Operational Evaluation of Pulse Oximetry in NICU Patients with Arterial Access

Original ...... Article Operational Evaluation of Pulse Oximetry in NICU Patients with Arterial Access

Dale Gerstmann, MD CONCLUSION: Ryan Berg, RRT These operational data suggest that with the methodology and devices

Ron Haskell, RRT currently in use, SpO2 values in most all neonates who require arterial Cathy Brower, RRT lines inaccurately correlate with measured arterial saturation. Kari Wood, MSN Journal of Perinatology (2003) 23, 378–383. doi:10.1038/sj.jp.7210944 Brad Yoder, MD Loren Greenway, PhD Gordon Lassen, RRT Robert Ogden, RRT INTRODUCTION Ronald Stoddard, MD Stephen Minton, MD The use of pulse oximetry in the neonatal intensive care unit (NICU) to monitor oxygenation has been a standard of practice for many years, gaining widespread use in the mid-1980s. Studies on the accuracy and reliability of neonatal pulse oximetry have 1–4 OBJECTIVE: been favorable. Although pulse oximetry oxygen saturation To investigate pulse oximetry in neonates who require arterial access as (SpO2) readings have not been accurate predictors of arterial 5–8 represented by the clinical data recorded to manage their care. oxygen tension (PaO2), pulse oximetry in our NICU and in probably most NICUs remains a critical detector for desaturation STUDY DESIGN: and hypoxemia events, and as an oxygenation monitor during Analysis of simultaneous SpO2 and SaO2 from: 7-year historical NICU oxygen therapy and assisted ventilation. data (N ¼ 31,905); 4-month prospective NICU data (N ¼ 566); The issue at hand developed from an index case in July 2001 veriﬁcation data using two hemoximeters (N ¼ 52); and NICU data from in our NICU. An enterprising neonatal respiratory therapist two collaborating centers (N ¼ 95 and 168). The bias function presented the last shift’s worth of blood gas and hemoximetry slips (SpO2ÀSaO2) was regressed against the measured ‘‘gold’’ standard, SaO2. from a sick, ventilated, preterm infant, which were overwritten RESULTS: with corresponding SpO2 values from the bedside monitor. The A signiﬁcant negative correlation was found for each of the data sets therapist commented, ‘‘The saturations really don’t match.

between the bias function and SaO2. This bias was similar for devices from What do you want me to do about it?’’ A tabulation of several manufacturers (Datex-Ohmeda, Masimo, Nellcor, and Spacelabs). all of the patient’s arterial hemoximetry measured saturations, Maximum operational performance occurred with peaks between 92 and corresponding pulse oximetry saturation readings, and blood

97% SaO2, but declined markedly above and below this narrow range. In gas PaO2 values demonstrated what appeared to be a systematic

all, 71 to 95% of patients exhibited data with significant bias. saturation-dependent difference between pulse and hemoximeter saturations, worse at low PaO2 values where SpO2 seemed to markedly overestimate the arterial measured saturation (SaO2). We worried that some procedural issue in our use of pulse oximetry might have occurred, or that this patient was illustrating a Neonatology Department (D.G., R.B., R.H., C.B., K.W., G.L.,R.O., R.S., S.M.), Utah Valley Regional Medical Center, Provo, UT, USA; Santa Rosa Children’s Hospital (B.Y.), San Antonio, TX, US; and measurement bias that we had not previously recognized. Hence, ALDS Hospital (L.G.), Salt Lake City, UT, USA. we began a systematic review of preanalytic, analytic, and These data were presented in part as a Late Breaker Abstract at the Pediatric Academic Societies postanalytic process and procedures to try and elucidate the mdash; Society for Pediatric Research Meeting, May 2002, Baltimore, MD, USA. An FDA Medical problem. We did not wish to perform a controlled scientific Device Report (#1024269) was filed with Med Watch on March 5, 2002. experiment on pulse oximetry function, but rather sought to Address correspondence and reprint requests to Dale R. Gerstmann, MD, Neonatology analyze available data recorded during the process of providing Department, Utah Valley Regional Medical Center, 1040 North 500 West, Provo, UT 84604, USA. Dale R. Gerstmann can also be contacted at: Pediatrix Medical Group, Office of Research and routine care, which might illustrate the operational performance of Education, 1301 Concord Terrace, Sunrise, FL 33323-2825, USA. this technology.

Presented here are our findings from a historical evaluation, a a couple of minutes of the blood gas draw. Quality control prospective evaluation, a procedural evaluation, a verification calibrations were performed every 8 hours. Since a calculated evaluation, and an evaluation of corroborating data from NICUs at SaO2 via a blood gas analyzer cannot be used as a point of two other institutions. comparing SpO2 accuracy, measured, not calculated, SaO2 was reported with the corresponding pre-blood draw baseline SpO2.No METHODS comparison of pulse oximeter pulse rate performance was Saturation Bias Function undertaken. Hemoximeter saturation values can be represented as fractional oxyhemoglobin saturation (O2Hb/THb) or functional saturation Historical Evaluation (O2Hb/(O2Hb+HHb)), where O2Hb ¼ oxyhemoglobin, An archived respiratory care blood gas data file covering the years HHb ¼ deoxyhemoglobin, THb ¼ total hemoglobin. The difference 1991 to 1997 was de-identified and queried for all NICU records. As being that THb ¼ O2Hb+HHb+(other Hb moieties). In neonates, a matter of course, each record represented blood gas analyzer ‘‘other Hb moieties’’ such as carboxyhemoglobin or values and also contained ventilator settings, inspired oxygen methemoglobin are uncommon. In the Verification evaluation fraction, SpO2 reading (Datex-Ohmeda 3700, Madison, WI) at the presented below, other hemoglobin moieties constituted <3% of time of blood gas draw, and hemoximeter (OSM-3, Radiometer, THb. Although elevated carboxyhemoglobin, as may occur with Copenhagen) measured oxyhemoglobin fraction. The queried file extreme hyperbilirubinemia, would render this approximation less was then screened and cleaned for missing values and data entry accurate, bilirubin levels were not available in the collected data errors. Functional SaO2 was derived by converting the available and so no patients were excluded on this basis. arterial fractional oxyhemoglobin value, then the bias term The relation between functional saturation and fractional (SpO2ÀSaO2) was calculated and regressed against SaO2. oxyhemoglobin in neonates can thus be approximated as: functional saturation (fractional oxyhemoglobin/0.97) ¼ 1.03 Prospective Evaluation fractional oxyhemoglobin). Pulse oximetry signal processing During July–October 2001, SpO2 and SaO2 values were collected algorithms produce a value which is equivalent to a ‘‘functional’’ and recorded on NICU patients during routine blood gas analysis. hemoximeter saturation determination. Thus, the appropriate Pulse oximeter models included: Datex-Ohmeda 3700 & 3740, comparison for SpO2 is SaO2, the functional hemoximeter Masimo 2000 (Masimo Corp., Irvine, CA), and Nellcor N-200 saturation. In some data sets examined for this work, the fractional (Nellcor, Pleasanton, CA). SaO2 was measured by integrated oxyhemoglobin value was the one available and it was ‘‘converted’’ hemoximeters in the ABL 725 (Radiometer, Copenhagen) and to functional saturation using the above formula. Chiron CIBA-Corning 865 (Bayer Diagnostics, Tarrytown, NY) The relation between SpO2 and SaO2 was evaluated by blood gas analyzers. The bias function was calculated and plotted. calculating the bias function (SpO2ÀSaO2) after Severinghaus Subanalyses were conducted by pulse oximeter type. No infants et al9 and others,10,11 and this was plotted against the ‘‘gold’’ were receiving nitric oxide, which could have caused increased standard, measured SaO2. Precision was indicated by the methemoglobin levels. 10th and 90th data percentiles in the range of the bias. Procedural Evaluation Pulse Oximeters and Hemoximeters In November 2001, an in-depth procedural evaluation was The pulse oximeters for which SpO2 data could be provided were performed in the NICU to determine whether the methods being the stand-alone or modular units available for use at the individual used by staff for the selection, placement, attachment and use of hospitals, except in the Prospective evaluation, where the Nellcor pulse oximeter sensors and monitors were appropriate and correct. N-200 was provided by one of the manufacturers. The following Observations on the timing of pulse oximeter readings in sensors were reported as being used with these pulse oximeters: correspondence to the drawing of blood for blood gases were made. Flex II or Oxytip+ OX-AF with Datex-Ohmeda models 3700 & 3740; Appropriate processing of blood gas samples and use of the unit’s Nellcor Oxisensor II, I-20 or N-25 with Nellcor model N200; Nellcor blood gas analyzer and integrated hemoximeter were also checked. Oxisensor II, I-20 or N-25 with Space Labs; LNOP-Neo with Masimo model 2000. Similarly, the blood gas analyzer or Verification Evaluation hemoximeter used for SaO2 measurement was the one available in Also in November 2001, an IRB approved study protocol was the NICU for patient care, except in the Verification evaluation initiated, which allowed for the drawing of a small number of where the IL 682 was provided by one of the manufacturers. duplicate arterial blood gas samples in order to verify readings All sites indicated that their procedure for recording the SpO2 value between two different hemoximeters (ABL 725 and IL 682). This was to choose the stable baseline value just prior to initiating was performed in order to rule out a hemoximeter measurement the arterial blood gas draw. Blood gas analysis or hemoximetry error. The paired t-test was used to compare values between was conducted in a dedicated NICU blood gas laboratory within hemoximeters with statistical significance determined at the

Journal of Perinatology 2003; 23:378–383 379 Gerstmann et al. SpO2 in NICU Patients

pr0.05 level. The bias function was calculated for simultaneous SpO2 readings with the Datex-Ohmeda Model 3740 pulse oximeter. No infants were receiving nitric oxide, which could have caused increased methemoglobin levels.

Evaluation of Data from Other Sites For comparison and corroboration, colleagues at two other institutions (Primary Children’s Medical Center, Salt Lake City, UT and Santa Rosa Children’s Hospital, San Antonio, TX) agreed to share de-identiﬁed data collected from routine neonatal blood gas analysis and hemoximetry. The bias function was calculated and plotted for data from the two sites.

Individual Patient Analyses Data collected for the Historical and Prospective evaluations included an unlinked patient study number. In patients where Figure 1. Pulse oximeter bias function: 7-year historical data. Shown sufficient data existed that a calculation seemed feasible, individual are the 10th, 50th, and 90th percentiles for 31,905 NICU records that correlation coefficients were determined. Sufficient data were compare (SpO2ÀSaO2) versus SaO2 obtained in our NICU between 1991 and 1997. The bias function demonstrates a significant negative defined as at least five data pairs of SpO2 and SaO2, and a correlation with SaO2, r ¼À0.67 (p<0.001). The bias function, difference between maximum and minimum SaO2 of at least 5. For (SpO2ÀSaO2), and the data variation (distance between 10th and 90th the historical data set, the last 130 patients who met these criteria percentiles), that is, precision, worsen as saturation values decrease. were used. In the prospective data set, 28 patients met this criterion. Individual correlation coefficients were considered significant at the pr0.05 level. collected data than with the historical data. There were 566 Operational Performance samples available for analysis on 43 patients (see Figure 2). In order to contrast the neonatal performance of the models of Correlation coefficients for the three pulse oximeters used during pulse oximeter for which data were available, operational this period were: r ¼À0.85 (p<0.001) Datex-Ohmeda (Models performance was defined as the frequency with which at any given 3700, 3740), N ¼ 278; r ¼À0.84 (p<0.001) Masimo (Model 2000), N ¼ 139; and r ¼À0.81 (p<0.001) Nellcor (Model SaO2 value the corresponding SpO2 reading was within the technical accuracy specified for neonatal pulse oximetry. In N-200), N ¼ 149. Pulse oximeters from all three manufacturers general, technical specifications for pulse oximeters used in appeared to perform similarly. neonates suggest an accuracy of ±3 digits (1 SD) between 70 and 100% saturation. The accuracy under 70% saturation is Procedural Evaluation unspecified. Operational performance (% of SpO No systematic variances were identified in the NICU staff’s 2 performance in the selection and application of pulse oximetry values ¼ SaO2±3) was determined for the Datex-Ohmeda 3700/ 3740, Masimo 2000, Nellcor N-200, and Spacelabs pulse oximeters sensors, in the maintenance and setting of monitors, in the using data from the Prospective evaluation and the Evaluation of recording of appropriate SpO2 reading in conjunction with arterial data from other sites. blood gas draws, nor in the processing of the arterial blood sample for analysis and hemoximetry. Observations were made by respiratory therapy supervisors, outside respiratory therapy directors, RESULTS and industry engineers and scientists. Historical Evaluation There were 31,905 records for analysis with SaO2 values 57 to Verification Evaluation 100%. Figure 1 shows the 10th, 50th, and 90th percentiles for the In total, 52 arterial blood samples were drawn from 10 NICU saturation bias function (SpO2ÀSaO2) versus SaO2 for the 7-year patients as part of routine blood gas monitoring and these were historical data set. Although the spread in data increases with lower run simultaneously on ABL 725 and IL 682 blood gas analyzers. All SaO2, the bias function shows a significant negative correlation SpO2 readings were taken using the Datex-Ohmeda 3740 pulse with SaO2 (r ¼À0.67, p<0.001). oximeter. The average difference in hemoximeter HbO2 paired values between the two devices was 0.07, with mean values (95% Prospective Evaluation confidence limits), being 88.3 (86.4, 90.3) and 88.4 (86.9, 89.9) A somewhat stronger negative correlation (r ¼À0.86, p<0.001) for the ABL 725 and IL 682 devices, respectively. The difference was of (SpO2ÀSaO2) with SaO2 was seen with the 4-month prospective not statistically significant. (SpO2ÀSaO2) was regressed against

380 Journal of Perinatology 2003; 23:378–383 SpO2 in NICU Patients Gerstmann et al.

Figure 3. Pulse oximeter bias function: other nicus. NICU ‘‘A’’ provided 95 paired SpO2 (Spacelabs) and SaO2 values. NICU ‘‘B’’ Figure 2. Pulse oximeter bias function: 4-month prospective data. provided 168 paired values (Datex-Ohmeda 3740). The correlation Shown is the bias function (SpO2ÀSaO2) data for three pulse oximeters coefficients for the pulse oximetry bias function for NICU ‘‘A’’ and ‘‘B’’ plotted against SaO2. Correlation coefficients for the pulse oximeters were: r ¼À0.92 (p<0.001) and r ¼À0.72 (p<0.001), respectively. used during this period were: Datex-Ohmeda Models 3700 and 3740, At these sites as well, the pulse oximetry bias is seen to worsen as N ¼ 278, r ¼À0.85 (p<0.001); Masimo Model 2000, N ¼ 139, saturation values decrease. r ¼À0.84 (p<0.001); and Nellcor Model N-200, N ¼ 149, r ¼À0.81 (p ¼ 0.001). Pulse oximeters from all three manufacturers appeared to percentages for the three (3) pulse oximeters used in this data set perform in a similar fashion. Overall, 566 samples were available for were: Datex-Ohmeda 3700/3740, 70% (7/10); Masimo 2000, 60% analysis on 43 patients yielding a negative correlation coefficient of (3/5); and Nellcor N-200, 77% (10/13). r ¼À0.86 (p<0.001). The pulse oximetry bias is seen to worsen at both high and low saturation values. Operational Performance Shown in Figure 4 is the operational performance of the four models of pulse oximeter that were utilized in the three NICUs. SaO for these samples, yielding regression coefficients of À0.90 2 None of the models exhibit consistent operational performance (p<0.001) and À0.81 (p<0.001) for the ABL 725 and IL 682 according to specifications across the range of SaO of 70 to 100%. devices, respectively. 2 SaO2 at peak performance varied depending on the device and Evaluation of Data from Other Sites ranged between an SaO2 of 92 to 97%. Operational performance At one center, NICU ‘‘A’’, 95 samples were provided using data declined at SaO2 both above and below this level. A rapid drop from the Spacelabs pulse oximetry module (Spacelabs Medical, in operational performance was seen in all pulse oximeters as Redmond, WA) and an AVL Omni 1-9 (Roche Diagnostics, Basel, saturations decreased. For SaO2<88 to 90%, operational Switzerland) blood gas analyzer/hemoximeter. At the second performance of the pulse oximeters was less than 50%. center, NICU ‘‘B’’, 168 samples were provided using data from Datex-Ohmeda series 3700 pulse oximeters and a Radiometer ABL DISCUSSION 725 blood gas analyzer/hemoximeter. These data are presented in Figure 3. The correlation coefficients for pulse oximetry biases in Data presented here clearly indicate a significant SpO2 bias when NICU ‘‘A’’ and ‘‘B’’ were: r ¼À0.92 (p<0.001) and r ¼À0.72 pulse oximetry is used in NICU patients who require arterial access. (p<0.001), respectively. The bias worsens as true saturation values deviate from a small range (92 to 97%). As an example, at an arterial saturation of 80%, Individual Patient Analyses SpO2 will typically read approximately 90%. The bias does not In the subset of the last 130 patients with sufficient data in the appear to be procedural, or related to variation in hemoximeters, or Historical data set, 124 (95%) demonstrated significant negative related to variation between patients. The bias does appear to be correlations between the bias function and SaO2. The pulse consistent across the pulse oximeter models which were used in the oximeters used in these patients were the Datex-Ohmeda 3700 and three NICUs, including a new generation model with advanced low 3740 models. For the 28 patients in the Prospective data set perfusion and motion detection algorithms that has recently been that met the data criteria, 20 (71%) demonstrated significant clinically evaluated.12 The poor operational performance of pulse negative correlations between the bias function and SaO2. The oximetry seen in these data seems at a level inconsistent with their

Journal of Perinatology 2003; 23:378–383 381 Gerstmann et al. SpO2 in NICU Patients

patient sample (i.e., single patient, blood pressure, heart rate, temperature, or medical therapy). Our data would agree with this last observation, as nearly all of the patients who had five or more blood gases demonstrated the poor correlation. A 1989 report by Praud et al.,11 analyzed 112 comparisons (60 neonates and 11 infants) of pulse oximeter and arterial oxygen saturation values 11 where the latter ranged from 80 to 100%. SpO2 overestimated SaO2 as SaO2 decreased and the bias function (SpO2ÀSaO2) exhibited a significant negative correlation to SaO2, r ¼À0.64 (p<0.01). Although the Praud et al., report does not contain arterial saturation values as low as seen in the Fanconi data or in the current data, the magnitude of the negative correlation is similar. Of concern is that the bias reported in these early papers for first-generation pulse oximeters appears to continue to be Figure 4. Pulse oximeter operational performance. Operational present in second- and third-generation devices as seen in our performance was defined as the percent of pulse oximetry readings at current data. each SaO2 value where the pulse oximeter value is within the device- The etiology of this pulse oximetry bias is unclear. Although specified accuracy for neonates, that is, SpO2 ¼ SaO2±3. Displayed here our Procedural evaluation did not result in any obvious are polynomial curves fitted to the operational performance data for the explanations, there may be some not-yet understood correlation four models of pulse oximeters for which data are available. Operational between our procedural methods and the observed bias that will performance as a percentage is plotted against SaO . Since 2 require further study. Bias from patient motion artefact is SaO2±3¼ SaO2±1SD, operational performance for the devices should commonly attributed to signal loss or corruption and often causes be at or above 68% (horizontal line) across the range of SaO2 of values. No pulse oximeter model shows instrument performance which is SpO2 to read low rather than high. Hypoperfusion can 17 consistent except within a narrow range. Peak performance occurs independently also introduce a pulse oximetry bias. Like motion between 92 and 97%, depending on device. There is declining artefact, hypoperfusion will generally cause the SpO2 to read low performance both above and below this narrow range. Operational rather than high. Bias from motion artefact and hypoperfusion performance falls to 50% by an SaO2 of 88 to 90% in all four models, would occur in a direction opposite to that seen in the current meaning that one-half the time the SpO2 reading deviates from the true study. Lastly, inconsistencies in readings between hemoximeters saturation by more than 3 units of saturation percentage. have been reported at very low saturation ranges.18 Since different hemoximeters/blood gas analyzers were used at the three sites for the current data, it is possible that part of the noted bias might be current perceived importance, role and use in the NICU. These because of this type of systematic bias, and also might help explain results suggest that in the range of SaO2 approaching 100%, use why the slope of the bias appears to vary between the various data of the pulse oximeters tested could lead to overtreatment with sets. However, within site, the bias existed regardless of the model of supplemental oxygen, and in the range of SaO2 approaching 70%, blood analyzer. a neonate could be under treated. Inaccuracy at low saturations may be an inherent problem with Although there are not a large number of publications that the standard dual wavelength pulse oximeter when attempting to address the topic, pulse oximeter readings at low arterial saturation measure very low saturations, such as with fetal oxygen saturation values have previously been reported in different patient monitoring.19 Adjustments have been proposed in emitter populations to either underestimate, overestimate, or accurately wavelength selection to improve device performance within these represent true saturation. In healthy adult volunteers breathing sub low oxygen saturation ranges.20 Similar modifications may be ambient oxygen concentrations, pulse oximetry underestimated required in order to improve performance of pulse oximetry for arterial saturation.9 For pediatric patients with cyanotic heart neonatal use. Pulse oximeters used in the NICU have historically disease, pulse oximetry overestimated true saturations at levels been calibrated by industry using healthy adults. However, below 80%.13,14 Other reports suggest that infants and children with compared to a healthy adult, blood characteristics in peripheral low saturation levels have SpO2 values that appeared to adequately tissue may be quite different in the neonatal patient with 15,16 represent SaO2. In the neonatal population, a 1988 report by cardiopulmonary disease who requires arterial access. These Fanconi10 analyzed 160 paired values in 20 mostly term/near-term conditions may render the standard calibration inappropriate. In NICU patients who had initial SaO2< 65% using a first-generation fact, in situ calibration technology may need to be developed in pulse oximeter. He found that SpO2 overestimated SaO2 at low order to provide patient specific calibrations that can change with saturations and underestimated at high saturations. In addition, the disease course. If more than two wavelength emitters become the bias did not appear to be correlated to characteristics of the necessary in order to broaden the range of pulse oximetry accuracy,

382 Journal of Perinatology 2003; 23:378–383 SpO2 in NICU Patients Gerstmann et al.

pulse spectrophotometry could become the basis for future 4. Zubrow AB, Henderson GW, Imaizumi SO, Pleasure JR. Pulse oximetry in generations of devices.21 sick premature infants and the effects of phototherapy and radiant warmers This project by no means was intended to be an exhaustive on the oxygen saturation readout. Am J Perinatol 1990;7:75–8. study of neonatal pulse oximetry, but rather a ‘‘snapshot’’ of 5. Walsh MC, Noble LM, Carlo WA, Martin RJ. Relationship of pulse oximetry clinical performance from available clinical data. Other to arterial oxygen tension in infants. Crit Care Med 1987;15:1102–5. manufacturers and models of pulse oximeters are routinely used in 6. Thilo EH, Andersen D, Wasserstein ML, Schmidt J, Luckey D. Saturation by pulse oximetry: comparison of the results obtained by instruments of NICUs and these devices may have operational characteristics different brands. J Pediatr 1993;122:620–6. different than seen in these data. Pulse oximetry experience in 7. Brockway J, Hay Jr WW. Prediction of arterial partial pressure of oxygen other NICUs may also be different. However, the question of with pulse oxygen saturation measurements. J Pediatr 1998;133:63–6. operational performance of pulse oximetry is a very important one, 8. Gupta R, Yoxall CW, Subhedar N, Shaw NJ. Individualised pulse oximetry and one that must be examined in much more depth by limits in neonatal intensive care. Arch Dis Child Fetal Neonatal Ed manufacturers. Isolated calibration and small quality control 1999;81:F194–6. studies apparently do not correlate with the operational 9. Severinghaus JW, Naifeh KH, Koh SO. Errors in 14 pulse oximeters during performance that clinicians must rely upon. One would also hope profound hypoxia. J Clin Monit 1989;5:72–81. 10. Fanconi S. Reliability of pulse oximetry in hypoxic infants. J Pediatr that the accuracy of SpO2 could be improved to much better than ±3 (1 SD). Values within ±2 occurring 90% of the time 1988;112:424–7. across the entire 70 to 100% saturation range would be a much 11. Praud JP, Caroﬁlis A, Bridey F, Lacaille F, Dehan M, Gaultier CL. Accuracy more clinically appropriate goal for neonatal pulse oximetry of two wavelength pulse oximetry in neonates and infants. Pediatr Pulmonol 1989;6:180–2. monitoring. 12. Hay Jr WW, Rodden DJ, Collins SM, Melara DL, Hale KA, Fashaw LM. In summary, a discrepancy between pulse oximetry and Reliability of conventional and new pulse oximetry in neonatal patients. hemoximetry derived arterial saturation measurements in an NICU J Perinatol 2002;22:360–6. patient prompted an evaluation of our pulse oximetry use and the 13. Lebecque P, Shango P, Stijns M, Vliers A, Coates AL. Pulse oximetry versus accuracy of these readings. All data examined indicate a systematic measured arterial oxygen saturation: a comparison of the Nellcor N100 and bias in currently used dual wavelength pulse oximetry devices, with the Biox III. Pediatr Pulmonol 1991;10:132–5. pulse oximetry generally overestimating hemoximetry arterial 14. Schmitt HJ, Schuetz WH, Proeschel PA, Jaklin C. Accuracy of pulse oximetry saturations at low saturations and underestimating at high in children with cyanotic congenital heart disease. J Cardiothorac Vasc saturations. The bias increases as saturations decrease. NICU Anesth 1993;7:61–5. patients who appear to have adequate arterial saturation by pulse 15. Boxer RA, Gottesfeld I, Singh S, LaCorte MA, Parnell Jr VA, Walker P. oximetry may in fact have actual saturations which are outside of Noninvasive pulse oximetry in children with cyanotic congenital heart acceptable limits. Until improvements in pulse oximetry disease. Crit Care Med 1987;15:1062–4. 16. Carter BG, Carlin JB, Tibballs J, Mead H, Hochmann M, Osborne A. Accuracy monitoring can be made that correct this measurement bias, of two pulse oximeters at low arterial hemoglobin–oxygen saturation. Crit adjustments to supplemental oxygen and ventilator settings in the Care Med 1998;26:1128–33. NICU patient must be based on and re-evaluated by arterial blood 17. Gehring H, Hornberger C, Matz H, Konecny E, Schmucker P. The effects of gas analysis and hemoximetry. motion artefact and low perfusion on the performance of a new generation of pulse oximeters in volunteers undergoing hypoxemia. Respir Care 2002;47:48–60. 18. Parath M, Sinha P, Dudenhoused JW, Lurrkus AK. Systematic instrumental References errors between oxygen saturation analysers in fetal blood during deep 1. Hay Jr WW, Brockway JM, Eyzaguirre M. Neonatal pulse oximetry: accuracy hypoxemia. Clin Chim Acta 2001;307:151–7. and reliability. Pediatrics 1989;83:717–22. 19. Mannheimer PD. Design and validation of pulse oximetry for low 2. Sardesai S, Durand M, McEvoy C, Johnson C, Maarek JM. Pulse oximetry in saturation. Anesth Analg 2002;94( Suppl 1 ):S21–5. newborn infants with birth weights of 620 to 4285 grams receiving 20. Mannheimer PD, Casciani JR, Fein ME, Nierlich SL. Wavelength selection dopamine and dobutamine. J Perinatol 1996;16:31–4. for low-saturation pulse oximetry. IEEE Trans Biomed Eng 1997;44: 3. Rajadurai VS, Walker AM, Yu VY, Oates A. Effect of fetal haemoglobin on the 148–58. accuracy of pulse oximetry in preterm infants. J Paediatr Child Health 21. Aoyagi T, Miyasaka K. The theory and applications of pulse spectro- 1992;28:43–6. photometry. Anesth Analg 2002;94( Suppl 1 ):S93–5.

Journal of Perinatology 2003; 23:378–383 383