This retrospective cohort study used the Healthcare Cost and Utilization Project (HCUP) New York State Inpatient Database covering 2017–2021. The dataset served as the source of patient- and hospital-level variables analyzed to evaluate relationships between institutional sleep medicine capacity and inpatient resource use for acute respiratory illness.
Adults admitted with acute hypercapnic respiratory failure (AHRF) were identified in the HCUP New York State Inpatient Database. The analytic cohort comprised 3,247 adults. The analysis was restricted to hospitals outside New York City to account for differences in hospital density, referral patterns, and access to specialty services relative to the metropolitan area.
Hospitals were classified by presence or absence of an on-site American Academy of Sleep Medicine (AASM-accredited) sleep facility. The authors used accreditation status as a structural proxy for institutional inpatient sleep medicine capacity. No additional process-level measures of inpatient sleep care were reported in the source.
The primary outcomes were hospital length of stay (LOS) and hospitalization cost. Both outcomes were log10-transformed before analysis to address skew and facilitate parametric modeling.
Analyses included bivariate comparisons and hierarchical multivariable linear regression models. The multivariable models adjusted for a range of demographic, clinical, socioeconomic, and hospital-level covariates. Model fit and the independent association of on-site AASM-accredited sleep facility status with outcomes were assessed. The source reports use of hierarchical models but does not provide full variable lists or coefficient estimates in the abstract.
In unadjusted analyses, admission to hospitals that had an on-site AASM-accredited sleep facility was associated with shorter LOS and lower hospitalization cost. The abstract does not report the magnitude of these unadjusted differences or p-values.
After adjustment for covariates in hierarchical multivariable models, the presence of an on-site AASM-accredited sleep facility was not independently associated with either LOS or hospitalization cost. The accreditation variable also did not improve model fit according to the authors' analyses. Specific adjusted effect sizes and measures of model fit are not reported in the abstract.
The strongest independent predictor of both longer LOS and higher hospitalization cost was greater comorbidity burden. Socioeconomic disadvantage indicators—specifically housing instability—and race categorized as other than White were independently associated with increased resource use. By contrast, a diagnosis of chronic obstructive pulmonary disease (COPD) was independently associated with shorter LOS and lower hospitalization cost in this cohort.
Among adults hospitalized with AHRF in non–New York City hospitals within New York State, LOS and hospitalization cost were more strongly related to patient-level comorbidity burden and markers of socioeconomic disadvantage than to the structural presence of an on-site AASM-accredited sleep facility. The authors interpret these findings to suggest that accreditation status alone may not capture the inpatient sleep medicine processes that matter for acute respiratory care. In other words, facility accreditation may be an incomplete proxy for the clinical services or care pathways that influence outcomes and resource use for AHRF.
Clinicians and health system leaders interpreting these results should note that process-level measures of inpatient sleep care, care coordination, or respiratory support practices were not detailed in the source. The findings emphasize the influence of patient complexity and social determinants on LOS and cost and suggest that investments targeting clinical processes and social needs may be more relevant to resource use than accreditation status per se.
The study was exempted by the University at Buffalo Institutional Review Board because it used de-identified administrative data; informed consent was waived. The analytic data derive from the HCUP New York State Inpatient Database and are subject to HCUP data-use restrictions; the analytic dataset is not publicly available. No competing interests were declared.
The abstract indicates limitations inherent to using accreditation status as a structural proxy: accreditation may not reflect the specific inpatient sleep medicine processes relevant to AHRF management. The abstract does not provide detailed model coefficients, confidence intervals, or sensitivity analyses, and it does not report outcomes stratified by specific process measures. As this is a preprint, the findings have not been peer reviewed and the authors caution against using the results to guide clinical practice without further validation.
In unadjusted comparisons, hospitals with on-site AASM-accredited sleep facilities had shorter LOS and lower hospitalization cost for adults admitted with AHRF.
After adjustment, accreditation status was not independently associated with LOS or cost; comorbidity burden and socioeconomic factors were stronger predictors of resource use.
Accreditation status alone may be an inadequate marker of inpatient sleep medicine capacity relevant to acute respiratory care; process-level and social-determinant factors warrant attention in efforts to reduce LOS and cost.
The study used HCUP NYS data (2017–2021), included 3,247 admissions outside New York City, and was exempted by the institutional IRB. The underlying HCUP dataset remains subject to data-use restrictions and is not publicly available from the authors.