Andrés Reyes Valdivia1, Arindam Chaudhuri2. 1. Department of Vascular and Endovascular Surgery, Ramón y Cajaĺs University Hospital, Madrid, Spain. 2. Bedfordshire - Milton Keynes Vascular Centre, Bedford, United Kingdom.
As the severe acute respiratory syndrome coronvirus-2 continues to take more lives, quite a few of these will be from the vascular population,1, 2, 3 typically older and with multiple comorbidities, recognised risk factors for coronavirus disease (COVID)-19–related mortality. Some such patients will be involved in ongoing vascular trials and registries.Presenting an “overall mortality rate” is obligatory at registry/trial reporting, particularly specifying the cause of death (COD). There may be some “unknown COD,” but these should be a minority, particularly in robust prospective studies. Some unknown COD may be attributable to COVID-19, with many dying without COD confirmation. This is related to lack of testing, or high false negative rates, with controversy surrounding the accuracy of oropharyngeal vs nasopharyngeal swabs particularly for late testing.The issues arising from COVID-19-related deaths are two-fold: first, to do with accurate capture of COD (reporting issues), and second, the influence of increased deaths on the completeness of data in ongoing studies (outcome issues). This is a research concern,
,
with calls to extend trial durations, and may necessitate post hoc/retrospective power calculations to reassess statistical validity of studies.We therefore hypothesize four possible scenarios related to mortality reporting: (1) patients with accurate categorization of COVID-19-related COD; (2) patients with prior confirmed COVID-19 infection who recover but die later with unknown COD; (3) patients dying from an unknown cause during the pandemic, where COD is uncertain; and finally (4) accurate capture of non-COVID-19-related COD. Options 1 and 4 are qualitatively most desirable in terms of data capture.Global clinical uncertainty has implications for registries that report on mortality. This concern pertains to both ongoing and to future study design, as we cannot predict patterns of disease chronicity or repetitiveness. This may lead to similar outcomes as indicated: (1) accurate COD capture, (2) inaccurate data capture leading to higher censoring at survival analysis, and (3) loss of patients in smaller studies due to high unexpected mortality, rendering them underpowered and redundant.Some editorial consensus is needed to guide adequate mortality reporting, to avoid misguided assessments that may lead to misleading conclusions.