Corrected End-Tidal Carbon Monoxide Closely Correlates With the Corrected Reticulocyte Count in Coombs’ Test-Positive Term Neonates
From the Division of Neonatology, Department of Pediatrics, Weill Medical College of Cornell University, New York, New York
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Objective. To evaluate clinical usefulness of corrected end-tidal carbon monoxide (ETCOc) measurements in healthy, term, Coombs’ test-positive neonates and correlate it to the corrected reticulocyte count (RC).
Methods. ETCOc and RC were determined (at 36 ± 12 hours of age) in 50 Coombs’ test-positive neonates and compared with the ETCOc values of 50 Coombs’ test-negative neonates.
Results. Fifty percent of Coombs’ test-positive infants had RCs <5% (within a normal range for a healthy newborn) and ETCOc = 1.8 ± 0.34 parts per million (ppm) and likely did not exhibit hemolysis. Among infants with elevated RCs, 72% had RCs between 5% and 8% and ETCOc = 2.77 ± 0.68 ppm, and 28% had RCs >8% and ETCOc = 4.52 ± 0.83 ppm. There was an almost linear correlation (r = 0.86) between the RC and the ETCOc among Coombs’ test-positive infants. The 50 Coombs’ test-negative infants had ETCOc = 1.6 ± 0.45 ppm. Serial ETCOc measurements were performed in 14 Coombs’ test-positive infants: in all but 1 infant ETCOc values declined over time.
Conclusions. There is a good correlation between ETCOc and RC in Coombs’ test-positive infants. ETCOc >2.5 ppm predicts a significant elevation of RC in 90% of Coombs’ test-positive infants.
Key Words: end-tidal carbon monoxide • reticulocyte count • infants • Coombs test
Abbreviations: RC, reticulocyte count • CO, carbon monoxide • ETCO, end-tidal CO • ETCOc, corrected ETCO • ppm, parts per million • ROC, receiver operator curve
Hemolytic disease caused by ABO incompatibility may exaggerate physiologic hyperbilirubinemia of the newborn and even cause anemia in affected infants.1 The incidence of ABO mismatch between the mother and her fetus is 
15%, and 10% to 20% of these neonates become significantly jaundiced (requiring therapy).2 Thus, it is important to identify and closely monitor the infants at risk for moderate to severe hemolysis among the ABO-incompatible infants. In the absence of a direct measurement of the degree of hemolysis, clinicians frequently relied on the direct Coombs’ test, serum total bilirubin, and corrected reticulocyte count (RC) (a sign of bone marrow compensation for the anemia)3 as screening tools for identification of such infants. However, the Coombs’ test may be negative even in the presence of hemolysis,4 clinical jaundice is sometimes absent at the time of the infant’s discharge from the nursery,5 and, before the corrected RC could be correlated with hemolysis, other causes of pathologic anemia have to be ruled out.
Carbon monoxide (CO) is produced in equimolar amounts with bilirubin during hemoglobin degradation, and 80% to 90% of endogenous CO is a by-product of red blood cell destruction. Concentration of CO in the expired breath, the end-tidal CO (ETCO), correlates directly with the degree of hemolysis and can be measured by gas chromatography.6 It is a highly sensitive and specific test but is time-consuming and fairly complicated for routine use in clinical settings. Recently, the CO-Stat End-Tidal Breath Analyzer (Natus Medical Inc, San Carlos, CA) became commercially available. It determines the ETCO in breath, corrected for ambient CO (ETCOc), at the bedside noninvasively within minutes.7
The objective of this study was to determine if there is a correlation between ETCOc and the corrected RC among ABO-incompatible neonates with a positive direct Coombs’ test.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Study Design
This was a prospective study approved by the New York Presbyterian Hospital Committee on Human Rights in Research. It took place between September 2000 and August 2001 in the newborn nursery of the New York Presbyterian Hospital (Weill Medical College of Cornell University).
Contemporaneously born, healthy, term infants who were admitted to the well nursery and met all inclusion and exclusion criteria were eligible for enrollment into this research study. The intention was to enroll 50 consecutive Coombs’ test-positive ABO-incompatible infants in group A and 50 consecutive Coombs’ test-negative ABO-compatible infants in group B. The inclusion criteria were:
- gestational age >37 weeks;
- size appropriate for gestational age;
- healthy (a work-up for sepsis in an infant who appeared well was not an exclusion criteria); and
- signed consent.
The exclusion criteria were:
- ABO incompatibility and a negative Coombs’ test;
- Rh disease;
- minor blood-group incompatibility;
- major congenital anomalies (minor congenital anomalies were not exclusion factors);
- perinatal asphyxia (Apgar score <6 at 5 minutes of age);
- respiratory rate >60/minute or oxygen requirement;
- large cephalohematoma or severe bruising;
- red blood cell abnormalities (spherocytosis, elliptocytosis, etc) determined on examination of blood smears;
- suspected or confirmed perinatal history of blood loss;
- family history of hemolytic anemia (glucose-6-phosphate dehydrogenase deficiency, pyruvate kinase deficiency, etc); or
- maternal smoking (maternal smoking affects the ETCOc of their infants during the first 48 to 72 hours of life8; in addition, conflicting results were published about the effect of maternal smoking on infant’s RC9–11).
The following tests and procedures are routinely performed according to the hospital policy and were not influenced by an infant’s participation in the study:
- ABO, Rh D typing, and a serum screen for unusual isoimmune antibodies (indirect Coombs’ test) are performed on the mother’s blood during the pregnancy;
- ABO, Rh D typing, and a direct Coombs’ test are performed on the infant’s cord blood; and
- at 24 to 48 hours of age, complete blood count and RC are performed on the blood from all infants who had a positive direct Coombs’ test at birth.
The following additional tests were performed in the group of the enrolled infants for the purpose of this study:
- ETCOc was measured within 24 to 48 hours of age;
- the corrected RC was calculated as a reticulocyte percentage adjusted to a standard hematocrit (45%) (corrected RC = patient’s RC [%] x patient’s hematocrit [L/L]/0.45 [L/L]); and
- medical records were reviewed to confirm that the subject met all inclusion and exclusion criteria and determine if the study participants developed hyperbilirubinemia and/or anemia later in life, indicating that the hemolysis during the neonatal period was not related to ABO incompatibility.
The ETCOc was determined with a single-use disposable nasal sampler CO-Stat End-Tidal Breath Analyzer (Natus Medical Inc). The device was calibrated every 30 days (according to manufacturer specifications). The measurements were performed in the nursery after a 15-minute "warm-up" period. Because a fraction of ETCO is derived from the ambient CO, the infants were allowed to equilibrate in the testing environment for at least 20 minutes. CO in room air measured by the device is then subtracted from the ETCO measured from the infant to give an ETCOc value. The noninvasive breath sampling took 2 to 3 minutes and did not disturb the infant in any way. A printed record was obtained for each measurement.
Repeated testing was performed in the following circumstances:
- The first 10 subjects were retested to establish the reproducibility of the ETCOc measurements. It was within ±0.3 parts per million (ppm) (similar to the accuracy of the instrument reported by other investigators). The mean value was used for the statistical analysis.
- Sampling had to be repeated with breath hydrogen interference (especially if sampling was performed immediately after feeding) or inadequate sampling or irregular breathing were reported. In most cases a repeated testing, after a delay, was successful.
- To determine a natural course of ETCOc values after birth, serial ETCOc measurements were performed every 12 hours (until discharge form the nursery) in a subgroup of patients with a positive direct Coombs’ test who met the following criteria:
- early development of hyperbilirubinemia requiring phototherapy (within 24–48 hours of age) (serum total bilirubin levels were determined in all clinically jaundiced infants, the results were plotted on the hour-specific bilirubin nomogram,12 and phototherapy was initiated according to the guidelines of the American Academy of Pediatrics13);
- presence of anemia within 24 to 48 hours of age (hemoglobin <11 mg/dL); or
- family history of a sibling with a significant hyperbilirubinemia requiring treatment with phototherapy or exchange transfusion.
- early development of hyperbilirubinemia requiring phototherapy (within 24–48 hours of age) (serum total bilirubin levels were determined in all clinically jaundiced infants, the results were plotted on the hour-specific bilirubin nomogram,12 and phototherapy was initiated according to the guidelines of the American Academy of Pediatrics13);
Statistical Analysis
All ETCOc values obtained from groups A and B were expressed in ppm and considered to follow a pattern of normal distribution. The ETCOc range, median, and mean ± standard deviation were calculated for both groups.
The corrected RCs were compared with the ETCOc values. A linear regression analysis was used to construct a regression curve and determine a regression coefficient between ETCOc and corrected RC for Coombs’ test-positive infants.
A receiver operator curve (ROC) was constructed by using the ETCOc values and corrected RCs to determine the cutoff value at which significant hemolysis would occur.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Among the 119 enrolled subjects, 50 Coombs’ test-positive infants and 50 Coombs’ test-negative infants were included in the study. Of the 19 excluded infants, 10 showed a hydrogen interference, so no ETCOc measurement was available, and 9 had ETCOc measurements taken beyond 48 hours of life (
72–84 hours of life). There were no significant differences between the infants in the 2 groups (Table 1) or in their age (in hours) when the ETCOc determinations were performed.
|
Based on the corrected RCs, 50 Coombs test-positive infants from group A were divided into 3 subgroups14:
- infants with a normal corrected RC (<5%);
- infants with a slightly elevated RC (5%–8%); and
- infants with a significantly elevated RC (>8%).
Fifty percent (25 of 50) of Coombs’ test-positive ABO-incompatible infants had a normal corrected RC at birth. It thus was unlikely that these infants were undergoing an active bone marrow response as a consequence of hemolysis. The ETCOc values in this subgroup of Coombs’ test-positive infants ranged between 1.2 and 2.5 ppm (mean: 1.8 ± 0.34 standard deviation). Six infants in this subgroup (6 of 25) had ETCOc values >2.0 ppm.
Seventy-two percent (18 of 25) of the remaining Coombs’ test-positive infants had a corrected RC between 5% and 8%. These infants had ETCOc values between 1.9 and 4.2 ppm (2.77 ± 0.68 ppm). Seven Coombs’ test-positive infants (7 of 50 = 28%) showed a corrected RC >8%. These infants had ETCOc values ranging from 3.5 to 6.1 ppm (4.52 ± 0.83 ppm). If other causes of hemolysis were ruled out (as per exclusion criteria), the elevated ETCOc in this subgroup of Coombs’ test-positive infants may indicate that ABO incompatibility was the cause of hemolysis.
The ETCOc values for the entire group A were 2.58 ± 1.2 ppm (range: 1.2–6.1 ppm; median: 2.1 ppm). There is almost a linear correlation (R = 0.86) between the corrected RC and the ETCOc (Fig 1) in these Coombs’ test-positive infants.
|
The ETCOc values for the 50 Coombs’ test-negative ABO-compatible infants (group B) ranged between 0.6 and 2.5 ppm (1.6 ± 0.45 ppm; median: 1.6 ppm). Twenty-eight infants in this group had ETCOc values >1.5 ppm, and among them 2 had ETCOc values >2 ppm.
The distribution of ETCOc values for age in groups A and B is shown in Figs 2 and 3. Because the nomogram only had tracks up to 36 hours of life,15 only ETCOc values taken during this time period could be plotted. The ETCOc values plotted higher in the Coombs’ test-positive infants. Some overlap between the ETCOc values of the group A (ABO-incompatible) and group B (ABO-compatible) infants has been noted. This could be explained by a slight asynchrony of the 2 measured processes: hemolysis occurs first, and then the bone marrow responds to it.
|
|
The ROC was constructed for ETCOc as a predictor of a corrected RC for determining hemolysis among the Coombs’ test-positive infants (Fig 4). For an ETCOc cutoff value of 1.5 ppm, there was 100% negative predictive value for elevated RC, with a relatively low specificity of 20%. Thus, an ETCOc measurement of <1.5 ppm at 36 ± 12 hours of age in a healthy term infant gives a reliable prediction that hyperbilirubinemia developing in these patients will not be caused by hemolysis. At an ETCOc value of 2.5 ppm in Coombs’ test-positive infants, a significant hemolysis would occur with a positive predictive value of 90% and a negative predictive value of 80%.
|
Serial ETCOc measurements were performed in 14 of the 50 Coombs’ test-positive infants. The indications for repeated ETCOc measurements were: early onset of hyperbilirubinemia (11 infants); a sibling that required phototherapy (2 infants); and a significant drop in the hemoglobin (1 infant). Ten of these infants had 2 ETCOc determinations, and 4 of them had 3 determinations. In all but 1 infant (with an initial ETCO of 1.7 ppm, which rose to 2.2 ppm 24 hours later), ETCOc values declined over time (in general within 18–24 hours).
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Results of this study indicate that there is an excellent correlation between a corrected RC and ETCOc values measured in the group of infants at risk for hemolysis (ABO-incompatible Coombs’ test-positive infants) during the first 48 hours of life (Fig 1). Several groups of investigators measured the bilirubin and ETCOc values in newborns and suggested that the ETCOc is a useful test for identification of infants with hemolysis.16,17 In contrast, even when hemolytic anemia is strongly suspected there is only sparse information in neonatal literature correlating the corrected RC and the severity of hemolysis.1,18 Some authors considered an RC >6% as laboratory evidence of severe hemolysis, because 10 of 11 infants who required exchange transfusion had RCs >6%;1 the others considered an RC >10% as being a sign of significant hemolysis.2,3 There are multiple difficulties in correlating the corrected RC with the degree of hemolysis: it depends on the type of antigen (reticulocytopenia is seen in infants immunized with the Kell antigen18), on the bone marrow’s capacity to respond, and on the timing of the blood sample (there is a lag period between the peak of hemolysis and the bone marrow’s response to it), and it may be difficult or expensive to rule out other causes of anemia. Thus, ETCOc determination may represent a more accurate and noninvasive test to identify infants at risk for hemolysis than a corrected RC.
The difference between the ETCOc values for groups A (2.58 ± 1.2 ppm for Coombs’ test-positive, ABO-incompatible infants) and B (1.6 ± 0.45 ppm for Coombs’ test-negative ABO-compatible infants) is statistically significant (P = .03). Some overlap among the ETCOc values between the groups (Fig 5) was expected, because hemolysis does not occur in 50% of Coombs’ test-positive infants,4 and nonantibody-mediated hemolysis may occur among Coombs’ test-negative neonates. However, the overlap between these groups raises a concern about the usefulness of a single ETCOc value in an individual infant. Based on the ETCOc values in groups A and B, the ROC constructed for ETCOc and corrected RC (Fig 4), and the presence of jaundice among infants in each group (Figs 2 and 3), it is possible to conclude that: the ETCOc value >2 ppm is extremely unlikely (2 of 50) among the ABO-compatible infants and warrants additional investigation; jaundice in Coombs’ test-positive infants with ETCOc <1.5 ppm (19 of 50) is unlikely caused by hemolysis; and ETCOc >2.5 ppm predicts a significant elevation of RC in 90% of Coombs’ test-positive infants. These results are in accordance with the results of a large international study that showed ETCOc values of 1.81 ± 0.59 ppm among hyperbilirubinemic infants and ETCOc values of 1.45 ± 0.47 ppm among the infants who did not develop hyperbilirubinemia.17
|
Serial measurements of the ETCOc may be used to follow the rate of hemolysis in an individual infant over time and his/her relative risk for later development of hyperbilirubinemia or anemia. The results of serial ETCOc determinations obtained during this study reaffirm that a natural course of ABO-related hemolysis is to decline after birth. A persistently elevated or increasing ETCOc value may indicate a different etiology of the hemolytic process and warrant additional investigation of its cause even among the ABO-incompatible infants. A sample size evaluated here is not large enough to determine the incidence of an increased ETCOc that occurred at a later time. Further experience with serial determinations of the rates of hemolysis especially during illnesses may provide additional information about the subpopulation of infants at increased risk for hemolysis.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The ETCOc, although not an independent predictor of hyperbilirubinemia or anemia, represents a useful, simple, and noninvasive tool in the evaluation of infants at risk for hemolysis.
|
FOOTNOTES |
|---|
Received for publication Aug 21, 2002; Accepted Mar 7, 2003.
Reprint requests to (M.N.) Perinatology Center, Weill Medical College of Cornell University, 525 E 68th St, N-506, New York, NY 10021. E-mail: mnesin{at}med.cornell.edu
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
- Kanto WP Jr, Marino B, Godwin AS, Bunyapen C. ABO hemolytic disease: a comparative study of clinical severity and delayed anemia.
Pediatrics.1978; 62
:365
–369
[Abstract/Free Full Text] - Desjardins L, Blajchman MA, Chintu C, et al. The spectrum of ABO hemolytic disease of the newborn. J Pediatr.1979; 95 :447 –457[CrossRef][Web of Science][Medline]
- Zipursky A. Isoimmune hemolytic diseases. In: Nahtan DG, Oski FA, eds. Hematology of Infancy and Childhood. Philadelphia, PA: WB Saunders; 1987:44–87
- Koenig JM. Evaluation and treatment of erythroblastosis fetalis in the neonate. In: Christensen RD, ed. Hematologic Problems of the Neonate. Philadelphia, PA: WB Saunders; 2000:185–207
- Catz C, Hanson JW, Simpson L, Jaffe SJ. Summary of workshop: early discharge and neonatal hyperbilirubinemia.
Pediatrics.1995; 96
:743
–745
[Abstract/Free Full Text] - Rodgers PA, Vreman HJ, Dennery PA, Stevenson DK. Sources of carbon monoxide in biological systems and applications of CO detection technologies. Semin Perinatol.1994; 18 :2 –10[Web of Science][Medline]
- Vreman HJ, Baxter LM, Stone R, Stevenson DK. Evaluation of a fully automated end-tidal carbon monoxide instrument for breath analysis.
Clin Chem.1996; 42
:50
–56
[Abstract/Free Full Text] - Seidman DS, Shiloh M, Stevenson DK, Vreman HJ, Paz I, Gale R. Role of hemolysis in neonatal jaundice associated with G6PD deficiency. J Pediatr.1995; 127 :804 –806[CrossRef][Web of Science][Medline]
- Meberg A, Haga P, Sande H, Foss OP. Smoking during pregnancy: haematological observations in the newborn. Acta Paediatr Scand.1979; 68 :731 –734[Web of Science][Medline]
- Knudsen A. Maternal smoking and the bilirubin concentration in the first three days of life. Eur J Obstet Gynecol Reprod Biol.1991; 40 :123 –127[CrossRef][Web of Science][Medline]
- Mercelina-Roumans PE, Breukers RB, Ubachs JM, van Werch JW. Hematological variables in the cord blood of neonates of smoking and non-smoking mothers. J Clin Epidemiol.1996; 49 :449 –454[CrossRef][Web of Science][Medline]
- Bhutani VK, Johnson L, Sivieri EM. Predictive ability of a predischarge hour-specific serum bilirubin for subsequent significant hyperbilirubinemia in healthy term and near-term newborns.
Pediatrics.1999; 103
:6
–14
[Abstract/Free Full Text] - American Academy of Pediatrics, Subcommittee on Neonatal Hyperbilirubinemia. Neonatal jaundice and kernicterus [commentary].
Pediatrics.2001; 108
:763
–765
[Abstract/Free Full Text] - Walka MM, Sonntag J, Kage A, Dudenhausen JW, Obladen M. Complete blood counts from umbilical cords of healthy term newborns by two automated cytometers. Acta Haematol.1998; 100 :167 –173[CrossRef][Web of Science][Medline]
- Bhutani VK, Herschel M, Stevenson DK. End tidal carbon monoxide hour-specific nomogram: for early and pre-discharge identification of babies with increased bilirubin production. J Perinatol.2001; 21 :501 –516
- Kaplan M, Vreman HJ, Hammerman C, et al. Combination of ABO blood group incompatibility and G6PD deficiency: effect on hemolysis and neonatal hyperbilirubinemia. Acta Pediatr.1998; 87 :455 –457[CrossRef][Web of Science][Medline]
- Stevenson DK, Fanaroff AA, Maisels MJ, et al. Prediction of hyperbilirubinemia in near-term and term infants.
Pediatrics.2001; 108
:31
–39
[Abstract/Free Full Text] - Vaughan JI, Manning M, Warwick RM, et al. Inhibition of erythroid progenitor cells by anti-Kell antibodies in fetal alloimmune anemia.
N Engl J Med.1998; 338
:798
–823
[Abstract/Free Full Text] - American Academy of Pediatrics, Provisional Committee for Quality Improvement and Subcommittee on Hyperbilirubinemia. Practice parameter: management of hyperbilirubinemia in the healthy term newborn. [published correction appears in Pediatrics. 1995;95:458–461].
Pediatrics.1994; 94
:558
–565
[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2003 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Morimatsu, T. Takahashi, K. Maeshima, K. Inoue, T. Kawakami, H. Shimizu, M. Takeuchi, M. Yokoyama, H. Katayama, and K. Morita Increased heme catabolism in critically ill patients: correlation among exhaled carbon monoxide, arterial carboxyhemoglobin, and serum bilirubin IX{alpha} concentrations Am J Physiol Lung Cell Mol Physiol, January 1, 2006; 290(1): L114 - L119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C M Harrison and C C Andersen Exhaled breath measures of inflammation: are they useful in neonatal chronic lung disease? Arch. Dis. Child. Fetal Neonatal Ed., January 1, 2005; 90(1): F6 - F10. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kaplan, M. Herschel, C. Hammerman, J. D. Hoyer, and D. K. Stevenson Hyperbilirubinemia Among African American, Glucose-6-Phosphate Dehydrogenase-Deficient Neonates Pediatrics, August 1, 2004; 114(2): e213 - e219. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
