Evaluating the role of inpatient sleep studies in hospitalized adults in relation to 1-year mortality and healthcare utilization: a population-based cohort study

RATIONALE/RESEARCH OBJECTIVE: Sleep-disordered breathing (SDB) is common among hospitalized individuals and may contribute to increased health utilization and mortality after discharge, yet the optimal timing and setting for sleep testing remain unclear. We compared 1-year healthcare utilization and mortality among hospitalized individuals undergoing (1) inpatient polysomnography (PSG), (2) delayed post-discharge PSG within 1 month of discharge (primary analysis), or (3) no PSG during hospitalization or within one year after discharge (matched non-PSG controls); secondary analyses examined delayed PSG completed within 3 and 6 months post-discharge. OBJECTIVES/METHODS: We conducted a retrospective province-based study using health administrative databases on all adults aged ≥18 years (Ontario, Canada), hospitalized between 2012 and 2018, and followed until the earliest of death, loss to follow-up, or December 31, 2019. Outcomes within 1 year after hospitalization included all-cause death and healthcare costs. To balance baseline characteristics across study groups, we used overlap propensity score weighting to create similar weighted populations for pairwise comparisons. RESULTS: We identified 748 individuals in the inpatient PSG group, 9310 in the delayed PSG group, and 7480 in the matched non-PSG controls by exact age, sex, health region and calendar year to the inpatient PSG group. In weighted populations, compared to delayed PSG (within the first month), inpatient PSG group was associated with increased 1-year mortality (HR: 2.18, 95% CI: 1.68-2.82) and higher cost (Ratio of Means [RoM]: 1.82, 95% CI: 1.64-2.03). Compared to non-PSG controls, inpatient PSG was associated with decreased mortality (HR: 0.72, 95% CI: 0.55-0.94) and higher cost (RoM: 1.73, 95% CI: 1.50-1.99). In the secondary analyses, delayed PSG, compared to non-PSG controls, was associated with lower cost when conducted within 3 and 6 months, but not within one month. CONCLUSIONS: In this population-based study of propensity score-weighted hospitalized individuals, inpatient PSG was associated with higher mortality and costs compared to delayed PSG, and with lower mortality, but greater resource use than non-PSG controls. Although weighting reduced measured differences, unmeasured clinical complexity may still influence these associations. Among hospitalized individuals with suspected SDB, delaying PSG until after discharge may achieve similar or better outcomes at lower cost, warranting confirmation in prospective studies.