Thread by @jpogue1

Guys. We need to talk about this Hydroxychloroquine + Azithromycin thing. It is out of hand. It all stems from this study that came out today. The study design: Comparative viral eradication on day 6 between HCQ, HCQ + Azithro, and control (not treated) COVID-19 patients.

The gist of the excitement is this figure describing viral "eradication" rates between the groups at day 6:
HCQ + Azithro group 6/6 (100%) by day 6
HCQ mono therapy 8/14 (57%)
Control 2/16 (13%)
Sounds amazing right? I mean this kinda stuff gets Dr. Oz all worked up.

But as always. The devil is in the details. Per the methods the study started with 26 pts in the HCQ group and 16 control. Then why only 20 HCQ pts in the analysis? Because in order to be in the cohort you had to be evaluable (swab) at day 6. That meant 6 HCQ pts were ineligible

What happened to those six HCQ patients?
- 3 were transferred to the ICU while still PCR positive
- 1 died (PCR negative)
- 1 left the hospital (PCR negative)
- 1 withdrew due to nausea (PCR positive)
I dunno- seems to me like 5 failures, but anyway lets move past this for now.

The real good stuff is in this table. Just to orient you. Top 16 patients control, next 14 HCQ monotherapy, last 6 HCQ+AZ. I want you to pay attention to the last 7 columns. These are Ct values at baseline (day zero) and through the six days of the study

What's a Ct value? Yeah great question. I've known for exactly 1 day.The cycle threshold (Ct) is the number of cycles to be run for the PCR test to turn positive. Bottom line the LOWER the number the more virus that is present; that is, the less number of cycles to hit threshold

Now look closely at the Ct values for HCQ monotherapy and HCQ+AZ. What do you see? For combo therapy all patients have baseline Ct values of 24 or higher, whereas 5 patients in the monotherapy arm have values < 23 (in particular 15, 17, 19, 22, 22).

Why do I bring this up. Because it means there was a chunk of pts in the monotherapy arm who would need a greater antiviral effect to reach "negative" or undetectable virus. In this analysis negative was defined as Ct>35. It's also notable that negative is often defined at Ct>40.

So did this impact the outcomes between monotherapy and combo therapy? You tell me. Eradication rates at day 6:
HCQ monotherapy (Ct <23): 1/5 (20%)
HCQ monotherapy (Ct 23+): 7/9 (78%)
HCQ + Azithro (all Ct 24+): 6/6 (100%)

I also failed to mention that a recent publication suggested BAL > sputum > nasopharyngeal swab (used in this study) for sensitivity in detecting SARS CoV-2. jamanetwork.com/journals/jama/…

So to recap:
1) This study had a lower threshold for "negative" than most and used as less sensitive swab sample
2) There were a decent number (23%) of total HCQ patients who were not eligible for analysis, but at least five could be considered failures.

2a) These failures could lessen the spread between HCQ and control
2b) These failures would impact monotherapy vs combination therapy (unclear how as they are not described)
3) When correcting for burden of disease/viral load, HCQ and HCQ + AZ look extremely similar

So what does it all mean? I don't know. Limitations aside I still think these data are encouraging as even w/ the failures added HCQ would look good. That said they most certainly do NOT tell us that HCQ works nor do they suggest that HCQ + Azithromycin is some special cure all.

So let's all slow down before we drastically add an antibiotic (and a potentially additively toxic one) to HCQ based on six isolates in this study. We are all trying to do the best we can for our patients and I get this, but we need to stop, read, and think through this stuff.