dissenting.
I dissent. I would reverse the judgment and remand the case for a new trial on the ground that the trial court erred in admitting evidence of other burn incidents allegedly caused by Johnson & Johnson plaster of Paris products.
The specific factual background pertinent to this issue is as follows. At the close of appellees’ case, counsel for appellees sought to admit evidence consisting of a written report prepared by a Johnson & Johnson salesman. The report, which is referred to throughout the record as the “Texas report,” consisted of information written by the salesman on a Johnson & Johnson complaint form. The information on the form was provided to the salesman by a nurse at a hospital in Texas. The nurse apparently told the salesman that on three occasions, Johnson & Johnson plaster of Paris splints had caused burns to patients of the *441hospital. The nurse also apparently stated that in two of the cases, the plaster of Paris had been dipped in cold water before applying it. The report contained no other information as to the circumstances of the burn incidents reported. The thickness of the splints applied to the patients, the limb to which they were applied or the manner of air circulation around the splints during the drying process were not stated. Other evidence produced at trial indicated that all of these factors might be relevant to the degree to which the plaster of Paris would heat up during the process of drying into a hardened form.
However, the report did contain certain other relevant information. The report indicated that the plaster of Paris used in Texas was extra-fast setting plaster of Paris. The product that was used in the splint applied to Mr. Rogers was simply fast setting plaster of Paris. The report also indicated that the plaster of Paris used in Texas bore the Johnson & Johnson code number 3296N219R. The relevance of this number is that the plaster of Paris code number on the material that burned Mr. Rogers was 3296N355. The first 4 digits and the letter N are the same. The first four digits mean that the products were packaged and or shipped by Johnson & Johnson on March 29, 1976. The “N” indicates that they were shipped from Johnson & Johnson’s New Brunswick plant. The last digits are the batch number of the product. Plaster of Paris is apparently manufactured by mixing various chemicals in large batches and then applying the wet plaster of Paris material to thin bandage material which is rolled up, packaged and shipped. Thus, the batch number indicates the chemical mixture from which this product was ultimately manufactured.
When counsel for appellees sought to admit this report, counsel for appellant strongly objected. He argued that it had not been demonstrated that the product involved in or the circumstances surrounding the Texas incidents were sufficiently similar to the instant case. Therefore, the probative value of the Texas report was far less than its potential for confusing and prejudicing the jury. The trial *442court nevertheless admitted the report, with the following instructions to the jury:
The sole purpose of reading that complaint [the Texas report] to you is for you to use as to whether or not Johnson & Johnson had knowledge that certain reactions did result from a shipment of Johnson & Johnson products made on the same day as the shipment in this case, if the code that will be read to you means what it’s supposed to mean, what it has been testified it means. Now, if that is not accurate, you will hear more testimony on it. But the sole purpose is for you to determine from it whether or not Johnson & Johnson had been notified, was on notice that this particular shipment had caused problems somewhere else. And it goes to the issue whether or not the methods used by Johnson & Johnson to keep track of any possible complaint that may arise was a method which is a carefully drawn one and not a negligently drawn one.
When I say this is the purpose you can use it for, it means for no other purpose, which means you cannot use this incident somewhere else about this other incident from this shipment on the same date, which doesn’t necessarily mean the same package, or same batch, or same group of material was shipped the same day. If the testimony was heard, it means what we heard. That is not evidence you can use to determine there was any defect in the product here in this particular case.
Despite this instruction, however, the report was ultimately used by counsel for appellees as evidence of a defect. An enlargement of the report with the code numbers highlighted was circulated to the jury and counsel for appellees repeatedly referred to it in closing arguments. These references were not confined to statements concerning the relevance of the report to the notice issue. Rather counsel argued that the report was proof of the inconsistent nature of plaster of Paris concerning its tendency to heat up, which was the central allegation of defect in this case. The argument also stressed the similarity of the code dates on *443the two products.1 The same arguments are repeated in appellees’ brief to this court.
Under these circumstances, it is apparent that all of the requirements for the grant of a new trial because of an improper evidentiary ruling are present. The admission of the evidence was erroneous and its admission was harmful to the appellant. See Whitman v. Riddell, 324 Pa.Super. 177, 471 A.2d 521 (1984).
Evidence like that involved herein is commonly referred to as evidence of similar accidents or occurrences. As the en banc panel in Madjic v. Cincinnati Machine Company, 370 Pa.Super. 611, 537 A.2d 334, 341, allocatur denied, 520 Pa. 594, 552 A.2d 249 (1988), recently reiterated, the admissibility of such evidence is limited. Such evidence may often confuse the jury and the true issues before it by turning the jury’s attention to collateral matters, i.e., the allegedly similar accidents and the circumstances under which they occurred.
When such evidence is proffered, the burden is on the proponent thereof to demonstrate that the evidence is “... comprised of similar accidents occurring at substantially the same place and under the same or similar circumstances.” Id., 370 Pa.Superior Ct. at 625, 537 A.2d at 341. Moreover, even where such evidence is admitted with a limiting instruction as to its use, the admission may nevertheless be regarded as reversible error. See Whitman v. Riddell, supra.
*444As appellant persuasively argues, the limits on the admissibility of evidence of similar accidents are equally applicable in a malfunction theory case such as that at bar. Uitts v. General Motors Corp., 411 F.Supp. 1380 (E.D.Pa.1974), aff'd mem., 513 F.2d 626 (3d Cir.1975). In Uitts, plaintiffs proceeded under the malfunction theory to establish a defect in a car manufactured by defendant. They attempted to introduce evidence of records of complaints to defendant concerning other accidents involving identical or substantially similar vehicles. The court noted the general rule that such evidence is not easily admitted because of its potential for prejudice and confusion of the jury. The court excluded the evidence of prior complaints, stating:
If plaintiffs were attempting to prove the existence of a specific defect or malfunction it is clear that the admission into evidence of the occurrence of similar accidents would require a showing that those accidents were caused by the same malfunction or defect. Simply because plaintiffs are proceeding under the MacDougall [malfunction] theory we do not believe they can introduce evidence of accidents which may involve a variety of causes, and then let the jury guess which of those causes was responsible for the accident in the present case.
Id. at 1383.
Clearly, appellees did not demonstrate a sufficient similarity between the incidents reported in the Texas report and the incident at issue in this case to warrant admission of the Texas report. First, the Texas report did not involve the application of the same type of plaster of Paris as was applied to Mr. Rogers. Second, the report did not reveal all of the circumstances of the reported incidents which might be pertinent to whether the plaster of Paris used in Texas had actually malfunctioned or, in fact, had simply overheated because of improper use or other secondary causes. Lastly, the potential for prejudice and confusion of the jury posed by this report was indeed large.
Although the report was admitted with a limiting instruction, it was not presented to the jury in that light. Indeed, *445the report was not even necessary or probative as to the issues for which it was admitted, i.e., notice or ability to recall or trace. Johnson & Johnson had already admitted that it was aware of prior complaints of burning incidents resulting from plaster of Paris. Thus, there was no need for such potentially prejudicial evidence to be admitted concerning notice.
Moreover, the report was not relevant to the issue of whether Johnson & Johnson had notice of a problem with the product involved here or should have recalled or traced it. The Texas report involved extra fast setting plaster of Paris, a different product than that involved herein. Thus, if it put Johnson & Johnson on notice of anything, or if it imposed on Johnson & Johnson any post-sale duties, such related to a different product. Finally, it is difficult to see how the mere existence of a prior complaint goes to prove anything concerning whether Johnson & Johnson could have traced and/or recalled a defective product.
Furthermore, even with the limiting instruction and assuming that the report was probative and non-cumulative as to notice or post-sale duties, the clear possibility that the jury’s attention would be diverted by the report outweighs the value of the report. The jury might well have focused on the Texas report and the similarity between the code numbers it bore and the numbers on the product used on Mr. Rogers. The jury could easily have regarded this report as evidence of otherwise inexplicable “malfunctions” of Johnson & Johnson plaster of Paris, which would reinforce appellees’ theory that Mr. Rogers’ burns were also caused by such a malfunction and, therefore, by a presumably defective product. The jury should not have been presented with this option as the Texas report was not probative of anything regarding malfunctions of fast setting plaster of Paris under the circumstances involved in this case. Avoidance of this potential misuse of evidence of similar accidents and this type of diversion of the jury’s attention is precisely the rationale that underlies the limits on the admission of such evidence. These risks were clear*446ly present in this case and clearly outweighed the probative value of the Texas report, which was never established.
Since I find this error controlling, I would not address the remaining issues raised on appeal. I would find that the judgment against Johnson & Johnson should be reversed and the case remanded for a new trial.
. The majority states that appellant never objected to the use of the Texas report by counsel for appellees in his closing arguments and that this objection is, therefore, waived. I strongly disagree. Appellant preserved its objection to the admission of this evidence throughout the trial, arguing its objection at every opportunity and reiterating it in post-trial motions. To suggest that it was waived because not raised again in response to closing arguments is untenable. Appellant had already objected to the evidence and the trial court was clearly informed of appellant’s position on the issue. In any event, appellant makes a point of the use of the Texas report in appellees’ counsel’s closing argument only to indicate how prejudicial the admission of the report actually was in the context of this trial. The error alleged was the admission of the evidence by the trial court, not the use of it by appellees’ counsel, and appellant’s objection to this error was fully preserved.