• Nizar Z Nader, Jan A Steinel, and Dennis H Auckley.
• Division of Pulmonary, Critical Care, and Sleep Medicine, MetroHealth Medical Center, Cleveland, OH 44109, USA.
• J Clin Sleep Med. 2006 Oct 15;2(4):431-7.

Study ObjectivesObstructive sleep apnea (OSA) case finding in hospitalized patients occurs commonly. In some institutions, the wait time to polysomnography (PSG) may be long. We investigated the impact of a protocol utilizing autoadjusting continuous positive airway pressure (CPAP) for early initiation of therapy in hospitalized patients suspected of having OSA.MethodsA retrospective chart review of patients with likely OSA (oxygen desaturation index > or =10 desaturations per hour with symptoms of OSA) hospitalized between 1999 and 2004 was performed. Subjects were split into intervention and control groups. The intervention group (n=62) underwent autoadjusting CPAP titration while hospitalized and were discharged on fixed CPAP. The control group subjects (n=62) had no autoadjusting CPAP titration and were discharged without CPAP (90% on nocturnal oxygen). Both groups then had a follow-up PSG.ResultsThe 2 groups were matched for baseline characteristics, admission diagnoses, and oximetry parameters. There were no significant differences in time to PSG or apnea-hypopnea index at PSG. In the intervention group, autoadjusting CPAP improved, but did not normalize, oximetry parameters. No significant differences were found in length of hospital stay or in number of urgent care visits, emergency department visits, or hospital readmissions pending PSG. Compared with PSG-determined CPAP pressures, autoadjusting CPAP underestimated the pressure in 60% of the intervention group, whereas 21% required bilevel positive airway pressure for optimal control.ConclusionsCompared with oxygen support or no therapy, an autoadjusting CPAP-titration protocol did not improve short-term outcomes in hospitalized patients with symptoms suggestive of OSA. Autoadjusting CPAP may underestimate optimal treatment settings.

17 keywords...

hide…