William T. Basco, Jr., MD, MSDisclosures

March 26, 2014

17 comments

PrintEmail

EDITORS' RECOMMENDATIONS

Pulse Oximetry and Congenital Heart Disease in Newborns

Pulse Oximetry Less Accurate in Lower Saturation Ranges in Kids

Hospitalized Kids: First, Do No Harm

Accuracy of Pulse Oximetry in Children

 

Ross PA, Newth CJ, Khemani RG

 

Pediatrics. 2014;133:22-29

 

Study Summary

 

Pulse oximetry equipment is largely calibrated using adult patients. This study represents a comprehensive attempt to gain information on the accuracy of pulse oximetry devices in infants and children across a large range of oxygen saturations, child ages, and skin colors. The gold standard with which the instruments were compared was arterial oxygenation readings obtained by arterial blood gas measurements.

 

The children were admitted to 5 US pediatric intensive care units (ICUs) in 2009 and 2010. All children had both pulse oximetry readings and arterial oxygen saturation readings taken at the same time, were younger than 18 years, and were intubated and mechanically ventilated. In addition to classification by demographic variables and diagnoses, children were also classified as to whether they had cyanotic congenital heart disease (CHD) or were admitted for acute respiratory failure.

 

Measurements were completed whenever the children had an arterial saturation value obtained. At that time, the bedside providers also recorded the patient's corresponding pulse oximetry reading, temperature, capillary refill time, most recent hemoglobin level, ventilator settings, and type or manufacturer of the pulse oximeter.

 

The investigators focused on several measurements, including the "mean bias" -- the average difference between the pulse oximetry reading and the arterial oxygen saturation reading, given that pulse oximetry overestimates oxygen saturation. They also evaluated the "local bias," which is the mean bias of the measurements within specific ranges (eg, 91%-95%). Finally, they calculated a measure of accuracy that combined bias, the precision of the instruments, and the number of samples obtained. An acceptable accuracy was defined as variation of 3% less.

 

Analyses accounted for disease category (respiratory or cardiac); patient sex, hemoglobin level, and temperature; ventilator settings; and manufacturer of oximeter and sensor.

 

Study Findings

 

In this study, 1980 comparisons between the arterial oxygen saturation and pulse oximetry reading were obtained from 225 children. Each patient provided a median of 5 paired readings. In general, the children with cyanotic CHD were much younger than those with acute respiratory failure.

 

Three different combinations of oximeter and sensor equipment were used. When all values were examined, two thirds of the readings exhibited a positive bias, in which the pulse oximetry reading was greater than the measured arterial oxygen saturation. The mean bias was 3.3%, with a median of 2%. The local bias was negligible in the 2 upper saturation ranges (91%-95% and 96%-97%), as well as at the low end of the spectrum (65%-70%). However, the other ranges exhibited biases that varied from 2% to 6%, with the highest mean bias (6%) occurring in the range of 81%-85% pulse oxygen saturation. This means that in the range of 81%-85% of pulse oximeter readings, the pulse oximeter overestimated the percent saturation by approximately 6%.

 

In multivariate modeling, several factors were associated with having a bias greater than 3%; these included the presence of cyanotic CHD, prolonged capillary refill, and an oxygen saturation < 96%. Black race, male sex, and certain combinations of sensors and equipment were associated with a lower likelihood of bias greater than 3%. Factors that were not contributory to the degree of bias were hemoglobin concentration, temperature, the fraction of inspired oxygen, and age younger than 2 months. Even in the subgroup of children with cyanotic CHD, similar patterns emerged.

 

Ross and colleagues concluded that there is significant variability in the bias of measurement between pulse oximetry and arterial oxygen saturation. They comment that bias is most notable in the pulse oximetry range of 81%-85%, although to some degree a bias exists in all ranges between 76% and 95%. They suggest that manufacturers should consider additional calibration against infants and other children to provide clinicians with more accurate readings upon which to base clinical decisions.

 

Viewpoint

 

I applaud the investigators for conducting what was probably a very difficult study to complete -- especially given that local providers had to be trained in the data and sample collection protocols, which took place in busy ICUs. However, they provide a very rich picture of what is going on.

 

The good news for most clinicians who might read this column is that pulse oximetry accurately reflects arterial oxygenation in most of the ranges (≥ 90%) that we deal with outside of the ICU. However, the findings suggest that clinicians should be more cautious in the 80%-90% range and consider making decisions more on the basis of arterial oxygen readings if patient saturations are < 90% on pulse oximetry. It is also worth remembering that pulse oximetry is affected by perfusion and may differ among patients of different skin hues.

 

 

 

 

 

 

MEDLINE Abstract

Print ThisPrinter-FriendlyEmail This Email ThisDownload ThisDownload This

 

View OnlineView This Abstract Online

Accuracy of pulse oximetry in children.

Pediatrics.  2014; 133(1):22-9 (ISSN: 1098-4275)

 

Ross PA; Newth CJ; Khemani RG

 

OBJECTIVE: For children with cyanotic congenital heart disease or acute hypoxemic respiratory failure, providers frequently make decisions based on pulse oximetry, in the absence of an arterial blood gas. The study objective was to measure the accuracy of pulse oximetry in the saturations from pulse oximetry (SpO2) range of 65% to 97%.

 

METHODS: This institutional review board-approved prospective, multicenter observational study in 5 PICUs included 225 mechanically ventilated children with an arterial catheter. With each arterial blood gas sample, SpO2 from pulse oximetry and arterial oxygen saturations from CO-oximetry (SaO2) were simultaneously obtained if the SpO2 was ≤ 97%.

 

RESULTS: The lowest SpO2 obtained in the study was 65%. In the range of SpO2 65% to 97%, 1980 simultaneous values for SpO2 and SaO2 were obtained. The bias (SpO2 - SaO2) varied through the range of SpO2 values. The bias was greatest in the SpO2 range 81% to 85% (336 samples, median 6%, mean 6.6%, accuracy root mean squared 9.1%). SpO2 measurements were close to SaO2 in the SpO2 range 91% to 97% (901 samples, median 1%, mean 1.5%, accuracy root mean squared 4.2%).

 

CONCLUSIONS: Previous studies on pulse oximeter accuracy in children present a single number for bias. This study identified that the accuracy of pulse oximetry varies significantly as a function of the SpO2 range. Saturations measured by pulse oximetry on average overestimate SaO2 from CO-oximetry in the SpO2 range of 76% to 90%. Better pulse oximetry algorithms are needed for accurate assessment of children with saturations in the hypoxemic range.