PURPOSE: Continuous positive airway pressure (CPAP) is the preferred treatment for obstructive sleep apnea. Starting at low CPAP levels during a therapeutic polysomnogram, especially split-night studies, can lead to occasional failure when the optimum CPAP pressure is not reached prior to the end of sleep. Using a CPAP prediction algorithm may achieve optimum CPAP pressure with fewer pressure changes and, thus, in a shorter period of time. One such algorithm was previously evaluated (Oliver. Chest 2000;117:1061-4) but the study was subject to bias as technicians performing the study were aware of the predicted CPAP pressure (CPAP-p), and gender information was not provided. Our study compares CPAP-p with measured optimum CPAP pressure (CPAP-m) in a cohort where the technicians were unaware of CPAP-p.
METHODS: This retrospective study included 105 patients who had a therapeutic polysomonogram between January 2005 and January 2007 at a single accredited sleep laboratory. CPAP-p was derived using the formula: (0.16xBMI) + (0.13xNC) + (0.04xAHI) –5.12; where BMI is the body mass index, NC the neck circumference in centimeters and AHI the apnea/hypopnea index. Pearson's correlation was used to compare CPAP-p with CPAP-m.
RESULTS: Our study included 50 male and 55 female patients, with a mean age of 52.8±3 years (table). The overall correlation coefficient of CPAP-p to CPAP-m was 0.5, with a higher correlation in males (r=0.63) than females (r=0.3) (figure). The prediction algorithm accurately predicted the effective CPAP pressure in 10 patients, was within ± 2 cm of the effective pressure in another 42 patients and outside ± 2 cm in 53 patients. The CPAP-p was higher than CPAP-m in only 7 patients.
CONCLUSION: Our study shows good correlation between the predicted and measured CPAP pressures in males but not in females. Compared to the prior study, our target AHI was 5 (instead of 10) and there was no technician bias.
CLINICAL IMPLICATIONS: This CPAP prediction algorithm may be useful in males as a starting pressure from which the effective CPAP pressure can be further titrated.
DISCLOSURE: Karim Djekidel, No Financial Disclosure Information; No Product/Research Disclosure Information