Search

In an attempt to identify predictors of long-term compliance with nasal continuous positive airway pressure (CPAP), we reviewed the records of 125 patients with obstructive sleep apnea (OSA) referred to our center for nasal CPAP trials. Severity of sleep apnea, sleep staging, daytime hypersomnolence, effectiveness of nasal CPAP, previous palatal surgery, and adverse reactions were compared in compliant and noncompliant patients. Nineteen patients did not tolerate a nasal CPAP trial in the laboratory or refused home nasal CPAP therapy. Ten patients were unavailable for follow-up. Of the remaining 96 patients, 23 (24 percent) had discontinued therapy, while 73 (76 percent) were still using nasal CPAP at 14.5 +/- 10.7 months (mean +/- SD). There were no statistically significant differences between the compliant and noncompliant patients in baseline apnea plus hypopnea index (AHI), baseline sleep staging, AHI while receiving nasal CPAP, sleep staging while receiving nasal CPAP, or frequency of adverse reactions during therapy. Severe daytime sleepiness was present in 65 of the 73 compliant patients and in 12 of the 23 noncompliant patients (p less than 0.05). Ten of 43 in the compliant group had previous palatal surgery compared with ten of 23 noncompliant patients (p less than 0.05). Our data confirm earlier observations in smaller samples that compliant and noncompliant patients have equally severe sleep apnea and good initial responses to nasal CPAP. Long-term compliance with nasal CPAP may be associated with the severity of daytime hypersomnolence on presentation. Previous palatal surgery was more frequent in patients who did not tolerate long-term nasal CPAP therapy.

A SPICE model of the foils and overbraiding of an integrated wiring system (IWS) harness is created to characterize their combined impedance as a function of frequency. Comparison of predicted values of impedance to measured data for a modeled cable harness is accomplished, in which fair agreement is seen to exist. A straightforward empirical adjustment is then made to the SPICE model, which results in bringing the model-predicted values into very close agreement with the measured data. Possible reasons for initial model-predicted value disagreement with empirical data are discussed, and form the basis for future research.