The authors of a recent article in the Canadian Medical Association Journal studied all possible complications of COVID-19. Their aim was to confirm which previously reported complications have the strongest association and which are the highest risk.
The authors write:
“This study provides estimates of absolute risk and relative odds for all identified diagnoses related to COVID-19, which are needed to help providers, patients, and policymakers understand the likelihood of complications.”
The article is the result of a collaboration between researchers from the Jacobs School of Medicine & Biological Sciences at the University at Buffalo in New York, Aetion Inc., also in New York, the Department of Medicine at the University of Toronto in Canada, and HealthVerity Inc. in Philadelphia, PA.
Stay informed with live updates on the current COVID-19 outbreak and visit our coronavirus hub for more advice on prevention and treatment.
The researchers used the anonymous health claims data of 70,288 people who had received a diagnosis of COVID-19 between March 1 and April 30, 2020.
“Health claims data” refers to information created during healthcare encounters, covering nationally representative health plans in the U.S.
The researchers divided people’s data into three categories — outpatients, inpatients, and patients admitted to the intensive care unit (ICU) — to examine any differences between healthcare settings.
Of these 70,288 people, 53.4% were admitted to the hospital, including 4.7% who were admitted to the ICU. The remaining 46.6% were outpatients. Overall, 55.8% were female, and the median age was 65 years.
The researchers defined a “hazard period” (7 days before to 30 days after the diagnosis) and a “baseline period” (120 days before to 30 days before the diagnosis).
They used the baseline period to identify and eliminate any chronic conditions that existed before the person developed COVID-19. A comparison of the two periods revealed which conditions were most likely to occur after the onset of COVID-19.
The researchers analyzed all 1,724 codes in the dataset and identified 69 as being significantly associated with COVID-19. These relative odds mean that conditions corresponding to these codes are likely to be attributable to COVID-19 or its treatment.
They also calculated the overall risk of receiving a new diagnosis with each of the codes upon getting a COVID-19 diagnosis. The risk is the proportion of people who received a diagnosis with a code during the hazard period among those who did not have the coded condition in the baseline period.
The analysis confirmed that many previously identified conditions are among the most common complications.
According to this study, the conditions most strongly associated with COVID-19 include:
A strong association does not mean that a given condition is common (high risk).
The conditions with the highest risk (the most common complications of COVID-19) included:
pneumonia, which around 27.6% of all and 81% of people in the ICU had
respiratory failure, which around 22.6% of all and 75.3% of people in the ICU had
acute kidney failure, which around 11.8% of all and 50.7% of people in the ICU had
other sepsis, which around 10.4% of all and 54.1% of people in the ICU had
The most high risk and highly associated complications included viral pneumonia, respiratory failure, sepsis, acute kidney failure, and ARDS.
One interesting finding was that the analysis did not confirm previous suggestions that COVID-19 significantly increases the risk of stroke.
The researchers carried out this study on a large data sample that allowed them to make statistically significant observations. They were also able to work with a detailed medical history of many people, allowing them to make sure that they did not count any pre-COVID-19 diagnoses as a complication of the disease.
The study authors highlight that their data and analysis have some limitations that are important to keep in mind.
For example, milder conditions are less likely to be recorded in health claims. This may explain why strongly associated and commonly reported complications, such as cough and loss of smell and taste, might appear in the analysis as low risk.
Also, the study does not capture the risk of the exacerbation of preexisting conditions, as it focused on newly diagnosed conditions.
In addition, the use of the ICD-10-CM potentially means that the team may have misclassified and excluded people who had developed COVID-19. Not all the codes are well-defined, so there is a possibility that medical professionals do not use them consistently. However, the authors believe that the large size of the dataset has balanced this.
Furthermore, the analysis included only people who had at least one medical claim, which means that estimated risks might be higher in this cohort than in the general population.
There is also a possibility that some of the complications may result from the treatment of COVID-19, not from COVID-19 itself.
The article has undergone peer review, but it is currently available as an early release, which means that it may receive minor revisions in the future.
For live updates on the latest developments regarding the novel coronavirus and COVID-19, click here.