Abstract
To analyze whether there is a good correlation between the pressure setting by autoCPAP and the one calculated by mathematical prediction formulas proposed by Hoffstein et al.([(0.16xBMI)+(0.13xNC)+(0.04xAHI)-5.12], (AmJRespirCritCareMed1994) and Séries et al [(0.193xBMI)+(0.077xNC)+(0.02xAHI)–0,611] (Chest2008).
Consecutive patients with OSAS diagnosis after a sleep study between December-2003 to December-2012. Patients underwent an attended sleep study for CPAP titration with an autoCPAP device (REMstar ®). The pressure was set from 4 to 15 cmH20. We analyze the correlations (Pearson r coefficient and Bland-Altman plots) between the pressure calculated by the formulas and the one obtained by autoCPAP.
We included 2514 male patients. Mean values were: age 54 years±12.5, neck circumference(NC) 43.8±3.6 cm, BMI 33±6kg/m2; AHI 49±22. Although there was a good correlation between the pressures calculated by both formulas (r=0.97, p<0.0001), was lower among the pressures obtained by autoCPAP and Hoffstein (r=0.62, P<0.001) or Series (r=0.54, p<0.001). Comparing with the optimum pressure by autoCPAP, we found a difference ≥2cmH2O in 220 (19.8%) patients respect to Hoffstein formula (in 24 was higher and lower in the other 196 patients). By contrast, with the Séries formula there was a difference ≥2 cmH2O in 393 patients (in 356 was higher).
When comparing with autoCPAP titration, Hoffstein formula underestimates CPAP pressure values >7cmH2O. The opposite happens with the Séries formula, which overestimates the pressure throughout the usual range of values except for pressures >12cmH2O.
- © 2013 ERS