Background: Timely discharge of medically-ready patients from acute care hospitalization can reduce strain on hospital resources and limit patient exposure to iatrogenic harm [1-2]. The smooth transition of these patients to post-acute care has been critical during the COVID-19 pandemic to increase acute care bed availability and reduce the potential of hospital-based viral transmission [3]. This study analyzed differences in barriers to discharge and hospital length of stay (LOS) between two time periods: before and during the COVID-19 pandemic. The investigation focused on a subset of acute care patients at Johns Hopkins Hospital (JHH), an urban academic medical center, preparing for discharge from medical, surgical, and neurology units and identified as having a barrier to appropriate discharge on multi-disciplinary rounds.
Methods: Barriers to discharge and discharge disposition data were collected by the Department of Care Coordination and Utilization Management at JHH in a six-week baseline period before COVID-19 became pervasive in the hospitalized population (February 1 – March 15, 2020) and in a six-week period when the prevalence of COVID-19 had risen among hospitalized patients (April 1 – May 15, 2020). Patients were medically ready for discharge but experienced delays due to extenuating barriers. This analysis included 64 patients in the baseline period and 68 patients in the COVID-19 period. ANOVA tests assessed LOS differences between specific barriers to discharge and discharge dispositions. Two-sample unequal variance t-tests examined LOS differences between the two time periods.
Results: Average LOS for the subset of acute care patients with identified barriers to discharge was significantly higher before COVID-19 (42.5 vs 28.1 days, p<0.05). For both periods, “High Cost/Complex Care Needs” was the most common barrier to discharge (40.6% before vs 42.6% during COVID-19) and “Use of Restraints” was the barrier to discharge with the longest average LOS. However, the second and third most common barrier to discharge differed between periods: “Patient/Family” (21.9%) and “Unfunded/Uninsured” (18.8%) were the next most common before, while “Insurance Company” (20.6%) and “COVID-19” (11.8%) were elevated during COVID-19. For both periods, “Skilled Nursing Facility,” “Self-Care,” and “Home Health Services” were the three most common discharge dispositions (32.3%, 17.7%, 14.5% before and 35.8%, 23.9%, 16.4% during COVID-19, respectively). Despite no statistically significant difference in average LOS for barriers to discharge and discharge dispositions between the periods, the discharge disposition with the longest average LOS during COVID-19, “Rehab,” had LOS < 40 days while the five discharge dispositions with the longest average LOS before COVID-19 all had LOS > 40 days.
Conclusions: For this subset of acute care patients with identified barriers to discharge, the three most common discharge dispositions and barriers to discharge were the same before and during COVID-19. However, there were differences in other common barriers to discharge, and average LOS was significantly lower during the pandemic. This decrease may have been influenced by elective surgery postponement, offering alternative arrangements to grow inpatient capacity, emergency waivers, and admission requirement changes at post-acute facilities due to COVID-19. Further assessment of barriers during the pandemic and routine circumstances could elucidate causality and additional discharge strategies.