Trends in COVID-19-related in-hospital mortality: lessons learned from nationwide samples.

SARS-CoV-2 infectivity remains widespread across the world, with the resulting disease, COVID-19, causing devastating sequelae. With disease-modifying therapy but no cure, and a long road to developing immunity through vaccination, understanding and identifying risk factors contributing to mortality must remain a priority. In The Lancet Respiratory Medicine, two Articles—one from England, the other from Brazil—offer insights into nationwide trends for inpatient mortality due to COVID-19.

These Articles contribute considerably to the growing literature on the markedly diverse inpatient mortality due to COVID-19 across jurisdictions, by providing nationwide, high-quality, population-level health-system data. Although hospital admission rates and mortality were high in both studies during the early stage of the pandemic, mortality declined notably only in England over the study period (March 1 to May 31, 2020). Older age, male sex, and the presence of comorbidities were identified as significant risk factors for inpatient mortality in both studies. In Brazil, the findings suggest that a strained health-care system—indicated by regional differences in access to resources, compounded by overburdened hospital systems—contributed to greater in-hospital mortality during the study period (Feb 16 to Aug 15, 2020, with mortality peaking between March and May). In England, there was a reduction in mortality over time in older patients, particularly those aged 70 years and older. Contextualising these findings related to in-hospital mortality would depend on a number of factors, including socioeconomic status, ethnicity, and literacy, as well as other community and in-hospital factors (figure ), which neither study fully addressed.

Figure

Risk factors associated with transmission of SARS-CoV-2 and the development of severe COVID-19 leading to in-hospital mortality

Social determinants of health contributed to differential mortality rates in both countries. In Brazil, although older age portended high mortality—greater than 55% in those older than 70 years—mortality was 12% in patients younger than 40 years. Comparing death rates across populations is difficult, but this figure is unexpectedly high compared with previously published data for this age demographic, including the findings from England, which show 2·9% mortality in those younger than 40 years. In the Brazilian study, there were higher death rates in those who were illiterate, those located in the resource-limited North and Northeast regions of Brazil, and those who identified as Black or Brown. By contrast, in England, deprivation scores had only a modest effect on inpatient mortality, and only Asian ethnicity was associated with increased in-hospital mortality. It is possible that socioeconomic status and other factors might have a different role across different jurisdictions, dependent on the adequacy of and access to the health-care system.

The high death rate during the initial phase of the pandemic in both countries might have been due, in part, to the high prevalence of SARS-CoV-2 infection throughout the community, with initial low levels of physical distancing and mask wearing before such practices became more widely adopted, leading to more severe disease in the setting of potentially greater viral inocula or multiple-hit exposures. In addition, host genetic factors could have contributed to the high age-specific infection fatality rates.4, 5

England's temporal mortality reductions in older adults across the early pandemic were impressive, with more than a 50% decrement in patients older than 80 years. Although the magnitude of this trend might seem difficult to explain, a study from the UK showed an 11·8% reduction in adjusted 28-day inpatient mortality over time for critically ill patients, with fewer intensive care unit admissions for those aged 75 years and older. This raises the possibility that older adults might have remained in the community longer earlier in the pandemic, perhaps not reaching the hospital. For example, it appears that patients residing in care homes might have been less likely to be transported to a hospital setting, potentially accounting for a portion of the high mortality seen in care homes during the peak of the pandemic.7, 8 Additionally, improvements in in-hospital mortality over time might reflect decreased strain on hospital systems and changes in clinical practice as providers learned how best to manage the disease.

Both studies highlight the importance of ascertaining the true prevalence of COVID-19 in the community to obtain accurate rates of morbidity and mortality. Even the perception of a strained health system can lead to unintentional excess deaths, from COVID-19 and other conditions, because individuals might avoid seeking care until later stages of disease or might die at home, leading to underestimates of the true mortality burden attributable to COVID-19. Cultural preferences related to treatment intensity, social determinants of health affecting the ability to access health care, health literacy, and implementation and adherence to public health policy standards all affect community trends in disease morbidity and mortality, and need to be considered in the interpretation of in-hospital mortality data.

Future studies should focus on linking community SARS-CoV-2 seroprevalence data, community viral load (for example, using wastewater-based screening), COVID-19 surveillance data, patient viral load data, and the role of nosocomial transmission from health-care workers acquiring infection in the community with hospital admission outcomes to model the progression of the pandemic and determine the concurrent infection fatality risk.5, 9, 10 For example, in areas of high community transmission and high coinciding hospital admission and death rates, interventions to improve self-quarantining, health literacy, access to health care and full protective equipment, and frequent testing for hospital staff (to prevent nosocomial infection) could be prioritised and implemented. To be successful, this approach, in turn, would need better understanding of community transmission dynamics. As we continue to battle COVID-19, identifying high-risk patients in hospital and community settings will be crucial, as will insights from population-based studies, helping to focus our community-based and hospital-based public health initiatives.