Telemedicine video duration as a predictor of 30 day hospitalization: A cohort study of skilled nursing facility patients

Background: With the rapid implementation of telemedicine during COVID 19, the quality of the medical encounters remains unclear One measure of the telemedicine visit is the length of the visit As a part of medical billing, the telemedicine visit length is documented Our aim is to determine the relationship between telemedicine video length and adverse health outcomes of 30 day hospital readmission Methods: This was a cohort study of patients dismissed from the hospital to a skilled nursing facility (SNF) from March 2020 to July 2020 in Olmsted County, MN The residents were in nursing homes which used iPad telemedicine during COVID 19 The primary outcome was 30 day hospital readmission The primary predictor was telemedicine video time as documented in the record The other predictors included age, sex, race, elder risk assessment (ERA) index, heart failure, dementia, diabetes, and educational status We analyzed the outcomes using chi square or logistic regression for unadjusted analysis We created a multivariable model using conditional logistic regression for age, sex, ERA, video time and significant variables Results: In this study of 722 SNF residents, the median age was 84 (IQR 77-90) with 65 % female Overall, we found 76/722 (10 5%) of the SNF residents were readmitted The average video visit length was 34 min (95% CI 27-41) in admitted patients versus 30 min (95% CI 29-31) in patients without readmission (p=0 11) We found that increased age (p=0 0007), history of diabetes (p=0 02) lack of dementia (p=0 002) predicted hospital readmission Sex, race, elder risk assessment index, educational level, and heart failure were not associated with hospital readmission After adjustment for age, sex, ERA, dementia and diabetes, video visit time remained insignificant Conclusion: In this study of 722 SNF residents, video visit length was not associated with 30 day hospital readmission The finding of 10 5% 30 day hospital readmission was below our average nursing home readmission rates Further studies on telemedicine quality metrics in SNF will provide further insights into this growing practice model