Today, the conventional rules that medical experts use to evaluate the merits of a malpractice lawsuit are established by attorneys, not doctors. These rules are based on inductive reasoning. Inductive reasoning is acceptable except for one thing; it forgives threats to validity. Although the premises are true, the conclusions are influenced by bias. The rules are as follows:

There are four categories of evidence:

• The disputed treatment represents a medical error.
• The standard of care represents a treatment free of error.
• Duty is the duty to do no harm and considers that, sometimes, an acceptable calculated risk is necessary.
• The complication is either an injury or an error of nature. An injury is preventable and is caused by a medical error that produces a measurable risk of harm. An error of nature is unpreventable and is caused by circumstances under no one’s control that produce a known background risk in the population. These are all true premises.

The hypothesis is “The disputed treatment either departs from or comports with the standard of care.” Any difference between the standard of care and the disputed treatment is a departure from the standard of care. Likewise, an associated complication is an injury until proven otherwise.

Inductive reasoning compares the disputed treatment to the standard of care. Inductive reasoning does not test the hypothesis; herein lies the problem. A complication remains an injury and is part of the evidence until trial.

Proof occurs during the trial. The preponderance of evidence is the criterion for proof—50 percent probability plus a scintilla. Scintilla is in the eyes of the beholder and has any value.

At trial, the plaintiff’s attorney has the burden of proof. Defense attorneys do not need to prove anything; they need only to cast doubt on the proof. Inductive reasoning and preponderance of evidence are used by both. Conclusions are nothing more than educated guesses having slightly more confidence than flipping a coin.

Finders of fact are the decision-makers. Because of inductive reasoning and the arbitrary nature of scintilla, mistakes in their decision-making are very possible, which is often favorable to the one side that exploits threats to validity the best.

These are today’s rules. There are other rules, but they are for scientists. Medical experts are scientists, not lawyers. These other rules are based on deductive reasoning – the scientific method. Deductive reasoning is not too dissimilar from inductive reasoning with one notable exception. When adapted to a lawsuit, these rules are as follows:

The categories of evidence and the premises are the same as in inductive reasoning.

The hypothesis is “The standard of care is not significantly different from the disputed treatment.” In the scientific method, this is called the null hypothesis. The proposition is that a difference between the standard of care and the disputed treatment only infers a departure from the standards of care. A complication in the disputed treatment is presumed to be an error of nature until proven otherwise.

Deductive reasoning compares the standard of care to the disputed treatment. Deductive reasoning tests the null hypothesis. Therefore, an error of nature can be proven otherwise. This is the notable exception from inductive reasoning.

In the interest of the test, the disputed treatment and the standard of care are separated into ten corresponding phases in which the duty to do no harm applies. As each of the 10 phases is compared, differences are measured as a risk of harm. In the sample of 10 results for the disputed treatment, some are the same and have no risk of harm. The occurrence of a complication corresponds to the background risk of an error of nature. Others are different and have a risk of harm. The occurrence of a complication exceeds the background risk. In fact, the occurrence of a complication equals the risk of harm multiplied by background risk.

A statistical test analyzes these ten results in the disputed treatment. The proof is a preponderance of evidence; however, scintilla, which is accepted as having any value, is 45 percent, making the criterion for proof 95 percent confident. This corresponds to a level of significance (alpha) of 0.05 – standard in the scientific method. The background risk represents the standard of care. The result of the test is the p-value. If p < 0.05, the difference is statistically significant, and the null hypothesis is rejected. The complication in the disputed treatment is caused by a medical error. If p ≥ 0.05, the difference is not statistically significant, and the null hypothesis is retained.

The test of the null hypothesis is part of the evidence.

At trial, the plaintiff’s attorney has the burden of proof. This is satisfied by rejecting the null hypothesis. The defense attorney casts doubt on this proof by retaining the null hypothesis. Conclusions are irrefutable proof, not educated guesses.

Finders of fact are the decision-makers. While finders-of-fact are not scientists, they are intelligent. They know proof when they see it. They hear all the evidence. Not testing the null hypothesis and, worse yet, manipulating the test is obvious to finders-of-fact and expose bias by medical experts and attorneys, who exploit threats to validity. They also know that 95 percent confidence stands in stark contrast to 50 percent confidence plus some ill-defined scintilla.

Medical experts are ethically obligated to objectivity. As scientists, they know or should know how to do this. If they do not, their opinions are not objective. Attorneys are ethically obligated to advocate. They will use whatever reasoning most benefits their client. Advocacy and objectivity do not necessarily conflict. This is the game changer that distinguishes an ethical medical expert from a hired gun and a principled attorney from a rainmaker, and sooner or later, it will become the new normal.

Howard Smith is an obstetrics-gynecology physician.