Published online May 1, 2008
PEDIATRICS Vol. 121 No. 5 May 2008, pp. 882-889 (doi:10.1542/peds.2007-0117)
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Castillo, A.
Right arrow Articles by Critz, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Castillo, A.
Right arrow Articles by Critz, A.
Related Collections
Right arrow Premature & Newborn
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Facebook   Add to Reddit   Add to Technorati   Add to Twitter  
What's this?

ARTICLE

Pulse Oxygen Saturation Levels and Arterial Oxygen Tension Values in Newborns Receiving Oxygen Therapy in the Neonatal Intensive Care Unit: Is 85% to 93% an Acceptable Range?

Armando Castillo, MDa, Augusto Sola, MDb, Hernando Baquero, MDc, Freddy Neira, MDc, Ramiro Alvis, MDc, Richard Deulofeut, MD, MPHd and Ann Critz, MDa

a Division of Neonatal-Perinatal Medicine, Emory University, Atlanta, Georgia
b Mid-Atlantic Neonatology Associates and Atlantic Neonatal Research Institute, Morristown Memorial Hospital, Morristown, New Jersey
c Department of Pediatrics, University of the North, Barranquilla, Colombia
d Pediatrix Medical Group, Neonatology, Dallas, Texas


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
OBJECTIVE. Our aim was to define the relationship of PaO2 and pulse oxygen saturation values during routine clinical practice and to evaluate whether pulse oxygen saturation values between 85% and 93% were associated with PaO2 levels of <40 mmHg.

METHODS. Prospective comparison of PaO2 and pulse oxygen saturation values in 7 NICUs at sea level in 2 countries was performed. The PaO2 measurements were obtained from indwelling arterial catheters; simultaneous pulse oxygen saturation values were recorded if the pulse oxygen saturation values changed <1% before, during, and after the arterial gas sample was obtained.

RESULTS. We evaluated 976 paired PaO2/pulse oxygen saturation values in 122 neonates. Of the 976 samples, 176 (18%) from infants breathing room air had a mean pulse oxygen saturation of 93.9 ± 4.3% and a median of 95.5%. The analysis of 800 samples from infants breathing supplemental oxygen revealed that, when pulse oxygen saturation values were 85% to 93%, the mean PaO2 was 56 ± 14.7 mmHg and the median 54 mmHg. At this pulse oxygen saturation level, 86.8% of the samples had PaO2 values of 40 to 80 mmHg, 8.6% had values of <40 mmHg, and 4.6% had values of >80 mmHg. When the pulse oxygen saturation values were >93%, the mean PaO2 was 107.3 ± 59.3 mmHg and the median 91 mmHg. At this pulse oxygen saturation level, 39.5% of the samples had PaO2 values of 40 to 80 mmHg and 59.5% had values of >80 mmHg.

CONCLUSIONS. High PaO2 occurs very rarely in neonates breathing supplemental oxygen when their pulse oxygen saturation values are 85% to 93%. This pulse oxygen saturation range also is infrequently associated with low PaO2 values. Pulse oxygen saturation values of >93% are frequently associated with PaO2 values of >80 mmHg, which may be of risk for some newborns receiving supplemental oxygen.

Key Words: oxygen therapy • pulse oxygen saturation • normoxemia • hypoxemia • hyperoxemia • oxygen arterial tension • newborns • oxyhemoglobin dissociation curve

Abbreviations: SpO2—pulse oxygen saturation • P50—partial pressure of a gas required to achieve 50% saturation • FIO2—fraction of inspired oxygen

Pulse oxygen saturation (SpO2) monitors have been in use throughout the world since the middle to late 1980s, and the results have been considered the "fifth vital sign."1–3 SpO2 monitors were designed to detect hypoxemia and potential hypoxia and unfortunately were introduced into practice without adequate education of health care providers responsible for data interpretation and administration of the fraction of inspired oxygen (FIO2) in NICUs.1,4 It has been estimated that >75% of neonatal health care workers have insufficient knowledge of neonatal oxygenation.4 The majority do not know how SpO2 monitors work or the differences between SpO2 monitors. Moreover, they have limited understanding of factors that affect the arterial oxygen saturation of the hemoglobin molecule, such as shifts in the oxyhemoglobin dissociation curve according to 2,3-diphosphoglycerate content, hemoglobin F concentration, and the Bohr effect. Many do not fully master the concepts of the partial pressure of a gas required to achieve 50% enzyme saturation (P50), oxygen content, oxygen delivery, and the alveolar gas equation.4 Therefore, many neonatal care providers do not know how high the PaO2 can be when the SpO2 reads 97% to 100% for newborns breathing supplemental oxygen. Furthermore, the performance of the current SpO2 monitors is not uniform, as described elsewhere.1,5,6

The normal oxygen saturation in healthy newborns breathing room air is ≥93% and varies according to postnatal age.7–12 In addition, in vivo intravascular arterial saturation was reported by Wilkinson et al13 to be between 85% and 95% in infants with pulmonary disease who were in stable condition and were breathing spontaneously. The same authors found a wide variation in intravascular arterial saturation values with PaO2 values between 30 and 90 mmHg in transfusion-treated and non–transfusion-treated infants.14 However, currently we do not know the optimal SpO2 levels for newborns who receive oxygen therapy. In many NICUs, the FIO2 administered is aimed to maintain SpO2 values equal to those of healthy newborns in room air, but this may be unnecessary and/or may lead to periods of hyperoxemia.1

In addition, we do not know exactly what "normal" PaO2 is. On the basis of P50, saturation, and oxygen content, PaO2 of >40 mmHg should be adequate for tissue needs during early neonatal life, given normal hemoglobin concentrations, cardiac output, blood flow, and cellular conditions. A number of publications and the American Academy of Pediatrics suggest that PaO2 values above 80 to 90 mmHg may be considered hyperoxemia, but these values are not based on systematically performed studies.15–17 On the basis of the known characteristics of the oxyhemoglobin dissociation curve and the alveolar gas equation, it is assumed that SpO2 values between 95% and 100% may be associated with hyperoxemia in infants receiving oxygen therapy. Because of the paucity of studies, however, it is still not known how frequently this actually occurs in current clinical practice.

Twenty years ago, Reynolds and Yu18 suggested guidelines for the use of SpO2 values, which included setting the lower alarm limits at 85% and the upper limits at 90% for newborns with acute respiratory distress, but the safety and potential benefits of such recommendations were not evaluated. Several studies suggested that avoiding SpO2 values of >95% may be beneficial for preterm infants breathing supplemental oxygen. In 2003, we reported that the use of adequate SpO2-monitoring technology, with the aim of avoiding SpO2 levels between 95% and 100% and wide fluctuations in SpO2 levels, was associated with lower morbidity rates in infants of <1500 g.19 Saugstad20 reported that lower SpO2 targets during the first week of life for preterm infants reduced complication rates and might improve growth. In 2006, we reported that a practice aimed at avoiding high SpO2 levels was associated with improved short- and long-term outcomes in very low birth weight infants.21 A systematic review of the literature indicated that liberal versus restricted oxygen exposure among low birth weight infants was associated with potential harms and no clear benefit.22 Most recently, it was reported that setting lower SpO2 alarm limits decreased the severity of retinopathy of prematurity.23 Furthermore, not only do we not know the normal SpO2 for infants breathing supplemental oxygen but also, as recently shown,24 success in maintaining the intended SpO2 range varies substantially among centers, among infants within centers, and in individual infants over time. It is significant that none of those investigations reported PaO2 measurements or compared the SpO2 values with PaO2 measurements.

The concept of normoxemia is not well defined in neonatal medicine, and this study was not designed to identify this. However, a valid concern of many neonatal care providers is the risk of possible persistence of significant hypoxemia at low SpO2 ranges in newborns undergoing oxygen therapy. An important objective in neonatal care would be to better understand which SpO2 ranges are associated with greater or lesser potential risk of PaO2 values linked with the potential for hypoxia and hyperoxia.

With the objective of better defining the relationship between PaO2 and SpO2 during routine clinical practice in NICUs, we designed this prospective study to analyze PaO2 values at different SpO2 ranges among newborns with arterial catheters. Our hypotheses were that, during routine clinical practice in NICUs, the SpO2 range of 85% to 93% would not be associated with PaO2 levels of <40 mmHg and that SpO2 values between 94% and 100% would pose a risk for PaO2 levels that might lead to oxidative stress in infants breathing supplemental oxygen.


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
After approval from the institutional review boards of Emory University and the University of the North, we performed a prospective, noninterventional comparison between noninvasive SpO2 readings and PaO2 values obtained from arterial samples for newborn infants in 7 different NICUs, between July 2005 and November 2006. Informed consent requirements were waived because nothing was changed in clinical practice, no additional blood samples were obtained, and the study subjects did not have any identifiers. In this multicenter international study, the samples were collected in 5 NICUs in the United States (Egleston Hospital, Grady Memorial Hospital, Emory Crawford Long Hospital, and Hillandale Hospital in Atlanta, Georgia, and Morristown Memorial Hospital in Morristown, New Jersey) and in 2 NICUs in Barranquilla, Colombia (Clinic of the Sea and Hospital of the Infant Jesus). All centers are located at or near sea level.

During this study, we did not modify any clinical practices or management or care protocols. The arterial samples were obtained, as indicated by the clinicians caring for the infants, from arterial catheters that had been placed previously on the basis of indications for clinical management. For the samples to be included, the infants needed to be without rapid deterioration during sampling. The following steps were taken to include the paired SpO2/PaO2 samples in the study. (1) The sensors of the SpO2 monitor had to be in the same territory as the arterial catheters (ie, both preductal or both postductal). (2) Arterial blood gas samples were ordered for clinical indications (no additional blood samples were obtained). (3) The SpO2 monitor reading was observed and recorded with accuracy by one of the investigators who was present during sampling. (4) SpO2 readings had to remain stable for 60 seconds before and 60 seconds after the blood gas sampling, with an accepted maximal variation of SpO2 of no more than 1%. Finally, each subject included in the study could have ≥1 paired PaO2/SpO2 sample over the days, but no more than 10 total samples were included for each individual.

To increase the generalizability of the findings into clinical practice, the SpO2 and PaO2 values were obtained by using the saturation monitors and blood gas analyzers used routinely in each NICU. The SpO2 monitors included Nellcor N300 and Nellcor N395 (Nellcor, Pleasanton, CA), Ohmeda Bios 3700 (Datex-Ohmeda, Madison, WI), Novametrix Oxypleth (Novametrix Medical Systems, Wallingford, CT), and Masimo SET series (Masimo, Irvine, CA) monitors. The arterial blood gas analyzers were from Abbott Laboratories (Abbott Park, IL), Datex-Ohmeda (Madison, WI), and Radiometer (Copenhagen, Denmark).

We decided a priori to include ≥100 different infants and as close as possible to 1000 paired PaO2/SpO2 samples (ie, a maximum of 10 samples per infant). For the analyses and comparisons of the samples, we selected PaO2 of <40 mmHg as a "low" PaO2 value. We based this on P50, percentage saturation, and oxygen content calculations. With a PaO2 value of 40 mmHg, hemoglobin may be 85% saturated. If the hemoglobin concentration is 14 g/dL, then the oxygen content would be ~16.2 mL/dL. This oxygen content is likely to be sufficient to avoid neonatal tissue hypoxia at usual ranges and interactions with neonatal cardiac output, regional blood flow, oxygen consumption, and oxygen delivery. In addition, all neonatal clinicians we surveyed would not allow a PaO2 of <40 mmHg for long, but varying proportions would accept PaO2 values of 40 to 45 mmHg and more would accept PaO2 values of 45 to 50 mmHg without additional increases in the FIO2 if the infant was otherwise well. We chose PaO2 of >80 mmHg as a "high" PaO2. Above a certain PaO2 (not precisely known in all circumstances), the dissolved PaO2 is unnecessary and adds very little to the oxygen content of the blood; in addition, it is potentially harmful. How frequently this happens with PaO2 values of >80 mmHg is not known, as it is with PaO2 values of >70 mmHg, >75 mmHg, or >90 mmHg.

Exclusion criteria included major congenital anomalies and acute clinical changes or changes in SpO2 of >1% during arterial blood sampling. Echocardiographic confirmation of patent ductus arteriosus and patent ductus arteriosus under treatment were not exclusion criteria.

Demographic data included gestational age, birth weight, gender, diagnosis, and postnatal age at the time of the study. In addition, we collected data on FIO2 at the same time of the arterial PaO2 and SpO2 sampling. Discrete demographic data were summarized with proportions and frequencies, and continuous data were summarized with means, medians, SDs, and ranges. Repeated-measures analysis of variance was used for comparisons of PaO2 and SpO2 within and between groups. Significance was accepted at the P < .05 level. We also performed graphic representations for the PaO2 and SpO2 data and calculated the best linear and curvilinear relationships. Statistical analyses were performed by using SPSS 13.0 for Windows (SPSS, Chicago, IL).


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
During the study period, 976 paired SpO2/PaO2 samples were obtained from 122 subjects. The gestational age was 29.2 ± 5.2 weeks and the birth weight 1338 ± 871.5 g. The median postnatal age at the time of sampling was 3 days (range: 1–38 days). The location of arterial catheters was postductal for 88.2% of the samples. The median FIO2 at the time of sampling was 0.34 (range: 0.21–1.0). In 41% of the cases, the FIO2 was between 0.25 and 0.40; in 14%, it was >0.70.

Of the 976 paired SpO2/PaO2 samples, 18% (176 samples) were obtained from newborns breathing room air (FIO2: 0.21) and 82% (800 samples) were obtained from newborns receiving oxygen therapy. The mean number of samples per subject was 8 ± 2.9, with a median of 9 samples per infant. The median numbers of samples and study subjects per center were 141 and 15, respectively.

Of all 976 PaO2 samples, SpO2 was <85% in 56 samples, 85% to 93% in 390, and >93% in 530. As expected, the mean PaO2 values (43.5, 58.5, and 94 mmHg, respectively) were significantly different (P < .001, analysis of variance for repeated measures). Table 1 shows the means, medians, and interquartile ranges for PaO2 and SpO2 in infants breathing room air or supplemental oxygen. Table 1 also shows the proportions of all samples with PaO2 values of >80 mmHg or <40 mmHg and with SpO2 values of 85% to 93% or >93%. In room air, 3 infants had SpO2 values of <85%, and they were then treated with supplemental oxygen.


View this table:
[in this window]
[in a new window]

  		TABLE 1 PaO2 and SpO2 in Room Air and With Supplemental Oxygen

 
Table 2 summarizes the means, medians, interquartile ranges, and ranges for newborns receiving supplemental oxygen. In the 325 samples (40.6%) with SpO2 values of 85% to 93%, the mean and median PaO2 values and the proportions of samples with PaO2 of >80 mmHg were statistically different from those for the 422 samples (52.7%) with SpO2 values of >93% (P < .001, repeated-measures analysis of variance) (Table 2).


View this table:
[in this window]
[in a new window]

  		TABLE 2 PaO2 Values According to SpO2 Range for Infants Breathing Supplemental Oxygen

 
The results from the paired PaO2/SpO2 samples from this clinical study are plotted in Figs 1 and 2. The graphs show a linear correlation and a more-physiologic curvilinear relationship, resembling the oxyhemoglobin dissociation "sigmoid" curve. Figure 1 shows the analyses for samples collected from newborns breathing room air. In these 176 samples, when the SpO2 values were 85% to 93%, the mean PaO2 was 58.3 ± 14.2 mmHg and the median was 57 mmHg (interquartile range: 15 mmHg; range: 36–113 mmHg); the PaO2 values were between 40 and 80 mmHg in 86% of the samples. When the SpO2 values were >93%, the mean PaO2 was 68.8 ± 15.9 mmHg and the median was 66 mmHg (interquartile range: 23 mmHg; range: 34–120 mmHg).


Figure 1
View larger version (16K):
[in this window]
[in a new window]

  		FIGURE 1 Oxyhemoglobin dissociation curve for newborns breathing room air. O indicates each observed paired PaO2 and SpO2 value; Linear; linear relation of the samples; S, curvilinear (sigmoid) relation of the samples.

 

Figure 2
View larger version (21K):
[in this window]
[in a new window]

  		FIGURE 2 Oxyhemoglobin dissociation curve for newborns under oxygen therapy (FIO2 > 0.21). O indicates each observed paired PaO2 and SpO2 value; Linear, linear relation of the samples; S, curvilinear (sigmoid) relation of the samples.

 
Figure 2 depicts the oxyhemoglobin dissociation curve for the samples obtained from newborns receiving supplemental oxygen. As shown, analyses of the samples with SpO2 values between 85% and 93% revealed that the majority (86.8%) were associated with PaO2 values of 40 to 80 mmHg, 4.6% with PaO2 values of >80 mmHg, and 8.6% with PaO2 values of <40 mmHg. Figure 2 also shows that, when the SpO2 values were >93%, 59.5% of the PaO2 values were >80 mmHg (Table 2) and <40% of the PaO2 values were between 40 and 80 mmHg (P < .001, compared with samples with SpO2 values of 85%–93%). Furthermore, Fig 3 shows a receiver operating characteristic curve for infants breathing supplemental oxygen and for samples with SpO2 values of >93%, considering PaO2 of >80 mmHg as positive. The area under the curve was 0.74, with an asymptotic 95% confidence interval of 0.68 to 0.80.


Figure 3
View larger version (7K):
[in this window]
[in a new window]

  		FIGURE 3 ROC curve for samples with SpO2 of >93% in relation to PaO2 of >80 mmHg in infants breathing supplemental oxygen. Area under the curve: 0.74; Asymptotic 95% confidence intervals: lower bound 0.68, upper bound 0.80.

 

   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
On the basis of the findings of this prospective, descriptive, noninterventional clinical study, we can conclude that, among newborns receiving oxygen therapy, SpO2 values of 85% to 93% are associated very infrequently with PaO2 values of <40 mmHg. Furthermore, this SpO2 range is associated with PaO2 values of >80 mmHg much less frequently than when SpO2 levels are 94% to 100%. To our knowledge, this is the first description of the relationship between PaO2 and SpO2 values in a multicenter, prospective, neonatal clinical study using different saturation monitors at sea level.

A few studies published since the late 1980s15,16,25–28 compared PaO2 values with SpO2 values. Their main objective was to assess the reliability of SpO2 values in the detection of hyperoxemia (defined in different studies as >80 or >90 mmHg). The authors of the largest study, which was performed in a single center with 792 readings, reported concern about using SpO2 values as the sole means of oxygen monitoring for preterm infants receiving supplementary oxygen and considered mandatory the combination of SpO2 monitoring with other methods of blood gas monitoring.15 Other, similar, single-center studies had only 46 to 291 measurements.16,25–28 All studies suggested the potential benefits of setting high alarm limits at 94% or 95%. The methods and goals of our study were different from those of all of those studies. In this study, we aimed to define more completely the relationship of PaO2 and SpO2 values during routine clinical practice in NICU and to define the risk of a PaO2 value that most clinicians currently would consider too low (<40 mmHg) when SpO2 values were 85% to 93%. This information may prove to be valuable, because PaO2 values are measured fairly infrequently in current clinical practice.

In this study, of the 325 samples obtained from infants breathing oxygen with SpO2 values between 85% and 93%, PaO2 values were 40 to 80 mmHg in 86.8% of the samples and >80 mmHg in 4.6%. Twenty-eight samples (8.6%) had PaO2 values of <40 mmHg. Of those 28 samples, 82% had SpO2 values of 85% to 90%. Although this event was not frequent, this PaO2 may in fact be too low for specific infants and of concern to clinicians. All of the infants who had PaO2 values of <40 mmHg during this study were clinically reevaluated, just as it would be done routinely. In all cases, either the PaO2 measurement was repeated and the result was found to be >40 mmHg in the new sample or the infant's PaO2 improved after a minimal increase of the FIO2.

At the high end of the SpO2 range (>93%), PaO2 values that may be considered high by most neonatal clinicians (>80 mmHg) occurred in almost 60% of the samples from newborns receiving oxygen therapy. In most cases, such PaO2 levels are likely to be unnecessary. Of course, PaO2 values of >80 mmHg can (and do) occur sometimes among infants breathing room air. This can happen when alveolar carbon dioxide pressures decrease (during tachypnea, crying, or intermittent mandatory ventilation), the alveolar oxygen pressure increases, and there is adequate pulmonary blood flow and minimal extrapulmonary and intrapulmonary shunting (as can occur in NICUs with continuous positive airway pressure or positive end-expiratory pressure ventilation). In this study, 29 samples (16.5%) from infants breathing room air had PaO2 values of >80 mmHg (Table 1); 25 of these samples were with SpO2 values of >93%. Whether a PaO2 level of >80 mmHg is always associated with more oxidant damage and, if so, to what degree remain to be fully investigated.

Our study has the strength of having been performed in various centers, which increases the generalizability of the results, but it also has several weaknesses. For example, we elected not to evaluate the effects that PaCO2, pH, body temperature, fetal hemoglobin concentration, adult hemoglobin concentration, 2,3-diphosphoglycerate level, and number of transfusions might have on the relationship between PaO2 and SpO2 in clinical practice. Furthermore, we did not evaluate the possible relationship between the time after transfusion and the time of PaO2 and SpO2 sampling. All of these factors, by themselves and in combination, could have potential effects on hemoglobin affinity for oxygen and change the P50. In daily clinical practice, however, most clinicians follow the values shown by the SpO2 monitors and do not analyze the aforementioned factors and their potential involvement in shifts of the dissociation curve. In addition, we elected not to calculate or to estimate the P50 in these samples, because no SpO2 value in this study was actually ≤50%, which would limit the validity of performing such calculations. Not verifying or modifying any setting of the blood gas machines used routinely in the different participating units might be another weakness, from a strictly physiologic point of view. However, we purposely elected not to make any modifications, to increase the generalizability of the findings to routine NICU clinical practice at sea level. We elected not to include any NICUs at higher altitudes but, on the basis of physiologic concepts and the alveolar gas equation, it is very likely that the results would be similar for neonates breathing supplemental oxygen at high altitude. Finally, by design, the subjects in this study were in relatively stable condition hemodynamically. The goal of our study was not to study the effects on PaO2 values and their relationship to SpO2 values during situations of acute desaturation, low perfusion, or motion or to analyze response times or numbers of false alarms or false readings for different monitors. This was studied previously, and the advantages of SpO2 monitors with signal extraction technology (eg, Masimo SET series) in such situations were described previously.1,27,29,30 In brief, pulse oximeters are calibrated to display functional saturation, as required by the Food and Drug Administration. However, some monitors are better than others and the bias, accuracy, and precision of the equipment differ, more so during unstable conditions and at lower SpO2 values. This could lead to differences with the arterial saturation measured with a CO-oximeter of even >3% and to a similar disparity in simultaneous SpO2 readings with different SpO2 monitors. Therefore, rigid arguments regarding whether it is better to use a SpO2 value of 84% or 86% (or 83% or 85%) as the lower limit is a bit presumptuous. We are planning to perform a new study to explore whether there are differences in the PaO2/SpO2 relationship with different monitors under fairly stable neonatal clinical conditions and to perform a similar analysis at higher altitude.

As mentioned previously, it is not scientifically known what the optimal PaO2 value is at all times for each condition in each infant. Therefore, we are not suggesting that the values we chose are the best. Also, we do not know what the high PaO2 associated with oxidative stress is. Accepting this ignorance in clinical care, we cannot ignore that dissolved oxygen not only is unnecessary but adds very little to the blood oxygen content. As would be expected, the receiver operating characteristic curve demonstrates that SpO2 values of >93% are not 100% specific or sensitive for PaO2 levels of >80 mmHg. However, the area of 0.74 found supports the concern raised by several investigators regarding potential hyperoxemia with SpO2 values of ≥94%. After the study was completed, we performed a posthoc analysis of the PaO2 levels measured at different SpO2 readings in infants breathing supplemental oxygen. The PaO2 values were >80 mmHg in 36.7% of the samples when the SpO2 values were 94% to 96%, but this occurred in 80% of the samples when the SpO2 values were ≥97%. These results suggest that the occurrence of PaO2 levels of >80 mmHg is markedly elevated when SpO2 values are >96%. Also, we analyzed the PaO2 values of <40 mmHg when the SpO2 values were 85% to 93% in infants breathing supplemental oxygen. Only 5 samples had hypoxemic PaO2 values of <40 mmHg when SpO2 values were 91% to 93%. It seems unlikely that we will be able to identify a single precise ideal SpO2 or PaO2 value for each individual infant at all times by using currently available technology. The optimal SpO2 value is likely to vary on the basis of many factors, including the monitor used, gestational age, postnatal age, clinical condition and diagnosis, pH, PaCO2, body temperature, total and fetal hemoglobin concentrations, 2,3-diphosphoglycerate content, transfusions, and other factors. In addition, oxygen delivery and tissue oxygenation depend much less on the PaO2 level than on other factors such as hemoglobin concentration, hemoglobin shifts (which are frequent in ill neonates), oxygen content, cardiac output, and regional blood flow, among others. In addition, the known or potential benefits of reductions in hyperoxemia and hypoxemia need to be balanced against the challenges of maintaining a specific PaO2 range.

Although many things remain to be learned regarding oxidative damage in newborns, we as health care providers are likely the only cause of hyperoxemia and hyperoxia, unlike hypoxia, which can be caused by many medical conditions. Better methods of assessing tissue oxygenation and altered redox status in clinical practice may someday become available. Until then, identifying practices that reduce the risk and/or likely will not be associated with a potential for hyperoxia and hyperoxic damage can only be beneficial for many infants.

With this in mind, the findings of this study demonstrate that in the great majority of cases there is no need to maintain SpO2 values of >93% in clinical practice to preserve a PaO2 level that may be sufficient for tissue needs in preterm infants breathing supplemental oxygen during the initial periods of their life and critical illness. Doing so may unnecessarily increase the exposure to hyperoxemia and increase the risk of potential deleterious effects of radical oxygen species, which were previously described by several authors31–40 and we recently summarized.41 We do not advocate 85% to 93% as the "best" SpO2 range. We aimed to determine whether such a range posed a high risk for clear and persistent hypoxia, a concern in clinical practice because we reported previously that this range is associated with improvement in important outcomes.19 After those initial results, we investigated whether short- and/or long-term outcomes were adversely affected when these arbitrarily chosen lower SpO2 targets were implemented. There were no adverse effects found and important outcomes improved in that study.21 Also, we very recently reported that there are significant gender-specific differences favoring female patients in the beneficial effects42; this adds support to the fact that there is no single best range for all infants at all times.


   CONCLUSIONS
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
On the basis of these data, we conclude that use of SpO2 values between 85% and 93% markedly decreases rates of PaO2 values of >80 mmHg and is not associated with significant hypoxemia in neonates breathing supplemental oxygen in NICUs. The data also show that there is a large proportion of samples with PaO2 values of >80 mmHg when the SpO2 values are >93%. Although SpO2 ranges between 85% and 93% seem sufficient to maintain normoxemia most of the time during the early NICU course, a future goal would be to find even safer SpO2 limits or improved technology that can assist clinicians in eliminating both hyperoxemia and hypoxemia during the early neonatal period.


   ACKNOWLEDGMENTS
 
The study was supported in part by a Goodard Scholarship (Dr Sola) and by funds from the Mid-Atlantic Neonatology Associates and Atlantic Neonatal Research Institute (Dr Sola).

We thank the anonymous reviewers who contributed to substantial improvements in the manuscript. We give special thanks to Dr Marta Rogido for her outstanding help during this study and to Dr Ronald Goldberg for his assistance in the preparation of the revised manuscript. We are thankful to the nurses and respiratory therapists who cared for these infants and notified us at the time of planned blood gas measurements.


   FOOTNOTES
 
Accepted Sep 17, 2007.

Address correspondence to Augusto Sola, MD, Division of Neonatology, Morristown Memorial Hospital, 100 Madison Ave, Morristown, NJ 07960. E-mail: augustosolaneo{at}gmail.com

The authors have indicated they have no financial relationships relevant to this article to disclose.


What's Known on This Subject

We and others have reported on oxygen monitoring and the risks of oxidative stress. We showed that clinical practices that aim to avoid SpO2 ranges that may be associated with hyperoxemia are associated with less morbidity in preterm infants.

 

What This Study Adds

We evaluated the risk of potential tissue hypoxia with SpO2 ranges used. PaO2 levels of <40 mm Hg were very infrequent with SpO2 targets of 85% to 93%. Hyperoxemia (with potential oxidative damage) was frequent with higher SpO2 targets.

 


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 

  1. Sola A, Chow L, Rogido M. Pulse oximetry in neonatal care in 2005: a comprehensive state of the art review [in Spanish]. An Pediatr (Barc). 2005;62 (3):266 –281[CrossRef][Medline]
  2. Mower WR, Sachs C, Nicklin E, et al. Pulse oximetry as a fifth pediatric vital sign. Pediatrics. 1997;99 (5):681 –686[Abstract/Free Full Text]
  3. Neff TA. Routine oximetry: a fifth vital sign? Chest. 1988;94 (2):227
  4. Sola A. Education in neonatal oxygenation has been insufficient: a need for darning. Presented at the Pediatric Academic Societies meeting; May 5–8, 2007; Toronto, ON, Canada
  5. Salyer JW. Neonatal and pediatric pulse oximetry. Respir Care. 2003;48 (4):386 –396[Medline]
  6. Jubran A. Pulse oximetry. Crit Care. 1999;3 (2):R11 –R17[CrossRef][Web of Science][Medline]
  7. Richard D, Poets CF, Neale S, Stebbens VA, Alexander JR, Southall DP. Arterial oxygen saturation in preterm neonates without respiratory failure. J Pediatr. 1993;123 (6):963 –968[CrossRef][Web of Science][Medline]
  8. Poets CF, Stebbens VA, Lang JA, O'Brien LM, Boon AW, Southall DP. Arterial oxygen saturation in healthy term neonates. Eur J Pediatr. 1996;155 (3):219 –223[CrossRef][Web of Science][Medline]
  9. Toth B, Becker A, Seelbach-Gobel B. Oxygen saturation in healthy newborn infants immediately after birth measured by pulse oximetry. Arch Gynecol Obstet. 2002;266 (2):105 –107[CrossRef][Medline]
  10. Levesque BM, Pollack P, Griffin BE, Nielsen HC. Pulse oximetry: what's normal in the newborn nursery? Pediatr Pulmonol. 2000;30 (5):406 –412[CrossRef][Web of Science][Medline]
  11. Rabi Y, Yee W, Chen SY, Singhal N. Oxygen saturation trends immediately after birth. J Pediatr. 2006;148 (5):590 –594[CrossRef][Web of Science][Medline]
  12. Kamlin CO, O'Donnell CP, Davis PG, Morley CJ. Oxygen saturation in healthy infants immediately after birth. J Pediatr. 2006;148 (5):585 –589[CrossRef][Web of Science][Medline]
  13. Wilkinson AR, Phibbs RH, Gregory GA. Continuous in vivo oxygen saturation in newborn infants with pulmonary disease: a new fiberoptic catheter oximeter. Crit Care Med. 1979;7 (5):232 –236[Web of Science][Medline]
  14. Wilkinson AR, Phibbs RH, Heilbron DC, Gregory GA, Versmold HT. In vivo oxygen dissociation curves in transfused and untransfused newborns with cardiopulmonary disease. Am Rev Respir Dis. 1980;122 (4):629 –634[Web of Science][Medline]
  15. Paky F, Koeck CM. Pulse oximetry in ventilated preterm newborns: reliability of detection of hyperoxaemia and hypoxaemia, and feasibility of alarm settings. Acta Paediatr. 1995;84 (6):613 –616[Web of Science][Medline]
  16. Poets CF, Wilken M, Seidenberg J, Southall DP, von der Hardt H. Reliability of a pulse oximeter in the detection of hyperoxemia. J Pediatr. 1993;122 (1):87 –90[Web of Science][Medline]
  17. American Academy of Pediatrics, American College of Obstetricians and Gynecologists. Guidelines for Perinatal Care. 5th ed. Elk Grove Village, IL: American Academy of Pediatrics; 2002:246
  18. Reynolds GJ, Yu VY. Guidelines for the use of pulse oximetry in the non-invasive estimation of oxygen saturation in oxygen-dependent newborn infants. Aust Paediatr J. 1988;24 (6):346 –350[Web of Science][Medline]
  19. Chow LC, Wright KW, Sola A, CSMC Oxygen Administration Study Group. Can changes in clinical practice decrease the incidence of severe retinopathy of prematurity in very low birth weight infants? Pediatrics. 2003;111 (2):339 –345[Abstract/Free Full Text]
  20. Saugstad OD. Oxygen for newborns: how much is too much? J Perinatol. 2005;25 (suppl 2):S45 –S49[CrossRef][Medline]
  21. Deulofeut R, Critz A, Adams-Chapman I, Sola A. Avoiding hyperoxia in infants <1250 g is associated with improved short- and long-term outcomes. J Perinatol. 2006;26 (11):700 –705[CrossRef][Web of Science][Medline]
  22. Askie LM, Henderson-Smart DJ. Restricted versus liberal oxygen exposure for preventing morbidity and mortality in preterm or low birth weight infants. Cochrane Database Syst Rev. 2001;(4):CD001077
  23. Vanderveen DK, Mansfield TA, Eichenwald EC. Lower oxygen saturation alarm limits decrease the severity of retinopathy of prematurity. J AAPOS. 2006;10 (5):445 –448[CrossRef][Medline]
  24. Hagadorn JI, Furey AM, Nghiem TH, et al. Achieved versus intended pulse oximeter saturation in infants born less than 28 weeks' gestation: the AVIOx study. Pediatrics. 2006;118 (4):1574 –1582[Abstract/Free Full Text]
  25. Baeckert P, Bucher HU, Fallenstein F, Fanconi S, Huch R, Duc G. Is pulse oximetry reliable in detecting hyperoxemia in the neonate? Adv Exp Med Biol. 1987;220 :165 –169[Medline]
  26. Adams JM, Murfin K, Gullikson M. Detection of hyperoxemia in neonates by a new pulse oximeter. Neonatal Intensive Care. 1994;7 (3):42 –45[Medline]
  27. Bohnhorst B, Peter CS, Poets CF. Detection of hyperoxaemia in neonates: data from three new pulse oximeters. Arch Dis Child Fetal Neonatal Ed. 2002;87 (3):F217 –F219[Abstract/Free Full Text]
  28. Bucher HU, Fanconi S, Baeckert P, Duc G. Hyperoxemia in newborn infants: detection by pulse oximetry. Pediatrics. 1989;84 (2):226 –230[Abstract/Free Full Text]
  29. Workie FA, Rais-Bahrami K, Short BL. Clinical use of new-generation pulse oximeters in the neonatal intensive care unit. Am J Perinatol. 2005;22 (7):357 –360[CrossRef][Web of Science][Medline]
  30. Robertson FA, Hoffman GM. Clinical evaluation of the effects of signal integrity and saturation on data availability and accuracy of Masimo SE and Nellcor N-395 oximeters in children. Anesth Analg. 2004;98 (3):617 –622[Abstract/Free Full Text]
  31. Sola A, Chow L, Rogido M. Retinopathy of prematurity and oxygen therapy: a changing relationship [in Spanish]. An Pediatr (Barc). 2005;62 (1):48 –63[CrossRef][Medline]
  32. Suzuki YJ, Forman HJ, Sevanian A. Oxidants as stimulators of signal transduction. Free Radic Biol Med. 1997;22 (1–2):269 –285[CrossRef][Web of Science][Medline]
  33. Felderhoff-Mueser U, Bittigau P, Sifringer M, et al. Oxygen causes cell death in the developing brain. Neurobiol Dis. 2004;17 (2):273 –282[CrossRef][Web of Science][Medline]
  34. Handley S, Wang XP, Sola A, et al. Hyperoxia causes oxidative stress and increases neural stem cells proliferation in the murine developing brain. E-PAS. 2006;59 :5168.6
  35. Spector LG, Klebanoff MA, Feusner JH, Georgieff MK, Ross JA. Childhood cancer following neonatal oxygen supplementation. J Pediatr. 2005;147 (1):27 –31[CrossRef][Web of Science][Medline]
  36. Sola A, Deulofeut R, Rogido M. Oxygen and oxygenation in the delivery room. J Pediatr. 2006;148 (4):564 –565[Web of Science][Medline]
  37. Døhlen G, Carlsen H, Blomhoff R, Thaulow E, Saugstad OD. Reoxygenation of hypoxic mice with 100% oxygen induces brain nuclear factor-{kappa}B. Pediatr Res. 2005;58 (5):941 –945[CrossRef][Web of Science][Medline]
  38. Vento M, Asensi M, Sastre J, et al. Hyperoxemia caused by resuscitation with pure oxygen may alter intracellular redox status by increasing oxidized glutathione in asphyxiated newly born infants. Semin Perinatol. 2002;26 (6):406 –410[CrossRef][Web of Science][Medline]
  39. Saugstad OD. Oxygen toxicity at birth: the pieces are put together. Pediatr Res. 2003;54 (6):789[CrossRef][Web of Science]
  40. Maniscalco WM, Watkins RH, Roper JM, Staversky R, O'Reilly MA. Hyperoxic ventilated premature baboons have increased p53, oxidant DNA damage and decreased VEGF expression. Pediatr Res. 2005;58 (3):549 –556[CrossRef][Web of Science][Medline]
  41. Sola A, Rogido MR, Deulofeut R. Oxygen as a neonatal health hazard: call for détente in clinical practice. Acta Paediatr. 2007;96 (6):801 –812[Web of Science][Medline]
  42. Deulofeut R, Golde D, Sola A. Treatment-by-gender effect when aiming to avoid hyperoxia in preterm infants in the NICU. Acta Paediatr. 2007;96 (7):990 –994[Web of Science][Medline]
PEDIATRICS (ISSN 1098-4275). ©2008 by the American Academy of Pediatrics

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Facebook Facebook   Add to Reddit Reddit   Add to Technorati Technorati   Add to Twitter Twitter    What's this?


This article has been cited by other articles:


Home page
 PediatricsHome page
A. Sola
Oxygen for the Preterm Newborn: One Infant at a Time
Pediatrics, June 1, 2008; 121(6): 1257 - 1257.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Castillo, A.
Right arrow Articles by Critz, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Castillo, A.
Right arrow Articles by Critz, A.
Related Collections
Right arrow Premature & Newborn
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Facebook   Add to Reddit   Add to Technorati   Add to Twitter  
What's this?