dissenting.

I disagree with Landau, J., that the trial court should have dismissed Edwards from the case and should have admitted the redacted agreement into evidence. However, I agree with him that the trial court should have permitted cross-examination of Edwards concerning the nature of the agreement, so that the jury could understand the extent of Edwards’s bias and interest. That error substantially prejudiced Key, in my view, and justifies reversal. For that reason, I dissent.

pro tempore, concurring.

Defendant Key Pharmaceuticals, Inc. (Key)1 appeals from a jury verdict in favor of defendant/cross-claim plaintiff Dr. Frederick Edwards (Edwards), that awarded a total of $23 million in compensatory and punitive damages. Key manufactures the prescription drug Theo-Dur, an asthma medication containing theophylline. This action was brought on behalf of Paul Bocci, III (Bocci), a long-time user of Theo-Dur, against Key for negligence and strict products liability after he suffered permanent brain damage as a result of theophylline toxicity. A claim was also brought against Edwards for medical malpractice for failing to diagnose and treat Bocci for theophylline toxicity. Edwards cross-claimed against Key for negligence and fraud, alleging that Key failed to advise physicians appropriately concerning theophylline toxicity caused by Theo-Dur. At trial, Bocci’s claim against Edwards was dismissed, and the jury returned verdicts in favor of Bocci and Edwards against Key. After trial, Key settled with Bocci.2 Key appeals from the judgment in favor of Edwards.

Because Edwards prevailed by jury verdict below, we state the facts in the light most favorable to him. Baker v. English, 324 Or 585, 587, 932 P2d 57 (1997).

The drug theophylline is a bronchodilator that has been used to treat asthma for many decades. Theophylline has a narrow therapeutic range: In order to prevent asthma symptoms, the serum levels of the drug in the blood generally must be at least 10 micrograms per milliliter (mcg/ml), but serum levels above 20 mcg/ml can be toxic. Saturation kinetics play a part in the way in which this drug may be metabolized; when the level of the drug in the body increases but the liver’s ability to metabolize it does not increase, or when the amount of the drug entering the body stays the same but the liver’s ability to metabolize it decreases for some reason, saturation can occur. This causes the serum levels of theophylline in the blood to increase. The ability of a *524body to metabolize theophylline may be affected by many things, such as smoking or the presence of a virus. Interactions between theophylline and other drugs may cause a body to metabolize theophylline at a slower rate, thus increasing the serum levels of theophylline in the blood and leading to theophylline toxicity. Theophylline toxicity can cause nausea, vomiting, headaches, diarrhea, tachycardia, seizures, and death.

Before the 1970s, theophylline therapy was difficult because a great deal of monitoring and adjustment of dosage was necessary to keep stable the amount of the drug in a patient’s blood at any given time. In the 1970s, Key introduced a new theophylline product, Theo-Dur, a timed-release capsule that it claimed had zero-order absorption. Zero-order absorption occurs when a drug is constantly absorbed by the system and eliminated at the same rate, thus keeping the serum levels of the drug in the blood stable. Key promoted Theo-Dur as being safer than other theophylline products; because of its zero-rate absorption, Key claimed, the risk of “toxic peaks” could be avoided. Key aggressively promoted Theo-Dur to physicians through sales representatives, journal advertising, and direct-mail campaigns. Promotional materials also urged patients, physicians and pharmacists not to accept generic or brand-name substitute theophylline products, as switching from Theo-Dur could cause “excessive toxicity.” In 1987, the Food and Drug Administration (FDA) informed Key that it must cease claiming that Theo-Dur was superior to other theophylline products due to zero-order absorption, because the claim was not sufficiently supported by clinical data. In 1989, the FDA again found that Key was making false and misleading claims that Theo-Dur was superior to other theophylline products.

In October 1987, Key became aware of several medical journal articles reporting a drug interaction causing theophylline toxicity when both theophylline and ciprofloxacin, a newly available antibiotic often used to treat respiratory tract infections common among asthmatics, were administered. A Key internal memorandum dated October 30,1987, stated: “Ciprofloxacin produces a 30-113% decrease in theophylline clearance. These effects are significant enough to *525cause a patient that is stabilized on theophylline to potentially become toxic.” In March 1988, Key proposed to the FDA that a warning be added to its Theo-Dur package insert concerning the interaction between theophylline and ciprofloxacin. Key’s application for the labeling change, however, was not submitted in proper form, and it was not until March 1989 that the labeling change was approved. The labeling change was not reflected in the 1990 volume of the Physician’s Desk Reference (PDR), although it was included in a May 1990 PDR supplement. Information such as a warning about theophylline-ciprofloxacin interaction is generally imparted to the medical community by drug manufacturers through “Dear Doctor” and “Dear Pharmacist” letters or statograms, through promotional materials sent to physicians, or through representatives of pharmaceutical companies (detailers) who call on physicians. The Code of Federal Regulations (CFR) allows for distribution of such warnings to physicians before revised labeling is approved by the FDA. The CFRs also allow a pharmaceutical company to change its labeling without preauthorization from the FDA in order to add warnings, precautions, or information concerning adverse reactions.

Between October 1987, when Key became aware of the theophylline-ciprofloxacin interaction, and Bocci’s injury in October 1990, Key took no steps to inform physicians or patients of this potential toxicity problem, although it knew of several cases of serious toxicity and one death caused by the interaction of theophylline and ciprofloxacin. During this period, Key continued to promote Theo-Dur as the only theophylline product that “protects against toxicity.”

Bocci began taking Theo-Dur when he was seven years old. On October 21,1990, when Bocci was 20 years old, he went to a medical clinic for treatment of a skin rash and was treated by Dr. Davis. Bocci indicated to Davis that he was not taking any medications, although he was taking 900 milligrams of Theo-Dur a day. Davis prescribed ciprofloxacin for the skin rash. Bocci took both ciprofloxacin and Theo-Dur from October 21 to October 26, 1990. On October 27, 1990, Bocci went to an urgent care clinic, The Eugene Clinic, with symptoms including nausea, vomiting and diarrhea. He was *526seen by Dr. Edwards. Edwards discovered that Bocci had been on a stable dose of Theo-Dur for a long time, and, although he considered a diagnosis of theophylline toxicity, he did not diagnose that condition. Edwards did not consult the PDR or the PDR supplement. Edwards did not think that a patient on a stable dose of Theo-Dur could experience severe theophylline toxicity that could lead to brain damage unless the patient had taken an overdose, because Theo-Dur had been marketed and promoted to him as a “safe” drug. The detailer who had promoted the drug to Edwards and the Eugene Clinic in 1989 and 1990 testified that he would tell physicians that Theo-Dur had zero-order absorption and that it was safe. Edwards testified that he did not make a connection between a patient on a stable dose of a safe drug such as Theo-Dur and a serious toxicity problem. He therefore diagnosed gastroenteritis, treated Bocci with antinausea medication and intravenous fluids, and then sent him home.

Several hours later, Bocci had violent seizures and was taken to a hospital emergency room. Anticonvulsant medications were administered but failed to control the seizures, and the physicians attending him had difficulty diagnosing the cause of the seizures. Late that evening, after learning that Bocci had been taking asthma medication, they tested for theophylline toxicity and discovered that Bocci’s theophylline level was 63 mcg/ml, well into the toxic range. The theophylline was removed from Bocci’s blood through dialysis and the seizures ceased. However, Bocci suffered permanent brain damage as a result of the seizures.

A number of physicians testified about their knowledge and understanding of Theo-Dur as of October 1990, when Bocci sustained his injury. Those physicians did not know about the theophylline-ciprofloxacin interaction and did not know that patients stabilized on standard doses of Theo-Dur could experience toxicity problems causing severe seizures such as those experienced by Bocci. Those physicians did not know that blood levels of theophylline should be obtained on patients using theophylline products if the patients experienced nausea and vomiting.

On Bocci’s claims, the jury returned a verdict finding that Key was 65 percent at fault and Bocci was 35 percent at *527fault. On Edwards’s cross-claims, the jury found that Key was negligent and that Key’s negligence caused Edwards’s damages and found that Key had made fraudulent misrepresentations to Edwards that caused him damage. The jury further found that Key caused 100 percent of Edwards’s damages. The jury also found clear and convincing evidence that Key had acted with wanton disregard for the health and safety of others and had knowingly withheld from or misrepresented to the FDA or prescribing physicians information known to be material and relevant to theophylline toxicity, in violation of applicable FDA regulations. The court entered a judgment in Bocci’s favor including compensatory damages of $5,621,648.203 and punitive damages of $35 million, and in Edwards’s favor for compensatory damages of $500,000 and $22.5 million in punitive damages. On appeal, the only issues before us concern the verdict and damage award in Edwards’s favor.

Mary Carter Agreement

A number of Key’s assignments of error relate to a “Covenant Not-to-Enforce-Judgment and Separate Loan Agreement” entered into by Bocci, Edwards, and Edwards’s insurer. This type of agreement is commonly known as a “Mary Carter agreement,” and we will use that term when discussing it.4 Through this agreement, which contemplated a lawsuit by Bocci against Edwards and other defendants, Edwards guaranteed Bocci a minimum recovery of at least one million dollars and, at the same time, limited any recovery against Edwards to one million dollars. The agreement contained the following terms: In consideration for $200,000, Bocci agreed not to enforce any judgment against Edwards based on the events of October 27, 1990; Edwards loaned *528Bocci an additional $800,000, the repayment of which depended on Bocci’s recovery, if any, from other defendants. In the event that Bocci recovered against no other defendants, no repayment of the loan was required. In the event that Bocci recovered an amount in excess of $3 million from other defendants, full repayment of the $800,000 loan was required.5 The existence of this agreement was known to Key before trial.

As an initial matter, we note that most of Key’s arguments pertaining to the Mary Carter agreement focus on whether Edwards’s participation in the case as a defendant allowed Bocci an unfair procedural advantage. As noted, see note 2, above, no issues relating to Bocci’s recovery against Key are currently before this court, because Key settled with Bocci after the appeal was commenced. After careful review of the briefs, however, we conclude that Key’s arguments concerning the Mary Carter agreement remain viable on appeal because, if Key is correct, the procedural advantages that it has identified relating to the Mary Carter agreement might indirectly have benefitted Edwards in his cross-claim as well as benefiting Bocci in his claim. Although Key’s primary argument was that the Mary Carter agreement provided Edwards a motive to maximize Bocci’s recovery against Key, as discussed below, other advantages relating to the way the case was presented to the jury, the way witnesses were examined, and similar matters, could have benefitted Edwards as well as Bocci.

Key asserts on appeal that the trial court erred in denying its pretrial motions concerning the Mary Carter agreement. Key argued in those motions that Edwards should be dismissed from the case because the agreement settled Bocci’s claim against Edwards, or alternatively, that the agreement should be declared void or that Edwards should be realigned as a plaintiff rather than a defendant.

On appeal, Key argues that the Mary Carter agreement should have been declared void, as a violation of public policy. In Grillo v. Burke’s Paint Co., 275 Or 421, 427, 551 *529P2d 449 (1976), the court rejected the notion that Mary Carter agreements are per se invalid under Oregon law. Moreover, ORS 18.455, which recognizes covenants not to enforce judgments, demonstrates that Mary Carter agreements do not violate the public policy of Oregon. The trial court did not err in denying Key’s motion to declare the agreement void on the basis of public policy.

Key next argues that Bocci’s claim against Edwards should have been dismissed before trial on the ground that the Mary Carter agreement had fully settled those claims and, thus, no justiciable controversy existed between those parties. Key argues, in essence, that Bocci’s claim against Edwards became moot when they entered into the Mary Carter agreement. We disagree. While the agreement had the effect of limiting Edwards’s liability for Bocci’s injury to an amount between $200,000 and $1 million, it did not establish Edwards’s liability. As the agreement itself states, one of its purposes was to permit Edwards to establish that he was not liable for Bocci’s injury, while limiting his financial exposure and allowing him an opportunity to cross-claim against Key. A justiciable controversy exists when “the interests of the parties to the action are adverse,” and “the court’s decision in the matter will have some practical effect on the rights of the parties to the controversy.” Brumnett v. PSRB, 315 Or 402, 405-06, 848 P2d 1194 (1993). Bocci’s and Edwards’s interests were sufficiently adverse to survive the first prong of that test: Bocci asserted that Edwards was negligent, and Edwards asserted that he was not negligent. How much Edwards would pay Bocci depended on whether, or to what extent, the jury determined Edwards to be negligent. That both Bocci and Edwards had an interest in establishing that Key was at fault does not destroy that adversity. A decision would have “some practical effect,” in that it would determine whether Edwards would pay Bocci $200,000, $1 million, or some amount in between.

That conclusion is consistent with our holding in Stephens v. Bohlman, 138 Or App 381, 909 P2d 208, rev allowed 324 Or 176 (1996). In Stephens, one of the defendants settled with the plaintiff for a flat sum, but agreed to remain in the case as a defendant. Distinguishing that kind of arrangement from a Mary Carter agreement, we concluded *530that no justiciable controversy existed between the plaintiff and the settling defendant because the settling defendant “had no interest in the outcome of the case against [the non-settling] defendant because it could neither gain nor lose anything as a result of the trial.” Id. at 385. We distinguished the agreement in Stephens from the “true” Mary Carter agreement at issue in Grillo, noting that in Grillo, while “the settling defendant’s interest in the case became adverse to that of the other defendant, the settling defendant clearly retained an interest in the outcome” due to the repayment provisions of the agreement. Id. (emphasis added). This case involves a “true” Mary Carter agreement like the one at issue in Grillo, and not a settlement agreement of the type at issue in Stephens. The trial court properly denied Key’s pretrial motion to dismiss Edwards as a defendant.

The dissent’s position — that no justiciable controversy existed between Bocci and Edwards after they entered into the agreement — is untenable. 158 Or App at 556. As an initial matter, neither defendants nor the dissent explain why the resolution of this question would justify overturning a verdict in Edwards’s favor on an entirely different claim. In any event, were the dissent’s analysis of this issue correct, it would lead to an inevitable conclusion that Grillo and Stephens were incorrectly decided. While the dissent narrowly frames the issue as whether the trial court in this case erred by failing to dismiss Edwards from the case, such a holding would have the much wider implication that no justiciable controversy ever exists between a plaintiff and a defendant who are parties to a Mary Carter agreement. Thus, if the dissent were correct, Mary Carter agreements would not be valid and enforceable under Oregon law, despite what we have said in Stephens, and despite what the Supreme Court has said in Grillo.

In any case, any error in failing to dismiss Edwards as a defendant before trial was not prejudicial. Bocci’s claim against Edwards was dismissed before the case was submitted to the jury, and Edwards would have remained a party in any event due to his claim against Key. Defendants offer no explanation of why the trial court’s failure to dismiss Bocci’s claim against Edwards would justify setting aside Edwards’s verdict against Key, and we can think of none. See Stephens, *531138 Or App at 386 (where settling defendant was dismissed before the case was submitted to the jury, error in failing to dismiss settling defendant before trial was harmless, given that the same evidence would have been admitted had that defendant been dismissed earlier).

Key argues alternatively that the trial court should have realigned Edwards as a plaintiff, citing authority from other jurisdictions. The gist of Key’s argument is that Edwards’s participation as a defendant created an unfair procedural advantage. Key notes that Edwards’s testimony (unsurprisingly) indicated that Key’s failure to warn, rather than Edwards’s failure to diagnose, was the cause of Bocci’s injuries, and further points out that Bocci’s attorney did not cross-examine Edwards vigorously to try to establish that Edwards was negligent. Key implies that the jury was misled by this alignment of the parties, to Key’s disadvantage.

Key’s arguments would be equally applicable to any case in which a plaintiff enters into a covenant not to enforce a judgment with a defendant who remains party to a lawsuit. ORS 18.455, however, specifically recognizes the validity and enforceability of such covenants6 and does not preclude defendants who enter into such agreements with plaintiffs from participating as defendants in a lawsuit. As to Key’s argument that the jury was likely to be misled if Edwards was not realigned as a plaintiff, we are unconvinced that juries are so naive as to assume that parties’ interests in a civil case are defined solely by whether they are labeled “defendants” or “plaintiffs.” Key had ample opportunity to— and in fact did — argue extensively to the jury about Edwards’s and Bocci’s mutual interest in verdicts against Key. Juries are regularly asked to make factual decisions concerning liability where numerous defendants are pointing fingers at each other in an effort to establish that they are not responsible for an injury. Juries are regularly asked to decide the merits of defendants’ cross-claims against one another. Juries, when properly instructed, can assess the interests of *532the parties, whether those parties are labeled as plaintiffs or defendants.

That conclusion brings us to Key’s next assignment of error, that the trial court abused its discretion in its instructions to the jury concerning the Mary Carter agreement. The trial court gave the jury the following prehminary instruction regarding the Mary Carter agreement:

“I do need to inform you that settlement agreement has been reached between the plaintiff and Dr. Edwards. Dr. Edwards, nevertheless, remains a party to this lawsuit under the terms of the settlement agreement and he will be offering evidence that he was not negligent in his involvement in this case and he also has filed what we call a cross-claim. He is making a claim himself against Key Pharmaceuticals.

“The settlement agreement contains a schedule for repayment, however, to Dr. Edwards by the plaintiff if the plaintiff is awarded a verdict against the defendant Key Pharmaceuticals Company. So there is that relationship that you need to be aware of, but the jury that is selected in this case is to consider the fact of the settlement only as it might bear on the issue of credibility or believability of the witnesses who testify, and it is not to be considered in any way in determining the amount of the verdict or damages, if any, that the jury should award at the end of the case.”

Key argues on appeal that the jury instruction should have informed the jury of the specific terms of the agreement, so that it could have pointed out to the jury the covert cooperation between Edwards and Bocci. In the trial court, however, Key argued that the statement that Edwards remained in the case to establish that he was not negligent was “absolutely ludicrous,” that Edwards had no financial stake in the outcome of the case, and that the reason Edwards was a party was “so that his insurance carrier can recoup some of the money that they have agreed to pay towards the million dollar guarantee.”7 We consider only the arguments that Key adequately preserved in the trial court. ORAP 5.45.

*533Key sought to have the court instruct the jury that Edwards had no financial stake in the outcome of Bocci’s claim. As discussed above, Edwards had a financial stake in the outcome of that claim. Key sought to have the court inform the jury about Edwards’s insurance coverage. Giving such an instruction would have been error, because “the injection of insurance into the trial of a case where it is irrelevant is ground for reversal.” Waterway Terminals v. P. S. Lord, 242 Or 1, 37, 406 P2d 556 (1965). See also OEC 411 (limits on admissibility of evidence concerning liability insurance). Edwards’s malpractice insurance situation had no independent relevance in this case and almost certainly would have been prejudicial to Edwards in the jury’s deliberation of liability as between Bocci and Edwards. The court did not err in giving the preliminary instruction quoted above. The instruction gave a general overview of the Mary Carter agreement that accurately, if briefly, summarized the document. Particularly in light of the fact that Key dees not point to any specific error in the trial court’s later, more in-depth instruction concerning the Mary Carter agreement, see note 7 above, we are unable to see how the court erred in giving the preliminary instruction.

Key next argues that the trial court erred in limiting what it could say about the Mary Carter agreement during its opening statement. Key makes no specific arguments concerning this assignment of error, and it is unclear from the excerpt of the record provided by Key exactly what Key wished to tell the jury about the agreement, except that the agreement “created a fund for litigation.” However, the real problem with Key’s argument is that this issue was raised before trial, and the court did not make a ruling on what Key could or could not tell the jury during opening argument. Key’s counsel asked for guidance as to what he could say about the agreement, and the court stated that “for the moment, I don’t have an answer to your question,” although the court did suggest that counsel should not go into too much detail about the Mary Carter ágreement in its opening. Trial *534courts have a great deal of discretion to control how parties make their arguments. See generally Troutman v. Erlandson, 279 Or 595, 605, 569 P2d 575 (1977). The trial court’s suggestion to Key’s attorney that he should not go into too much detail about the agreement during opening arguments was not an abuse of discretion, in light of the court’s substantive decision about the admissibility of the terms of the agreement, discussed below.

Key’s central argument concerning the Mary Carter agreement is that the trial court erred in refusing to allow Key to introduce the agreement itself into evidence. In Grillo, 275 Or at 427, the court stated that such agreements are valid and enforceable, but went on to note that the agreement in that case “would have been subject to pretrial discovery and, upon request of defendant, would have been admissible into evidence.” Key relies on that statement in support of its contention that the trial court was required to admit the agreement into evidence, rather than simply to describe the relevant portions of the agreement to the jury.

The trial court did not ignore the above-quoted statement from the Grillo case. Rather, the court recognized that tension existed between that dictum from Grillo and a more recent case, Holger v. Irish, 316 Or 402, 851 P2d 1122 (1993). In Grillo, the plaintiff and a defendant entered into a Mary Carter agreement that was similar in substance to the one at issue here. The nonsettling defendant, however, did not discover the existence of the agreement until after the trial.8 The question in Grillo was whether the defendant was entitled to a new trial based on newly discovered evidence. 275 Or at 423. The court concluded that the defendant was not entitled to a new trial because, although it did not know specifically of the Mary Carter agreement, it knew that the other defendant, who was a party to that agreement, intended to settle with the plaintiff, and in fact knew that the defendant had advanced money to the plaintiff. Id. at 427-28. Thus, the court concluded that the defendant who was not party to the agreement had not exercised due diligence and was not entitled to a new trial. Id. at 429. Holger, by contrast, concerned *535the propriety of informing a jury about a plaintiffs settlement with a defendant that had resulted in the dismissal of that defendant from the case before trial. The trial court told the jury of the settlement, on the ground that the jury would “wonder why the [settling defendant was] not involved in the case.” Holger, 316 Or at 413. The Supreme Court held that telling the jury of the settlement was error, because “in the usual case, it is not proper to inform the jury concerning a plaintiffs remedies or potential remedies against persons who are not parties in the dispute that the jury is to decide, unless that information has independent relevance.” Id. at 414.

We agree with the trial court that some tension exists between Grillo and Holger. Grillo declares, albeit in dictum, that a Mary Carter agreement would have been admissible into evidence. Holger indicates that a jury generally should not be informed of settlements unless the information has independent relevance. The trial court attempted to accommodate those competing interests by allowing the jury to know as much about the Mary Carter agreement as had “independent relevance.” See also OEC 408, 411 (evidence concerning settlements and liability insurance not admissible to demonstrate liability but may be admissible for other purposes, such as proving bias or prejudice of a witness). We agree with the trial court that the dictum in Grillo should not be read expansively to mean that irrelevant or prejudicial material contained in a Mary Carter agreement always must be admitted into evidence. The more reasonable reading of Grillo, and a reading that is easily reconcilable with Holger, is that the existence and terms of a Mary Carter agreement may be relevant, but that the agreement’s admissibility is subject to the strictures otherwise placed on relevant evidence.

Although on appeal Key appears to be arguing that the entire agreement should have been admitted into evidence, in the trial court Key argued that the agreement should be admitted into evidence in a redacted form, deleting portions that Key believed were “self-serving,” i.e., recitals that Bocci and Edwards agreed that Key failed to reasonably warn physicians concerning Theo-Dur’s toxicity, that *536Edwards and Bocci agreed that Key was primarily responsible for Bocci’s injury, and that Edwards was not admitting any liability by entering into the agreement. Key did not, however, propose to redact the information concerning Edwards’s liability insurance from the agreement. As noted above, admission of evidence pertaining to Edwards’s liability insurance would have been error. Key argues on appeal that Bocci and Edwards had the “burden to object to the portions containing inadmissible evidence such as the amount of the settlement and insurance.” That is incorrect. Where a party attacks the exclusion of an exhibit that contains some irrelevant material, that party has “the burden of excising the irrelevant portions of the exhibit to preserve the claimed error.” Fazzolari v. Portland School Dist. No. 1J, 78 Or App 608, 614, 717 P2d 1210 (1986), aff'd on other grounds 306 Or 1 (1987). See also State v. Thomas, 149 Or App 557, 561, 945 P2d 1056 (1997) (if any part of an offer of proof was properly excluded, there is no reversible error); Smith v. White, 231 Or 425, 435, 372 P2d 483 (1962) (‘Where there is a single offer of proof which includes matter that is objectionable, the entire offer may be rejected.”). Given the nature of Key’s offer of proof, the trial court did not err in failing to admit the Mary Carter agreement, as redacted by Key, into evidence.

Key also argues that the trial court erred in limiting its cross-examination of Edwards about the Mary Carter agreement. On cross-examination, counsel for Key asked Edwards whether, in view of the Mary Carter agreement, he had any financial exposure in this case. Edwards replied that he did. Despite that answer, Key’s counsel asked: “And you do not have such exposure in this case, at all, do you, sir?” The court sustained the objection to that question. Key’s counsel then asked the court if he could “explore that,” and the court said that he could not. On appeal, Key argues that the extent of Edwards’s financial exposure under the Mary Carter agreement was relevant to demonstrate Edwards’s bias. Because Key did not make an offer of proof concerning that issue, we are able to infer from the record only that Key was not allowed to bring out on cross-examination the actual amount of the settlement and loan as set forth in the Mary *537Carter agreement, and Edwards’s insurer’s participation in the agreement.9

OEC 609 provides, in part:

“(1) The credibility of a witness may be attacked by evidence that the witness engaged in conduct or made statements showing bias or interest. * * *

“(2) If a witness fully admits the facts claimed to show the bias or interest of the witness, additional evidence of that bias or interest shall not be admitted. * *

Impeachment for bias or interest may be appropriate in a situation such as this, where Edwards not only had an interest in proving his own claim against Key, but the Mary Carter agreement gave him an interest in maximizing Bocci’s recovery against Key as well. However, for several reasons, we reject Key’s arguments that the trial court committed reversible error by limiting the cross-examination of Edwards.

A court has discretion to limit the cross-examination of a witness for bias, particularly in situations in which the information sought would be prejudicial. The information Key sought from Edwards on cross-examination apparently was the exact amount that Edwards’s malpractice insurer would pay should Bocci recover particular sums from Key. The potential for prejudice was significant because the information Key was trying to elicit could have distracted or confused the jury, which had been properly instructed not to consider that settlement when determining damages. Moreover, Key makes no argument that the insurance information was relevant to demonstrating Edwards’s bias. Cf. Holger, 316 Or at 415 (even a legally correct instruction concerning the effect of settlement could be error if it could distract jury from appropriate analysis); Johnson v. Hansen, 237 Or 1, 4, 389 P2d 330 (1964) (“In the ordinary case, the presence or absence of insurance is not only irrelevant, but the unnecessary injection of the subject into the trial is prejudicial.”).

*538The need for evidence demonstrating bias must be weighed against the danger of unfair prejudice. OEC 403. Key has failed to articulate how the cross-examination it wished to undertake would have demonstrated further bias than was already apparent from the nature of the Mary Carter agreement itself. Given the questions asked, it is clear that Key was attempting to show that Edwards’s malpractice insurer, rather than Edwards himself, had the only financial exposure in the case. Key does not explain how that fact somehow made Edwards more biased against Key and in favor of Bocci than would have been the case if Edwards had no malpractice insurance. In fact, the opposite would be true: If Edwards rather than his insurer were potentially responsible for paying for the harm to Bocci, then he would have all the more reason to be taking the position that Key caused Bocci’s harm. Given the doubtful relevance of this evidence, the trial court did not err in excluding the evidence in light of the danger of unfair prejudice.

For essentially the same reason, any error in failing to allow Key to pursue its cross-examination on this point undoubtedly was harmless error. As noted above, the jury knew of the existence of the Mary Carter agreement and knew that Bocci’s recovery against Key could affect how much Edwards would pay Bocci. Aside from the fact that the information that Key was apparently trying to solicit did not demonstrate bias, any possible bias would relate to Bocci’s claim; it would not relate to Edwards’s own claim against Key. The existence of Edwards’s claim against Key told the jury what it needed to know about Edwards’s “bias” in this case, at least as it pertains to Edwards’s own claims against Key which are the only claims at issue on appeal. To the extent that this evidence could be said to have any marginal relevance whatsoever, it would be relevant only to Edwards’s bias in favor of Bocci’s claims, not to Edwards’s bias in favor of his own claims. Evidential error is not presumed to be prejudicial. OEC 103(1). We are unable to conclude that the trial comb’s limitation on Key’s cross-examination of Edwards, if error at all, was reversible error as to Edwards’s claim against Key.

The trial comb did not err in limiting Key’s cross-examination concerning the amount of the settlement and *539Edwards’s malpractice insurer’s participation in the Mary Carter agreement.

Jury Deliberations

After the verdict was received, the trial court notified the parties by letter that a copy of the current issue of Prevention magazine was found in the jury room and that it contained an article entitled “Get Off the Asthma Tightrope.” The author of that article advocated asthma control via inhaled steroids as safer than oral medications such as theophylline. The article did not discuss Theo-Dur or drug interactions. Key moved to have the court recall the jurors and examine each one under oath concerning that magazine. The court denied Key’s motion, as well as Key’s motion to set aside the verdict and motion for a new trial. The trial court held that Key had failed to make an adequate showing of juror misconduct. The court found that, even if one or more of the jurors had read the article in question, no misconduct had occurred, and there was nothing in the article that was pivotal to a fair resolution of the case: “The Prevention article could not possibly have affected the outcome of the case.”

On appeal, Key argues that the trial court abused its discretion in refusing to recall the jury to interrogate them about the magazine and in denying its motion for a new trial. For the following reasons, we disagree that the court abused its discretion.

UTCR 3.120 provides, in part:

“(2) After a sufficient showing to the court and on order of the court, a party may have contact with a juror in the presence of the court and opposing parties when:

“* * * * *

“(b) there is a reasonable ground to believe that a juror or the jury has been guilty of fraud or misconduct sufficient to justify setting aside or modifying the verdict or judgment.”

“There is a strong policy in Oregon to protect jury verdicts from attack, and courts are hesitant to interrogate jurors after they have reached a verdict in order to probe for potential misconduct.” Koennecke v. State of Oregon, 122 Or *540App 100, 103, 857 P2d 148 (1993). See also Erstgaard v. Beard, 310 Or 486, 498, 800 P2d 759 (1990) (same). In the present case, not only is there no evidence of actual jury misconduct, but there is virtually no potential for misconduct. The jurors were not instructed to avoid bringing reading materials into the jury room and were not forbidden to read articles such as an article on asthma in the current issue of a popular magazine. The court did not abuse its discretion in denying Key’s motion to interrogate the jurors.

The same is true for the court’s denial of Key’s motion for a new trial. “Only the clearest kind of juror misconduct can trigger the trial court’s right to exercise its discretion and order a new trial.” Erstgaard, 310 Or at 497. A comparison of the present case and Erstgaard is instructive. In Erstgaard, a malpractice action, juror Barrett stated during voir dire that she had been treated by defendant Dr. Allen, that she had stopped seeing Allen due to a change in insurance, and that she could be objective in deciding the case. Id. at 490. After trial, several other jurors complained that Barrett had stated that Allen had saved her niece’s life, and had stated that finding the doctor negligent would ruin the doctor’s reputation. Id. at 492-93. The Oregon Supreme Court found that the trial court had abused its discretion in granting a new trial under those circumstances:

“In the relatively few cases in which this court has either permitted or required a new trial for juror misconduct that occurred during the deliberating process, we have found none in which the misconduct consisted solely of juror argument. All the cases have involved specific acts by jurors designed (and later claimed, either explicitly or implicitly) by the particular offending jurors to give them special knowledge concerning one of the disputed facts in the case then under consideration. See, e.g., Sanders v. Curry County, 253 Or 578, 456 P2d 493 (1969) (unauthorized inspection of premises); Wolfe v. Union Pacific R. Co., [230 Or 119, 368 P2d 622 (1962)] (unauthorized visit to accident scene); Thomas v. Dad’s Root Beer, Etc., 225 Or 166, 356 P2d 418, 357 P2d 418 (1960) (view of accident scene and unauthorized experiment); Eckel v. Breeze, 221 Or 572, 577, 352 P2d 460 (1960) (view of scene); Schneider v. Moe, 151 Or 353, 50 P2d 577 (1935) (view of accident scene). Barrett’s actions were different. She did not obtain new information *541relating to Allen’s care for the plaintiff child. She simply disclosed the basis of her pre-existing bias. That is argument, not superior knowledge of a pivotal fact concerning some issue in the case actually being decided by the jury.” Id. at 497-98.

The present case presents significantly fewer reasons for granting a new trial. In Erstgaard, the court recognized that juror Barrett’s remarks were inappropriate, and “of doubtful merit.” Id. at 497. Here, all that can be said is that a juror may have carried a current issue of a popular magazine into the jury room. No misconduct has been demonstrated. The magazine contained an article by a doctor who advocated treating asthma by careful monitoring of breath flow and the use of inhaled steroids. That article, even if read by any of the jurors, could not have provided “superior knowledge of a pivotal fact concerning some issue in the case actually being decided by the jury.” Id. at 498. The trial court did not abuse its discretion in denying a new trial.10

Motion for Directed Verdict

Key next argues that the trial court erred in failing to direct a verdict in its favor on Edwards’s negligence and fraud cross-claims. We consider only those arguments that Key preserved in the trial court and presented properly on appeal. ORAP 5.45.11

*542Key acknowledges that, under Oksenholt v. Lederle Labs, 294 Or 213, 215, 656 P2d 293 (1982), a physician can “maintain an action for misrepresentation and negligence against a prescription drug manufacturer that misrepresents information about its drug to the doctor[.]” In Oksenholt, the plaintiff, a physician, relied on information provided by the defendant drug manufacturer and prescribed a drug that caused a patient to go blind. Key seeks to distinguish Oksenholt on the ground that the physician in that case actually prescribed the drug involved, whereas Edwards did not prescribe Theo-Dur to Bocci. Nothing in Oksenholt indicates that such a distinction would defeat a physician’s claim such as this, where the physician alleges that the manufacturer’s negligence and fraud caused the physician to misdiagnose a condition caused by the manufacturer’s drug.

Key further argues that the trial court erred in denying its motion for a directed verdict on the cross-claim because it satisfied its duty to warn and because Edwards failed to demonstrate causation. Key points out that Edwards did not prescribe the Theo-Dur and that he did not know that Bocci had been taking ciprofloxacin. Key therefore concludes that its failure to warn of the drug interaction had no causal connection to Bocci’s injuries. We disagree. There is evidence that Edwards considered a diagnosis of theophylline toxicity when he examined Bocci. There also is evidence from which a jury could infer that the reason Edwards did not correctly diagnose Bocci’s condition of theophylline toxicity is because he relied on Key’s representations that Theo-Dur was “safe” because it had “zero-order absorption.” The trial court properly denied Key’s motion for a directed verdict on Edwards’s negligence cross-claim.

Concerning Edwards’s fraud claim, Key asserts that, as a matter of law, Edwards had no right to rely on its representations about the safety of Theo-Dur, because Edwards had a duty to exercise his own professional judgment in the treatment of his patients. The only case on which Key relies for its assertion is McEwen v. Ortho Pharmaceutical, 270 Or 375, 528 P2d 522 (1974). McEwen, however, leads to the *543opposite conclusion. In McEwen, the plaintiff sued the manufacturers of oral contraceptives that caused blindness. The drug manufacturers had a duty to warn prescribing and treating physicians of dangerous side effects produced by the drugs. “The warning should be sufficient to apprise the general practitioner as well as the unusually sophisticated medical man of the dangerous propensities of the drug. * * * In short, it is incumbent upon the manufacturer to bring the warning home to the doctor.” Id. at 388 (citations and internal quotation marks omitted). While McEwen involved neither a claim by a physician against a pharmaceutical company nor a claim for fraud, its rationale supports Edwards’s position that Key violated its duty to warn physicians adequately about the dangerous propensities of Theo-Dur. McEwen does not support Key’s contention that the trial court should have granted its motion for a directed verdict on Edwards’s claims.

Damages

Key assigns as error the trial court’s decision to allow Edwards to amend his complaint at trial to increase his prayer for punitive damages. We review a trial court’s decision to permit amendments to pleadings for abuse of discretion. Key asserts that the trial court abused its discretion in allowing the amendment.12 The motion to amend was based on evidence discovered on the first day of trial, i.e., a notice of adverse findings from the FDA, indicating that Key “has disseminated false and misleading labeling” for Theo-Dur. That evidence had been sought earlier through discovery but was not produced until trial. Motions for leave to amend are to be “freely given when justice so requires.” ORCP 23 A. A court has “ample discretionary authority to allow amendments, provided the proffered amendment ~ues not substantially change the cause of action or interject an entire new element of damage.” Cutsforth v. Kinzua Corp., 267 Or 423, 433, 517 P2d 640 (1973) (citation omitted). Edwards already had prayed for punitive damages, and the amendment merely *544modified the amounts prayed for. The trial court did not abuse its discretion by allowing the amendment.

Key also assigns error to the trial court’s admission of the FDA notice of adverse findings mentioned above, on the ground that it is not relevant. We reject Key’s argument without discussion.

Key next argues that the trial court should have dismissed Edwards’s punitive damages claim on the ground that Edwards failed to present sufficient evidence to create a jury question as to whether punitive damages were recoverable in this case.

ORS 30.927 provides:

“(1) Where a drug allegedly caused the plaintiff harm, the manufacturer of the drug shall not be liable for punitive damages if the drug product alleged to have caused the harm:

“(a) Was manufactured and labeled in relevant and material respects in accordance with the terms of an approval or license issued by the Federal Food and Drug Administration under the Federal Food, Drug and Cosmetic Act of the Public Health Service Act; or

“(b) Is generally recognized as safe and effective pursuant to conditions established by the Federal Food and Drug administration and applicable regulations, including packaging and labeling regulations.

“(2) Subsection (1) of this section does not apply if the plaintiff proves [by clear and convincing evidence] that the defendant, either before or after making the drug available for public use, knowingly in violation of applicable Federal Food and Drug Administration regulations withheld from or misrepresented to the agency or prescribing physician information known to be material and relevant to the harm which the plaintiff allegedly suffered.”

Key argues that Edwards failed to prove that it had withheld from or misrepresented to the FDA information that was known to be material and relevant to the harms Bocci and Edwards suffered.13 As noted above, an FDA notice *545that Key “has disseminated false and misleading labeling” for Theo-Dur, was introduced into evidence. An FDA finding that Theo-Dur’s labeling was misleading, particularly considered in conjunction with the evidence concerning what Key knew about theophylline toxicity, is sufficient to create a jury question as to whether Key “knowingly in violation of applicable Federal Food and Drug Administration regulations withheld from or misrepresented to the agency * * * information known to be material and relevant to the harm which the plaintiff allegedly suffered.” ORS 30.927(2). Nonetheless, Key argues that the information in question — its false claims concerning zero-order absorption — were not “material and relevant to the harm which plaintiff allegedly suffered,” because Bocci’s harm was caused by a theophylline-ciprofloxacin interaction of which Edwards was not aware. We reject Key’s argument. Bocci’s and Edwards’s harms were caused by Edwards’s failure to diagnose theophylline toxicity, which in turn was caused by Key’s promotion of Theo-Dur as a “safe” drug with “zero-order absorption.” That type of promotion of Theo-Dur was the subject of the FDA letter of adverse findings. The trial court did not err in allowing this issue to be decided by the jury.

Key also asserts that the jury should not have been permitted to consider punitive damages pursuant to ORS 30.927(2), because Edwards was not the “prescribing physician.” We need not address whether the legislature intended the term “prescribing physician” to cover physicians in the position of Edwards, i.e., those who treat patients who are taking drugs prescribed by other physicians, because, as noted above, the jury found clear and convincing evidence that Key had withheld relevant information from the FDA. Edwards was not required to prove both that the FDA had been misled and that a “prescribing physician” had been misled. Proof on the FDA question was sufficient for purposes of ORS 30.927(2).

Key next argues that the punitive damage award in this case violates the Due Process Clause of the Fourteenth Amendment under the standard set forth by the United States Supreme Court in Honda Motor Co. v. Oberg, 512 US 415, 114 S Ct 2331, 129 L Ed 2d 336 (1995), and Pacific Mut. Life Ins. Co. v. Haslip, 499 US 1, 111 S Ct 1032, 113 L Ed 2d *5461 (1991). Key first argues that ORS 30.925, ORS 30.927, and the Oregon Supreme Court’s decision on remand in Oberg do not provide sufficiently specific standards by which the propriety of a jury punitive damage award may be reviewed.14 We reject Key’s argument for several reasons.

In its post-trial motion for remittitur, Key argued that the trial court’s “review of the punitive damage awards should be guided by the criteria in ORS 30.925, which resemble those in Haslip[.] Under Oberg, courts review the jury’s punitive damage awards to determine if they are “within the range that a rational juror would be entitled to award in the light of the record as a whole,” depending on the “statutory and common law factors that allow an award of punitive damages for the specific kind of claim at issue.” Oberg v. Honda Motor Co., 320 Or 544, 549, 888 P2d 8 (1995).

On appeal, Key contends that judicial review in the light of statutory and common-law factors pertaining to punitive damages may satisfy due process in some cases, but in other cases it will not. However, Key does not suggest anything that a reviewing court should review for other than statutory and common-law factors, and we are unable to discern any other factors that the United States Supreme Court has deemed to be required in order to satisfy due process. See Honda Motor Co., 114 S Ct at 2338-42, 129 L Ed 2d at 343-50 *547(discussing factors relevant to review of punitive damage awards). Haslip, 499 US at 20-22 (same). Given that Key urged the trial court to review the jury’s punitive damage awards to both Bocci and Edwards based on the statutory factors set forth in ORS 30.925, Key’s position on appeal is untenable. Key argues for the first time on appeal that the trial court erred in failing to review the punitive damage award to Edwards independently of the punitive damage award to Bocci, because the statutory criteria from ORS 30.925 apply to strict liability punitive damage awards, and Edwards’s claims involved negligence and fraud. Not only did Key fail to preserve this issue in the trial court, but it waived any argument on this point by specifically requesting the court to review the punitive damage awards in their entirety under the criteria listed in ORS 30.925.

Finally, Key argues that the trial court erred in denying its post-verdict motion for remittitur or a new trial on the ground that the punitive damage award was unconstitutionally excessive. As noted above, Key asked the court to review the punitive damages in light of the factors listed in ORS 30.925, arguing that the jury improperly relied exclusively on one of those factors — the financial condition of the defendant — and that the large compensatory damage awards satisfied the goal of deterrence. Key argues that the punitive damage award was unconstitutionally excessive because it bears no relationship to Key’s culpability as compared with Edwards’s culpability. Key points out that Edwards failed to diagnose Bocci’s theophylline toxicity. That fact does not assist Key here. The jury found that Edwards was not at fault for failing to diagnose theophylline toxicity under the circumstances. The primary focus of punitive damages in this context is the misconduct of the defendant against whom those damages are being assessed. We reject Key’s argument that the punitive damage award was unconstitutionally excessive because Edwards was somehow culpable despite the jury’s explicit conclusion otherwise.

Judgment in favor of cross-claim plaintiff - respondent is affirmed.

dissenting.

The trial court erred in resolving the parties’ disputes about the Mary Carter agreement in three respects. First, the trial court should have dismissed Edwards from the case, as the settlement rendered plaintiffs claims against him nonjusticiable. Maintaining Edwards in the case distorted the very nature of the trial and substantially prejudiced Key Pharmaceuticals, Inc. (Key) in the process. Second, the trial court should have admitted the redacted agreement itself to inform the jury that Edwards — nominally a defendant — actually had every interest in seeing plaintiff prevail against Key. Third, at the very least, the trial court should have permitted cross-examination of Edwards concerning the nature of the agreement, so that the jury could understand the extent of his bias and interest. The concurrence reaches a contrary conclusion with respect to each of those three matters. I disagree and, on those bases, respectfully dissent.

Before trial, plaintiff, Edwards and Edwards’ss insurer entered into a “Covenant Not-to-Enforce-Judgment and Separate Loan Agreement.” Under the terms of that agreement, Edwards agreed to pay plaintiff $200,000 up front and to loan plaintiff an additional $800,000. Plaintiff agreed not to enforce any judgment against Edwards arising out of the pending claims. He also agreed that, if he obtained judgment against Key in excess of $3 million, he would repay Edwards the $800,000 loan. If plaintiff recovered a smaller amount from Key, the agreement provided for a smaller return of the loaned amount. The agreement is, as the *549concurrence correctly observes, a typical Mary Carter agreement.

Mary Carter agreements are not new, having been a regular feature of the litigation landscape since at least the Florida appellate court decision from which the name was derived in 1967. Booth v. Mary Carter Paint Co., 202 So 2d 8 (Fla Dist Ct App 1967).1 They are, however, quite controversial. From the beginning, they have been subject to criticism that they are champertous, that they mislead juries, promote unethical collusion, and create the potential for an allocation of a larger share of liability to less culpable — but deeper pocketed — defendants, thus frustrating the operation of equitable contribution rules. In particular, courts and commentators around the nation have expressed concern about the effect of Mary Carter agreements on the integrity of the adversarial process.2

Mary Carter agreements create the potential to distort the adversarial process in a number of ways, by virtue of *550their creation of a substantial incentive for the settling defendant to cooperate fully with the plaintiff to shift liability to the remaining defendants, all the while appearing to the jury as a codefendant. The Mary Carter defendant’s cooperation with the plaintiff may take the form of contributing discovery materials to the plaintiff, cooperating during voir dire by using peremptory challenges allocated among the defendants for the benefit of the plaintiff, permitting the settling defendant to be subject to leading direct examination as if the witness were hostile to the plaintiff, providing the plaintiff with essentially two opportunities to cross-examine the defendants’ witnesses, and permitting the plaintiff to reap the benefit of additional argument in support of his or her cause. After engaging in the support of the plaintiffs efforts, Mary Carter defendants often then will be dismissed from the case, further demonstrating the essentially sham nature of their participation in the trial.

Because of such problems, some courts have declared that Mary Carter agreements simply are void as against public policy. In Elbaor v. Smith, 845 SW2d 240, 250 (Tex 1992), for example, the Supreme Court of Texas held that such agreements are per se invalid:

“Remedial measures cannot overcome nor sufficiently alleviate the malignant effects that Mary Carter agreements inflict upon our adversarial system. No persuasive public policy justifies them, and they are not legitimized simply because this practice may continue in the absence of these agreements. The Mary Carter agreement is simply an unwise and champertous device that has failed to achieve its intended purpose.”

Id. at 249. The Nevada Supreme Court reached the same conclusion, but by the rationale that Mary Carter agreements are essentially champertous, because they permit a person not a party to the lawsuit to retain a financial interest in its outcome. Lum v. Stinnett, 488 P2d 347, 351 (Nev 1971). See also Dosdourian v. Carsten, 624 So 2d 241, 243-46 (Fla 1993) (“We are convinced that the only effective way to eliminate the sinister influence of Mary Carter agreements is to outlaw their use.”); In re Guardianship of Babb, 642 NE2d 1195, *5511204-07 (Ill 1994) (Mary Carter agreements violate contribution statute); Trampe, 252 NW at 678 (undisclosed agreement “clearly imposed a fictitious suit upon the court” resulting “in the trial of issues which were not real”).

Other courts have concluded that Maxy Carter agreements are void unless the settling defendant is dismissed from the lawsuit or realigned as a plaintiff. The Oklahoma Supreme Court, for example, has held that:

“If the agreement does not absolutely settle the conflict, but rather hinges on the amount of the verdict [against non-settling defendants], the trial court should review the circumstances of the agreement and either hold that portion of the agreement granting agreeing defendant an interest in a large plaintiffs verdict unenforceable as against public policy, or dismiss the agreeing defendant from the suit. In no circumstances should a defendant who will profit from a large plaintiffs verdict be allowed to remain in the suit as an ostensible defendant.”

Cox v. Kelsey-Hayes Co., 594 P2d 354, 359 (Okla 1978) (emphasis in original).

Still others have held that, because of the potential for variation in the exact terms of Mary Carter agreements and in their effect on the litigation process, their validity must be examined on a case-by-case basis. The California Supreme Court’s decision in Abbott Ford, Inc. v. Superior Court, 741 P2d 124, 132 (Cal 1987) (en banc) is illustrative:

“[A] broad ruling on the inherent validity or invalidity of sliding scale agreements ‘in general’ is inappropriate because such agreements may have a variety of effects at different stages of the litigation process — discovery, settlement, trial or appeal. The potential problems posed by a particular provision in such an agreement may call for one remedy — e.g., disclosure of the agreement to the nonagreeing parties or to the jury — in one context, and another remedy — e.g., invalidation of a specific provision, or the agreement as a whole — in a different context. Thus, analysis requires close attention to the specific provisions of the agreement itself, the factual setting in which the agreement is entered into, and the agreement’s effect on the particular aspect of the judicial process at issue.”

*552A majority of those courts addressing the matter have held, guardedly, that Mary Carter agreements are not invalid per se. As the Maryland Court of Appeals observed, “Nearly every court which has considered the matter has recognized that agreements such as that in the case at bar have a potential for skewing the posture of the trial[.]” General Motors Corp. v. Lahocki, 410 A2d 1039, 1046 (Md 1980). In recognition of the distortive effects of the agreements, a majority of the courts hold that the agreements are subject to disclosure to the parties, to the court, and to the jury. Bohna v. Hughes, Thorsness, Gantz, et al., 828 P2d 745, 757 n 30 (Alaska 1992) (juries should be informed when settlements change the normal interests of a party); Sequoia Mfg. Co., Inc. v. Halec Const. Co., 570 P2d 782, 792-95 (Ariz Ct App 1977) (agreement must be disclosed to the court and remaining parties at the earliest opportunity; trial court has discretion to present the agreement to the jury); Firestone Tire & Rubber Co. v. Little, 639 SW2d 726, 728 (Ark 1982) (agreement must be disclosed to the nonagreeing party and may be admitted into evidence); Gatto v. Walgreen Drug Company, 337 NE2d 23, 29 (Ill 1975), cert den sub nom Gatto, Administrator v. Calumet Flexicore Corp. et al., 425 US 936 (1975) (settlement agreement must be disclosed to the court and opposing parties; disclosure to the jury might protect the integrity of the proceedings); Ratterree v. Bartlett, 707 P2d 1063, 1074-76 (Kan 1985) (agreement must be disclosed to court, and the general terms of the financial interest of a settling defendant in the outcome of the case should be disclosed to the jury); Lahocki, 410 A2d at 1046-47 (agreement must be disclosed to the court and opposing parties and admitted into evidence, unless there is danger of self-serving statements, in which case a statement of the terms of the settlement must be presented to the jury); Johnson v. Moberg, 334 NW2d 411, 415 (Minn 1983) (agreement must be disclosed promptly to court and remaining parties); Hegarty v. Campbell Soup Co., 335 NW2d 758, 765 (Neb 1983) (agreement must be disclosed upon proper motion and admitted into evidence); Cox, 594 P2d at 360 (pretrial agreement must be disclosed, and trial court erred in not admitting at least part of agreement into evidence); Hatfield v. Continental Imports, Inc., 610 A2d 446, 451-52 (Pa 1992) (agreement or at least the existence of the reason for the potential bias must be conveyed to the jury); *553Poston by Poston v. Barnes, 363 SE2d 888, 890 (SC 1987) (agreement must be disclosed to the jury); Corn Exchange Bank v. Tri-State Livestock, 368 NW2d 596, 600 (SD 1985) (if an agreeing defendant stands to gain financially from a plaintiffs verdict, jury must be informed of the contents of the agreement); Slusher v. Ospital by Ospital, 777 P2d 437, 444 (Utah 1989) (agreement must be disclosed to court and parties, then court should disclose the existence and basic contents of agreement to jury, which may include admission of the agreement into evidence); State ex rel Vapor Corp. v. Narick, 320 SE2d 345, 348 (W Va 1984) (agreement must be promptly disclosed to court and opposing counsel so that trial court can decide whether agreement is valid and determine what measures should be taken to insure that the nonsettling party will not be prejudiced).

Oregon stands with the majority. In Grillo v. Burke’s Paint Co., 275 Or 421, 551 P2d 449 (1976), the court acknowledged the criticisms of Mary Carter agreements in general— that they “distort[] the relationship between plaintiffs and defendants, resulting in a non-adversary and possibly collusive proceeding[.]” Id. at 426. The court also noted that

“the majority of jurisdictions which have considered this question do not condemn the agreements as invalid per se but instead require that the agreements be subject to pretrial discovery procedure and be admissible into evidence on request of any non-settling defendant.”

Id. After reviewing those decisions, the court expressly aligned itself with them and concluded that, although the Mary Carter agreement in that case was not per se invalid, “it would have been subject to pretrial discovery and, upon request of [the] defendant, would have been admissible in evidence.” Id. at 427.

My point in setting forth a summary of the case law is this: It is well-nigh universally recognized that Mary Carter agreements are dangerous. Even in jurisdictions declining to rule on the validity of the agreements per se, courts are careful to explain that parties, the court, and jurors should be fully informed of the agreements to avoid the distorting effects that they commonly have upon the adversarial process.

*554With that in mind, I turn to the manner in which the agreement was handled in this case. There is no question that plaintiff informed Key of the existence of the Mary Carter agreement and provided Key with a copy of it. Key then moved to have the agreement itself admitted into evidence, citing Grillo. The trial court acknowledged Grillo and its holding that Mary Carter agreements are admissible upon request. The court nevertheless denied Key’s motion on the ground that Grillo had been implicitly limited by a subsequent decision of the Oregon Supreme Court, Holger v. Irish, 316 Or 402, 851 P2d 1122 (1993). The trial court characterized the Holger decision as holding that, unless there is an overriding or compelling reason, the jury is not to be told that there has been a settlement.” The trial court, however, did agree to deliver a preliminary statement to the jury pool very briefly summarizing the agreement as follows:

“The settlement agreement contains a schedule for repayment, however, to Dr. Edwards by the plaintiff if the plaintiff is awarded a verdict against the defendant Key Pharmaceuticals Company. So there is that relationship that you need to be aware of, but the jury that is selected in this case is to consider the fact of the settlement only as it might bear on the issue of credibility or believability of the witnesses who testify, and it is not to be considered in any way in determining the amount of the verdict or damages, if any, that the jury should award at the end of the case.”

Key also moved to dismiss Edwards from the case. According to Key, because of the settlement agreement, Edwards no longer stood in an adversarial relationship with respect to plaintiff, and plaintiffs claims against him were nonjusticiable. Key complained that, by virtue of his remaining in the case in the posture of a defendant, Edwards would prejudice its ability to obtain a fair trial, among other things, by unfairly consuming peremptory challenges ostensibly allocated to the defense, by permitting plaintiff to conduct direct examination of Edwards by leading questions as if he were a hostile witness, and by effectively awarding plaintiff the strategic advantage of having additional objections, argument, and cross-examination. The trial court denied the motions.

*555At trial, Edwards in fact participated as a “stealth plaintiff.” He participated in jury voir dire3 opening statement and examination of witnesses ostensibly as a defendant, but in reality as a coplaintiff. For example, plaintiff was permitted to call Edwards as a “hostile” witness, and conducted the direct examination by extensive leading questions as to Edwards’s own blamelessness and Key’s culpability. When Key attempted to cross-examine Edwards about the true nature of his interest in the litigation, the trial court precluded Key from pursuing the matter. When Key asked Edwards if he had “exposure in this case, at all,” the trial court sustained plaintiffs objection, commenting only: “There is financial exposure. Sustained.” Key asked: “Can I explore that, your honor?” The trial court said, “No, you may not.” Then, at the end of trial — and true to the pattern of Mary Carter cases — plaintiff voluntarily dismissed Edwards from the case. The trial court informed the jury only that:

“[P]laintiff s claims against Dr. Edwards have been withdrawn and you are not to concern — you are not to concern yourself with any reasons about why the court has withdrawn those allegations.”

In its instructions to the jury, the court ultimately gave a slightly more complete summary of the Mary Carter agreement:

“Under the terms of the agreement, defendants Dr. Edwards and the Clinic paid a certain amount of money to Plaintiff. In return, Plaintiff agreed not to enforce any judgment that might be entered in this lawsuit against Dr. Edwards or the Eugene Clinic.

“* * * * *

“All or none or a portion of the money paid by Dr. Edwards and the Clinic to Plaintiff will be repaid, depending upon the amount of any recovery that Plaintiff obtains from the non-agreeing defendants.”

*556The first of the trial court’s mistakes with respect to the Mary Carter agreement was its denial of Key’s motion to dismiss plaintiffs claims against Edwards on the ground that they no longer presented a justiciable controversy. As the Supreme Court explained in Brumnett v. PSRB, 315 Or 402, 405, 848 P2d 1194 (1993), a justiciable controversy exists when “the interests of the parties to the action are adverse” and “the court’s decision in the matter will have some practical effect on the rights of the parties to the controversy.” In this case, Edwards and plaintiff settled their differences and completely realigned their interests. If plaintiff established at trial that Edwards was negligent, the percentage of fault that the jury assigned to Edwards would only lessen any recovery from Key. Thus, plaintiffs sole objective at trial was to establish Key’s — not Edwards’s — liability. Similarly, Edwards’s only interest at trial was in establishing Key’s liability, for if the jury awarded more than $3 million against Key, Edwards stood to reap the return of $800,000 from plaintiff. That this complete coincidence of interests existed is nowhere better illustrated than in plaintiffs direct examination of Edwards. Called as an adverse witness, Edwards was not subjected to a single adverse question. The entire focus of the supposedly hostile examination was the culpability of Key, not of Edwards. The fact that plaintiff had no real interest in pursuing a verdict against Edwards — and, indeed, had every interest in preventing the jury from finding against Edwards and thus threatening the potential recovery from Key — is further borne out by the fact that, at the close of the evidence, plaintiff dismissed Edwards from the case.4 Plaintiff and Edwards were not, in any sense of the term, adverse, and their alignment as plaintiff and defendant was completely illusory.

The concurrence insists that the case against Edwards was justiciable, because the complaint alleges that the parties are adverse and because a jury verdict against *557Edwards hypothetically could have affected his liability to plaintiff. As to the first assertion, the concurrence is simply incorrect. Parties cannot create adversity by declaration when there is none in reality. In Oregon Medical Association v. Rawls, 281 Or 293, 295, 574 P2d 1103 (1978), for example, the plaintiff and the defendant sought a declaratory judgment as to the constitutionality of a statute. The complaint listed the parties as adverse, but they took identical positions as to the constitutionality of the challenged law. On its own motion, the Supreme Court dismissed the appeal, because the parties, in fact, were not adverse concerning the issue before the court. Id. at 299.

As to the second assertion, the concurrence certainly is correct that it is hypothetically possible for a jury verdict against Edwards to have affected Edwards’s ultimate liability to plaintiff. The jury — at least before Edwards was dismissed — conceivably could have awarded $3 million in damages but allocated some portion of fault to Edwards. In that event, the verdict against Key would have been insufficient to trigger plaintiffs obligation to return some or all of the $800,000 loan to Edwards. I think it is debatable whether such a hypothetical contingency is sufficient to satisfy the requirements of justiciability. Courts in other jurisdictions have questioned whether a case is justiciable when a Mary Carter agreement has determined the limits of the settling defendant’s liability. See, e.g., Shelby v. Keck, 541 P2d 365, 370 (Wash 1975) (en banc); Schell v. Albrecht, 383 NE2d 15, 17-19 (Ill App Ct 1978). The concurrence fairly relies on dictum in Stephens v. Bohlman, 138 Or App 381, 385, 909 P2d 208, rev dismissed 324 Or 177, 925 P2d 907 (1996), for a contrary conclusion. We do not have to resolve that matter, however, because the two requirements of justiciability are conjunctive. Thus, whether or not a jury verdict hypothetically could have affected Edwards’s liability, the fact remains that plaintiff and Edwards were not adverse, and the justiciability of plaintiffs claims against Edwards fail on that ground alone.

The concurrence suggests that, even if the claims against Edwards were not justiciable, any error was harmless, because Edwards had asserted a cross-claim against Key and would have remained in the case in any event. Of *558course, the concurrence glosses over the fact that it was Edwards’s status as a named defendant that permitted him to maintain the cross-claim in the first place. Indeed, the Mary Carter agreement itself recites that one reason for the agreement was to allow Edwards to pursue a cross-claim against defendant. If plaintiffs claim against Edwards had been dismissed, as it should have been at the beginning of trial, then Edwards would have been required to file a separate claim against Key and then move for consolidation of the two separate actions. Whether the trial court would have granted such a motion is pure speculation, and such speculation cannot form the basis for establishing the justiciability of a controversy. Brumnett, 315 Or at 406-07.

Aside from that, the concurrence misses the point as to the nature of the harm that flowed from failing to dismiss the claim against Edwards. The point is not that Edwards remained in the trial, but that he remained in the trial as a defendant, with all the advantageous consequences as to tactics, voir dire, witness examination, and argument that such alignment affords. To be sure, the harm that flowed from the misalignment of the parties is most clear with respect to plaintiffs case against Key. Plaintiff was allowed to try much of his case through Edwards, whose nominal status as a defendant conferred benefits on plaintiff that he otherwise would not have enjoyed. But that misalignment infected Edwards’s cross-claim against Key, as well. The concurrence, in fact, concedes that the same procedural advantages that plaintiff obtained from the Mary Carter agreement inured to Edwards on his cross-claim. 158 Or App at 528. In my view, the trial court should have dismissed plaintiffs claims against Edwards at the beginning of trial, and I would reverse on that basis alone.

The trial court’s second mistake with respect to the Mary Carter agreement was its failure to admit the agreement itself into evidence. Key asked that the agreement be admitted to show Edwards’s bias and interest in the outcome, and for the purpose of impeachment on his claim damages for emotional distress arising out of the filing of plaintiffs lawsuit against him. Key contended that the Mary Carter agreement — which was negotiated before the filing of the lawsuit and which expressly advised Edwards that plaintiff would be *559filing claims against him — demonstrated Edwards’s lack of credibility as to his emotional distress.

Grillo plainly requires the admission of the agreement itself. The Supreme Court clearly held that the agreement is “admissible in evidence.” Grillo, 275 Or at 427. A blenderized summary of such agreements is not something that may be admitted “in evidence.” Nor is anything in Holger to the contrary. In that case, the Supreme Court said nothing about Mary Carter agreements or about its earlier decision in Grillo. The court held merely that, under OEC 408, “evidence of a compromise is not admissible unless it is offered for a purpose having independent relevance.” Holger, 316 Or at 414 (emphasis added). Mary Carter agreements always are independently relevant, given the fact that they are direct evidence of the settling defendant’s substantial interest in seeing the plaintiff prevail, notwithstanding that defendant’s alignment as a nominal defendant. OEC 408, in fact, expressly provides that evidence of compromise or offers of compromise is admissible “when the evidence is offered for another purpose, such as proving bias or prejudice of a witness.” Thus, there is no “tension” between the two decisions, as the trial court and the concurrence suggest.

Even assuming, for the sake of argument, that a summary of the contents of a Mary Carter agreement could suffice under Grillo, the fact remains that the summary the court gave in this case was inadequate. Before trial, all that the court told the jury was that there was a settlement, which contained a “schedule of repayment * * * if the plaintiff is awarded a verdict against the defendant Key Pharmaceuticals Company,” and that there was, as a result, a “relationship that you need to be aware of.” I fail to see how that abbreviated — and inaccurate — summary provided the jury sufficient information about the nature of the terms of the agreement to enable it to appreciate fully the extent to which Edwards’s interests were aligned with plaintiffs.

I note that not even the concurrence suggests that the preliminary instruction sufficed. The concurrence maintains that any deficiencies in the initial overview of the agreement were remedied by the court’s more detailed, post-trial instruction to the jury. I respectfully disagree. To begin *560with, I find the trial court’s subsequent instruction to the jury only slightly more detailed than the initial one. In any event, whatever the merits of the post-trial instruction, the point is that, at that juncture, the damage was done, and the instruction was pointless. The purpose of the majority rule to which the court in Grillo adhered is to provide the jury full disclosure of the true relationship of the parties during the trial, so that the jury can evaluate the evidence as it is offered in the light of the bias that the Mary Carter agreements typically reveal. In this case, by the time the court delivered the post-trial instruction on which the majority relies, Edwards already had exercised his peremptories in voir dire, already had been subject to leading direct testimony in the guise of “hostile” examination by plaintiff, and already had been protected from cross-examination on the nature and extent of his bias in favor of plaintiff. The post-trial instruction was simply too little, too late.

The concurrence alternatively complains that, because the Mary Carter agreement contains references to Edwards’s insurance carrier, it would have been improper for the trial court to have admitted the entire agreement into evidence. In my view, the observation is somewhat beside the point. No one raised that argument at trial5 and — particularly given that the matter easily could have been remedied by redaction had plaintiff or Edwards raised the issue — I think it is inappropriate for us to dispose of Key’s arguments on appeal on that ground. Aside from that, I think it is fairly debatable whether the mention of insurance in this context is so clearly inappropriate as the concurrence assumes. OEC 411(1) provides that evidence of liability insurance is not admissible “upon the issue whether the person acted negligently or otherwise wrongfully.” That is not the purpose for which the Mary Carter agreement was offered. It was instead *561offered as evidence of bias and pecuniary interest in the outcome, and OEC 411(2) expressly permits evidence of insurance to show “bias, prejudice or motive of a witness.”

In short, the Mary Carter agreement itself was admissible. It was relevant to the issue of Edwards’s bias and interest. The trial court’s sole justification for excluding the document — its reading of the Holger decision — was incorrect.

The trial court’s final mistake concerning the Mary Carter agreement was its refusal to allow Key to cross-examine Edwards concerning his interest in seeing plaintiff prevail, notwithstanding his alignment as a codefendant and his direct examination as an ostensibly adverse witness. OEC 609-1 provides that “[t]he credibility of a witness may be attacked by evidence * * * showing bias or interest.” Consistent with that rule is the familiar refrain in the case law that “ ‘[i]t is always permissible to show the interest or bias of an adverse witness.’ ” State v. Hubbard, 297 Or 789, 796, 688 P2d 1311 (1984) (quoting Clevenger v. Schallhorn, 205 Or 209, 215, 286 P2d 651 (1955)). That is not to say that parties have unlimited leeway in establishing the bias or interest of adverse witnesses. As the Supreme Court explained in Hubbard, the trial courts have some discretion in limiting the extent of the inquiry. 297 Or at 798. That discretion itself, however, is subject to limits:

“The discretion of the trial judge to exclude evidence relevant to bias or interest only obtains once sufficient facts have been established from which the jury may infer that bias or interest. Typically, this would require wide latitude be given to the cross-examiner to ask and receive answers to questions sufficient to demonstrate to the jury the nature of the bias or interest of the witness.”

Id.

In this case, the trial court did not merely limit Key’s inquiry into the evidence establishing the nature and extent of Edwards’s bias. The court precluded such inquiry altogether. The transcript reveals the following colloquy:

“Q. (by Mr. Gamier, for Key): Is it your view that by virtue of the agreement that you have entered into with the *562plaintiff in this case you have any financial exposure of the outcome of this case?

“Mr. Holland (for Edwards): Your honor, I object. That flies in the face of the court’s ruling.

“The Court: Objection overruled.

“A. You’re asking if I have an understanding that I have financial exposure in this case?

“Q. Yes, sir.

“A. Yes, I do.

“Q. And you do not have such exposure in this case at all, do you, sir?

“Mr. Holland: I object.

“The Court: Sustained. There is financial exposure. Sustained.

“Mr. Gamier: Can I explore that, your honor?

“The Court: No, you may not.”

Thus, not a single question was permitted as to the facts revealing Edwards’s bias in this case. Under OEC 609 and Hubbard, that was reversible error.

The concurrence concludes that the trial court did not abuse its discretion in ruling on the objections to the cross-examination of Edwards because it allowed Key at least some inquiry into the fact of Edwards’s financial exposure. I disagree. The sole question that the court allowed Edwards to answer was whether he had an understanding as to whether he had any financial exposure in the case. I submit that permitting that single question as to Edwards’s understanding of the consequences of his agreement is not what the Supreme Court had in mind in requiring “wide latitude * * * to ask and receive answers to questions sufficient to demonstrate to the jury the nature of the bias or interest of the witness.” Hubbard, 297 Or at 798.

The concurrence contends that its review of the matter is hampered by Key’s failure to make an offer of proof as to what exactly it intended to ask Edwards, and that the only arguably preserved contention is that Key should have been *563allowed to inquire about the amount of Edwards’s financial stake in the outcome of the trial. It is true that Key did not make an offer of proof. But in light of the record of this case, I find that no great impediment.

OEC 103(1)(b) provides that an offer of proof is unnecessary when “the substance of the evidence [sought to be admitted] was made known to the court by offer or was apparent from the context within which questions were asked.” That is the case here. Key’s arguments in support of its efforts to get the Mary Carter agreement before the jury were voluminous in the extreme. Key raised the matter of admitting the agreement and permitting cross-examination on it before trial, during trial, and during cross-examination of Edwards. The parties actively debated the extent to which Key should be permitted to examine Edwards concerning his relationship to plaintiff in light of the agreement, and the court acknowledged in its comments during the cross-examination of Edwards both the scope of Key’s intended cross-examination and the fact that its prior ruling on the admissibility of the agreement “ties your hands pretty substantially in terms of cross examination.” In Davis v. O’Brien, 320 Or 729, 737, 891 P2d 1307 (1995), the Supreme Court held that pretrial contentions may be sufficient to preserve an error for appeal where the contentions were “presented clearly to the initial tribunal” and the “parties are not taken by surprise, misled, or denied opportunities” to meet the contentions. In light of the hundreds of pages of briefing and argument on the Mary Carter issues in this case, I find it impossible to believe that anyone was caught off guard as to Key’s intentions concerning its cross-examination of Edwards.

Even assuming, for the sake of argument, that the only matter that Key preserved was whether it could inquire into the amount of Edwards’s interests in the outcome of the case, I still conclude that the trial court erred in excluding the cross-examination. The amount of Edwards’s interest is directly relevant to the nature and extent of his bias. Parties routinely are allowed to examine expert witnesses as to the amount of their compensation to establish their bias. State v. Brown, 299 Or 143, 150, 699 P2d 1122 (1985) (amount of expert witness fees relevant to show bias). See also Laird C. *564Kirkpatrick, Oregon Evidence, 362 (3d ed 1996) (“Generally, the pecuniary interest of an expert witness and any bias of the expert in favor of the calling party can always be shown.”). I can envision no good reason to treat other witnesses differently.

The concurrence attempts to bolster its assessment of the trial court’s ruling by reference to OEC 403, reasoning that the trial court’s decision represents a sound effort to exclude unduly prejudicial information that was only of “marginal value” in establishing Edwards’s bias. To begin with, no party at trial or on appeal made an argument in support of the trial court under OEC 403. Aside from that, I disagree with the concurrence’s evaluation of the amount of Edwards’s direct pecuniary interest in the outcome of the trial. I think the jury would have been very interested to know that he stood to gain $800,000 if plaintiff recovered a sufficient verdict against Key, and I daresay that might have had something to do with the vigor with which plaintiff and Edwards attempted to exclude any reference to that fact from the jury’s knowledge. Further, I do not understand how the disclosure of the amount of Edwards’s financial interest in the outcome of the trial would be unduly prejudicial. The fact is that Edwards had a substantial financial stake in the outcome of the trial. That clearly is relevant. OEC 609. The concurrence suggests that the jury could have been “distracted or confused.” I do not understand why. The point of the cross-examination is simple and straightforward: If the jury returns a verdict against Key of over $3 million, Edwards reaps the return of $800,000 from plaintiff. I do not see what is the least bit confusing about that.

In short, the trial court erred in its handling of the Mary Carter agreement in three ways. Each of the court’s errors substantially prejudiced Key, and each would justify reversal. The concurrence errs in concluding otherwise.