COVID-19 Watch + COVID-19 Pulse

The clinical guidance for 90 percent of infected COVID-19 adult patients who do not meet eligibility for inpatient admission is to self-isolate. To support these patients, alternatives to in-person care are needed to manage an unpredictable clinical course; identify and intercept patients rapidly deteriorating at home, prevent viral spread during in-person visits; and minimize future surges in emergency departments (EDs). In addition, fingertip pulse oximeters have been proposed to improve in-home early detection of respiratory deteriorations but are untested and the operational infrastructure to support large-scale monitoring is limited. While telemedicine has been widely adopted during the pandemic as an alternative to conventional outpatient care, limited telemedicine access may be exacerbating observed disparities for Black and Latino patients. In our health system, Black and Latino patients used video-visits 15 percent less often than white patients. Text messaging and phone calls may improve healthcare access for communities of color, but the evidence for these telecommunication modalities to be effective and improve equity are limited. The University of Pennsylvania Health System (UPHS) developed and deployed COVID Watch to improve access to health care for COVID-19 patients who are self-isolating at home. COVID Watch sends twice-daily, scheduled text messages to assess patients for shortness of breath using a clinical algorithm to determine whether patients need an urgent escalation to a team of dedicated, on-call nurses within one hour. These nurses are supported by an on-call team of clinicians who can conduct urgent phone or video assessments. Patients can also trigger the algorithmic assessment independent of the scheduled messages. As of May 21, 2020, COVID Watch has managed 3,628 COVID-19 patients at home, of which 1,295 are confirmed COVID-19 positive; of these, 61 percent are Black or Latino, higher than the proportion of all UPHS COVID-19 positive patients that are Black or Latino (55 percent).