On 2020-04-20 15:36:38, user Philip Davies wrote:
Interesting study, thank you.
This is another study that attempts to ascertain if oral HCQ tablets can be of clinical use in patients more than one week into symptomatic disease, hospitalized with bilateral pneumonia and with evidence of established inflammatory reaction (cytokine storm). That's a big ask for any oral medication.
The study is again small (both arms have less than 100 patients). The most significant outcome measured (death) is realized in very small numbers (3 and 4). The confidence levels are extremely wide.
The are several problems with this study. There are marked differences in the two populations. The study honestly attempts to accommodate these confounding factors using a propensity score method (IPTW). Normally this method is valuable but here I can’t see that it has been well applied.
It pays to look at the raw data. There is a significant difference (between the two arms) in the initial intensity of disease.
At baseline (admission), HCQ arm comprises 78.3% men (>20% more of these higher risk patients than control arm with 64.9%); HCQ arm has 21.9% patients with more severe disease in the form of CT showing >50% lung affected). This is >80% more than in control arm (12.1%). HCQ arm has 90.5% patients with CRP > 40mg/l (CRP is a good indicator of impending/current severity). This is 10% higher than control arm (81.9%). HCQ arm had median O2 flow on admission = 3 litres/minute (50% higher than control arm at 2 litres / minute).
So, at baseline, the HCQ arm had significantly more patients with severe disease than control arm. The O2 flow is actually more significant than first sight would suggest. 2 l/m is always the first step in O2 therapy. The data shows us that most patients in the control arm could hold their sats on this first step therapy. This also means they may have been OK on just 1 l/m. We don't know. But we do know that most patients in the HQN could not hold their sats at that first step and needed an increase (3 l/m ... so that's 50-300% more O2 than control arm).
Admittedly there were other confounding factors which compromised the control arm more than HCQ arm (some chronic disease elements). But it's clear to me that disease severity was markedly more established in the HCQ arm.
Another factor to note: the HCQ treatment was not initiated at the moment those baseline values were obtained (on admission). The HCQ was initiated within 48 hours. So let’s look again at the timelines. The median duration of symptoms at admission shows that the HCQ arm comprised patients who were further into worsening illness: they were admitted on D8 compared to control, D7. They may not have had HCQ initiated until D10.
Then we look at outcomes: the raw data shows that the disadvantaged HCQ arm actually does better in the two most important outcomes, death and ICU admission. The HCQ delivers 12% less death and ICU admissions than the control arm. Admittedly the numbers are small so the confidence levels are very wide.
So what does that tell us? The answer is not much. But even accepting the poorly aligned baseline for disease severity, the outcomes with their wide 95% confidence levels do deliver a mildly promising indication on the 'swingometer'. They point more towards benefit than harm when using HCQ in this advanced disease state.
As a final comment on significant side effects (increased QT interval) from the use of HCQ. Once again, this trial used a particularly high dose of HCQ (600mg/day...right at ceiling dose for rheumatological use and much higher than the total antimalarial treatment dose). They also added azithromycin (another QT lengthening drug) to 20% of the HCQ patients. It’s not surprising at all to find such QT lengthening in a sick, more elderly population taking these medications in particularly high doses).
Further trials should utilize conservative doses of CQ/HCQ which have been proven safe in many millions of patients.
We don't yet know how this will pan out. We urgently need proper evidence. Statistically robust studies into prophylaxis and early intervention are likely to deliver the most interesting results.
Dr Philip Davies<br /> Aldershot Centre For Health<br /> http://thevirus.uk