More lonely deaths in hospitals and nursing homes from COVID

Patients who died from COVID in 2020 were almost 12 times more likely to die in a medical facility than patients who died from any cause in 2018, reports a new Northwestern Medicine study.

This the first study to look at place of death for patients with COVID-19 and how these distributions compare to previous trends in location of death for non-COVID-19 illnesses. 

The paper was published July 9 in the Journal of the American Geriatrics Society.

“Where you die is important and reflects end-of-life quality for the patient and the family,” said lead author Dr. Sadiya Khan, assistant professor of preventive medicine in epidemiology at Northwestern University Feinberg School of Medicine and a Northwestern Medicine physician. “The patients dying of COVID-19 in medical facilities may not have any family with them because of visitor restrictions. 

“A loved one dying alone takes a huge mental toll on families,” Khan said. “It impairs the family’s ability to grieve and cope with the loss. For patients, we’ve all thought about how terrible it would be to have to die alone. This is the horror happening to thousands of people in medical facilities where no family member or loved one is able to be present with them during their final moments on earth.”

The new study analyzed data from the Centers for Disease Control and Prevention for deaths related to COVID-19 from February 1, 2020, to May 23, 2020, and found 68.7% of patients who die of COVID-19 died in medical facilities, 22.7% in nursing homes, 5.2% at home and 1.9% in hospice facilities. When compared with 2018 deaths due to all causes over a similar time period, 35.7% of deaths took place in medical facilities, 19.1% in nursing homes, 31.1% at home, and 7.9% in hospice facilities.  

There was significant variability across states, with some states having a much higher proportion of nursing home deaths (e.g. Minnesota, 60%) and home deaths (e.g. New York, 8%) deaths. 

“High rates of nursing home deaths in several states reveal a highly vulnerable population and the inability to optimize resources such as PPE (personal protection equipment) to prevent infection transmission these high-risk locations,” Khan said. “It’s especially important as nursing homes are reopening to visitors and may be exposing residents, especially in areas where there are increasing rates of cases.”

But nursing home statistics in the study only capture a glimpse of the high proportion of deaths linked to these facilities. It doesn’t include people who contracted COVID-19 in a nursing home and were transferred to a hospital or staff who got it working there.

To address the heightened risk in nursing homes, Khan suggests access to adequate PPE for staff and universal testing/screening of people before they are allowed to enter the facilities, even if they don’t have symptoms. 

To support lonely COVID-19 patients in hospital beds and nursing homes, Khan said these facilities need a virtual infrastructure.

“We can’t just rely on individual’s iPhones and iPads,” Khan said. “There is a land line phone in every room, why couldn’t we have a virtual phone in every room or access to face-to-face communication for each patient and their families?”

“These results highlight yet another way that COVID-19 has impacted the health care system,” said first author Dr. Sarah Chuzi, a Northwestern Medicine fellow in cardiovascular diseases. “While recent research shows U.S. deaths in medical facilities are decreasing and deaths at home and in hospice facilities are increasing, the burden of deaths attributed to COVID-19 may reverse these overall trends. 

“End-of-life care is a hugely important but understudied aspect of medicine. We wanted to ensure this aspect of patient-centered care was acknowledged and studied in order to motivate efforts to improve our current system,” Chuzi added.

Other Northwestern authors include Dr. Sarah Chuzi, Rebecca Molsberry, Megan McCabe, Dr. Clyde Yancy, Dr. Adeboye Ogunseitan, Norrina Allen, Rebecca Molsberry and Megan McCabe. 

The research was supported in part, by the National Institutes of Health's National Center for Advancing Translational Sciences, grant KL2TR001424.