How We’re Going to Solve the Nursing Home Pandemic with Data

Nursing home residents are being hit hard by the Covid-19 pandemic. Very hard. There have been over 27,000 deaths in nursing homes compared to about 80,000 deaths overall in the U.S. That means roughly one third of the fatalities in the U.S. are occurring in nursing homes.

Much of the coverage has been about the number of deaths and whether nursing homes are going to liable to get sued. Instead of looking for someone to blame, perhaps there is hope in stopping or at least drastically slowing the spread.

I own and operate a health data company providing operational and clinical analytics for nursing homes. In early March, we refocused our business on helping our clients stop the spread of the disease using data. We started by reporting on residents with high temperatures and alerting for possible Covid-19 symptoms written in the medical records.

When the virus hit, it was clear that nursing homes were going to be severely affected. The nursing home population is elderly, and many residents have pre-existing conditions. This is also an environment where it is extremely difficult to ensure proper social distancing especially considering that these residents need nursing care and thus interaction with staff members. It was inevitable that the virus would penetrate the nursing homes as the employees live outside the facility and were going to bring the disease in from the community each day as they traveled back and forth from home.

The CDC and the Centers for Medicare and Medicaid Services (CMS) – the federal agency which regulates nursing homes – put several measures in place to protect the frail residents. Visitors were disallowed. Staff members were required to wear personal protective equipment, and no employee was to be allowed into a facility without having their temperature taken. Residents needed to have their temperatures taken every day and monitored for symptoms such as cough, sore throat and shortness of breath. This is where we focused our analytical efforts to stop the spread.

However, the emphasis on temperatures was a mistake. It is becoming clear now that temperatures and outward symptoms are very late developing symptoms. People are contagious from Covid-19 for as much as two weeks before they show any of these signs. Therefore, taking temperatures was minimally effective if at all.

So, what is the earliest method of detecting Covid-19 in a nursing home?

The ultimate answer is going to be widely available rapid testing that can be done daily for every resident and staff member even if they are asymptomatic. But this has been slow to materialize.

Dr. Richard Levitan in his article titled “The Infection That's Silently Killing Coronavirus Patients” published in the NY Times on April 20th clearly explains how the earliest sign for Covid-19 is “silent hypoxia” or a drop in blood oxygen saturation with no other outward symptoms.

Dr. Levitan writes “There is a way we could identify more patients who have Covid pneumonia sooner and treat them more effectively — and it would not require waiting for a coronavirus test at a hospital or doctor’s office. It requires detecting silent hypoxia early through a common medical device that can be purchased without a prescription at most pharmacies: a pulse oximeter.

“Pulse oximetry is no more complicated than using a thermometer. These small devices turn on with one button and are placed on a fingertip. In a few seconds, two numbers are displayed: oxygen saturation and pulse rate. Pulse oximeters are extremely reliable in detecting oxygenation problems and elevated heart rates.”

Dr. Levitan was exploring how people can detect Covid symptoms at home and potentially go to the hospital before severe disease sets in. There has been debate in the medical community whether pulse oximeters can be relied upon to be used at home. It is important to understand your baseline oxygen level and the accuracy of the reading may vary by manufacturer. Some suggest it may be better to rely on your doctor to do the reading.

But what about nursing homes? A nursing home is a setting that has an extremely high risk for Covid outbreak. Pulse-ox readings are being done by trained nurses and early detection really matters in preventing an outbreak.

This was the inspiration for us to conduct our own study. What we discovered was fascinating. We studied the data of three hundred nursing homes and found that in most cases the oxygen saturation numbers were flagging the problem as much as two weeks before high temperatures! In cases where high temperatures were spiking first, it was mainly because daily pulse-ox readings were not being done on every resident.

The CDC’s guidance does not mandate a proactive pulse-ox be done on every resident – only temperatures. Some states have mandated that a full set of vital checks be done preemptively. However, the rules vary state by state. Even in states where oxygen readings are mandatory, it is used as a tool to monitor an individual resident but not as a way to monitor and track the spread.

Sample chart demonstrating Covid-19 spread in a nursing facility

Until widespread testing is available, oxygen saturation levels can be used to detect the virus early. These are the protocols that need to be put in place:

First, a pulse ox should be done on every resident every day. Then we can look at the number of residents with a reading of less than 95% oxygenation who are also more than 1% below their baseline. The baseline is important as many residents with pre-existing conditions – such as COPD – may start out with an oxygen saturation level below 95%.

We then count the number of residents who flagged this way daily and track it day by day. We found that as soon as the number increases to ten residents or higher in most facilities, this was an indicator of possible outbreak.

At this point, residents can be tested for Covid-19 and isolated in a separate unit. Stronger infection control measures can be put in place. They can get a chest x-ray and considered for discharge to the hospital. This can be done about two weeks earlier than if we would just be relying on temperature checks.

I hope that a strong focus will be placed on performing proactive pulse-ox readings and that we can further the discussion on how we can stop the spread in nursing homes.

This is how we can save lives.