Their methodology is based on an assumption that healthcare workers are as likely to die of covid as the general population, that is, their job does not expose them to elevated risk. You can see the start of their methodology as:
"As a start, the number of deaths among HCWs was simply estimated by applying the crude mortality rate from each country (namely, the number of deaths reported to the WHO COVID-19 Dashboard divided by the population size) to the estimated number of HCWs in each country derived from ILOSTAT . This simple estimation considers HCWs to have a similar exposure to SARS-CoV-2 infection and risk of death to that of the general population..."
From there they do adjust for some things, but they are the ways that healthcare workers are different, demographically, from the general population. They are not looking at occupational risk.
I could still totally believe that healthcare workers are at elevated occupational risk, but a paper that assumes their risk is what you would expect from their demographics isn't going to help us answer that question.