Patients hospitalized with acute exacerbation of chronic obstructive pulmonary disease (AECOPD) who have obstructive sleep apnea (OSA) have a higher cost of care but lower mortality rates than patients with AECOPD without OSA, according to study findings published in the Journal of Clinical Sleep Medicine.
Investigators sought to characterize in-hospital outcomes among patients with COPD-OSA overlap who are hospitalized with AECOPD. Primary outcomes included length of stay, associated hospitalization costs, and in-hospital mortality. Secondary outcomes included use of positive pressure ventilation therapies.
The investigators used data from the United States National Inpatient Sample (NIS) (part of the Healthcare Cost and Utilization Project databases) to conduct a retrospective, observational, case-controlled design analysis of patients hospitalized with AECOPD. The NIS includes estimated costs and outcomes from several million inpatient stays per year in hospitals across the US. The investigators matched patients with COPD-OSA overlap 1:1 with patients who were hospitalized with AECOPD but did not have OSA.
Diagnosis codes were used to identify adult patients admitted from January 2007 through December 2017 with a primary diagnosis of AECOPD, as well as those with the additional diagnosis of OSA, with patient-matching based on body mass index (BMI), race/ethnicity, age, and sex.
“
The diagnosis of OSA is associated with reduced mortality in these hospitalizations, which may be related to greater utilization of supportive ventilation when OSA is recognized.
Overall, the AECOPD study sample included 508,165 patients with COPD-OSA overlap (mean age, 64.77 years; 52.37% women; 77.82% White, 15.38% Black, 4.51% Hispanic), and 505,682 patients without OSA (well-matched) in analysis. After matching, the OSA group had a slightly higher propensity to illness vs the non-OSA group (Charlson Comorbidity Index 4.76 vs 4.43), though not a clinically significant difference.
The investigators found that the COPD-OSA-overlap cohort vs the no-OSA cohort had longer median lengths of stay (4 days vs 3 days), increased utilization of noninvasive positive pressure ventilation (NIPPV) (13.92% vs 6.78%), and higher mean costs ($32,197 vs $29,011) (all P <.001). When mechanical ventilation was required for more than 4 days, COPD-OSA overlap was associated with earlier initiation of mechanical ventilation (2.53 days vs 3.35 days from admission to procedure; P =.001).
In subset analysis stratified by duration of continuous NIPPV (<24 hours, 24-96 hours, >96 hours), NIPPV usage showed statistically significant differences, with the COPD-OSA-overlap cohort vs the non-OSA cohort showing higher utilization.
The investigators noted COPD-OSA overlap was associated with reduced mortality compared with no-OSA (0.81% vs 1.05%; P <.001), possibly related to earlier use of NIPPV at higher frequencies. After adjusting for confounders, the differences remained in NIPPV use (adjusted OR [aOR], 1.998; 95% CI, 1.970-2.026) and in mortality (aOR, 0.650; 95% CI, 0.624-0.678).
Study limitations include missing information/coding errors in the data set, the study’s retrospective observational design, and a lack of data regarding COPD or OSA severity.
“Patients with COPD-OSA overlap have higher utilization of supportive ventilation and longer length of stay during AECOPD hospitalizations, contributing to higher costs,” the investigators concluded, adding, “The diagnosis of OSA is associated with reduced mortality in these hospitalizations, which may be related to greater utilization of supportive ventilation when OSA is recognized.”
