Multi-analyte calibration and verification of a multi-parameter laser-based pulse oximeter.

Almost since its introduction pulse oximetry was known to overestimate oxygen saturation in cases of carbon monoxide poisoning or elevated methemoglobin (metHb) levels. To eliminate this dangerous behavior some manufacturers have added additional LED emitters to try to increase the number of measured hemoglobin species and to improve measurement accuracy, but have not been very successful. We hypothesized that the use of narrow-band laser light sources would make accurate and precise measurement of the four primary species of hemoglobin possible, even in cases of elevated levels of carboxyhemoglobin (COHb). Calibration and verification studies were performed on a tissue simulator that employed an artificial finger pulsating with whole human blood. This simulator allowed safe generation of 165 different combinations of the levels of oxyhemoglobin (O2Hb), COHb, metHb, and reduced hemoglobin (RHb) for calibration of the laser-based pulse oximeter. A follow-on study used 56 mixed hemoglobin levels for verification and statistical analysis of the performance of this device. This laser-based pulse oximeter measured all four species of hemoglobin accurately and precisely (ARMS ≤ 1.8%) for metHb levels in the clinically normal range. At elevated metHb levels the device continued to provide accurate and precise measurements of metHb and RHb (ARMS ≤ 1.7%). The use of monochromatic laser light sources can create a new generation of highly accurate, multi-parameter, pulse oximeters.

A characterization of motion affecting pulse oximetry in 350 patients.

As part of an oximetry research effort in which plethysmographic data were collected from moving patients, a wide variety of patient motion affecting pulse oximetry was observed and characterized by the clinical incidence, type, severity and duration of patient motion. 350 patients were observed for movement in clinical settings, including ICU, SICU, MICU, PICU, NICU, OR and PACU, at 4 hospitals and on an ambulance. 20% (70) of the patients exhibited motion. Half (35) of the moving patients were instrumented to record oximetric, plethysmographic and/or acceleration information. 31% of NICU infants moved compared to only 7% of adults in ICUs. The most common noisy oximetry signals were caused by motion characterized by extending/flexing (/kicking in infants) and by clenching/pressing/rubbing. In infants, these motion types accounted for 53% and 11% of motion, respectively. Less common infant motion types were patient cares, shifting body position and cough/cry. The most common adult motions were equally divided among extend/flex, clench/press/rub, twitch/shake and transport motion types. Extend/flex motions typically demonstrated high plethysmographic waveform modulation (71.5% maximum) and high acceleration. Clench/press/rub motions typically also had high modulations, but low G-force. With one high-G exception, twitch/shake motions had little or no effect on oximetry readings. Less common adult motion types affecting pulse oximetry included cough/cry, strain/posture, tremors and tap/bump. Most recorded motions were aperiodic and short-lived, 62% lasting less than 10 seconds, only 5% lasting over 1 minute. NICU patients made the longest lasting continuous series of motions, while adults made 86% of the motions lasting a second or less.