This study is getting some play, purporting to show that a large percentage of medical-malpractice tort suits are ill-grounded. According to these reviewers, 40% of all malpractice suits filed are “frivolous”, and some actually come from patients who were not even injured! Commentators are already touting it as the smoking gun for malpractice “tort reform”. I think the reasonable conclusion is exactly the opposite.

About 40 percent of the medical malpractice cases filed in the United States are groundless, according to a Harvard analysis of the hotly debated issue that pits trial lawyers against doctors, with lawmakers in the middle.

Many of the lawsuits analyzed contained no evidence that a medical error was committed or that the patient suffered any injury, the researchers reported.

The vast majority of those dubious cases were dismissed with no payout to the patient. However, groundless lawsuits still accounted for 15 percent of the money paid out in settlements or verdicts. . . .

The study found 3 percent of claims analyzed were filed by patients who had no injury. Of the claims that involved injuries, two-thirds were caused by medical error. But the remaining injury claims, or 37 percent, lacked evidence of a medical mistake, and most of those - 72 percent - were thrown out or otherwise resolved without a payout to the patient.

Altogether, the Harvard researchers reviewed 1,452 malpractice claims randomly selected from five insurance companies. The cases were resolved - meaning they ended in a verdict, a settlement or a dismissal - between 1984 and 2004. The claims resulted in a combined $449 million in verdicts and settlements. 

Forty percent frivolous claims! But take a look at those numbers. Only 3% of cases filed contained no injury - in the view of the doctors analyzing the cases. Within the 40% allegedly frivolous cases, 37% were ones where the patient was in fact injured but the medical reviewers, in their judgment, absolved the doctor. The report also notes that 72% of even these cases with real injuries were thrown out of court. So, in the end, 85% of successful claims in malpractice trials were for cases in which documentable medical error had occurred - according to these physician reviewers; at the same time, 90% even of claims dismissed at trial are cases of real harm to patients, in which there is a dispute whether the treating professional is at fault.

Some cite this as evidence of a tort system badly in need of “reform”, but I am not sure on what grounds. Ignoring structural or procedural issues (how these cases should be handled), what outcome profile would be expected from a good tort system? What improvements could we expect to see?

You could, I suppose, move the percentage of “correct” verdicts against doctors (in the judgment of doctors) from 85% to some higher figure, but, realistically, how high can it go? 95%? Certainly not 100%, in the best of cases. A 10% improvement would be good, but if the maximal improvement you could ever hope to get by any means is 10% - 15% more correct verdicts, is that really grounds for revising the entire system?

More to the point, what evidence is there from this study that the system is not actually working optimally now? The evidence shows that about 40% of cases filed against doctors are “frivolous” (again, in the judgment of doctors), but 85% of cases resolved with an award to the patient involve real error. It might be better to discourage filing wrongly-conceived suits in the first place, but if we had a perfect system for weeding out “frivolous” suits, we wouldn’t need trials - we could just weed out the wrongful claims and settle all the rest of the, by definition, well-grounded claims.  We don’t have such a system. Our mechanism for separating well-grounded claims of harm from ungrounded ones is the malpractice trial - that’s what the trial is for. And by the evidence of this study, these trials are serving to concentrate the percentage of well-grounded claims, among all claims, by almost 50% (from 60% of cases filed to almost 90% of cases settled), while reducing the concentration of unfounded claims by almost two-thirds (from 40% of cases filed to 15% of cases settled). Is that really so bad? Again, it might be nice to bump both those numbers up a bit, but there’s a practical limit to how far we can do so, and it hardly seems necessary to throw out the entire system in search of an incremental benefit of maybe 10% of cases each way (good or bad).

Aside from the practical limitations of improving a system that is already working remarkably well at doing exactly what it is intended to do, critics of this system seem to misunderstand its purpose in the first place. Civil trials are not, primarily, places for people to receive compensation for injury; they are places for resolving disputes over whether - or how much - compensation is even deserved. They are, by definition and by design, adversarial environments. The litgants are expected to have competing claims and interpretations of fact. Necessarily, the trial accepts (more or less) one litigant’s claims and interpretations, and rejects the other’s as false or unproven. In other words, 100% of all trials result in a determination that one part has asserted fact-based claims that are not justifiable on the evidence. This is not evidence of fraud, disingenuity, or abuse of the system - it is merely the way the adversarial system works. And, in the nature of things, it is inevitable that at least part of the time these judgments will go against the patient-plaintiffs in malpractice suits. Some patients will lose their suits - there’s no avoiding it. This is not evidence that those suits should not have been litigated at all; it is merely evidence that they were litigated and the patient lost, as one party or the other must inevitably do in every suit. (Significantly, the critics of the tort system never claim that doctors are immoral or abusing the system by defending themselves - but surely, if the fact that a patient loses is evidence that that patient was abusing the system by asserting false claims, by exactly that argument the fact that a doctor loses a case is evidence that that doctor was abusing the system by asserting false defenses, and worse, by trying to cover up medical mistakes. Somehow, though, “tort reform” never involves pre-trial review of doctors’ defenses and a demand for capitulation in cases where the doctor really is at fault.)

Patients bring suits when they feel they have really been harmed. (I am suspicious of the claim of “frivolous” suits on logical grounds alone. It is hugely against patients’ and lawyers’ interests to bring a weak suit, since they will only hurt themselves. Inherently both patients and their lawyers have an incentive to bring suit only when they think they can win - and I am sure the lawyers, at least, are not surprised to find that 90% of frivolous suits are dismissed out of hand.) Not only are they entitled to make their case and seek compensation, but there is no other way to run a tort system than by allowing patients to file complaints and waiting for them to do so. Necessarily, some patients will feel aggrieved without good reason - it’s human nature to do so. And so some suits will be without merit, and presumably those suits will be dismissed before argument, or will lose at trial. The evidence suggests that that is, in fact, exactly what is happening. But it is both unrealistic and logically absurd to demand that patients only bring “good” suits - and grossly unfair to them either to insist they submit to a physician-driven review of their claims against physicians before even being allowed to state their case (as some “reformers” suggest), or accept lower awards on their proven claims simply because some other claims made by some other people are unproven (as is also often heard). Neither of these suggestions makes sense either ethically or by way of the inherent epistemology of the adversarial trial system. They are even less reasonable when the system so patently works as intended.

The simple truth is that the fact that “frivolous” suits are brought is not evidence the system does not work. The system inherently discourages frivolous filings, but it cannot prevent them before trial, because the purpose of the trial itself is to determine whether the suits are good or not. And that question is often open to reasonable disagreement. One example of a “frivolous” suit identified in the study was a patient whose metastatic breast cancer was repeatedly missed on exam over a period of years; the reviewers felt this was not a medical error given the patient’s clinical profile. Surely the patient is entitled to feel otherwise? And surely the factual questions in the case - the patient’s signs and symptoms, the doctor’s approach, the prevailing standard of care - are amenable to review and interpretation? This may very well have been an ungrounded suit - one the patient should not have won, on the facts - but it is hard to believe it was a frivolous suit - one that no reasonable person would think could prevail, or one that was so inherently baseless as to constitute an abuse simply in asking it to be heard. Remember that over 90% of the “frivolous” suits identified by this study were of this type - cases where the patient clearly suffered harm and reasonable people disagreed on the facts. Remember too that over 70% of those cases were dismissed outright before a jury even heard them. Add to this the fact that malpractice insurance - a supposed driver of the need for tort reform - is a miniscule part of overall healthcare costs to begin with, and it hardly sounds like we need tort “reform”. Maybe we just need more tort cases.