SESSION TYPE: Sleep I
PRESENTED ON: Sunday, October 21, 2012 at 10:30 AM - 11:45 AM
PURPOSE: Obstructive sleep apnea is often treated with continuous positive airway pressure (CPAP), which acts as a pneumatic splint to prevent airway collapse. The goal of CPAP titration polysomnography (PSG) is to ascertain a pressure calibrated in centimeters of water (cm H2O) that lowers the apnea-hypopnea index to normal during sleep. The use of prediction equations has been put forth as a cost-saving method for predicting optimal starting pressures. This study compared the accuracy of the 3 most commonly referenced clinical prediction equations in a large, consecutively acquired sample of multi-ethnic, urban city and suburban residents who presented for CPAP titration. This population is significantly different from populations previously studied.
METHODS: The Hoffstein, Laredo, and Stradling equation-predicted CPAP pressures (cm H20) were computed for 678 patients; 63% were African American, 33% Caucasian, 51% female, and 49% male. These were compared to pressures derived from the overnight PSG. This comparison was repeated for the full-night and split-night titration subgroups. Ranges around the predicted values were determined which captured the middle 90% of the patients. Median absolute difference between predicted values and PSG derived CPAP pressures was computed to determine which equation most closely predicted the PSG titration pressure results for this sample.
RESULTS: 350 patients underwent full-night titration, 328 split-night. Ranges (median absolute differences) were: Full night: Loredo -3 to 5 (1.82); Hoffstein -3 to 7 (1.88); Stradling 1 to 9 (3.57). Split night: Loredo -4 to 4 (1.95); Hoffstein -4 to 6 (1.95); Stradling -1 to 9 (3.19). All patients: Loredo -4 to 5 (1.89); Hoffstein -4 to 6 (1.94); Stradling 0 to 9 (3.44).
CONCLUSIONS: Thus, the Loredo equation achieved the lowest median absolute differences and consequently was the closest fit to this population’s PSG derived CPAP pressures.
CLINICAL IMPLICATIONS: This will help guide efforts to streamline care for OSA patients in an urban setting.
DISCLOSURE: The following authors have nothing to disclose: David Leszczyszyn, Jonathan Clark, Ann McNallen, Elsa Mathew, Jacob Wegelin, Samuel Taylor
No Product/Research Disclosure InformationVCU Medical Center, Richmond, VA