Study: Jeong, Han Eol, et al. “Association between NSAIDs use and adverse clinical outcomes among adults hospitalised with COVID-19 in South Korea: A nationwide study.” medRxiv (2020).
Method: “We conducted a population-based cohort study using South Korea nationwide healthcare database, which contains data of all subjects who received a test for COVID-19 (n=69,793) as of April 8, 2020. We identified a cohort of adults hospitalised with COVID-19, where cohort entry was the date of hospitalisation. NSAIDs users were those prescribed NSAIDs while hospitalised and non-users were those not prescribed NSAIDs.” 
South Korea has a universal health care system, so that researchers with government permission can access a virtual treasure trove of medical information from a large number of patients. This study began with data from almost 70,000 patients who tested positive for Covid-19 by April 8th, not too far into the pandemic in South Korea. They narrowed the data set to “1,824 adults hospitalized with COVID-19 (mean age 44.7 years; female 59%)”, of which 285 were NSAIDs users and 1,539 were non-users . The non-users were the control, non-randomized.
Yes, in some sense, an RCT (randomized controlled trial, often placebo controlled) is the “gold standard”. But an RCT is not always the best study to answer any question. Moreover, if we threw away all data from other types of studies, we’d lose a vast amount of useful information. This study has the advantages of a high enough number of patients, data set from one source, with a large control group, and it is a fairly straight-forward study design.
In addition, the study “compared NSAIDs to paracetamol (acetaminophen) to minimize confounding by indication.” What if the difference between the two groups was not the use of NSAIDS, but the indication that led to that use, such as pain or fever? To reduce confounding (a misunderstanding of the relation between the intervention and the results), the study authors also compared NSAID use to use of paracetamol (Tylenol), which is also given for pain and/or fever. So this is a well-thought out study.
Outcome: “Our primary outcome was a composite of death, intensive care unit admission, mechanical ventilation use, and sepsis; secondary outcome was cardiovascular or renal complications.” 
The study was mainly interested in finding out whether NSAID use affected the odds of death and/or admission to the ICU, and/or the use of mechanical ventilation, and/or a progression to sepsis (which is a life-threatening condition).
Findings: “Compared with non-users, NSAIDs users were associated with increased risks of the primary composite outcome (aOR 1.54, 95% CI 1.11−2.15) and cardiovascular or renal complications (aOR 2.64, 95% CI 1.67−4.16). The association with primary outcome remained consistent when comparing NSAIDs to paracetamol (aOR 1.31, 95% CI 0.89−1.95).” 
The adjusted odds ratio (aOR) was 1.54. The control is assigned the value of 1.00, so for those without NSAID use, their risk of the “primary composite outcome” (death, ICU admission, ventilation, sepsis). Then the comparative risk of that outcome is compared to those patients given NSAIDs. An odds ratio (OR, also called Risk Ratio, or Hazards Ratio) of less than 1.00 means that the intervention (in this case NSAIDs) reduced risk, and a value of greater than 1.00 means the intervention increased risk. A value of 1.54 means that there was a 54% increase in risk of the primary outcome (all those bad things) when patients are given NSAIDs.
The comparison with Tylenol showed that this difference in outcome was not due to the indication, to the fact that the patients had fever or pain. Tylenol had an odds ratio of 1.31; however, this was not statistically significant, as indicated by the 95% Confidence Interval of below and above 1.00.
“The 95% confidence interval is a range of values that you can be 95% certain contains the true mean of the population.”  The mean value is the average. On average, in this study, patients had a 54% increased risk with NSAID use. The 95% CI for NSAID use was 1.11 to 2.15, meaning that we can be quite certain that the true value is within the range of 11% increased risk to 115% increased risk. (A value of 2.15 is a 115% increase as the reference value is 1.00; so subtract the 1.00 from the 2.15 to get 1.15 or 115%.) If we did a series of studies like this one, we would get different values, not exactly 1.54 each time. But according to this statistical analysis, the value from repeated studies of this type should almost always be in that range.
When you have a 95% CI above and below 1.00, especially well above and well below, it usually means the finding is not statistically significant. In other words, the result cannot be distinguished from statistical noise, from the degree of random various found in any data set. So the Tylenol finding of (aOR 1.31, 95% CI 0.89 – 1.95) was not statistically significant, and therefore Tylenol did not have any effect (that this study could find) on the risk of the primary outcome. So we can then conclude that the reason NSAIDs increased risk of a worsening of the patient’s condition (in the stated outcomes) is not due to the underlying cause of pain or fever, but probably due to the medication itself.
Take Away: Given this study’s findings, it would be prudent to refrain from giving Covid-19 patients any type of NSAID. As far as we now know, NSAIDs may increase risk of a severe outcome in Covid-19 patients.
Persons who are self-quarantined because they might have Covid-19, or who have a mild case and were sent home should also not take NSAIDS. Do not take aspirin (Bayer, Excedrin), naproxen (Aleve), ibuprofen (Advil), or any prescription NSAID that you might have from a previous illness. Tylenol can be used with your doctor’s permission. Ask your doctor before taking any pain reliever, if you may have Covid-19.
Ronald L Conte Jr
Note: the author of this article is not a doctor, nurse, or healthcare provider.
Adjusted Odds Ratio – The risk or chance of a particular outcome for a particular group in the study. The intervention being tested might increase or decrease those odds. The adjustment takes into account various factors, such as age, gender, whether a person smokes, preexisting medical conditions, etc., so that these factors are statistically nullified in calculating the odds.
NSAID – Non-Steriodal Anti-Inflammatory Drug, a type of medication that reduces inflammation, not including steroids. These include: aspirin (Bayer, Excedrin), naproxen (Aleve), ibuprofen (Advil), and more than a few prescription medications.
The most common fever and pain reducer which is not an NSAID is paracetamol or acetaminophen (Tylenol). Before and after surgery, patients are often told to discontinue used of NSAIDs, leaving them with only one non-prescription option for pain relief: Tylenol. As Tylenol is not an NSAID, it does not increase risk of bleeding as all NSAIDs do, so it can be used before and after surgery (if approved by your physician or surgeon).
Sepsis – “a potentially life-threatening condition caused by the body’s response to an infection. The body normally releases chemicals into the bloodstream to fight an infection. Sepsis occurs when the body’s response to these chemicals is out of balance, triggering changes that can damage multiple organ systems. If sepsis progresses to septic shock, blood pressure drops dramatically. This may lead to death.” 
1. Jeong, Han Eol, et al. “Association between NSAIDs use and adverse clinical outcomes among adults hospitalised with COVID-19 in South Korea: A nationwide study.” medRxiv (2020).
2. Mayo Clinic, Patient Care & Health Information, Diseases & Conditions, Sepsis.
3. Simply Psychology, “What are Confidence Intervals in Statistics?,” By Saul McLeod, published June 10, 2019.