Study population and setting
This retrospective cohort used province-wide electronic health records from Ontario, Canada, to assess the relationship between ABO and Rh blood type and SARS-CoV-2 positivity and severity. A total of 225,556 among the 2.66 million Ontario residents with ABO blood group documented prior to December 2019 (8.5%) met the inclusion criteria of having had a SARS-CoV-2 PCR test between January 30 and June 30, 2020, documented sex and date of birth, and Ontario residence. Data on demographics, comorbidities (history of cardiac ischemia or arrhythmia, cancer or chronic kidney disease within 5 years of SARS-CoV-2 test, and history of congestive heart failure or diabetes), clinical conditions during in-hospital stays, vital status, residential income quintile, rural residency were ascertained through the linkage of the following electronic databases: Canadian Institute for Health Information’s Discharge Abstract Database, National Ambulatory Care Reporting System database, Ontario Health Insurance Plan claims database, and the Ministry of Health’s Registered Persons Database. The primary outcome of interest was a positive SARS-CoV-2 PCR test (using the first positive test for individuals with multiple tests). The secondary outcome was a composite endpoint of severe COVID-19 infection (intensive care unit admission, hospitalization for 7 days or more, myocardial infarction or viral pneumonia with 14 days of positive PCR test) or death. The association between the blood groups and SARS-CoV-2 test positivity and severe COVID-19 infection was assessed using multivariable modified logistic regression.
Summary of Main Findings
The mean age of the sample was 53.8 years, 71% were female, and the distribution of comorbidities was similar across blood groups: 11% chronic kidney disease, 27-29% cancer, 21% diabetic, 38-42% hypertensive, 10-12% heart failure, 13-17% COPD, and 34-38% dementia or frailty. In multivariable analysis, the adjusted relative risk of SARS-CoV-2 positivity was higher among B and AB blood groups (aRR 1.21, 95% CI: 1.13-1.29 and 1.15, 95% CI: 1.03-1.28 respectively), and 5% lower among O blood group (aRR 0.95, 95% CI: 0.91-1.01), relative to the blood group A. The blood groups demonstrated similar associations with COVID-19 severity/death. The protective association with the O blood group was more pronounced when comparing it to the other blood groups combined, a 12% lower risk of SARS-CoV-2 positivity (aRR 0.88, 95% CI: 0.84-0.92) and 13% lower risk of severity or death (aRR 0.87, 95% CI: 0.78-0.97). In a separate multivariable model not including ABO group, Rh- status was associated with 21% lower relative risk of positivity (aRR 0.79, 95% CI: 0.73-0.85) and 18% lower risk of severity or death (aRR 0.82, 0.68-0.96). The inverse association between the O blood group and Rh- group and SARS-CoV-2 positivity was more pronounced among those < 70 years-old in a multivariable regression that did not further adjust for age. However, neither blood group nor Rh status were associated with severity/death among the 7,071 individuals infected with SARS-CoV-2.
The study leveraged province-wide electronic health data to capture a large sample of individuals who had a SARS-CoV-2 PCR test over a five month period with documented ABO blood type in Ontario, Canada. The authors excluded individuals whose ABO blood group was documented after the start of the COVID-19 pandemic to avoid selection bias from blood testing performed as a result of COVID-19-related illness. The authors performed sensitivity analyses to assess the robustness of findings to potential misclassification of SARS-CoV-2 infections due to false negative PCR tests.
Although the authors refer to the primary outcome, positive PCR test result, as the risk of infection, because the denominator consists only of individuals who sought testing during the period, this outcome represents positivity in this sample. Those who sought tests were likely driven by a mix of indications including symptoms, contact with cases, or requirements for employment or travel and are not representative of the population-based sample at risk of infection. Limiting the study sample to the 24% of those tested whose ABO blood type was documented before December 2019 further selected a sample that was not representative of those tested or the underlying population at risk of infection. Specifically, older adults, women, those with comorbidities, and those with severe COVID-19 outcomes are over-represented among those with documented ABO. If getting tested or the indications for having ABO blood type documented were related to blood type (e.g. differential by ethnicity which is linked to blood group) in addition to the outcomes, this might have biased the observed associations. Data on ethnicity was not available, though marginalized ethnic groups have been disproportionately impacted by COVID-19 and the distribution of blood group types also differs by ethnicity; therefore it was not possible to account for these differences in this analysis. Finally, the authors noted that their definition of severe disease might have missed severe cases.
This was a large cohort study evaluating the association between blood group types and SARS-CoV-2 positivity or severe COVID-19 infection.
This review was posted on: 12 March 2021