Tracking Gender Data on COVID-19 – Part 4: The availability of sex-disaggregated information on healthcare worker cases and deaths

Lorenz Noe, Eric Swanson August 27, 2020

This blog series is part of an ongoing study by Data2X and Open Data Watch to track the gender impacts of COVID-19. We recently published a report on the availability of sex-disaggregated data for tracking the primary and secondary impacts of the pandemic. Learn more in our summary blog post and full technical report, Tracking the Gender Impact of COVID-19: An Indicator Framework. This blog post originally appeared on Open Data Watch’s website. Read the first, second, and third blog.

In this fourth post of a blog series, we examine what emerging measures of the direct impact of the pandemic on healthcare workers can tell us about the frontline capacity of countries and how it differs for men and women. In addition, we summarize the existing data on sex-disaggregated COVID-19 cases and deaths from Global Health 50/50 and ask how complete our picture is when compared to all reported cases and deaths, including by income group and region. Here are our top five takeaways with a more detailed explanation of each below:

  • According to preliminary global studies, women account for over 70 percent of infections among healthcare workers, which is in line with their share of the total healthcare workforce, yet they represent less than a third of all deaths among healthcare workers.
  • Nearly nine in ten people now live in a country with at least some sex-disaggregated data on COVID-19. For countries with data on cases and deaths, women make up 49 percent of cases but only 42 percent of deaths.
  • As of August 7, 2020, 16 low-income countries published sex-disaggregated data. Nearly two in three cases in low-income countries are now disaggregated by sex, compared to a fourth in June. In lower-middle-income countries, sex-disaggregated data are available for nine in ten cases, compared to around two in five in July.
  • Low reporting of sex-disaggregated cases in upper-middle-income countries is still driven by the absence of sex-disaggregated cases in Brazil, with one of the largest caseloads in the world. If Brazil’s 2.9 million cases were sex-disaggregated, the share of cases reported by sex for upper-middle-income countries would nearly double from 44 percent to 84 percent.
  • The capacity to provide sex-disaggregated information on cases and deaths varies widely across regions. Countries in the Middle East and North Africa, for example, provide information on only two-thirds of cases and just over half of deaths, highlighting the need for stronger case reporting and death registration systems. Europe and Central Asia, meanwhile, provide sex-disaggregated information on just over 60 percent of cases, but over 90 percent of deaths.

The availability of sex-disaggregated data on healthcare workers

The current series of blogs has thus far focused on global trends for men and women no matter their station in life or occupation. Yet it is intuitive that the jobs that men and women do will affect the risks that they run of infection and death due to COVID-19. We, therefore, wanted to take a closer look at the data around the effects of the pandemic on healthcare workers in particular.

Global Health 50/50 also tracks globally comparable data on this phenomenon, but regularly updated, comparable information is very hard to find. As it stands, only five countries are listed with sex-disaggregated incidence of COVID-19 among healthcare workers: the Dominican Republic, Germany, Italy, Spain, and the United States. The infections of female healthcare workers relative to men range from 65 percent in the Dominican Republic to 77 percent in Spain.

Comparing these numbers is difficult, as they do not take into consideration the differences by sex in the healthcare workforce overall, nor do they necessarily reflect the same understanding of who is employed in the healthcare profession. In addition, information on deaths by occupation are extremely hard to tally even for overall cases, and the counts often suffer from differing terminologies and lack of funding for more accurate reporting, as a recent investigation reveals in the United States.

Even in a recent global study across 130 countries on COVID-19 infections and mortality of healthcare workers, information on sex-disaggregated cases and deaths is available for only around 27 countries out of the whole sample. Nevertheless, this sub-sample allows us to approximate the disproportionate impact of the pandemic on healthcare workers by sex: In this review, women accounted for 71.6 percent of cases, but only 29.2 percent of deaths, compared to our picture of the global population as a whole, where women account for 49 percent of cases and 42 percent of deaths.

This disparity between cases and deaths is particularly striking given the fact that women account for 70 percent of employment in the health workforce, according to a WHO study across 104 countries. This phenomenon undoubtedly reflects the different jobs that men and women occupy within the healthcare workforce, mixed with physiological differences between men and women which make men much more susceptible to dying of COVID-19, for reasons that are still being explored.

In order to allow countries to adequately protect their first line of defense against the COVID-19 pandemic and ensure gender equality, we need better data on cases and deaths among healthcare workers. Yet even information on sex-disaggregated occupations is sparse itself. To address this dearth of information, we need better regular data collection instruments, such as labor force surveys and rapid assessment surveys that can help list current cases and deaths among the workers that are closest to the pandemic.

Table 1 summarizes the sex-disaggregated data available as of August 7, 2020. There were 72 countries reporting sex-disaggregated data for COVID-19 cases and deaths. (These include four nations of the United Kingdom – England, Wales, Scotland, and Northern Ireland – that are reported separately.) They represent over 70 percent of the world’s population (2019 data).

Seven countries report sex-disaggregated data for deaths only. Forty more countries report sex-disaggregated data on cases but lack data on deaths. In total, nearly nine in ten people now live in a country with at least some sex-disaggregated data on COVID-19. For countries reporting sex-disaggregated data on both cases and deaths, women make up 49 percent of cases but only 42 percent of deaths.

Although stories about reported cases and deaths caused by the COVID-19 virus have mainly covered high-income countries, information about the pandemic in low- and lower-middle-income countries is slowly emerging and will be crucial to monitoring the impacts of the pandemic on people now and the 2030 Agenda. As of August 7, 2020, 16 low-income countries published sex-disaggregated data (eleven of them for cases only and five, Afghanistan, Burkina Faso, Haiti, Liberia, and Uganda, for both cases and deaths). Sex-disaggregated data are available for 24 lower-middle-income countries.[1] See Table 3 below for a more in-depth discussion on the reporting of deaths and cases by income group.

Table 2 summarizes the number of cases and deaths compiled by the European CDC and reported by Our World in Data (OWID) for the countries with sex-disaggregated data and for the remaining 94 countries that currently lack any sex-disaggregated data.

The countries included in the GH5050 tracker account for 74 percent of global cases and 93 percent of deaths. Reporting on cases is slowly catching up to reporting on deaths, with nearly three in four cases containing information on sex, in stark contrast to four in ten cases in early June. Brazil by itself could do much to close the remaining gap between reporting on deaths and cases; if the country provided disaggregated information on its cases, the global share would increase to nearly 90 percent.

To arrive at a more nuanced understanding of the way in which countries report on COVID-19 cases and deaths, we analyzed the global shares according to World Bank FY2021 income groups in Table 3. Compared to previous posts, the availability of sex-disaggregated case reporting in low-income countries has increased. Over half of low-income countries report on cases in GH5050 and they provide sex-disaggregation on nearly 2 in 3 of their cases, compared to a fourth in June.[2] 

Just under half of lower-middle-income countries report sex-disaggregated cases, but this accounts for 9 in 10 cases in lower-middle-income countries. Upper-middle-income countries reverse these positive numbers: While about the same share of upper-middle-income countries publish sex-disaggregated data on cases as low-income countries, this accounts for just over forty percent of their cases. This is still driven by the absence of sex-disaggregation of cases by Brazil, with one of the largest caseloads in the world. If Brazil’s 2.9 million cases were sex-disaggregated, the share of cases reported by sex for upper-middle-income countries would increase by forty percentage points from 44 to 84 percent. The Center for Disease Control recently published long-overdue information on sex-disaggregated cases in the United States, which means the vast majority of cases in high-income-countries are sex-disaggregated. However, the future of U.S. case and death reporting is unclear, due to ongoing restructuring of reporting procedures that raises the worrying specter of political interference.

Reporting on deaths by sex, meanwhile, reveals the most straightforward correlation between statistical capacity and income. High-income countries report practically all their deaths by sex and this share declines with each step down the income group ladder.

Examining the availability of sex-disaggregated information by region gives further indication of where the statistical capacity to report needs to be bolstered to tackle the pandemic most effectively. As Table 4 below shows, the capacity to provide sex-disaggregated information varies widely across regions. Virtually all cases in East Asia & Pacific are sex-disaggregated, while just under two-thirds are disaggregated in Europe and Central Asia.

Sub-Saharan Africa has a higher sex-disaggregated share of cases than the global average, but disaggregates just two-thirds of all deaths, perhaps pointing to issues of death registration systems in these countries. South Asia joins North America in disaggregating all of its cases and deaths. The Middle East and North Africa have middling disaggregation of cases and, like Sub-Saharan Africa, the region needs more robust death registration systems. Latin America, where the pandemic rages most acutely as of the time of this writing, disaggregates less than half of its cases, again owing to the absence of sex-disaggregated case reporting from Brazil. However, its death reporting is close to complete.

As the COVID-19 pandemic continues to affect every aspect of our lives, we will continue to monitor the availability of sex-disaggregated data and look forward to sharing insights via blogs and briefs in the near future. Stay tuned!

Stay in touch. Sign up for gender data updates.