Doctors, nurses and other health care professionals battling COVID-19 are suffering from high rates of depression, anxiety and PTSD, with each of these conditions affecting roughly one in five workers, according to a comprehensive new review and meta-analysis of studies published Wednesday.

"The WHO estimates that the global population has a prevalence of depression and anxiety that floats around 4%," said Nathaniel Scherer, a researcher at the London School of Hygiene & Tropical Medicine's International Centre for Evidence in Disability and one of the first authors of the study in PLOS One. "We're finding through this that health care workers are floating around 20%. That's quite a big jump. That's something that needs to be looked at and addressed."

As a meta-analysis, the article combined the best data from earlier research to draw broader conclusions about mental illness among medical professionals throughout the world. The authors narrowed down 7,570 relevant articles to 65 studies deemed worthy of inclusion, improving on previous reviews by incorporating studies written in Chinese as well. The studies in total covered 97,333 health care workers from 21 countries across the planet. After combining and statistically analyzing this data, the team estimated the prevalence among health care workers of moderate depression at 21.7%, mild depression at 36.1%, moderate anxiety at 22.1%, mild anxiety at 38.3% and moderate PTSD at 21.5% — all well above the average occurrence in the general population.

"When we think about the hierarchy of evidence, systematic reviews and meta-analyses sit at the top," said Scherer. "Individual, single primary studies are useful. They are incredibly great additions to the evidence base, providing information for a given context and a given setting — but that's what they're doing. They're providing evidence for a single situation and a single setting."

Through a systematic review and meta-analysis, "You're seeing that this is consistent across this period," added Scherer. "It's giving you a fuller picture than a single study can ever do."

Scherer and his coauthors followed the established PRISMA guidelines for systematic reviews and meta-analyses. They also publicly registered the search criteria they used to find data. "The idea of a systematic review is that others could go and pick this up and get similar results," Scherer added.

Unlike most systematic reviews, the article included data from pre-prints and other papers that had not been peer-reviewed. "That peer review process really does strengthen the papers that are going through it," said Scherer. "We chose to include non-peer-reviewed work simply because COVID is such an ongoing emerging crisis. There's time sensitivity to this."

The authors took pains to ensure that the non-peer-reviewed papers did not distort their findings. "We compared the estimates coming from peer-reviewed journals and those that weren't peer-reviewed, just to see if there was any significant difference in those estimates," Scherer explained. "We didn't find any significant difference, so it means that the non-peer-reviewed articles aren't skewing this overall pooled estimate to either be an over- or an underestimate."

The team only included studies that used standardized screening tools to assess symptoms. This made it easier to compare data across cultures that may conceptualize mental health in different ways. "These tools don't ask about mental health, per se," explained Scherer. "They may ask, in the last month, to what extent did they feel their sleep was restless? They're trying to tease out these symptoms."

The authors also broke down their estimates by region. The highest rates of depression and anxiety came from the Middle East, while East Asia and North America reported lower rates of those conditions.

A large percentage of the studies came from East Asia, particularly China, and the only studies that used random sampling — the ideal for sample selection — came from the country. All of this affected the precision of Scherer's estimates. "The results will be more precise for regions that have more studies," said Scherer. "With non-random sampling, there may be some biases."

The paper offers much-needed insights into the mental health of workers who are under intense stress even during normal times. Physicians, for example, have the highest suicide rate of any profession. Yet the needs of professionals in this sector are still overlooked. Scherer noted that he could not find any good global estimates of mental illness among medical workers from before the pandemic. "I'd love it if someone knew of a paper and could direct me," added Scherer.

In addition, all the studies were cross-sectional, meaning that they only looked at one point in time. "We're presenting this as the prevalence during the pandemic," Scherer explained. "It may well be that a lot of these stressors may have been present prior to the pandemic. I think that would be a really valuable thing to find out."

Scherer believes that researchers must look into the specific causes of mental illness in these groups. "Is it to do with COVID? Is it to do with longer working hours? Is it to do with time spent away from loved ones? Is it to do with broader issues about systems of employment for health care workers? I don't know — and that's something that needs to be explored through more research," said Scherer.

He thinks we might get some answers through qualitative studies, including surveys administered by trained independent personnel. "It has to involve health care workers as active partners in that decision-making process," Scherer added. "It's going to impact their lives."

Scherer also sees a need for additional surveys of mental health in other countries, as well as for input from mental health professionals. Online therapy could be another powerful resource for socially distanced workers. But the immediate pressure from the COVID-19 pandemic can pose a challenge to thinking long term.

"It is a balancing act between needing a rapid response and it being based on good evidence," said Scherer. "To have that rapid response, you have to build off of what's been done in other settings."

"We need to speak to health care professionals," Scherer emphasized. "We need to get their voices heard."

The paper, "Prevalence of depression, anxiety and posttraumatic stress disorder in health care workers during the COVID-19 pandemic: A systematic review and meta-analysis," published March 10 by PLoS One, was authored by Yufei Li, Nathaniel Scherer, Lambert Felix, and Hannah Kuper, London School of Hygiene & Tropical Medicine, University of London.