LAUSR

Introduction/Hypothesis: Acute respiratory failure (ARF) with mechanical ventilation (MV) is a limited, labor-intensive resource that is associated with high mortality. Planning for seasonal variations in the utilization of this resource is essential for healthcare system preparation. We aim to describe the longterm national seasonal variation of ARF case volume. Methods: This is a longitudinal cross-sectional study using the 2002-2014 Healthcare Utilization Project’s National Inpatient Sample datasets, which contains data from approximately 7 million annual hospital discharges. We utilized 5 ICD-9-CM diagnostic codes (518.5, and 518.81-84), and 3 procedural codes for MV (96.70-72) to define ARF. Accounting for the complex NIS sample design, we describe the quarterly and annual national case volume for these codes. To examine inter-year variability, we calculate the difference between the maximum and minimum quarterly sum of discharges within a year relative to that year’s minimum quarter. Results: From 2002-2014 there were an estimated 482 million hospital discharges. We found that the most utilized diagnosis code was 518.81(Acute respiratory failure). When using the only diagnostic codes the annual incidence increased from 347 to 788 persons per year over the study period with an average hospital mortality of 20%. When using diagnostic plus procedural codes the annual incidence increased from 187 to 253 per 100,000 persons per year over the study period with an average hospital mortality of 30%. The within-year variations from maximum to minimum quarter are generally larger for diagnostic codes than for diagnostic codes with procedure codes for MV. The relative difference broken down by hospital size shows large hospitals at 17.5%, medium 20.7%, small at 23.5%. When broken down by type, rural non-teaching hospitals had an average relative difference of 29.4%, Urban non-teaching 21.5%, and urban teaching 14.8%. When examined by region the relative difference was 16.7% for the Midwest, 15% for the Northeast, 20% in the South, and 21.2% in the West. Conclusions: The seasonal variation of ARF was most was most pronounced in small, rural, southern, and western hospitals. Further Hospital-level studies may help to create predictive models for hospitals to predict seasonal needs for mechanical ventilation resources.