Skip to content

News & Knowledge

Which Product-Safety Data is Admissible?


Which Product-Safety Data is Admissible?

January 16, 2020

This article originally appeared in Goldberg Segalla’s Product Liability Playbook. Read the issue here.

It starts with an emergency room patient linking an injury to a certain product.

A Consumer Product Safety Commission coder assigns a code that best fits the accident.

Then the information is used to estimate the number of ER visits associated with the product.

Plaintiffs in some product-liability actions use such estimates and “other accident” data from the National Electronic Injury Surveillance System to show that manufacturers know their products were dangerous.

Such evidence is crucial in proving that manufacturers should have re- designed or stopped selling their products, but often is irrelevant and unreliable because it fails to consider the exceptional circumstances behind each individual’s use of the product. And even when such evidence is relevant its value is substantially outweighed by its prejudicial effect and the likelihood that it will confuse or mislead the jury. And then there are hearsay concerns.

With product-liability litigation often massive and expensive and capable of having far-reaching effects on defendants, attorneys should be aware of CPSC limitations. Those who are can more effectively and fairly argue their cases on the merits.

Here are some contingencies to keep in mind.


Rule 403 of Article IV of the Federal Rules of Evidence provides a balancing test for excluding relevant evidence in product liability cases.

“Relevant evidence” is defined as “evidence having any tendency to make the existence of any fact that is of consequence to the determination of the action more probably or less probable than it would be without the evidence.”

The 403 balancing test is strongly weighted toward the admission of evidence. There must be a danger of unfair prejudice that substantially outweighs the probative value of the evidence.

Unfair prejudice is not the same thing as prejudice. Almost all relevant evidence is prejudicial to one side or the other. Proving unfair prejudice is often difficult. Proving the balance is even more so.

Then there’s the fundamental problem with 403 objections: They concede the relevance of the evidence.

A 403 objection is usually the last objection a lawyer makes when seeking to exclude evidence.

When Neiss Data is Not Relevant

Federal Rule of Evidence 402 often renders “other accident” data irrelevant, making it inadmissible.

For example, one should be keenly aware that if no dispute as to a fact or issue exists, “other accident” data deployed as proof of that issue or fact is not relevant because it does not make the existence of the issue or fact “more probable or less probable than it would be without the evidence.” If the evidence is not relevant, courts should not find it admissible.

In the case of Rye v. Black & Decker Manufacturing Co., John Rye was injured when a portable 7¼-inch Black & Decker circular saw he was using became bound up and kicked back as he was trimming a piece of wood from the end of a treated pine board, amputating part of his right hand. The saw was not equipped with a riving knife, a device used to prevent such accidents.

Rye used 17 “prior complaints” to show Black & Decker knew of problems with the design of the chainsaw, including letters from customers, civil complaints, and published court opinions. But the defendant already had admitted that such a kickback could occur. So the court found that Rye “was not harmed by the [trial] court’s refusal to allow in the prior incidents as evidence that appellee had notice that its saws could kick back.”

A plaintiff’s desire to introduce into evidence proof of other accidents is circumstantial evidence, or evidence in which “even if the circumstances depicted are accepted as true, additional reasoning is required to reach the proposition to which it is directed.”

McCormick on Evidence notes that “circumstantial evidence … can be offered to help prove a material fact, yet be so unrevealing as to be irrelevant to that fact.”

In many cases, the dissimilar and exceptional nature of “other accidents” can make the evidence “unrevealing” in nature. As the court in Nachtsheim v. Beech Aircraft Corp. notes, “As the circumstances and conditions of the other accidents become less similar to the accident under consideration, the probative force of such evidence decreases.”

When Neiss Data Constitutes Hearsay

Hearsay is defined in Federal Rule of Evidence 801 as “a statement, other than one made by the declarant while testifying at the trial or hearing, offered in evidence to prove the truth of the matter asserted.”

Established hearsay principles likely render NEISS data inadmissible because of its untrustworthy nature.

NEISS data is untrustworthy because it involves multiple layers of hearsay. The first layer arises when a patient provides medical personnel information about the nature and cause of an accident. The questions a doctor or nurse asks a patient about an accident are not designed to elicit truthful statements for determining liability; they are instead designed to find out how the patient characterizes his ailments so that he may be treated accordingly. A patient’s account of an accident is often vague, self-serving, or inaccurate.

The second layer of hearsay arises when the hospital functionary arbitrarily chooses which of the patient’s information to truncate for purposes of furnishing a report to a NEISS coder. A hospital functionary’s choices of what information to exclude or include is not guided by expertise in law and fact analysis, but that person’s one-sided description. It is no more appropriate for a hospital official to discern and truth than it is a lawyer to pick up a scalpel.

A third layer of hearsay arises when the NEISS coder receives and interprets that information, paraphrasing an already restated account into the NEISS database. Here, the pursuit of trustworthiness is further compounded by a summary limit of 600 characters.

At its essence, the NEISS coder’s role is not designed to provide accuracy. Entries are often made up of incomplete sentences and keywords. This relay of information from patient to hospital functionary to coder is synonymous with the childhood game of telephone.

The abovementioned concerns were acknowledged by the court in McKinnon v. Skil Corp., where the First Circuit excluded evidence that was based on CPSC data on the grounds that the evidence “in the reports is simply a paraphrasing of versions of the accidents given by the victims themselves who surely cannot be regarded as disinterested observers.”

From the first description of the accident, the quest for reliable evidence is influenced by “double hearsay in many instances.” The court observed that CPSC data is derived from a CPSC investigator, which constitutes the first level of hearsay, and an accident victim interviewee, which constitutes the second hearsay level.

When “Other Accident” Evidence is Unduly Prejudicial

Under Federal Rule of Evidence 403, even evidence that is relevant “may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence.”

Unfair prejudice in this rule means “an undue tendency to suggest decision on an improper basis, commonly, though not necessarily, an emotional one.”

Uitts v. General Motors Corp. illustrates the potential for unfair prejudice from the introduction of “other accident” evidence. In this case, a plaintiff sought the introduction of 35 reports of supposedly similar accidents. The court stated that “[p]roof of prior accidents or occurrences are [sic] not easily admitted into evidence, since they [sic] can often result in unfair prejudice, consumption of time and distraction of the jury to collateral matters.”

The court continued that, to minimize the prejudicial effect of the r ports, the defendant “would have had to go through each one individually with the jury. The result would have been a mini-trial on each of the 35 reports offered by plaintiffs. This would lengthen the trial considerably and the minds of the jurors would be diverted from the claim of the plaintiffs to the claims contained in these reports.”

The court in Uitts also acknowledged hearsay concerns, noting that permitting reports containing hearsay into evidence “would be tantamount to allowing the persons making the statements to testify against defendant without being subject to cross-examination or required to take an oath. Therefore, any probative value these reports might have is outweighed by their prejudicial nature.”

The Product Liability Playbook is collaborative effort of Goldberg Segalla’s Product Liability practice group examining the latest practices, emerging developments, and influential court decisions in product liability law. View our most recent issue.