The Science Journal of the American Association for Respiratory Care

2004 OPEN FORUM Abstracts

COMPARISON OF TWO 3RD GENERATION PULSE OXIMETERS IN THE ADULT INTENSIVE CARE UNIT –RESPIRONICS NOVAMETRIX MARS AND NELLCOR N-595 PULSE OXIMETERS



Lori Mercuri, CRT
Manager of Respiratory Care
Richard A. Smith, MD, FCCP
Medical Director, Respiratory and Pulmonary Services
Midstate Medical Center, Meriden, CT

Summary: We compared the performance of two 3rd generation pulse oximeters, the Respironics Novametrix Model 2001 and Nellcor N-595 in 17 adult ICU patients. An acceptable ARMS for SpO2 of 01.41 was computed for this patient population. Based upon these measurements and our experience with both devices, we believe both devices offer equivalent clinical performance in this patient group.

Introduction:
Midstate Medical Center is community medical center in central Connecticut with a 9-bed adult ICU. We recently performed an inter-device comparison study between two 3rd generation pulse oximeters in 20 mechanically ventilated patients. We sought to judge the relative performance of the two monitors in this patient population.

Methods:
A Respironics Novametrix Model 2001 with Rev. 3.0 software and a Nellcor Model N-595 with 2.4.5.0 software were used in this study. Finger probes from each device were used in the study, and both probes were placed on the same hand. Each device was interfaced to a laptop computer that recorded one-second values from each. Averaging times for each monitor was set for the default. This allowed later comparison and computation of statistics in Excel.

Results:
Seventeen (17) adult (ages 54 to 77 years) ICU patients were studied with diagnoses including respiratory failure, pneumonia, COPD, etc. Problems with sensors or sensor sites, though very infrequent, were minimized or eliminated by removing data where either monitor displayed zeros. Data files were processed individually by calculating the difference between the SpO2 and pulse rate between the two monitors being compared. The difference was then averaged (bias) and the standard deviation of the population was calculated (precision). Then root mean square(ARMS) was calculated using the bias and precision numbers from the two parameters. Bias and precision for SpO2 and pulse rate between the Model 2001 Rev. 3.0 and the Nellcor N-595 was 0.13 ± 1.40% SpO2 and 0.14 ± 1.66 bpm (n=17 subjects). The ARMS for SpO2 was 1.41. These values demonstrate that Model 2001 Rev. 3.0 relative to the Model N-595 SpO2 values are quite close in the adult ICU data and can be judged to be both statistically and clinically equivalent.

Conclusion: The results of this clinical comparison provide support for the equivalent clinical performance of the MARS Model 2001 with Rev. 3.0 software and Model N-595 in the adult ICU population.