SARS-CoV-2 Antibody Formation Among Healthcare Workers, September 2, 2020

Interpretation

For most employees, it is not possible to determine whether they acquired infection in the hospital or in the community. When the rate for job classifications is much higher than the background community rate or when contact tracing yields a likely exposure for a group of individuals, the site of transmission can often be inferred. Early on, there was an excessive focus on the travel history from Asia and specifically China to identify risk. It was unrecognized that the pandemic had entered the New York City area from Europe and local transmission in our communities was rapidly accelerating. As a result, many staff were exposed without Personal Protective Equipment at that time. It was not uniform policy for all employees to wear masks when gathering outside of patient care areas such as the cafeteria, break rooms and conference rooms. There was no hospital associated infections transmitted to employees after April 15^th when the true character of the epidemic was realized and policies were improved.

Many employees had a COVID-19 syndrome with exposure in the midst of the epidemic but had negative nasopharyngeal swabs. This is likely due to the absence of the virus in the nasopharynx even as they had infection and went on to make antibodies. It is unlikely to be operator error in swabbing because a small cadre of trained technicians and nurses performed all of the swabbing. Many more had a mild COVID-19 syndrome with exposure in the midst of the epidemic and likely had true SARS-CoV-2 infection but did not make the antibody measured in this assay especially if they did not have fever. Fever appears to be a very important indicator of antibody response because it is an indicator of more severe and systemic disease. Milder and more local disease may be due to exposure to a low level inoculum. The severity of illness in COVID-19 has been shown to correlate with the use of face masks which likely reduce inoculum. (5) There were clusters of COVID-19 cases in Administration, Information Systems and Plant Maintenance which likely originated from the community with subsequent intrahospital spread.

There was an abrupt drop in hospital acquired COVID-19 infections among employees when the true sources and modes of infection were discovered. Among physicians, primary care doctors with unprotected exposures in their offices were disproportionately affected while all 9 critical care specialists with extensive but protected exposure in the ICU, remained uninfected. This demonstrates that Personal Protective Equipment is effective in preventing infection and it is the unrecognized infected and often asymptomatic or minimally symptomatic coworker or patient who represents the greatest risk for transmission.

The overall half-life for the SARS-CoV-2 IgG antibody measured with multiple determinations was 148 days with confidence range 131 to 171 days which is consistent with serologic responses after many viral infections. It is significantly longer than the 36 days, measured on two occasions, in a cohort of community patients. (6) Many had stable or rising antibody levels for unclear reasons.

Most theorists estimate that 16% immunity is inadequate to prevent a resurgence of infection in an uncontrolled environment. It is not possible to factor in the effect of careful mask wearing, hand hygiene and disinfection but these activities coupled with an additional increment of unmeasured immunity among those with infections too mild to elicit antibody will likely be sufficient to protect healthcare workers through any community resurgence that may occur in the fall or winter. Further research is needed into the full spectrum and duration of immunity to SARS-CoV-2.