0604 Comorbidities and Admission Rates in Inpatients Undergoing Sleep Studies

Abstract

Abstract Introduction Uncontrolled sleep-disordered breathing (SDB) and hypoventilation, which are common in COPD, CHF and obesity hypoventilation patients can lead to death and readmissions. It is unknown whether inpatient sleep studies to diagnose and optimize treatment improve care and prevent readmissions. Methods All patients > 18 years old with sleep studies while inpatient at Baystate Medical Center between October 2015 and September 2017 were included. Patient characteristics, comorbidities, sleep study diagnoses, and treatment recommendations were evaluated. Admission (inpatient or observation) and death rates were determined for 1-year before admit date and 1-year after discharge date of index admission. Results 326 adult inpatients had 120 portable and 304 in-laboratory tests performed. Average age was 62.9±14.4, mean BMI was 37.2±12.3 and 56% were male. Principal diagnoses were CHF (50%), COPD (39%), both COPD and CHF (20%) and obesity hypoventilation (27%). 31 used PAP and 71 used oxygen prior to admission. Sleep diagnoses included OSA (73%), central sleep apnea (CSA) (29%), treatment emergent CSA (8%), hypoxia (48%), hypoventilation (41%), and normal or non-diagnostic (6%). Treatment recommendations included CPAP (25%), BiPAP (18%), BiPAP ST (3%), ASV (4%), iVAPS (22%), oxygen only (5%) and further titration (20%). The average length of stay was 11.6 ± 9.6 days. There was no difference in the percentage of patients who had an admission before or after their sleep study (53% vs 56%, respectively). In addition, no difference was seen in the median number of admissions before and after the sleep study (median=1.0, IQI=0-2, p=0.77). 90-day readmission rate was 19%. 14% died. Conclusion SDB, hypoxia and hypoventilation were common in inpatients evaluated with sleep studies with PAP therapy recommended in most patients. Further research is needed to determine whether inpatient testing and subsequent treatment can result in decreased readmissions and death. Support None