OBJECTIVES: This study explored the variation in COVID-19 related average length of stay by Social Vulnerability Index (SVI) of patients in the US

METHODS: This study included patients diagnosed with COVID-19 infection between 1^st March 2020 to 28^th Feb 2021 with ICD-10-CM recorded in Optum’s de-identified Clinformatics® Data Mart Database. Only the patients having continuous eligibility of 3-months post (follow-up period) the first diagnosis of COVID-19 (index date) were included in the study. Zip code (if available) closest to the COVID-19 diagnosis index of those patients were joined with the Minority Health SVI (mhSVI) file to map patients’ SVI score ranging from 0 to 1. We have estimated the average length of stay across the four patient groups, categorized based on SVI score (Group 1 (least vulnerable): 0-0.25; Group 2: 0.251-50; Group 3: 0.501-.75; Group 4 (highly vulnerable): 0.751-1) and have applied the Wilcoxon statistical test to analyze the level of significance in cost variation across the four SVI groups. Further, we will also be looking at the region-wise trends in length of stay across the four groups.

RESULTS: Among 120,501 patients diagnosed with COVID-19, we were able to map the SVI score for 98% of patients. We observed a significant longer length of stay for patients as we moved from group 1 (8.5 days) to group 4 (9.7 days). The estimated length of hospitalization significantly increased by 8%, 9%, and 13% respectively from group 2 to group 4 when compared to group 1.

CONCLUSIONS: Vulnerable patients had more impact of COVID-19 as compared to less vulnerable group. Significant variation in length of stay might be correlated with difficult access to healthcare, severe COVID-19 presentation, and more co-morbid conditions. This study helps to determine the gaps in healthcare resource allocation as well as supports planning and preparedness for pandemic.