COVID-19 is occurring in clusters, making good data and resource allocation crucial

COVID-19 is occurring in clusters, making good data and resource allocation crucial
© Getty Images

Amid worldwide panic, the media are giving continual updates on COVID-19 cases and deaths with data feeds from websites including Johns Hopkins, the Centers for Disease Control and Prevention (CDC), and the World Health Organization (WHO). An additional site, Worldometer, integrates these feeds and additionally pulls in local sources of information to give country and state-level reports. As a result, Worldometer reports a greater number of both cases and deaths, and additionally attempts to report on severity when possible.  

None of these publicly available data feeds to the media report daily hospitalizations. Because COVID-19 results in a spectrum of illnesses, from asymptomatic to fulminant multi-organ failure, reporting the test-positive cases without any information on severity is resulting in paralysis of reallocation of critical resources such as protective face masks and equipment including ventilators. The rate of positive tests in any area is determined by the availability of COVID-19 testing and the proclivity of the population to get tested. It is not a reliable proxy for the COVID-19 pandemic.  

The deaths attributed to COVID-19 reported on media banners are also problematic. People who die outside of a hospital may not be swabbed for COVID-19, and even those deaths that occur in the hospital may be variably delayed in their reporting to departments of community health. The most important and actionable epidemiological data are the daily incidence and prevalence of hospitalized patients stratified by those admitted to hospital wards and those in intensive care units requiring supportive measures including mechanical ventilation. 

ADVERTISEMENT

Experts who advise governments and the public should be doing so based upon hospitalization data. An executive order is needed mandating that each U.S. hospital report its daily census of confirmed COVID-19 patients, admitted to medical floors and intensive care units, to a central data center such as the CDC, which has been receiving only voluntary reports and is out of date.  

The current thinking among hospital administrations is in line from official and expert messaging that COVID-19 will spread evenly across the U.S. and that each major city eventually will become like New York. As a result, each hospital is hunkering down for anticipated waves of COVID-19 patients and conserving resources with no thought of reallocation to hospitals in greater need. Many states have mandated the cessation of elective surgeries and procedures, freeing up many face masks, protective gowns, gloves, ventilators and anesthesia machines, which can be used for mechanical ventilation. In areas of COVID-19 inactivity, these resources are idle when some metro areas such as New York are overwhelmed with patients. A response to the disproportionate needs created by COVID-19, predicated on increased production of medical supplies and equipment, cannot be as rapidly responsive as reallocation of idle resources.  

An alternative view to the present epidemiological analysis influencing hospital administrations and the government is that COVID-19 will not spread uniformly in the United States. COVID-19 hospitalizations per 1,000 inhabitants will reveal important clustered patterns that are not discernible from the test-positive and mortality data currently shown. It appears that COVID-19 does not spread uniformly in any country or region, but rather clusters in hot spots according to principles of communicability. These principles include not only when the virus arrived in the community, but also the tendencies for close contacts living together in smaller apartments and homes; the use of elevators, public transportation, restaurants and stores; and on the prevailing weather.  

Viral epidemics dependent on human-to-human spread tend to settle down during the summer months, partially because of less indoor crowding but also because of less intra-individual reinfection with the virus. For example, with COVID-19, the virus is present in nasal secretions and presumably in expired/coughed/sneezed air for three to four weeks. Thus, the air density of viral particles, known as aerosols and droplets, could be reduced with fresh air through open windows or being outdoors.  

Municipalities are frightened that they will become the next COVID-19 hot spot on the scale of New York City. Our leaders deserve a chance to understand — from experts and the media — the best epidemiological data and modeling based on these data. Fortunately, to date the COVID-19 pandemic cannot be characterized as a sequential planting of equipotent viral seeds across a region. It did not spread uniformly to every region of China or Italy. For example, 60 percent of Italian deaths occurred in the Milan metro area, and this trend continues.

ADVERTISEMENT

We are seeing the same pattern of clustered spread in the U.S., with New York dominating the reports of hospitals being overwhelmed, relative quiet across the Midwest, and little or no activity in some areas of rural America. Hospitalizations on the news banner “scoreboard” by city and state, preferably reported per 1,000 inhabitants, would clarify the nation’s immediate needs.  

Reallocation of critical resources from centers with manageable COVID-19 activity to overwhelmed hospitals is the most rapid approach to this crisis. In short, the right data and the most informed analyses will save lives. We urgently need daily hospitalization data to perform the most informative epidemiological analyses, which will help us understand that this is a clustered, rather than uniformly spreading, outbreak.

Peter A. McCullough, MD, MPH, is vice chairman of medicine at Baylor University Medical Center and a professor of medicine at Texas A&M College of Medicine in Dallas. An internist, cardiologist and epidemiologist, he is the editor-in-chief of “Cardiorenal Medicine” and “Reviews in Cardiovascular Medicine.” He has authored over 500 cited works in the National Library of Medicine.