Clinical Performance Evaluation of Pediatric and Neonatal Low Saturation Oximetry Sensors

1.0 BACKGROUND

A common method for assessing the respiratory status of hospitalized pediatric and neonatal patients is the use of pulse oximetry. This noninvasive device uses a light emitting diode (LED) emitter, which is applied to an area of the body with good local blood flow. Red and infrared light is shined through the blood-perfused tissue under the sensor and the detected light information is sent back to a signal processing unit, or monitor, for calculation of oxygen saturation (SpO2)1. Typically the sensor may be placed either on the forehead, a finger, or in neonates - the foot, palm or toe. Use of this noninvasive method provides continuous information on the oxygenation status of patients and has greatly reduced the number of arterial blood gas punctures or samples required for patient care.

While the advantages of noninvasive pulse oximetry are significant, there are a number of factors that can negatively impact the performance of the device2. Patient monitoring during low saturation episodes has been one of the challenges to optimal device performance. Improvements to sensor technology over the last 20 years have progressively improved the accuracy and reliability of the device. Since their introduction to clinical care in the 1980's, a number of improvements have been made to increase the accuracy of pulse oximetry in oxygen saturations of 80% and below.

Nonetheless, while much of the clinical research on pulse oximetry sensors over the years has evaluated device accuracy and reliability, limited data are available in these conditions1,6.

The purpose of this study is to evaluate the accuracy of new pulse oximetry sensors specifically designed for low saturation episodes in pediatric and neonatal patients. Results of this study will provide basic information to drive further improvements in the technology.

2.0 PURPOSE AND OBJECTIVES

The general purpose of this study is to evaluate the feasibility, accuracy and performance of pediatric and neonatal oximetry sensors over clinically relevant ranges of arterial saturations between 60-100%.

The study objectives are as follows:

1. To evaluate the accuracy of a neonatal sensor in the saturation range of 70-80% in the following weight category: 0-5kg.
2. To evaluate the accuracy of neonatal and pediatric transmission (digit/foot/hand) sensors in the saturation range of 60-80% in the following weight categories: 0-5kg; and, 5-40kg.
3. To evaluate the accuracy of neonatal and pediatric transmission (digit/foot/hand) sensors in the saturation range of 80-100% in the following weight ranges: 0-5kg; and, 5-40kg.

These objectives will be achieved by enrolling hospitalized patients from several neonatal and pediatric units, including cardiac cath lab, OR and intensive care areas. Table 1 below outlines the specifics for data collection.

Table 1 60-80% Saturation Range 80-100% Saturation Range

Neonatal <5kg. Transmission sensor (20 data points targeted) Transmission sensor only (30 data points targeted)

Pediatric 5-40kg. Transmission sensor only (20 data points targeted) Transmission sensor only (30 data points targeted)

*Data required for 70-80% saturation range only

Data may be collected continuously from the test and reference devices by a validated computer data acquisition system and/or by Case Report Form (CRF) recording, for direct comparison to measurements taken from Co-Oximetry (or arterial blood gases (ABGs) if Co-Oximetry is not an option). Arterial blood sampling will be obtained as necessary for patient management as defined by the attending physician.

3.0 DEFINITION OF TERMS

3.1 Pulse oximetry sensors: Pulse oximetry sensors are devices, which are attached to the skin (hand, forehead, finger, toe, foot) with or without an adhesive material and connected to a signal-processing unit for the purpose of noninvasive oxygen saturation monitoring. For the purposes of this study, three different sensors will be simultaneously placed on each enrolled patient: a reference sensor and two test sensors (Nellcor Puritan Bennett LLC, Boulder, CO, and Masimo Corporation, Irvine, CA). The Masimo sensor will be a rainbow R25, R25-L, R20 or R20-L, and the Nellcor sensor will be an OxiMax MAX-A, MAX-P, MAX-I or MAX-N. All sensors are currently available on the market but the Masimo sensors have not been compared to the Nellcor sensors in head-to-head tests. During the study, a Masimo sensor will be connected to a Masimo bedside pulse oximetry monitor (the Radical-7) via a standard patient cable, and a Nellcor sensor will be connected to a Nellcor bedside pulse oximetry monitor (the N-600x) via a standard patient cable.

3.2 Sensor accuracy: Sensor accuracy is the capability of the tested sensor to reflect oxygen saturations that are consistent with those obtained directly from arterial blood sampling and evaluated with Co-Oximetry. 3.3 Skin color: For the purposes of this study, skin color will be assessed with Nellcor's visual inspection rating scale (i.e., 1 very light, 2 olive hue, 3 dark olive, 4 extremely dark).

3.4 Data point: Paired data value from Co-Oximetry and the test sensor. 3.5 Transmission sensor: A sensor where the emitter and photodetector are opposite of each other, with the measuring site (i.e., digit) in-between 3.6 Reflectance sensor: A sensor where the emitter and photodetector are next to each other and on top of the measuring site.

4.0 MATERIALS AND METHODS 4.1 Study Design. This is a multicenter observational study wherein the information collected on the test devices will not be used for clinical management. A repeated-measures experimental design will be used to compare the test sensors with conventional oximetry and Co-Oximetry, the defined "Gold Standard". Both of the test sensors and a reference sensor will be simultaneously applied to each subject. The reference sensor will be applied for the patient's clinical management, while the test sensors will be applied to an extremity as indicated.

A minimum of 50 subjects with a minimum of 100 data points across sites will be necessary to address subject-to-subject variability and to yield a statistically significant RMSD. Enrollment in the study will be suspended as soon as both minimums are achieved. A data point is defined as a paired data value from Co-Oximetry and the test sensors. If high variability is observed, a need for more subjects or data-points may be determined.

5.0 STUDY PROCEDURE

All patients who meet the eligibility criteria for the study, after having signed the Informed Consent will have the following procedures carried out:

5.1 The reference sensor will be applied according to the product Directions For Use7 (DFU), while the test sensor or sensors will be applied to a recommended site per the test sensor's DFU.

5.2 Demographic (the patient's physical characteristics) and baseline data will be collected. 5.3 Skin color. For the purposes of this study, skin color will be assessed with Nellcor's visual inspection rating scale (i.e., 1 very light, 2 olive hue, 3 dark olive, 4 extremely dark).

5.4 The date, time, and test sensor tracking numbers will be recorded on the CRF.

5.5 Each sensor will be connected to a patient cable, which will be connected to a SpO2 monitor. The monitors connected to the Nellcor and Masimo test sensors will not be used for patient care management decisions. Alarms on the test sensors/monitors will be inactivated prior to study use to avoid patient management confusion.

5.6 Test sensors and data collection will be discontinued five to fifteen minutes after the last anticipated arterial blood draw. The date and time of sensor discontinuance will be recorded on the CRF.

A sample study procedure can be found in Attachment A. A sample Adverse Event Form can be found in Attachment B.

6.0 DATA ANALYSIS

Pulse oximetry measurements of saturation from the test devices (sensors) will be compared to Co-Oximetry measurements of arterial oxygen saturation to demonstrate that the test sensors meet the oxygen saturation accuracy specifications for SpO2 when used on neonatal and pediatric populations.

SpO2 accuracy will be determined by calculating the root mean square of the differences (RMSD) between test sensor and SaO2 CO-oximeter (or ABG) value to the following equation:

RMSD = where d is the difference between SpO2 and SaO2. Paired data values may then be plotted comparing the Pulse-Oximetry values and Co-Oximetry values, across the saturation range of interest (i.e., ≤80%).

7.0 ADVERSE EVENT REPORTING All known and anticipated adverse events associated with this study are identified in Section 8.0, Risk Analysis. All reportable anticipated, and unanticipated, adverse events will be documented on the Adverse Event Form.

8.0 RISK ANALYSIS

The risks to participation in this study are primarily physical. Involvement in this study requires the addition of two additional pulse oximetry sensors to the extremities of the subjects. There are no psychological, social, economic, legal or other risks that have been identified. We believe that the risks from the devices and the study procedure are non-significant.

Physical risks from the use of the devices may consist principally of a burn to the skin. Pulse oximetry sensors are attached to the fingers or other skin surfaces, and use light to measure saturation, generating a small amount of heat. Because of the low amount of current required to power the LED's, risk of burn is minimal. The long history of safe performance of Nellcor and Masimo Pulse Oximetry in the marketplace is evidence that this risk is very low. In addition, the short duration of this study makes this risk extremely unlikely.

The application and removal of oximetry or other non-invasive tissue sensors present a minimal risk to the subject. There may be minor discomfort associated with removal of adhesive sensors or a reaction to the standard adhesive. Injury from a reaction to a sensor adhesive is rare.

9.0 BENEFITS

There are no direct benefits to subjects who participate in these studies. Benefits from the study will be the overall improvement of patient care, as a sensor can be developed which is expected to be accurate and reliable in reporting saturations below the 80% saturation range. Future patients will benefit from the accurate and reliable products that are developed using the procedure described above.

10.0 STATEMENT OF NON-SIGNIFICANT RISK

Nellcor believes that this is a "non-significant risk" device study due to the nature of the devices being tested. Utilizing the information presented in Section 9.0 and the FDA criteria listed below to distinguish between significant and non-significant risk devices, Nellcor has determined that the devices referred to in the Protocol present no potential for serious risk to the health, safety, or welfare of a subject and are NOT:

1. Implants; or
2. Used in supporting or sustaining human life; or
3. Substantially important in diagnosing, curing, mitigating or treating disease or in preventing impairment of human health.

Nellcor requests that the IRB indicate its agreement with this determination of risk in its letter of approval for this study.

11.0 ETHICAL CONSIDERATIONS

No therapeutic modifications are proposed in this study. Written informed consent will be obtained from each participating patient. If the patient is a minor, written informed consent will be obtained from the subject's parents or legal guardian(s) and their permission sought for participation in this study. Federal policies for protection of human subjects will be followed at all times8.

12.0 CONFIDENTIALITY

Subject confidentiality will be kept at all times. Patient records may be made available to employees from Nellcor and the United States of America Food and Drug Administration (FDA) for data review only. The results of this study may be presented at meetings or in publications; however, subject identity will not be disclosed.

13.0 COMPENSATION TO THE STUDY SUBJECTS

None

14.0 FINANCIAL OBLIGATIONS

None