Pages

25 March 2022

Getting Rich By Being Wrong: How Do We Prevent Stupid Patients From Unfairly Enriching Stupid Doctors?

This post sets forth the bare outlines of an economic paradox. I don't know if it is novel or has a name. I may research this and explore it further later.

Suppose that there are two ways to respond to a health condition and health care consumers are free to choose between them. Let's call one the "conventional approach" and the other the "novel approach."

Without being rigorous, we'll illustrate a set of facts where this paradox is present with an example, not the exact conditions when this paradox is present.

Suppose that 99% of the relevant healthcare practitioners favor the conventional approach because their greater experience and knowledge causes them to favor this conclusion, while only 95% of patients, who are less certain and less well informed favor the conventional approach. 

Thus, 1% of the relevant healthcare practitioners favor the novel approach, while 5% of patients favor the novel approach.

Further suppose that healthcare practitioners are ethical and do not offer an approach to their patients which they believe is not in the best interests of their patients. And, suppose that for every 100 relevant health care professionals there are 100,000 patients.

Now, run the numbers. There are 99 health care professionals using the conventional approach serving 95,000 patients (960 patients per professional). There is 1 health care professional using the novel approach serving 5,000 patients (5,000 patients per professional).

A health care professional with more than five times as many patients each can probably charge more and thus will make more profits by favoring the novel approach than by favoring the conventional approach, even if, as the facts suggest, the conventional approach is much more likely to be in the best interests of the patients, objectively, than the novel approach.

Certainly, in some rare cases, the health care professional using the novel approach is a genius with unique insight who is the one acting in the best interests of the patient. Indeed, given how new ideas are adopted, it will usually be the case that the economics will be similar at first whether or not the new idea is a good one, although later on, as the evidence becomes more clear, more practitioners will adopt the novel approach causing the economics to change.

We don't have a problem with early adopters of a good novel approach making an extraordinary profit. This is similar to what we already do mechanically with patent rights.

But we don't want early adopters of a bad novel approach to receive undue profit, particularly because there is some irreducible level of stupidity and distrust of valid evidence in the general population of patients, and so, the economic upside to continuing to provide a bad novel approach is likely to be very sustained and extreme especially as some early adopters of the novel approach abandon it as unwise.

There are two very familiar ways to discourage this possibility.

A Pre-Approval Policy

One is to require the novel approach to be approved by an agency like the Food and Drug Administration in advance to be safe and effective relative to the status quo, in ethically conducted clinical trials in advance before the novel approach can be used. 

The pre-approval policy imposes costs of obtaining approval that counterbalance the excess profits associated with being an early adopter of the novel approach if it is approved, and is a total loss not compensated by subsequent early adopter profits, if the novel approach is not approved. 

The pre-approval policy assures that the risk of a bad novel approach being permitted and causing harm is pretty much as low as it is humanly possible to be. But while it properly discourages bad novel approaches, it may provide too weak of an incentive to adopt a good novel approach, since the costs of approval are still significant, and the early adopter reward after the novel approach is permitted may be very modest if the successful clinical trial wins over the practitioners using the conventional approach greatly reducing the excess profits for being an early adopter in the absence of a patent.

Also, the pre-approval process denies the benefit of the good novel approach both to patients waiting during the pre-approval process, and if excess profits are boosted with a patent but the early adopters don't have the capacity to provide the patented profit widely enough, to post-approval patients who would benefit from the approved good novel approach who aren't able to access the novel approach either due to high cost or due to limited capacity to provide it due to intellectual property rights.

A Tort Liability Policy

The other option would be to simply impose malpractice liability on practitioners who take a bad novel approach to their patients that causes harm to the patients when the threshold determination is made that doing so constitutes malpractice.

Tort liability also serves to create a powerful incentive for practitioners using the conventional approach to transition to a good novel approach so that they don't face malpractice liability when the threshold determination is made that the good novel approach is better in the amount harm suffered by the patients.

So far, so good, but how does one determine what the threshold is in that circumstance, and under what circumstances should a patient be free to waive that liability with informed consent about the risks that the approach chosen is the wrong one.

A hybrid approach might be to have the government do clinical trials at its own expense of novel approaches. Once a clinical trial is completed, it becomes malpractice to use a bad novel approach, and it also becomes malpractice to fail to use a good novel approach.

This doesn't fully resolve the original problem in the time period from when the novel approach is discovered to when the clinical trial is completed. But it does lower the malpractice stakes greatly for informed practitioners, powerfully shuts down bad novel approaches, and powerfully encourages the rapid adoption of good novel approaches.

Dual Track Waivers and Clinical Trials

A waiver with informed consent while a clinical trial is pending isn't the same as a regular clinical trial. 

In a clinical trial, patients enroll hoping that they will get a good novel treatment (which prescreening non-human clinical trials make more likely), but knowing that they might end up in a control group. They are taking the risk that the novel approach, if they get it is a bad one, balanced against the possible good fortune that the novel approach is a bad one but they are assigned to the control group and don't suffer as a result. But, they are also taking the risk that the novel approach is good but they don't get it because they are in the control group, balanced against the hope driving people to volunteer that the novel approach is good and they receive the novel approach. When novel approaches are more likely to be good than bad, the clinical trial participants are at a disadvantage.

But, in a waiver with informed consent, the patient knows for a certainty that they are getting the novel treatment. The patient is taking the risk that the novel approach is a bad one, but not the risk that they aren't actually getting the novel approach. When the novel approaches are more likely to be good than bad, the non-participants in the trial benefit and no one will sign up for the trial absent non-intrinsic incentives.

Then we have to consider the patient stupidity factor. In this dual regime of clinical trial participants and waiver patients, when there is evidence that the novel treatment is likely to be bad, smart people won't do waivers and will not sign up for clinical trials, while stupid people will do waivers and the clinical trials may not have enough enrollees to proceed.

But when the novel treatment is likely to be good, smart people will want to do a waiver, smart people will not want to do the clinical trial, and stupid people, disproportionately will not want to do either a clinical trial or a waiver because they are more likely to wrongly think that the novel treatment will be bad.

In conclusion, allowing patient waivers when it is clearly more likely that a novel treatment will be good, or is clearly more likely that a novel treatment will be bad, due to some sort of prescreening procedure, doesn't really work.

Clinical Trials With Free Choice While The Trials Are Pending Without Waivers

Another option would be to not allow waivers of malpractice liability and instead impose strict liability on the practitioners for harm caused to patients by using a bad novel treatment that has not yet been clinically approved. Thus, instead of having patients who aren't in a good position to evaluate the risk assume that risk, the better informed and knowledgable practitioner bears the risk that the novel treatment will be a bad one.

This would discourage lots of practitioners from using unapproved novel approaches and would probably make the patient charges to administer unapproved novel approaches high, but will leave the patients who could afford to take that risk the ability to do so, knowing that if it doesn't work out that they will be compensated by the practitioner making them pay the high treatment fees. This could work out to be a back door waiver, in effect, but it would prevent clinical trials from going empty, and it would allow a bold genius practitioner and their patient to benefit from a good novel approach sooner. 

A Multi-Stage Rule Regarding Waivers With Special Clinical Trial Group Treatment

An even more intermediate position would prohibit waivers when there isn't enough preliminary clinical trial data available for a patient to make a decision with informed consent and instead impose strict liability on the practitioner at that stage. But, perhaps waivers could be allowed once the late stage clinical trial was fully enrolled, perhaps with a premium price, strict liability as to the clinical trial participants (who are also taking the control group risk) and no cost of participation for clinical trial participants, and a shift of the risk from practitioner to patient for people who waive with informed consent at that stage outside a clinical trial group. This sounds close to the correct rule.

No comments:

Post a Comment