Improved prediction of CPAP failure using T90, age and gender.

Sleep apnea syndrome is associated with increased risk of cardiovascular disease. In treating older patients, there is a special emphasis put on minimally invasive and conservative procedures and a simple method for predicting the potential for treatment success is essential. Continuous positive airway pressure (CPAP) is the first choice for treatment, however, it is not always successful. In cases where CPAP was unsuccessful, treatment with bilevel positive airway pressure (BiPAP) is the next treatment option. In this study, we examine commonly evaluated respiratory parameters, obesity, and age relative to their ability to predict CPAP failure. We also tried to find differences in the predictive ability of these parameters in older and younger patients. The predictive ability, relative to CPAP failure, was examined for each individual parameter as well as for combinations of parameters. All variables had a statistical association with CPAP failure; failure prediction reliability ranged from poor to moderate. Combining T90, age, and gender can be used to find patients who will benefit from BiPAP as the first choice for treatment. An initial BiPAP indication can produce relevant reductions in treatment cost.

Risk factors for failure of continuous positive airway pressure treatment in patients with ostructive sleep apnoea.

BACKGROUND AND AIMS: Obstructive sleep apnoea is a potentially serious sleep disorder associated with the risk of cardiovascular disease. It is treated with continuous airway pressure (CPAP) but this is not always successful. Unsuccessful cases should be treated by bilevel positive airway pressure (BiPAP). The aim of this study was to determine whether common respiratory parameters and/or body mass index (BMI) can be used to predict the probability CPAP failure and hence start such patients on BiPAP from the outset. METHODS: A sample of patients treated by CPAP for OSAS was evaluated a retrospective cohort study. The data measured in sleep monitoring of the successfully treated group and of the group where CPAP had failed were compared. Subsequently, the predictive abilities of BMI, Apnoea Index (AI), Apnoea-Hypopnea Index (AHI), percentage of sleep time in less than 90% oxygen saturation (T90), average oxygen saturation over the duration of sleep (SaO2) and average desaturation per hour of sleep (ODI) were assessed with respect to CPAP failure, both individually and in combination. RESULTS: A sample of 479 patients was included in the study. All of the recorded variables except AI were significantly associated with failure of CPAP and their ability to predict the failure ranged from poor to moderate. Since there was significant correlation among all the variables measured a two-variable prediction model combining T90 and BMI produced no significant improvement in the quality of CPAP failure prediction. CONCLUSIONS: BMI was a significant predictor of CPAP failure although it was slightly less predictive than T90. The set of monitored variables included in our study does not allow for CPAP failure to be predicted with clinically relevant reliability.