The study, based on self-reported data from users of the COVID Symptom Study smartphone app between 24 March and 23 April 2020, found the prevalence of COVID-19 was 2,747 per 100,000 app users among frontline care workers compared with 242 per 100,000 app users from the general community. After accounting for differences in testing for healthcare workers compared with the general community, the researchers estimate frontline workers are around 3.4 times more likely to test positive for COVID-19.
Professor Andrew Chan, senior author, from Massachusetts General Hospital, USA, said: "Previous reports from public health authorities suggest that around 10-20% of COVID-19 infections occur among health workers. Our study provides a more precise assessment of the magnitude of increased infection risk among healthcare workers compared to the general community. Many countries, including the US, continue to face vexing shortages of PPE. Our results underscore the importance of providing adequate access to PPE and also suggest that systemic racism associated with inequalities to access to PPE likely contribute to the disproportionate risk of infection among minority frontline healthcare workers." 
Gloves, gowns, and face masks are recommended for those caring for COVID-19 patients, but surging demand and supply chain disruptions have resulted in global shortages. Some areas have attempted to conserve PPE by reusing items or using them for longer periods of time, but data on the safety of such practices is scarce.
The latest study is based on data collected from the COVID Symptom Study smartphone app between 24 March and 23 April 2020. App users were asked to provide background information about themselves, such as age, race and whether they already have any medical conditions.
Participants were also asked if they worked in health care and, if yes, whether they had direct patient contact in their job. For the purposes of the study, frontline healthcare workers were defined as participants with direct patient contact, and this group was further subdivided according to whether they cared for patients with suspected or confirmed COVID-19 and the frequency with which they used PPE (always, sometimes, never). They were also asked to report if they had enough PPE when needed, if they had to reuse PPE, or if they did not have enough because of shortages. In addition, they were asked if they worked in inpatient care, nursing home, outpatient, home health, ambulatory clinic, or other, but they were not asked to give their specific role.
All participants were asked if they felt physically well at the outset of the study and again with daily reminders. If they reported not feeling well, they were asked about their symptoms. They were also asked if they had been tested for COVID-19 and what the result had been.
Some 2.6 million people from the UK (2,627,695) and 182,408 people from the USA were enrolled in the study. The researchers excluded 670,298 people who used the app for less than 24 hours and 4,615 people who tested positive for COVID-19 at the outset. This left 2,135,190 participants, of whom 99,795 people identified themselves as frontline healthcare workers. The participants reported symptoms for an average of 19 days.
Of those included in the study, there were 5,545 reports of a positive COVID-19 test over 34 435 272 person-days .
In secondary analyses, after accounting for pre-existing medical conditions, frontline healthcare workers who reported having inadequate PPE were 1.3 times more likely to report a positive COVID-19 test than those who said they had adequate equipment to protect themselves (inadequate PPE: 157 positive COVID-19 tests in 60,916 person days; adequate PPE: 592 positive tests in 332,901 person days). The increase in risk was similar for healthcare workers who reported reusing PPE, who were almost 1.5 times more likely to report a positive COVID-19 test than those with adequate PPE (reuse of PPE: 146 positive COVID-19 tests in 80,728; adequate PPE: 592 positive tests in 332,901 person days).
Risks were highest for healthcare workers caring for patients with confirmed COVID-19 without adequate PPE, at almost six times higher than healthcare workers with adequate PPE who were not exposed to COVID-19 patients (inadequate PPE caring for COVID-19 patients: 83 positive tests in 11,675 person days; adequate PPE not exposed to COVID-19 patients: 186 positive tests in 227,654).
Even with adequate PPE, however, the risk of testing positive for SARS-CoV-2 was almost 2.4 times greater for those caring for people with suspected COVID-19 and around 5 times greater for healthcare workers caring for people with a confirmed COVID-19 diagnosis, compared with those who were not exposed to COVID-19 patients (adequate PPE caring for suspected COVID-19 patients: 126 positive tests in 54,676 person days; adequate PPE caring for confirmed COVID-19: 280 positive tests in 50,571 person days; adequate PPE, no exposure to patients with COVID-19: 186 positive tests in 227,654 person days.)
However, the data was collected at a time of global shortages of PPE, so it is not clear if the risks would be the same for people working on the front lines of the COVID-19 medical response today. The researchers also note that they did not ask about the type of PPE used and the study was carried out at a time when disinfection protocols for PPE were not yet established. They caution that their findings relating to PPE reuse should not be extended to reflect the risk of PPE reuse after disinfection protocols, which have now been implemented in various settings.
In post-hoc analyses, healthcare workers from Black, Asian and minority ethnic (BAME) backgrounds had greater risk of testing positive for COVID-19 compared with non-Hispanic white healthcare workers. After accounting for pre-existing medical conditions, healthcare workers from BAME backgrounds were almost five times more likely to report a positive COVID-19 result than somebody from the general community (98 positive COVID-19 tests in 72,556 person days for BAME healthcare workers vs 1,498 positive tests in 23,941,092 person days for the general community). In comparison, non-Hispanic white healthcare workers were around 3.5 times more likely to report a positive COVID-19 test (726 positive tests in 935,860 person days).
There were also differences in the reported adequacy of PPE according to race. Around one in three BAME healthcare workers reported reuse of PPE or use of inadequate PPE (36.7%), compared with around one in four non-Hispanic white care workers (27.7%).
The prevalence of positive COVID-19 tests was higher in US healthcare workers compared with the UK (US: 461 positive tests per 100,000 app users; UK: 227 positive cases per 100,000 app users). However, after accounting for differences in access to testing, the researchers found the risk of predicted COVID-19 based on symptoms reported through the app was double that of the general population for UK-based healthcare workers compared with 1.3 times greater for healthcare workers in the US. The researchers say this higher infection rate could be the result of differences in the availability of PPE between the two countries.
Dr Erica Warner, a co-author of the study, from Harvard Medical School and Massachusetts General Hospital, said: "Our findings highlight structural inequities in COVID risk. BAME healthcare workers were more likely to work in high-risk clinical settings, with known or suspected COVID patients, and had less access to adequate PPE. Ensuring access to, and appropriate use of, high-quality PPE across care settings would help mitigate these disparities." 
The study did not ask people to give their specific job type, their level of experience or the frequency of their exposure to patients with COVID-19. This means that those identifying as frontline healthcare workers might include roles with limited but some contact with patients, such as receptionists, hospital porters or cleaners, and this might affect the findings. In addition, they were not asked about the type of PPE used and whether or not they received appropriate training. The researchers say they are planning additional questionnaires to probe these topics in more detail.
Writing in a linked Comment article, Prof Linda McCauley from Emory University, USA, who was not involved in the study, said: "If we are ever to outpace COVID-19, there must be accountability at every level, from the community to top government officials. By combining a centralised mechanism for supply chain oversight, with universal masking and data transparency at local levels, it is possible to afford health-care workers the protection they deserve."
Long H Nguyen, David A Drew, Mark S Graham, Amit D Joshi, Chuan-Guo Guo, Wenjie Ma, Raaj S Mehta, Erica T Warner, Daniel R Sikavi, Chun-Han Lo, Sohee Kwon, Mingyang Song, Lorelei A Mucci, Meir J Stampfer, Walter C Willett, A Heather Eliassen, Jaime E Hart, Jorge E Chavarro, Janet W Rich-Edwards, ScD
Richard Davies, Joan Capdevila, Karla A Lee, Mary Ni Lochlainn, Thomas Varsavsky, Carole H Sudre, M Jorge Cardoso, Jonathan Wolf, Tim D Spector, Sebastien Ourselin, Claire J Steves, Andrew T Chan on behalf of theCOronavirus Pandemic Epidemiology Consortium.
Risk of COVID-19 among front-line health-care workers and the general community: a prospective cohort study.
The Lancet, 2020. doi: 10.1016/S2468-2667(20)30164-X.
1. Quote direct from author and cannot be found in the text of the Article.
2. Person days are calculated as the number of participants multiplied by the number of days of data recorded. It allows researchers to take into account both the number of people in the study and the amount of time each person spends in the study.