OBJECTIVE: Postoperative patients exhibiting signs or symptoms of obstructive sleep apnea (OSA) have been identified to be at increased risk for respiratory compromise. One of the key markers associated with OSA is repetitive reductions in airflow (RRiA). A real-time pulse oximeter saturation pattern recognition algorithm (OxiMax SPD™ intended for adult in-hospital use only) designed to detect specific signatures in the SpO(2) trend associated with RRiA may provide caregivers early indication of its presence so they can treat the patient appropriately. The purpose of our study was to test the performance of saturation pattern detection (SPD) in a clinical study targeting subjects with a high prevalence of RRiA. METHODS: Overnight polysomnograph (PSG) recordings were collected on 104 sleep lab patients. RRiA was defined in terms of specific criteria from four PSG signals, evaluated in consecutive 10 min epochs. PSG scoring was conducted blind to calculation of SPD. Statistical measures of sensitivity, specificity and area under the receiver operating characteristic (ROC) curve were calculated for the detection of RRiA by SPD. RESULTS: Data were analyzed for 92 valid sets of patient recordings, encompassing 3,917 epochs. At the highest available SPD alert setting, the sensitivity was 80.2% (95% C.I. = 76.8-83.3%), the specificity was 88.3% (87.2-89.3). Area under the ROC curve was 0.87 (0.84-0.89). CONCLUSIONS: The real-time SPD algorithm was able to detect episodes of RRiA in sleep lab patients with a high degree of sensitivity and specificity.
OBJECTIVE: Postoperative patients exhibiting signs or symptoms of obstructive sleep apnea (OSA) have been identified to be at increased risk for respiratory compromise. One of the key markers associated with OSA is repetitive reductions in airflow (RRiA). A real-time pulse oximeter saturation pattern recognition algorithm (OxiMax SPD™ intended for adult in-hospital use only) designed to detect specific signatures in the SpO(2) trend associated with RRiA may provide caregivers early indication of its presence so they can treat the patient appropriately. The purpose of our study was to test the performance of saturation pattern detection (SPD) in a clinical study targeting subjects with a high prevalence of RRiA. METHODS: Overnight polysomnograph (PSG) recordings were collected on 104 sleep labpatients. RRiA was defined in terms of specific criteria from four PSG signals, evaluated in consecutive 10 min epochs. PSG scoring was conducted blind to calculation of SPD. Statistical measures of sensitivity, specificity and area under the receiver operating characteristic (ROC) curve were calculated for the detection of RRiA by SPD. RESULTS: Data were analyzed for 92 valid sets of patient recordings, encompassing 3,917 epochs. At the highest available SPD alert setting, the sensitivity was 80.2% (95% C.I. = 76.8-83.3%), the specificity was 88.3% (87.2-89.3). Area under the ROC curve was 0.87 (0.84-0.89). CONCLUSIONS: The real-time SPD algorithm was able to detect episodes of RRiA in sleep labpatients with a high degree of sensitivity and specificity.
Authors: Jeffrey B Gross; Kenneth L Bachenberg; Jonathan L Benumof; Robert A Caplan; Richard T Connis; Charles J Coté; David G Nickinovich; Vivek Prachand; Denham S Ward; Edward M Weaver; Lawrence Ydens; Song Yu Journal: Anesthesiology Date: 2006-05 Impact factor: 7.892