Residents of a nursing home in the mid-Atlantic region experienced declines in blood-oxygen saturation levels several days before developing symptoms of COVID-19, according to a pilot study co-authored by Mendel Lerner ’20K, a recent graduate of the quantitative economics program at Katz School of Science and Health.
“In this study, we retrospectively evaluated 84 nursing home residents to identify whether pulse-oximetry would be useful as an early indicator of a nascent COVID-19 outbreak in the facility,” said Lerner.
The study, “Oxyhemoglobin Saturation (SpO2) as an Early Indicator of COVID-19: A Pilot Study,” appears in SSRN, a platform for early-stage research. The other co-authors were Rabbi Dr. Yossi Ives, founder and chairman of Tag International Development and Tag Institute for Jewish Social Values, and Dr. Michael Akerman, director of pulmonology for both Highland Medical, P.C., and Montefiore Nyack Hospital.
Pulse-oximetry, or the measure of blood-oxygen levels, is an effective tool for detecting the attack on the lungs caused by the COVID-19 virus. The pulse-oximeter is a small device that attaches to the tip of a finger and in 15 seconds measures oxygen saturation in the blood. Symptoms from COVID-19 generally develop between five and 10 days after infection, but it doesn’t cause shortness of breath in most patients. Oxygen saturation can drop silently over days, and by the time patients feel shortness of breath or have evident trouble breathing, they may already have alarmingly low oxygen saturations, having been infected for over a week and spreading the virus to others.
Healthy people normally get a regular oxygen reading above 95%, though that number can be affected by exercise or other activity. Individuals with existing health conditions may have a lower normal reading, but a steadily falling number may indicate an infection or the development of a comorbidity, which is another chronic disease or condition that has reduced the body’s ability to maintain the baseline oxygen saturation.
“We analyzed trends in body temperature and blood-oxygen saturation in the overall population for the days surrounding the COVID-19 test results to determine the efficacy of measuring blood-oxygen levels as an early sign of an early-stage, possibly asymptomatic, outbreak of COVID-19 within a facility,” he said.
Current nursing home guidance from the(CDC) focuses on daily temperature monitoring to detect a COVID-19 infection, as typical under the general medical practice of identifying an ill patient.
“In regard to COVID-19,” said Lerner, “our findings indicated that temperature monitoring may not be the best or earliest indicator of an impending outbreak.”
Megadata Health Systems, a data analytics company in New Jersey, approached Lerner to conduct a study in order to test their assertion that hypoxemia was prevalent in the residents of its 300 nursing homes during tests for early detection of COVID-19.
Of the 84 residents in the study, 24 tested positive for the severe acute respiratory syndrome, SARS-CoV-2, in their first round of COVID-19 testing. Although the baseline blood-oxygen values were lower, as expected, than the average for healthy young adults, the mean values were above the 95% that is generally considered mild hypoxemia.
“Even with the comorbidities and lower oxygen saturation of the elderly population,” said Lerner, “this population was otherwise healthy and, therefore, the unhealthy effects of COVID wouldn’t be swallowed up by normal unhealthy variations due to the group’s comorbidities.”
An increase in individual temperature of at least 1 degree Fahrenheit and decreases in blood-oxygen levels of both 1% and below 95% were considered significant. From COVID-positive residents, they observed a mean decrease in blood-oxygen levels from 95.8% to 93.5% on the date of COVID-19 testing 14 days later.
“Our findings demonstrated that declines in blood-oxygen levels preceded temperature increases and COVID testing by many days,” said Lerner. “Therefore, we strongly recommend daily pulse-oximeter testing in all vulnerable populations and automated statistical trending and monitoring as an early warning system that could be provided by companies that manage the electronic medical records for these facilities.”