Despite the ever-present risk of a pandemic sweeping across the world, we were wholly unprepared for COVID-19 and it swiftly brought us to our knees. Stockholm suffered an intense wave of the virus beginning March through May and the pandemic still represents a global challenge.

Danderyd Hospital, an acute-care 530-bed hospital providing both general and specialized hospital care, admitted an escalating number of patients, culminating at a staggering 133 COVID-19 patients/day at the end of April. Faced with an unknown virus, we did the best we could considering the circumstances and in a matter of days, the whole hospital was reorganized, which was no small feat. To address the need for surge capacity not seen for decades, general medical wards switched to COVID-19 wards overnight, and medical staff were forced to adapt to new procedures and logistic challenges. Testing capacity and personal protective equipment was limited throughout the nation, and priorities had to be made on an hourly basis.

As limited social contacts were encouraged among the rest of society to reduce the spread of the infection, healthcare workers were expected to do the opposite. Although the occupational risk of healthcare workers to be infected with SARS is well documented, little was at the time known about the safety of COVID-19 front-line healthcare workers. Despite efforts to protect the health and safety of this essential workforce, many healthcare workers were anxious about both their own safety and about passing the infection to their families.

To address this uncertainty, we determined the seroprevalence of SARS-CoV-2 IgG antibodies among healthcare workers and the associations between COVID-19 infection and occupational exposure to SARS-CoV-2 at Danderyd Hospital. In line with reports from the SARS epidemic, our findings support an occupational risk of SARS-CoV-2 infection. Among the 2149 study participants, 19.1% had developed IgG antibodies against SARS-CoV-2 by early May. Not only was this substantially higher than the seroprevalence found in the general population of Stockholm during the same time period, but also associated to COVID-19 patient contact. The seroprevalence was furthermore significantly higher among healthcare workers in non-COVID wards compared to healthcare workers without patient contact.

Since our results were contradictory to several recent reports of seroprevalence among healthcare workers, we assessed possible divergencies in infection prevention and control precautions. Due to limited testing capacity, we did not RT-PCR test and subsequently isolate infected healthcare workers, and RT-PCR testing was limited to patients with typical COVID-19 symptoms. We argue that this may have contributed to a substantial number of SARS-CoV-2 positive cases among practicing healthcare workers as well as difficulties in distinguishing between COVID-19 and non-COVID-19 patients leading to transmission in non-COVID-19 wards. PPE was furthermore only worn by healthcare workers in contact with patients with known or suspected COVID-19, aerosol filtering face masks were restricted to aerosol-generating procedures, and we reused and shared manually-cleaned and disinfected face shields. Although the effect of restricted RT-PCR testing and PPE usage was not scientifically evaluated, these factors may have contributed to transmission to and among healthcare workers.

The state of knowledge about SARS-CoV-2 increases as the pandemic continues, and measures are being improved and adapted accordingly. Our self-investigating approach serves as an example of how strategies during novel situations must be under constant re-evaluation, including appropriate infection control measures to protect healthcare workers as our most valuable resource. Our findings support comprehensive and decisive measures to mitigate intra-hospital transmission. The principle of zero occupational infection must be strived for, starting with the recognition of COVID-19 as a possible occupational disease.

Collaborating authors: Sebastian Havervall, Ann-Sofie Rudberg, Peter Nilsson, Sophia Hober, Charlotte Thålin