There are 85,000 medical malpractice lawsuits filed annually. Among them, 52,190 are summarily dropped for reasons unknown; 26,860 are settled; 1,190 result in plaintiff verdicts, and 4,760 in defense verdicts. Only 33.3% of these lawsuits are likely to have merit, while 66.7% do not. To make matters worse, only one out of every 37.5 claims reviewed by attorneys is represented, meaning that 3,102,500 other claims are abandoned for reasons known only to those attorneys. Of these, one million may have merit. In fact, more potentially meritorious cases are rejected by attorneys than those that proceed. The total litigation cost is $55.6 billion, with two-thirds attributed to frivolous lawsuits. The average cost is approximately $700,000 per lawsuit, and each lawsuit takes about two years to litigate. If nothing else, this is a turbulent sea of uncertainty.
When two-thirds of all decisions are wrong, there is a problem. The issue lies in decision-making. Traditional decision-making in medical malpractice litigation relies on inductive reasoning, which uses generalities, resulting in qualitative outcomes. The fundamental principle here is the “preponderance of evidence.”
All decision-making principles have a “level of confidence,” representing the odds of being right. For the preponderance of evidence, the level of confidence is “50% probability plus a scintilla,” with scintilla being discretionary and typically “just enough to win.” A coin toss has a 50% probability.
Similarly, all decision-making principles also have a “type-1 error,” representing the odds of being wrong. For the preponderance of evidence, the type-1 error is 50% minus a scintilla, only slightly better than a coin toss.
The solution also lies in decision-making. Hypothesis testing is deductive reasoning, using specifics and resulting in quantitative outcomes. When hypothesis testing is adapted for medical malpractice, at the very least, scintilla is assigned a value of 45%. This adjustment gives the preponderance of evidence a 95% level of confidence and a 5% type-1 error.
Here’s how hypothesis testing works in medical malpractice. The process follows two rules, similar to traditional decision-making:
Rule I: Define the objective evidence, which includes the standard of care, medical intervention, harm, and proximate cause.
Rule II: Analyze the objective evidence by comparing the standard of care and medical intervention to determine harm and proximate cause.
In inductive reasoning, analysis uses the preponderance of evidence, making a general comparison between the standard of care and the medical intervention to conclude whether there is a departure from the standard of care. In deductive reasoning, however, the comparison is more structured. Both the standard of care and the medical intervention are separated into 10 phases, and corresponding phases are compared. This method also uses the preponderance of evidence, but with a scintilla of 45%, aligning the preponderance of evidence with hypothesis testing. Ninety-five percent confidence is the sine qua non of hypothesis testing. Hypothesis testing uses a statistical test, providing a quantitative 95% confidence in the conclusion, with the level of significance (alpha) set at 0.05.
The objective of hypothesis testing is to prove the “null hypothesis.” The result is the p-value. If the p-value is equal to or greater than 0.05, the null hypothesis is accepted, indicating no statistically significant difference between the standard of care and the medical intervention. If the p-value is less than 0.05, a statistically significant difference exists, indicating that the medical intervention departs from the standard of care.
Some triers-of-fact on either side of a case may object to hypothesis testing because a scintilla of 45% changes the burden of proof to “clear and convincing evidence.” Scintilla is supposed to be a smidgen.
They are free to change the value of scintilla because scintilla is discretionary. A level of significance of 0.5 keeps them faithful to a scintilla of a smidgen. However, the chance of rejecting a true null hypothesis (type-1 error) will be around 50% rather than 5%. As for finders-of-fact, this casts as much doubt on the conclusion as traditional decision-making does.
Some might claim that hypothesis testing is too complicated or confusing. However, there are only two rules, and they are no different from traditional decision-making.
Some might argue that hypothesis testing is untried. However, all the objective evidence remains the same. The only difference lies between deductive and inductive reasoning. Deductive reasoning is used in court regularly. Moreover, hypothesis testing aligns with the Supreme Court’s Daubert Decision.
Lastly, some may object because hypothesis testing, as a decision-making method, is irrelevant to the evidence in the case. However, according to the rules of evidence, how an expert on either side reviews evidence to reach an opinion is, itself, evidence. One side’s decision-making is no less relevant than the other side’s. If nothing else, hypothesis testing exposes their decision-making.
After adopting hypothesis testing, rather than 85,000 lawsuits per year, there may be one million, most of which will be meritorious. Rather than settling frivolous lawsuits to avoid the risk of losing at trial, it will be more cost-effective to defend such lawsuits until the defendant is exonerated, dismissed with prejudice, or the lawsuit is dropped. Likewise, rather than engaging in a protracted legal battle over a meritorious claim, it would be more cost-effective to negotiate an expedient settlement. While total costs may rise to $70 billion from $55.6 billion annually, this would amount to $70,000 per lawsuit, not $700,000. The time to adjudicate lawsuits would also be shorter.
Howard Smith is an obstetrics-gynecology physician.
