The fog of medical war

WIRED so often runs annoying agitpop articles that it's a pleasure to find another piece there without an axe to grind, this time examining the continuing confusion over whether chloroquine is effective in COVID-19.  The author manages to point out that the opponents of the drug aren't yet in any more position to be sure that it's a failure than its advocates are in a position to know that it's a success.  All this without taking more than a few tiny potshots at President Trump.

I admit, however, to some disappointment.  I hoped by now we'd have clear evidence, and obviously it would have been nice to get confirmation that the drug works.

14 comments:

douglas said...

You know, given all the seeming variations of symptoms and combinations thereof that people get from this (or don't in the case of asymptomatics), it seems likely to me that you'd expect that no treatment will be broadly successful. You'd expect that some treatments would be sometimes successful and sometimes not at all, and others simply useless. But that's just a guess, and I'm no microbiologist/virologist.

douglas said...

It's a pretty good article, but the final conclusion it has is questionable on it's assumptions- "Imagine the satisfaction of knowing a true fact—of being able to help sick people, or knowing that hydroxychloroquine doesn’t help and being able to move on."
Proven drugs work most of the time on most people- but also *don't* work on some people sometimes, or have limited efficacy on others, and sometimes have adverse effects (thus the legally mandated long list of side effects in every pharmaceutical commercial). Some proven drugs even kill people on occasion. The trials are mostly to show safety in use, and some efficacy that doctors would want to prescribe it. They're making it sound like the expectation is for a drug that always works on everyone, or forget it. That's overblown.

Texan99 said...

Of course I'd rather hear that we've found a treatment that's wildly successful, but I'd be happy even to know that HCQ was modestly and consistently helpful, on average, in a statistically meaningful way. Not that's it's a guaranteed cure for any individual, but at least that if you prescribe it broadly, you'll improve the case fatality rate for the population--because this is a cheap, widely available, well-tolerated drug.

For the whole range of patients, most are going to recover anyway, so it's hard to draw conclusions from the fact that, when you prescribe HCQ to a bunch of them, most of them survive. We need to know whether a statistically significant greater number of them survive, or at least enjoy shorter times to recovery, than we would have expected without the treatment. The differential effects would be easier to detect in the dire category of older patients who are much more likely to die, but that may be where the drug is riskier to take. We really need to know more.

Elise said...

Thanks for this - a friend and I were wondering about it this morning. I can understand that people are reluctant to possibly end up in the placebo half of a clinical trial but I'm not entirely convinced this reluctance can be blamed on politicians and media hyping HCQ. Many years ago, I was offered the chance to participate in a clinical trial that tested two different ways of administering a new, stronger chemotherapy. When I asked why this wasn't a 3-way trial - new drug 1 ways, new drug other way, old drug - I was told that very few people would agree to take the chance they'd get the old drug once they were informed the new drug was stronger. (Oddly enough, I would have been fine taking that chance then but probably not now.) So I suspect that even without the hype, if COVID-19 patients were told they would either get a drug that might work or get a placebo that definitely wouldn't work, most of them would refuse to be part of the trial and insist on getting the "might work".

I wonder if it's easier to spin up a national clinical trial in a country like the UK - perhaps getting treatment via the NHS there means it's difficult for patients to insist on getting the treatment they want.

E Hines said...

improve the case fatality rate

It also would help if the NLMSM--and experts discussing the matter with the NLMSM--were more precise in their terms of reference.

If we include the asymptomatics and those lightly enough symptomatic that they only see their GP and get sent home in the "case" population, the fatality rate already is very low, and going lower as we learn more about the denominator of those rates. It's tough for any new drug to improve on that.

If we limit ourselves to those sufficiently sick to warrant hospitalization, then we can see some serious improvement--or not. And we already have a fatality rate from non-HCQ treatment against which to bounce HCQ outcome rates. Here, too, though, it's critical to understand the populations being handled, and it's much harder to understand them. Of those hospitalized, how many have comorbidities, and of that population, which group has the seriousness of their condition driven by their comorbidity as potentiated by the presence of the Wuhan Virus, and which group has the seriousness of their condition driven by their Virus infection as potentiated by the presence of their comorbidity? This breakout is especially difficult when the only comorbidity is age.

Eric Hines

Aggie said...

So many of the studies have a flawed premise - for example, the VA study had big doses being given to aged vets that were already very sick. Others are late-stage treatment studies of patients in the highest-risk groups just before they are ready for a ventilator. I came across one news article that was headlined as 'Hydroxychlorquine study' only to find out that a medical team did a literature search, I kid you not. Anytime I come across an article with the word 'touted' I stop right there: It's going to be anti-Trump in its premise, and not about hydroxychloroquine.

I've taken the drug before, working in West Africa so I know its side-effects are negligible for me. If I catch COVID-19 and I notice it, I'll be going on the Hydroxychloroquine / Zpack / Zinc mix as soon and as quickly as possible, unless my physician team thinks Remdesevir would be a better choice, because everything that I've read about treatment, the success is greater when it starts earlier. I'm already taking vitamins that peg me at 100% or greater of the RDA for everything the medical world cares about, including zinc and Vitamin D.

E Hines said...

I take 1000mg of Vitamin C daily, too; this is another immune system booster. I've been on that dose for several years.

Regarding the hydroxychloroquine side effects, one in particular leads to heart problems. Folks on HCQ chronically for lupus suffered very small rates of heart problems after 10 years, and zero such problems after 5 years. The currently used dose for the Wuhan Virus is 200mg, an intermediate sized dose as I understand it. The currently used course of treatment is for two such doses the first day, then one dose per day for the next 5 days.

There wouldn't seem to be much of a heart problem side effect in this regime.

Eric Hines

douglas said...

Aggie and Mr. Hines- Apparently the issue is that Azithromycin and Hydroxychloroquine *both* have a small risk of heart rhythm issues, and in combination, the risks increase (as you'd imagine). I've also read that this was noticed far more in Europe where they were using a significantly larger dose than doctors here in the US. Just a little more info, for what it's worth.

Texan99 said...

The heart risks are pretty low, but then the dilemma is this: do you want to prescribe a drug that has even low risk, if you don't have a solid basis for believing it's effective? If your guess that it might be effective is reasonably well-grounded, and the risk of the patient's become extremely ill or even dying of WuFlu is higher, it still makes sense to prescribe it. But both risks are rather low, and fuzzy. Is the heart-risk rate 0.01%? Is it higher, because you're combining drugs whose combination hasn't been studied much yet? Is the death rate from WuFlu 0.05%? 0.01? Is it 8% or 50%, because you're dealing with a population that has age and co-morbidity against it, and you're concentrating on people who already show symptoms? Not a single one of these numbers is solid, but I have to say that the heart-risk number sounds like the last one that should be keeping us up nights. It's by far the best studied and understood. I'll bet you'd have to work hard to find a doctor with a bad memory of losing a patient over it after prescribing either HCQ or Z, whereas we now have a solid corps of doctors who know what it's like to lose a WuFlu patient.

On a more positive note, New Neo posted an article about a 100+-year-old woman who recently survived a bout. She'd also survived a bout with the Spanish Influenza when she was an infant, at the tail end of the 1918-1919 pandemic. I hope she passed that immune system down to lots of descendants.

Texan99 said...

PS, the problem of wanting to prescribe HCQ-Z early because it might be more effective in the early stages, vs. wanting to prescribe it later because that's when we can be more sure that the risks of a full-blown WuFlu case are greater than the minor risks of the heart problem, remind me of the problem doctors face in treating a suspected stroke. There's a drug they can give that's very effective, and whose considerable risks are outweighed by the potentially life-saving benefit, both only if they're quite sure not too much time has passed from onset. There's a strict protocol on the timeline, which sometimes is perfectly awful, because the patient's family may not be able to pinpoint the onset of symptoms. They might be willing to roll the dice, but the doctor won't. Everyone's making a decision based on a black box they can't open. The protocol was developed because cold-blooded researchers realized we don't make good decisions in that dilemma. The family will almost reliably say, "Go ahead, give it a try," not a bad choice if the main goal is narrowing the choices to a prompt return to health or a sudden death. The doctor, however, can't tolerate the risk of the sudden death, particularly when it's not clear yet whether the patient might slowly recover naturally (completely or at least enough) without treatment.

E Hines said...

Everyone's making a decision based on a black box they can't open.

Even Schroedinger's cat had a 50-50 chance. The problem here, though, at least from my perspective, is that the outcome isn't binary. Suppose the drug works, and the patient survives. But those side effects arise and leave the patient either trapped in a body he can't use or can't communicate from or leaves him functionally vegetative. In either of these outcomes, the patient becomes a terrible burden, economically and emotionally, on the spouse.

None of that is anything I'd wish on me or my wife. It's why we have DNRs and Medical Powers of Attorney. Of course, those don't take the emotion out of a decision that must be made in real time, including the doctor's emotion (which should be irrelevant; he's likely to get over it in a finite amount of time, but the spouse will be dealing for the rest of one life or another), but maybe they'll make the decisions a little easier.

With the Wuhan Virus, the decision is simple: hit me with the drugs. My heart will do what I tell it to do. Even if it disobeys, that's such a low probability occurrence as to be laughable. I'm more likely to get run over by a bus. I'd as soon get well quickly so I can go back to running that other risk.

Eric Hines

Texan99 said...

Agreed--and that's why patients and families are probably likely to ask for the post-stroke drug. They'd much rather convert the choice to full recovery or death, and skip the worst-case life-in-death scenario. But there are other possible outcomes, like a difficult recovery that still restores many years of life, with only minor disabilities. It's a great deal like the agonizing choice people face when told that a pregnancy is fraught with the risk of an awful disability--with the difference that I don't believe we have the same risk to make that choice for an infant that we do for ourselves or someone who has chosen us to be a surrogate.

Still, I agree, those trade-offs should be for the patient or his surrogates. The unwillingness of doctors and hospitals to go along with the family's wishes is a problem in this kind of case, just as it is in decisions whether to risk a surgery. I understand why hospital directors refuse to expose themselves to regulatory retribution and lawsuits and bad ratings, but none of that should override the patient's wishes.

E Hines said...

regulatory retribution and lawsuits and bad ratings

Those things, though, are political decisions beyond the doctor's ability to handle. Such regulatory and legal abuses are straightforwardly, if not politically easily, corrected, and it's the politicians we elect who fail on this--and so it's us who fail on this.

He'd still be stuck with the emotional business, but at least he'd be limited to that.

Eric Hines

Texan99 said...

Indeed, what's wrong with the system that induces hospitals to play it safe with ratings is only partly the fault of legislators or even regulators. A lot of it is how laws are written, defining liability, but frankly a lot of it is on the public attitude, which becomes the attitude of juries, that hospitals are piggybanks that should be raided, because they never should have become wealthy providing something as rare and valuable as technical medical interventions in the first place.

Then we wonder why medical care gets rarer and more expensive with every passing year, and why insurance costs so much.