Is Dexamethasone a COVID-19 'Breakthrough'?

The BBC reports, “low-dose steroid treatment dexamethasone is a major breakthrough in the fight against the deadly virus, UK experts say.” Before tossing aside your mask and booking a cruise, let’s take a deeper dive into the data.

The study they are discussing is Oxford’s Randomised Evaluation of COVID-19 Therapy (RECOVERY) trial, which is looking a hospitalized patients with confirmed cases of COVID-19. In general, patients are receiving the “usual” care provided during hospitalization and are randomized to have added other treatments, like hydroxychloroquine, where Oxford is reporting no benefit as of June 5^th. They are now reporting the results of giving patients dexamethasone.

Dexamethasone is a well-known, cheap, globally available, steroid that reduces or down-regulates the inflammatory response. Like other steroids, it has a checkered history for improving outcomes from sepsis with some studies showing benefits, others not so much. Because the initial clinical picture of COVID-19, especially the respiratory failure, appeared to be sepsis, steroids were thought to be possibly beneficial, and many protocols began to include a steroid along with the kitchen sink in treatment. 

Some physicians raised concerns about the timing of administration of an agent that reduces the inflammatory response. Given too soon, it might blunt the body’s defenses and make the situation worse; given too late, it had little effect on an out-of-control immune response like cytokine storm. There was some debate back and forth about where the Goldilocks moment to administer a steroid might be; when a patient felt out of breath, when the oxygen levels dropped, when supplemental oxygen was failing to improve the situation, or when the patient was being considered for ventilator life-support.

Like many of the COVID-19 studies, publication, let alone peer review is in the future, the dexamethasone results are provided only in a press release from Oxford. 

“A total of 2104 patients were randomised to receive dexamethasone 6 mg once per day (either by mouth or by intravenous injection) for ten days and were compared with 4321 patients randomised to usual care alone. … Dexamethasone reduced deaths by one-third in ventilated patients (rate ratio 0.65 [95% confidence interval 0.48 to 0.88]; p=0.0003) and by one fifth in other patients receiving oxygen only (0.80 [0.67 to 0.96]; p=0.0021). There was no benefit among those patients who did not require respiratory support (1.22 [0.86 to 1.75]; p=0.14).” [Emphasis added]

So yes, dexamethasone appears to save lives, but without more details, it is hard to measure its clinical impact. Remember remdesivir, the first drug shown to be effective in treating COVID-19 did not affect outcomes, it lowered hospital length of stay, which is of great value in not overwhelming the health systems carrying capacity, but is more like Tamiflu than a “cure.” What is most interesting in the press release is how they chose to describe dexamethasone’s benefit, a third – 33%, and a fifth – 20%. Pretty impressive numbers, but they illustrate the relative improvement. 

Those same results can be expressed in other ways. For example, those results, when described in absolute terms, show a 12% improvement for patients on ventilators and a 5% absolute improvement for those patients receiving supplemental oxygen. Another way to describe the efficacy is in numbers needed to treat, NNT - this is the number of patients who have to be treated to save one life. For patients on a ventilator, eight patients must be treated to save one life; for those receiving supplemental oxygen, you need to treat 25 patients. To be fair, the press release provided those numbers, but the media seems to be ignoring them. A moment of reflection will tell you that when it comes to COVID-19 and getting your attention, 33% sounds a lot bigger, than 12%, and who the hell understands reporting an NNT of 8. [1]

From a clinical perspective, the good news is that dexamethasone might be a useful tool in the therapeutic toolbox, mainly because it’s use is well understood, it is widely available, and cheap. Cheap is good when the NNT is high; you can understand that when you see, we spend more than $20 billion annually on statins. It has no benefit for those not requiring supplemental oxygen, roughly 80% of those hospitalized, so the patient population is small. It will be life-saving for a precious few, and there is value in that, but it is just one step in a long therapeutic journey, not the great leap forward that the media headlines would have us believe. 

[1] To provide a bit of context for NNT, statins perhaps the most widely prescribed medication in the US has an NNT of 104 and the Mediterranean diet, an NNT of 62 for cardiovascular disease.