CPR was rare and is consistent with previous reports from this institution.1 Thus, 1 in 1651 infants who were >34 weeks' GA required CPR ± medications for >1 minute. It is noteworthy that the majority of VI infants were hypotensive on arrival to the NICU, whereas the majority of no VI infants were normotensive. The cause for the hypotension in the VI infants is unclear but likely multifactorial. First, very few infants had evidence to support acute blood loss, suggesting that volume depletion is not a common determinant of hypotension. Moreover, several studies in newborn animals as well as in humans have shown that asphyxia can lead to a large placental transfusion into the infant.7–9 Thus, many asphyxiated infants are not hypovolemic except in the rare instance of massive fetal-maternal transfusion or a tight nuchal cord in which the vein but not the artery is occluded, causing fetal anemia.10 The initial HCT in the infants who did not receive VI was significantly higher than in infants who required VI. However, the HCT of the VI infants was bimodal in distribution, with the 3 hypovolemic infants having such a low HCT (23 ± 3 vs 47 ± 3 for nonhypovolemic infants; P = .001), that the mean for the entire group was lower. It is also possible that hemodilution by the infused volume contributed to lower HCTs in the VI infants. The clinical sign of pallor in the DR as a potential indicator of hypovolemia may be misleading in the presence of severe fetal acidemia. This reflects the redistribution of cardiac output in response to interruption of placental gas exchange with decreases in renal, gastrointestinal, and peripheral blood flow to preserve cardiac and cerebral blood flow. Asphyxial infants often appear pale but with normal blood pressures. Indeed only 3 of 13 of the infants who received VI did so out of concern for poor perfusion, although the heart rate was adequate.
Although VI infants had lower blood pressure on arrival to the NICU and the subsequent 2 hours compared with no VI infants, they did not demonstrate an increase in heart rate. This suggests that the normal compensatory mechanisms to maintain cardiac output in the face of hypotension were not present in the VI infants. It is possible that the hypotension was secondary to asphyxial myocardial injury. This is suggested by the greater derangements in placental gas exchange and larger base deficit on NICU admission in the VI-treated group. Multiple studies have demonstrated profound myocardial effects after perinatal asphyxia, including radiographic evidence of failure, ischemic changes on electrocardiogram, elevated troponin levels, and echocardiographic changes in contractility.11–14 These observations raise the possibility that VIs during DR resuscitation may be detrimental and exacerbate poor cardiac output. The neonatal heart operates at the top end of the Frank-Starling curve and is not as effective as the more mature heart at increasing cardiac output in the face of increased preload under normal conditions.2,15–18 Thus, it is much less likely to compensate in the face of compromised cardiac function from asphyxia. The potential contribution of myocardial dysfunction and/or poor vasomotor tone versus hypovolemia to early hypotension was evaluated in a randomized, controlled trial of dopamine versus colloid in very low birth weight infants in the NICU. It was noted that 90% of infants who were given dopamine responded with improved blood pressures as compared with only 45% of those who received colloid.19
Both groups demonstrated placental abnormalities in the majority of cases examined. Because of the small number of cases involved, it is possible that differences between groups may be detected with a larger sample size. Similarly, no differences in mortality were noted; however, this may also represent a type 2 error.
In conclusion, intensive CPR is a rare event in near-term and term infants; however, those who require intensive CPR frequently receive volume resuscitation. The indication for VI use seems to be related to the consequences of severe asphyxia more often than a clinical indication of hypovolemia. The persistent low blood pressure in infants who received volume resuscitation in the DR suggests that other factors (eg, myocardial dysfunction) may be important contributors to the lack of response to CPR. These observations raise important questions as to the efficacy and/or benefits of VI in this population. Randomized, controlled trials to evaluate risks and benefits of volume resuscitation in the context of severe asphyxia are warranted; however, given the difficulties in addressing these issues in infants in the DR, such initial studies may be better suited for neonatal animal models. Until such studies are completed, when balancing the lack of evidence for the use of volume and the critical concern for the adequate replacement of volume in infants who may have had a significant volume loss, the current guidelines represent a prudent therapeutic approach.
We gratefully acknowledge the efforts of the nurses and respiratory therapists of the Parkland Neonatal Resuscitation Team. We are indebted to Karen Kirby for preparing this manuscript.
- Accepted August 30, 2004.
- Reprint requests to (M.H.W.) Division of Neonatology, Department of Pediatrics, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390. E-mail:
No conflict of interest declared.
Dr Perlman’s current address is: Department of Pediatrics, Weill Cornell Medical Center, New York, NY.
Dr Laptook’s current address is: Women and Infants Hospital, Providence, RI.
- ↵Kattwinkel J. Textbook of Neonatal Resuscitation. 4th ed. Elk Grove, IL: American Academy of Pediatrics/American Heart Association; 2000
- ↵Anderson PAW, Kleinman CS, Lister G, Talner NS. Cardiovascular function during normal fetal and neonatal development and with hypoxic stress. In: Polin RA, Fox WW, eds. Fetal and Neonatal Physiology. Philadelphia, PA: W.B Saunders; 1998:861–865
- ↵Versmold HT, Kitterman JA, Phibbs RH, Gregory GA, Tooley WH. Aortic blood pressure during the first 12 hours of life in infants with birth weight 610 to 4220 grams. Pediatrics.1981;67 :607– 613
- ↵Van Haesebronck P, Vanneste K, Pretere C, Trapper Y de van, Theiry M. Tight nuchal cord and neonatal hypovolaemic shock. Arch Dis Child.1987;62 :1276– 1277
- ↵Bucciarelli RL, Nelson RM, Egan EA, Eitzman DV, Gessner IH. Transient tricuspid insufficiency of the newborn: a form of myocardial dysfunction in stressed newborns. Pediatrics.1977;59 :330– 337
- ↵Rudolph AM. Distribution and regulation of blood flow in the fetal and neonatal lamb. Circ Res.1985;57 :811– 821
- ↵Gill AB, Weindling AM. Randomised controlled trial of plasma protein fraction versus dopamine in hypotensive very low birthweight infants. Arch Dis Child.1993;69 :284– 287
- Copyright © 2005 by the American Academy of Pediatrics