consequences as delivered by an attending technician prior to polysomnography (PSG) hook up preparation, could improve the sleep quality and increase mask toler- ance during the titration night and long term adherence in patients. Materials and methods: Thirty consecutive patients referred for CPAP titration in our sleep laboratory were randomized into two groups: the study group of 15 patients who received information about OSAHS, its consequences and treatment by an attending technician, and a Control group of 15 patients who did not receive this information. All patients were submitted to a stan- dard titration protocol. After 17 months, all patients received one telephone call to evaluate CPAP adherence. Results: Comparing study group versus Control group, patients had similar age (54 ± 12 vs 51 ± 13 years), gen- der (87 vs 100% male) and BMI (31 ± 4 vs 31 ± 6 kg/ m 2 ). PSG data showed differences (p < 0.05) in: sleep efficiency (83 ± 12 vs 76 ± 13%), REM sleep latency (77 ± 45 vs 110 ± 53 min), wake after sleep onset (59 ± 40 vs 126 ± 111 min); stage 2 (46 ± 10 vs 68 ± 9%); stage 3 + 4 (24 ± 8 vs 10 ± 7%), and REM sleep (25 ± 9 vs 15 ± 6%). Mask acceptance rates were different (p = 0.01) between groups (100reported contin- uing use of CPAP. Conclusion: One information session on OSAHS, its consequences and treatment by an attending technician prior to the PSG hook up preparation highlighted improved sleep quality and mask tolerance during the titration night. However, long term adherence was not increased. Our data suggest that continuous support of an informative and educational nature could play a sig- nificant role during CPAP treatment. Financial support: AFIP, FAPESP. doi:10.1016/j.sleep.2006.07.198 P390 Positive impact of an educational retreat with sleep- technicians on the success of CPAP titrations Khurshid A Khurshid * , Jean-Paul Spire Sleep Disorders Center, Department of Neurology, Uni- versity of Chicago, Chicago, IL, USA Objective: To evaluate the effectiveness of an educational intervention on the magnitude of successful CPAP titra- tions in our sleep laboratory as an element of quality improvement project. Materials: 1. Review of sleep studies in the preceding three months to identify the modifiable causes of unsuc- cessful CPAP titrations; 2. Review of studies in the three months followings the educational intervention. Methods: 1. Modifiable causes of unsuccessful CPAP titrations were identified. 2. Educational retreat involving didactic presentation by a faculty member and a sleep fellow followed by discus- sion of various ways of correcting the causes of unsuc- cessful CPAP titrations. The factors involved in successful CPAP titrations were identified and high- lighted. All sleep technicians took part in this retreat. 3. The magnitude of successful CPAP titrations in the preceding three months was compared with the mag- nitude of successful CPAP titrations in the three months following the intervention. Results: Four hundred and twenty-two sleep studies done in the preceding three months in our sleep laboratory were reviewed. Of the total CPAP titrations and split-night studies 71% were successful CPAP titrations, 17% were partially successful titrations, and 12% were unsuccessful titrations. Of the CPAP titrations 68% were successful, 24% were partially successful, and 8%were unsuccessful titrations. Of the split-night studies 78% were successful, 11% were partially successful, and another 11% were unsuccessful titrations. In the three months following the intervention, there was a significant increase in the percentage of successful CPAP titrations, decrease in the percentage of partially successful CPAP titrations and no significant change in the percentage of unsuccess- ful CPAP titrations. doi:10.1016/j.sleep.2006.07.199 P391 Automated detection of irregular respiration: A marker of wakefulness I. Ayappa 1,* , R.G. Norman 1 , D. Whiting 2 , F. Anderson 2 , E. Donnelly 2 , D.M. Rapoport 1 1 NYU School of Medicine, New York, NY, USA 2 Fisher and Paykel Healthcare, Auckland, New Zealand Objectives: Sleep without sleep disordered breathing is marked by regular respiration. We have previously shown that manual identification of perturbations in ventilatory pattern were highly predictive of wakeful- ness, though not all wakefulness was associated with ventilatory irregularity. Based on these data, we devel- oped a neural net to identify these irregular patterns (IRREG). The present study evaluates the performance of this neural net in detecting episodes of wakefulness during nasal CPAP. Materials and methods: 17M/5F with OSAHS (9–119/h RDI untreated) underwent full NPSG during CPAP titration. Periods of IRREG were detected by the neural network from the flow signal after automated exclusion of regions of apnea/hypopnea. Sleep and AASM arous- als were manually scored. IRREG events were tabulated against sleep stage and arousal. All runs of wake greater than 2 min were tabulated for IRREG. Abstracts / Sleep Medicine 7 (2006) S1–S127 S83