Thread by @ajumathew_

Steroids in #COVID19 (thread)

Steroid therapy is one of the cheapest and most effective therapy for COVID pneumonia. But, confusions abound. Stay with me. It's a little complicated.

RECOVERY trial looked at short course steroid use for hospitalized patients with COVID. Patients lived longer when steroids were given for patients who needed supplemental oxygen or resp device support.

The problem: Steroids given to patients who did not require O2 or resp device support showed no benefit (and possible harm). See how red line is on the top now (?more chance for dying).

But... what is the level at which doctors started O2 therapy in this population? It's not the same as the definition we are setting in our guidelines (<92%). Randomization happened from March 19th to June 8th. O2 guidelines changed April 9th.

Point #1: Should we be strict in our definition for hypoxia (SpO2 <92%) patients who should be considered for timely steroids?

Or, should we go for a more dynamic definition. Eg; consistently dropping O2 levels; signif drop in O2 levels with walking test; worsening symptoms?

Key issue: Never start steroids early in the course of the disease in its viremia phase. In the subgroup of patients who did not get O2/device support, steroids were possibly started quite early in 50% of patients. Median and IQR: 3 (6-10).

Viremia phase in SARS-CoV2 is is short and usually concludes in first week. Whereas in SARS, viremia peaks in week 2. I asked one of World's foremost experts on SARS viruses: "Viremia lasts for just a few days. Will vary depending upon disease severity."

Did steroids hurt that subgroup of patients in the trial? I would say, yes. Could that be the reason there was a 'possibility for harm' noted in the trial? Yes.

Point # 2: Do not start steroids in the early phase of the disease. It causes more harm than you can possibly imagine.

Does short course of #steroids in a TIMELY manner harm patients? We have data from #RECOVERY trial. No serious adverse effects of concern. No mucormycosis.

#Mucormycosis happens with prolonged course of immunosuppression. Not with a short course of steroids given to the RIGHT patient in the RIGHT time. Let's not fear-monger and scare docs in peripheral centres from using it appropriately.

Point # 3: Indiscriminate use of steroids harm patients. Indiscriminate:
1, prolonged course for patients who don't need it.
2, short course of 'preventive' therapy
3, short course of therapy for patients who don't need it

Can we use lab tests (eg; CRP level) to figure out who would benefit? Possibly. But one absolute value on a lab test SHOULD NOT be used to make the decision to start steroids.

Don't treat a number. Treat the patient.

Let's not forget that RECOVERY trial was done in the West where there's no shortage of O2 or beds. In current situation in India, there's scarcity of healthcare resources. In such a situation, must we wait for patients to become hypoxic with a SpO2 of < 92% to intervene?

Experts would say that stating all of the above (that a hard endpoint of SpO2 <92%) would lead to indiscriminate use. Well, you tried that and it didn't help prevent indiscriminate use. Wd you use that definition if your loved one is in India in a place where there's no O2 beds?

What is the best public health practice in such a situation?

Issue clear statements. Provide guidelines on its use and non-use.

Evidence-Based Medicine is using the best available evidence and formulating policies around it.

TLDR: Timely and appropriate use of steroids in COVID19 saves lives. Don't use it early; don't use in mild disease. Don't wait for SpO2 to drop to <92% if healthcare resources are stretched. Instead, use a dynamic definition as stated above. Talk to a doctor. Don't self-medicate.