Opinion
No. 20160198-CA
01-25-2018
KYLIE ANN BEDDOES LEE, Appellant, v. KENNETH L. WILLIAMS AND MOAB FAMILY MEDICINE PC, Appellees.
Tyler S. Young and Allen K. Young, Attorneys for Appellant Catherine M. Larson and Kathleen J. Abke, Attorneys for Appellees
Opinion Seventh District Court, Moab Department
The Honorable Lyle R. Anderson
No. 130700019 Tyler S. Young and Allen K. Young, Attorneys for Appellant Catherine M. Larson and Kathleen J. Abke, Attorneys for Appellees JUDGE RYAN M. HARRIS authored this Opinion, in which JUDGES KATE A. TOOMEY and JILL M. POHLMAN concurred. HARRIS, Judge: ¶1 During her first pregnancy, Kylie Lee sustained a permanent injury while under the care of Dr. Kenneth Williams. This injury occurred because Lee did not receive a particular medication during her pregnancy that almost certainly would have prevented the injury. As a result of the injury, all of Lee's future pregnancies now carry a substantial risk of birth defects or miscarriage. In fact, Lee's subsequent pregnancy came with serious complications. After the birth of her second child, Lee sued Williams for negligence. ¶2 Williams defended the case, in part, on statute of limitations grounds. During pretrial proceedings, the trial court narrowed the issues to be tried by ruling on summary judgment that, while genuine issues of material fact precluded a summary ruling regarding when Lee learned that she was injured, Lee knew no later than March 2009 that Williams may have been negligent. A few months later, after a bifurcated trial, the issue of when Lee discovered her injury was submitted to a jury. The jury found that Lee knew that she "might have sustained an injury" prior to September 2010, more than two years before she filed her complaint. The court then dismissed Lee's lawsuit on the grounds that it was not timely filed. ¶3 Lee now appeals, and asks us to consider two main arguments. First, she contends that the trial court should not have determined, on summary judgment, that she knew of Williams's negligence by March 2009. Second, Lee contends that the trial court made several errors during the trial, including an error regarding a jury instruction and an erroneous decision to exclude certain evidence. We disagree with Lee's first argument, and conclude that the trial court's summary judgment ruling was sound. We agree, in large part, with Lee's second argument, however, and conclude that the trial court committed three errors in the course of the trial, and that at least two of these errors were not harmless. Accordingly, we affirm in part and reverse in part, and remand the case for a new trial.
BACKGROUND
Medical History
¶4 In 2008, Lee became pregnant and began receiving ongoing medical care from Williams, an employee of Moab Family Medicine, PC. During the course of this treatment, Williams determined that Lee's blood was "Type A" and "Rh-negative." ¶5 Human blood is categorized into types, denoted by a letter or letters (for example, Type A, Type AB, Type O). In addition, human blood is further categorized, in a binary way, relative to its "Rh-factor." A person's blood is "Rh-negative" if it does not have a specific particle (called the "D particle") on the surface of its red blood cells. Blood that has the D particle is "Rh-positive." When describing blood type, these terms are often shortened to simply "negative" and "positive," and are often combined with the letter category, so that a person's blood is referred to, for instance, as "A-negative" or "O-positive." Williams testified at trial that mixing blood of different types and Rh-factors can cause medical complications, and that concerns about such complications sometimes arise during pregnancy, because a pregnant woman may have a different blood type and Rh-factor than her fetus. ¶6 The medical complication pertinent to this case arises when a pregnant woman with Rh-negative blood mixes that blood with an Rh-positive fetus. In that circumstance, Williams testified that the woman may become "Rh-sensitized," which causes her body to react to the fetus's Rh-positive blood as if it is an infection or a foreign substance. Once a woman is Rh-sensitized, she is permanently Rh-sensitized. An Rh-sensitized woman has no symptoms until she has a fetus with Rh-positive blood, at which point the woman's body may break down the fetus's blood cells, potentially causing severe harm to the fetus and sometimes even resulting in a miscarriage. ¶7 In this case, as it turned out, Lee's blood type was Rh-negative and her first baby's blood type was Rh-positive, creating a risk for Rh-sensitization if their blood mixed. To protect against Rh-sensitization, doctors can administer an injectable medication called RhoGAM, which Williams testified significantly reduces the risk of Rh-sensitization. RhoGAM is preventative; that is, it can only prevent Rh-sensitization, and cannot undo the sensitization once it has occurred. Accordingly, RhoGAM has no beneficial effect if the woman who receives it is already Rh-sensitized. In cases like Lee's, where a pregnant woman is Rh-negative and her fetus is Rh-positive, a doctor would typically order that she receive RhoGAM twice, once between 26 and 28 weeks gestation, and once upon delivery. ¶8 At the outset of her first pregnancy, Lee was not yet Rh-sensitized, and so RhoGAM treatments to prevent sensitization would have been appropriate. While Williams maintains both (a) that he informed Lee that she needed a RhoGAM injection at 26-28 weeks and (b) that he ordered such an injection, there is no evidence to support these contentions in the contemporaneously created medical records. In any case, it is undisputed that Lee did not receive a RhoGAM injection at 26-28 weeks, or at any other time prior to delivery of her first child. ¶9 Lee gave birth to her first child on December 30, 2008. Williams would later testify that the infant suffered complications associated with Rh-positive infants born to Rh-sensitized mothers, prompting him to suspect that Lee had become Rh-sensitized. Accordingly, Williams ordered tests to determine whether Lee was Rh-sensitized. Despite Williams's suspicions that Lee may already have been Rh-sensitized, Lee received a RhoGAM injection—her first—after the baby's birth. Lee testified at trial that the nurse who gave her the injection indicated that Lee needed it because she was Rh-negative and would need additional injections during every subsequent pregnancy. As noted previously, RhoGAM has no beneficial effects for women who are already Rh-sensitized. ¶10 Williams maintains that, on December 31, 2008, the day after the first child's birth, he informed Lee that she had become Rh-sensitized and that she would potentially suffer complications and miscarriages in future pregnancies as a result. However, there is no confirmation of this in any medical record, and Williams acknowledges that, prior to January 5, 2009, he had not yet received lab results confirming that Lee was Rh-sensitized; he characterized his misgivings that Lee may have been sensitized as merely a "suspicion" prior to that date. Also, there is no evidence in any of Williams's records that Lee was informed of her sensitization after the lab results were received on January 5, 2009, and Rh-sensitization is not mentioned in the records for any of the ten doctor's visits she had after the test. ¶11 A few weeks later, in March 2009, Lee conducted some independent Internet searches regarding RhoGAM. Lee maintains that she engaged in these searches because she remembered the nurse giving her a RhoGAM injection following her delivery and became curious about what that meant. During her Internet searches, Lee learned that RhoGAM was meant to prevent Rh-sensitization. Lee claims that, at this time, she also learned for the first time that she should have received a RhoGAM injection during her pregnancy. Lee maintains, however, that because she received a RhoGAM injection after delivering her first child, she believed she was not Rh-sensitized at that point. Further, Williams later noted that Lee would have learned from her research that her chances of becoming Rh-sensitized—even without the RhoGAM injection—were slim, something on the order of 1 or 2 percent. Other than these March 2009 Internet searches, there is no evidence in the record that Lee conducted any further research into Rh-sensitization. ¶12 In 2011, Lee became pregnant again. During this pregnancy, Lee received treatment from a second doctor. Lee never told the second doctor that she knew she was Rh-sensitized, or that she was wary of potential problems associated with Rh-sensitization. This time, Lee's child was born with serious complications that required blood transfusions and were consistent with injuries caused by a mother's Rh-sensitization. At this point, Lee claims, the second doctor informed her she had become Rh-sensitized. ¶13 Subsequently, on September 27, 2012, Lee filed an action against Williams and Moab Family Medicine (collectively, Defendants), alleging that Williams negligently failed to order that she receive a RhoGAM injection during her 2008 pregnancy, and that this negligence caused her to become Rh-sensitized.
Pre-Trial Proceedings
¶14 In response to the lawsuit, Defendants moved for summary judgment, alleging that Lee's claims were untimely under the applicable two-year statute of limitations. Defendants pointed out that "[t]he statute of limitations [for medical malpractice] begins to run when a person knows or should have known that they have suffered an injury" attributable to negligence. Defendants argued that Lee knew or should have known about both her injury and Williams's alleged negligence in causing it "by March 2009" when she conducted Internet searches for "RhoGAM." After briefing and oral argument, the trial court partially granted the motion, and partially denied it. The court determined as a matter of law that Lee knew about Williams's alleged negligence no later than March 2009, when she learned that she should have been—but was not—given RhoGAM during her pregnancy. However, the court also ruled that it could not determine, on summary judgment, whether Lee also knew at that point that she had been injured (Rh-sensitized). ¶15 Following this ruling, Defendants asked the court to bifurcate the trial, with the first phase devoted entirely to the question of whether Lee's complaint was timely filed. In this first phase, the jury would be asked to determine whether Lee knew or should have known that she was injured by Williams's alleged negligence prior to September 2010 (two years before she filed suit). Only if the jury found Lee's complaint to be timely would the trial proceed to a second stage in which the jury would determine whether Williams had been negligent and, if so, whether that negligence had caused Lee to sustain harm. The court granted Defendants' motion and bifurcated the trial. ¶16 At a pre-trial proceeding immediately before the first phase of the trial, the court determined that it would not permit Lee—at least not during the first phase of the trial—to introduce a "discharge summary" medical record created by Williams on January 1, 2009, that stated that Lee "did miss her 26-week RhoGAM, which was quite unfortunate and despite having been ordered." The court also determined that it would not permit Lee to introduce other medical records maintained by Williams, none of which contained any order for a RhoGAM injection and therefore contradicted the aforementioned discharge summary. ¶17 Lee objected to the exclusion of this evidence, arguing that, while it was certainly relevant to the second-phase negligence question, it was also relevant to the first-phase question of when she knew she was injured by not receiving RhoGAM. In response, the court made several statements indicating it believed the evidence was irrelevant to the first-phase discovery issues. It ultimately conceded, however, that the evidence had at least some "probative value" but would nevertheless be excluded because of its "potential of confusing the jury about whether they're supposed to decide what happened with regard to [the RhoGAM injection] that should have been given during pregnancy." ¶18 Thereafter, the court instructed the parties to submit proposed jury instructions for the trial's first phase. Defendants submitted an instruction stating that "'[d]iscovery' of an injury from medical malpractice occurs when an ordinary person through reasonable diligence knows or should know that she might have sustained an injury" (emphasis added). Lee objected to this proposed instruction, arguing that the inclusion of the words "might have" misstated the law. Lee argued that the instruction should instead indicate that discovery of a legal injury occurs when both the injury and the negligence that caused it have been discovered, as opposed to when the injured person knows she "might have" been injured. Ultimately, the trial court overruled Lee's objection, and agreed to give Defendants' requested instruction. In addition, the court agreed to present the jury with a special verdict form that asked, "Do you find that Defendants have established, by a preponderance of the evidence, that Kylie Lee knew or should have known, by September 27, 2010, that she might have suffered an injury?" (Emphasis added.)
Trial
¶19 At trial, during jury selection, some issues arose with one of the potential jurors (Juror 26). First, Juror 26 stated that he had once been a patient at Moab Family Medicine, Williams's employer and co-defendant, although he had not been a patient of Williams. Second, he indicated that his wife had once been a patient of Williams himself. Third, he indicated that he had known of Williams and his wife for several years, although he did not know them personally. Fourth, Juror 26 indicated that he was involved in a scouting program and one of the boys involved in that program was the son of one of Williams's nurses (Nurse), who was identified as a trial witness. The trial court allowed Lee's attorney, without apparent restriction, to ask Juror 26 follow-up questions about these disclosures. After asking these questions, Lee raised a challenge to strike Juror 26 for cause, expressing concern that Juror 26's acquaintance with Williams and personal relationship with Nurse's son would bias him in favor of Defendants. In response, the trial court asked Juror 26 if he would "be affected if [Nurse] testifie[d] as a witness" in the case. When Juror 26 stated that he would not be, the court denied Lee's challenge for cause. Lee used all of her peremptory challenges during jury selection, and Juror 26 was seated on the jury. ¶20 During the trial, Williams called Nurse as a witness for the purpose of testifying generally about Williams's habits as a medical practitioner. Nurse testified that she frequently worked with Williams, and that as a matter of course Williams thoroughly explained medical injuries to his patients. Nurse was not asked any questions about Williams's care of Lee specifically. During the course of Nurse's testimony, Lee's counsel formed a suspicion that Nurse had discussed her testimony with Williams's counsel, and asked to examine Nurse on that basis. During this examination, which occurred outside the jury's presence, Nurse admitted that Williams's counsel contacted her the night before she testified, and that she discussed the substance of her testimony with Williams's counsel. Lee objected to Nurse's testimony and asked that it be stricken, arguing that Nurse was one of Lee's "treating medical providers" and that defense counsel could not have ex parte communications with such providers. The trial court denied the motion. ¶21 During the trial, Lee testified that Williams never told her she was Rh-sensitized. In contrast, Williams testified that he had, in fact, informed Lee of her injury the day after her first child was born. ¶22 At the conclusion of the evidence, the trial court instructed the jury as follows:
The issue for you to decide is whether Kylie Lee filed her claim more than two years after she should have discovered her legal injury. You shall consider and weigh all the evidence to determine whether Kylie Lee by the use of reasonable diligence had actual or constructive facts by which she knew or should have known, prior to September 27, 2010, that she might have suffered an injury. If the greater weight of the evidence supports the defense of Dr. Williams and Moab Family Medicine on this issue, Kylie Lee's claim against them is time barred, and your verdict is for Dr. Williams and Moab Family Medicine. If, however, the greater weight of the evidence does not support the defense of Dr. Williams and Moab Family Medicine on this issue, your verdict should be for Kylie Lee.(Emphasis added.) And:
"Discovery" of an injury from medical malpractice occurs when an ordinary person through reasonable diligence knows or should know that she might have sustained an injury.(Emphasis added.) In addition, the trial court provided the jury a special verdict form that asked them to answer only one question: "Do you find that Defendants have established, by a preponderance of the evidence, that Kylie Lee knew or should have known, by September 27, 2010, that she might have suffered an injury?" (Emphasis added.) ¶23 After deliberation, the jury answered the question in the affirmative, and found that Lee "knew or should have known, by September 27, 2010, that she might have suffered an injury." Based on this verdict, the trial court determined that Lee's complaint against Defendants was time-barred, and therefore dismissed her complaint with prejudice.
ISSUES AND STANDARDS OF REVIEW
¶24 Lee raises two main arguments on appeal. First, she contends that the trial court erred when it determined, as a matter of law on summary judgment, that she knew of Williams's potential negligence "by March 2009." We review a trial court's decision on a summary judgment motion for correctness, analyzing "the facts and all reasonable inferences drawn therefrom in the light most favorable to the nonmoving party." DOIT, Inc. v. Touche, Ross & Co., 926 P.2d 835, 841 (Utah 1996) (citation and internal quotation marks omitted). ¶25 Second, Lee contends that the trial court committed four errors during its management of the trial. Lee first argues that the trial court erred by incorrectly instructing the jury about when an injury is "discovered." "Whether a given jury instruction correctly states the law is reviewable under a correction of error standard, with no particular deference given to the district court's ruling." State v. Dozah, 2016 UT App 13, ¶ 13, 368 P.3d 863. ¶26 Next, Lee contends that the trial court erred by refusing to dismiss Juror 26 for cause. We review "the propriety of a denial or grant of a challenge for cause" for abuse of discretion, looking to "the entire voir dire exchange with the challenged juror." State v. Maestas, 2012 UT 46, ¶ 41, 299 P.3d 892 (citation and internal quotation marks omitted). ¶27 Lee also argues that the trial court erred in excluding certain items of evidence, specifically: (a) Williams's discharge summary stating that he had ordered a RhoGAM injection for Lee during her pregnancy; and (b) Williams's other medical records that gave no indication that a RhoGAM injection had ever been ordered. We review the admission or exclusion of evidence for abuse of discretion. State v. Ramirez, 924 P.2d 366, 369 (Utah Ct. App. 1996). Even if we determine that evidence was excluded in error, we will not disturb the outcome of the trial unless "it is reasonably likely a different outcome would result with the introduction of the evidence and confidence in the verdict is undermined." State v. Colwell, 2000 UT 8, ¶ 26, 994 P.2d 177. ¶28 Finally, Lee contends that the trial court erred when it denied her motion regarding defense counsel's ex parte communication with Nurse. "The permissibility of defense counsel's ex parte meetings with a plaintiff's treating physicians requires interpretation of" Utah appellate court case law. Wilson v. IHC Hospitals, Inc., 2012 UT 43, ¶ 24, 289 P.3d 369. Further, "[t]he interpretation of precedent is a question of law that we review for correctness." Id. (citation and internal quotation marks omitted).
ANALYSIS
I. The Summary Judgment Ruling
¶29 Lee first contends that the trial court erred when it determined on summary judgment that she knew of Williams's negligence no later than March 2009, even if she may not yet have known that she had been injured by any such negligence. ¶30 Under the Utah Health Care Malpractice Act (the Malpractice Act), "[a] malpractice action against a health care provider shall be commenced within two years after the plaintiff or patient discovers, or through the use of reasonable diligence should have discovered the [legal] injury, whichever first occurs." Utah Code Ann. § 78B-3-404(1) (LexisNexis 2012). In interpreting the Malpractice Act, our supreme court has stated that "the statute of limitations begins when exercising [reasonable] diligence a patient should have discovered" both (1) "his injury" and (2) "its possible negligent cause." Daniels v. Gamma West Brachytherapy, LLC, 2009 UT 66, ¶ 31, 221 P.3d 256. In this case, the trial court determined, as a matter of law, that Lee knew, no later than March 2009, that Williams was negligent, and that therefore the second requirement was met as a matter of law, even if the first one might not have been. ¶31 The trial court based this ruling on two facts that it considered undisputed: (1) that Lee knew, from her Internet searches in March 2009, that she should have been given a RhoGAM shot during 26-28 weeks gestation; and (2) that Lee knew that she had not in fact been given any such shot. From these two facts, the trial court reasoned that, as a matter of law, Lee "clearly knew that [Dr.] Williams might have been negligent" by no later than March 2009. ¶32 This reasoning is unassailable, and Lee wisely does not directly assail it. Instead, she argues that, as a legal matter, a person cannot be deemed to have discovered "negligence" unless and until that person also discovers that they have been injured as a result of that negligence. We disagree. ¶33 The two elements of a medical malpractice action against a health care provider do not have to be discovered at the same time. Certainly, as a factual matter, simultaneous discovery of injury and negligence can occur—for example, a patient who discovers that her doctor mistakenly has operated on the wrong limb will immediately understand that she has been injured by negligence. But the two elements certainly can be discovered independently from one another. In Roth v. Joseph, 2010 UT App 332, 244 P.3d 391, for instance, the patient immediately knew that he had been injured, but did not realize until some time thereafter that his injury may have been caused by negligence. See id. ¶¶ 26-27. There, we held that the statute of limitations had not begun to run until the patient discovered the possibility of negligence, the final necessary element for his claim. Id. ¶ 27; see also Daniels, 2009 UT 66, ¶ 31 (holding that the statute of limitations does not begin to run until a plaintiff discovers both "his injury and its possible negligent cause"). ¶34 While it is possible for a plaintiff to discover an injury before discovering the possibility that it was negligently caused, the opposite factual scenario can also occur, especially in medical situations where injuries are at least temporarily asymptomatic. Not all negligent actions lead to injury, and a plaintiff may very well learn that her health care provider acted negligently long before she learns that any injury resulted from that negligence. ¶35 Indeed, our supreme court previously contemplated that precise scenario in Seale v. Gowans, 923 P.2d 1361 (Utah 1996). In that case, a plaintiff alleged that her doctor was negligent when he failed to discover a cancerous mass in her breast during a routine mammogram. Id. at 1362. Although that negligence did not lead to immediate injury, the plaintiff later came to believe that the doctor's failure to detect the cancerous mass in her breast contributed at least in part to the recurrence of her cancer some years later. Id. At trial, the doctor presented the Malpractice Act's statute of limitations as a defense, arguing that the plaintiff discovered her legal injury when she learned that he had not correctly diagnosed her cancer after her first mammogram. Id. On appeal, our supreme court noted that the plaintiff had discovered her doctor's negligence when she learned that he had discovered the mass in her breast during her first mammogram but had failed to diagnose it as cancer, but held that the plaintiff had not discovered that she was injured by her doctor's negligence until her cancer finally recurred. Id. at 1365. In reaching its decision, the court explicitly discussed the possibility of discovering negligence before injury, and noted that because "the law does not recognize an inchoate wrong," "a claim for negligence is not actionable" until there is "actual loss or damage," even if the would-be plaintiff is already aware of the negligence. Id. at 1364 (citations and internal quotation marks omitted); see also Hunsaker v. State, 870 P.2d 893, 897 (Utah 1993) (noting that actual damages must be pled in addition to a breach of duty in order to assert a claim for negligence). Because the plaintiff in Seale had discovered her doctor's negligence, but had not yet realized an injury, the supreme court held that the statute of limitations did not begin to run until that injury manifested itself. See Seale, 923 P.2d at 1365. ¶36 This case presents one of those situations—like Seale—where a plaintiff has discovered that a medical provider may have been negligent, but where the results of any such negligence would not necessarily be immediately apparent. As noted, the odds of becoming Rh-sensitized, even without a RhoGAM shot, are something on the order of 1 or 2 percent and, moreover, a woman who has been Rh-sensitized may be entirely asymptomatic until her next pregnancy. In this case, Lee discovered the possibility of Williams's negligence when she learned she should have received RhoGAM during her first pregnancy and knew that she did not actually receive it. But discovering negligence, at least in this case, is not the same thing as discovering injury. ¶37 Accordingly, the trial court's summary judgment ruling was sound. The undisputed evidence demonstrated that Lee knew, no later than March 2009, that Williams might have been negligent. The trial court correctly entered summary judgment on this point, and appropriately narrowed the issues for trial—at least on the statute of limitations issue—to whether and when Lee discovered that Williams's negligence caused her injury.
At the time, the statute of limitations relevant to medical malpractice was codified at Utah Code section 78-14-4. The portion of that statute relevant to our analysis did not materially differ from the statute of limitations we consider in this case.
II. Trial Management Issues
¶38 Next, Lee argues that the trial court made four errors during the course of the trial. We agree with Lee on three of the trial issues raised. A. Jury Instructions and Verdict Form ¶39 Lee first argues that the trial court erred when it instructed the jury that a medical malpractice plaintiff discovers injury when "an ordinary person through reasonable diligence knows or should know that they might have sustained an injury." (Emphasis added.) Lee argues that the inclusion of the words "might have" in the jury instructions and in the special verdict form was error. We agree. ¶40 As discussed above, our supreme court has held that, to discover legal injury under the Malpractice Act, a plaintiff must discover "both the fact of injury and that it resulted from negligence." Daniels, 2009 UT 66, ¶ 1 (emphasis added); see also id. ¶ 31 (the statute of limitations begins to run when "a patient should have discovered his injury and its possible negligent cause"). With regard to the "negligence" element, our jurisprudence speaks in terms of discovering a "possibility of negligence." Roth, 2010 UT App 332, ¶ 21. But with regard to the "injury" element, a mere "possibility" of injury is not enough—a plaintiff must know (or at least should know, through reasonable diligence) that she actually sustained an injury in order for the statute of limitations to start running. See Seale, 923 P.2d at 1365 (noting that discovery of both negligence and the possibility of injury resulting from that negligence is not sufficient to trigger the statute of limitations); see also Roth, 2010 UT App 332, ¶ 21 (noting that a plaintiff must discover "the existence of an injury" before the statute of limitations begins to run (emphasis added)); Deschamps v. Pulley, 784 P.2d 471, 473 (Utah Ct. App. 1989) (noting that "the two-year provision does not commence to run until the injured person knew or should have known that he had sustained an injury and that the injury was caused by negligent action" (emphasis added) (citation and internal quotation marks omitted)); cf. Model Utah Jury Instructions 2d CV325 (2016), https://www.utcourts.gov/resources/muji/inc_list.asp?action=showRule&id=3#325 [https://perma.cc/PUP9-4CE2] (stating that "discovery" occurs when a plaintiff knows "that [she] sustained the injury" and also knows of "the possibility of a health care provider's fault in causing the injury" (emphasis added)). We are aware of no case law supporting the proposition that a medical malpractice plaintiff discovers legal injury upon realizing that she might have or could possibly have sustained an injury. ¶41 In this case, however, the trial court instructed the jury that "[d]iscovery of an injury from medical malpractice occurs when an ordinary person through reasonable diligence knows or should know that she might have sustained an injury" (emphasis added). The addition of the words "might have" was erroneous, and instructed the jury to apply the law incorrectly. By adding these two words to the relevant jury instructions (and to the special verdict form), the trial court impermissibly relaxed the burden of proof that Defendants are required to meet for their statute of limitations defense. Under applicable case law, Defendants would need to establish that Lee knew (or should have known) about her injury in order to trigger the statute of limitations, whereas under this instruction Defendants would need to establish only that Lee was (or should have been) aware of a possibility that she might be injured. ¶42 In their brief, Defendants attempt to defend the trial court's instruction by asserting that "the statute of limitations begins to run when a plaintiff actually discovers, or through the use of reasonable diligence should have discovered, her legal injury." Because of this, Defendants assert that "a plaintiff need not have actual knowledge of a physical injury in order for the time limitation to begin to run," but need only have "sufficient information from which a reasonable person exercising reasonable diligence would discover the injury." This argument is correct, as far as it goes. But Defendants overlook the fact that the thing that a plaintiff must (or reasonably should) know is that she is actually injured, not just that she might have been injured. Establishing actual or constructive knowledge of a potential circumstance is significantly easier than establishing actual or constructive knowledge of the circumstance itself. Cf. Luckau v. Board of Review of the Indus. Comm'n of Utah, 840 P.2d 811, 814-15 (Utah Ct. App. 1992) (contrasting the determination that a series of events could have caused a result from the determination that the events did cause the result, and noting that the latter is more difficult to prove). ¶43 In the end, we are persuaded that, by including the words "might have," the jury instruction and the special verdict form incorrectly stated the law and effectively lowered Defendants' burden for proving their affirmative defense. Accordingly, the trial court's decision to include those words in the jury instructions and the special verdict form was erroneous. B. Juror 26 ¶44 Lee next contends that the trial court abused its discretion when it did not dismiss Juror 26 after Lee's challenge for cause. We find this argument unpersuasive. ¶45 In general, a prospective juror is not presumed to be biased unless some portion of his or her voir dire responses reveal "evidence of bias or partiality." Butterfield v. Sevier Valley Hosp., 2010 UT App 357, ¶ 21, 246 P.3d 120 (citation and internal quotation marks omitted); see also West v. Holley, 2004 UT 97, ¶ 14, 103 P.3d 708 (stating that "[v]oir dire responses revealing evidence of bias or partiality give rise to a presumption that a potential juror is biased"). This can occur when a juror openly indicates that he or she would be biased, but can also occur when a juror's voir dire responses reveal that he or she has developed a personal "'relationship of affection, respect, or esteem'" with a witness or party "'that cannot be deemed disinterested, indifferent, or impartial.'" State v. Cox, 826 P.2d 656, 660 (Utah Ct. App. 1992) (quoting State v. Brooks, 563 P.2d 799, 802 (Utah 1977)). After a presumption of bias has been raised, the prospective juror must be dismissed for cause unless the presumption is rebutted. Id.; see also West, 2004 UT 97, ¶ 14 (stating that a juror who is presumed to be biased "must be dismissed unless that presumption is rebutted"). ¶46 Courts should not lightly presume, however, that a potential juror is biased. Here, Lee maintains that the court should have presumed that Juror 26 was biased following Juror 26's statements that he was generally acquainted (but not friends) with Williams, that he received medical care at Moab Family Medicine (but not from Williams), and that he was involved in a scouting program with Nurse's son. We do not agree that Juror 26's answers were sufficient to raise a presumption of bias. "Jurors are not biased merely because they are acquainted with a party or witness." Butterfield, 2010 UT App 357, ¶ 32. Juror 26's statements that he had met both Williams and Nurse and had been treated at the clinic which employed Williams, without more, do not indicate any relationship beyond the level of a mere acquaintance with Williams and Nurse. And bias cannot fairly be inferred from a mere acquaintanceship alone. ¶47 This conclusion is bolstered by the fact that, in this case, the trial court gave Lee ample opportunity to question Juror 26 to determine the extent and depth of any potential bias. Lee was permitted to question Juror 26 regarding his association with Nurse and Williams; Juror 26 answered Lee's questions; and Lee's counsel finished questioning Juror 26 in his own time, without being cut off by the trial court. While trial courts are by no means required to "'permit every question that might disclose some basis for counsel to favor or disfavor seating a particular juror,'" State v. Holm, 2017 UT App 148, ¶ 11, 402 P.3d 193 (quoting State v. Reece, 2015 UT 45, ¶ 45, 349 P.3d 712), the fact that the trial court in this case afforded that courtesy to Lee clearly satisfied the court's responsibility to allow Lee the opportunity to adduce facts that may support a presumption that Juror 26 was biased. Even after being allowed more or less unfettered permission to question Juror 26, Lee was unable to unearth anything more than the fact that Juror 26 was merely acquainted with both Williams and Nurse. The trial court followed up by asking Juror 26 if his relationship with Nurse would affect his ability to be fair and impartial, and Juror 26 answered in the negative. Certainly, this question would not by itself have been enough to rehabilitate Juror 26 if a presumption of bias had been raised. See West, 2004 UT 97, ¶ 15 (stating that "a presumption of bias cannot be rebutted solely by a juror's bare assurance of her own impartiality"). But here, because no presumption of bias was raised by Juror 26's disclosures, this question by the trial court was appropriate and bolstered its conclusion that Juror 26 could be fair and impartial. ¶48 Accordingly, the trial court did not abuse its discretion when it denied Lee's motion to dismiss Juror 26 for cause. C. Exclusion of Evidence ¶49 Lee next contends that the trial court erred when it refused to allow Lee, during the bifurcated trial, to introduce certain evidence derived from Williams's own medical records generated during his treatment of Lee. Specifically, Lee contends that the trial court erred when it excluded: (a) part of Williams's discharge summary medical record stating that he had ordered a RhoGAM injection for Lee during her pregnancy, and (b) some of Williams's other medical records that contradicted this assertion by indicating that no RhoGAM injection had ever been ordered for Lee. ¶50 The parties and the trial court offer four distinct characterizations of the trial court's reasoning in excluding the proffered evidence. Lee asserts that the trial court excluded the evidence purely on the grounds that the court did not find it to be "relevant" under rule 402 of the Utah Rules of Evidence, and argues that this rationale is incorrect. Defendants argue that the proffered evidence was indeed irrelevant, but also maintain, in the alternative, that the evidence was properly excluded under rule 403 of the Utah Rules of Evidence, because its probative value was substantially outweighed by either the danger of unfair prejudice or by the risk of "waste of time." For its own part, while the trial court discussed relevance at length, it stated that its ultimate decision to exclude the proffered evidence rested on its determination that the probative value of the evidence was outweighed by the risk that it would "confus[e] the jury," apparently also referencing rule 403. We find none of these justifications persuasive. ¶51 First, the evidence was relevant. "Relevant evidence is admissible," Utah R. Evid. 402, and evidence meets the definition of "relevant" if it has "any tendency" to make a fact "of consequence in determining the action" "more or less probable than it would be without the evidence," Utah R. Evid. 401(a)-(b). As our supreme court has noted, the concept of "relevance" is extremely broad. See State v. Reece, 2015 UT 45, ¶ 64, 349 P.3d 712 (stating that "[e]vidence that has even the slightest probative value is relevant under the rules of evidence" (alteration in original) (citation and internal quotation marks omitted)). ¶52 In deciding whether and to what extent the proffered evidence was relevant, the trial court seems to have been informed, at least in part, by its bifurcation of the trial into two phases: one phase concerning whether the statute of limitations barred Lee's claim and, if necessary, a second phase concerning the merits of Lee's medical malpractice claim, including the question of whether Williams was negligent. Certainly, the proffered evidence would have been relevant to whether Williams was negligent, and would arguably have been even more probative to that issue than to the determination of when Lee knew about her injury. However, the proffered evidence is nevertheless relevant to whether and when Lee knew about her injury. ¶53 Lee wanted to use the medical records to argue that Williams misrepresented in his records that he had ordered a RhoGAM injection for Lee but in fact had never done so, and hoped that, by so doing, she could cast doubt upon Williams's credibility, including specifically the credibility of his assertion that he had informed Lee of her Rh-sensitization on or about December 31, 2008. We agree with Lee that this evidence would have at least some tendency to make it more or less probable that Williams informed Lee of her Rh-sensitization as he claimed because it touched on whether Williams's testimony was generally credible. Evidence that "may shed light on the credibility of a witness [] does have probative value." See State v. Miranda, 2017 UT App 203, ¶ 35, 407 P.3d 1033. We therefore consider the proffered evidence to have been relevant, even during the first stage of the bifurcated proceedings. In fact, the trial court stated as much during its oral ruling, noting that the evidence "has probative value." Thus, the proffered evidence should not have been excluded pursuant to rule 402. ¶54 Similarly, we are not persuaded that the evidence should have been excluded pursuant to rule 403. That rule requires trial courts to balance the probative value of proffered evidence against countervailing factors, such as the potential for unfair prejudice, waste of time, or confusion, and instructs trial courts to "exclude relevant evidence if its probative value is substantially outweighed by" any of those factors. Utah R. Evid. 403. We are unpersuaded that the danger of any of those things "substantially outweighed" the probative value of the proffered evidence. ¶55 First, we do not see any appreciable risk of unfair prejudice to Williams by admission of the proffered evidence. Evidence is unfairly prejudicial when it has "the tendency to suggest a verdict on an improper, emotional basis." State v. Hildreth, 2010 UT App 209, ¶ 44, 238 P.3d 444. However, "[r]ule 403 does not require a trial court to dismiss all prejudicial evidence because all effective evidence is prejudicial in the sense of being damaging to the party against whom it is offered." State v. Marchet, 2009 UT App 262, ¶ 44, 219 P.3d 75 (second alteration in original) (citations, internal quotation marks, and brackets omitted). Here, we are unable to perceive any potential for unfair prejudice to Williams as a result of introducing his own medical records. Certainly, the proffered evidence may indicate to the jury that Lee did not receive a RhoGAM injection during her first pregnancy (a stipulated fact), and that Williams had no contemporaneous record of ordering one. Indeed, it may potentially even raise the possibility that Williams wrote that he ordered an injection in order to conceal his negligence. While this evidence could have reflected poorly on Williams's credibility, that is not a danger of unfair prejudice. Weighing the credibility of witnesses is one of a jury's principal functions. See Lyon v. Bryan, 2011 UT App 256, ¶ 10, 262 P.3d 1199. We perceive no appreciable danger of unfair prejudice to Williams from the introduction of his own medical records regarding his treatment of Lee. ¶56 Second, we do not perceive any appreciable risk that admission of Williams's own medical records would have constituted an undue waste of the jury's time. There are many ways in which evidence may waste a jury's time. For instance, evidence "wastes time" if it will "do little more than tell the jury what result to reach," see Davidson v. Prince, 813 P.2d 1225, 1232 n.7 (Utah Ct. App. 1991), or if it is "utterly unrelated" to the issues central to a proceeding, see State v. Martin, 2017 UT 63, ¶ 49. But those concerns are not present here. The proffered evidence was not voluminous, and would not have added significant time to the trial. Moreover, the evidence was not "utterly unrelated" to the question the jury was asked to answer. ¶57 Finally, we disagree that the probative value of the excluded evidence was substantially outweighed by the danger that it would have confused the jury. Without doubt, the proffered evidence also shed light on the negligence questions that were to be taken up, if at all, in the second phase of the bifurcated trial. But merely because a piece of evidence is relevant to more than one issue does not mean it is inherently confusing. A jury—especially a properly instructed one—can be expected to keep straight the questions it is asked to answer in a bifurcated trial, and we have no trouble concluding that it is improper to exclude relevant evidence in the first phase of a bifurcated trial simply because that evidence might be more relevant to the second phase. ¶58 Although trial courts generally have quite a bit of latitude to admit or exclude evidence pursuant to rule 403, see Diversified Holdings, LC v. Turner, 2002 UT 129, ¶ 38, 63 P.3d 686, that discretion is not unbounded. In our view, that discretion was exceeded here. The trial court should not have excluded Williams's own medical records from evidence during the first phase of the bifurcated trial. D. Counsel's Ex Parte Conversation with Nurse ¶59 Lee next contends that the trial court erred when it denied her motion regarding the ex parte conversation that occurred between Williams's counsel and Nurse. Relying on Sorensen v. Barbuto, 2008 UT 8, 177 P.3d 614, and its progeny, Lee maintains that any ex parte communication between Williams's counsel and Nurse that concerned Lee's case should have been deemed improper, and that the trial court should then have engaged in further deliberation to determine how to mitigate the impropriety. We find Lee's arguments persuasive. ¶60 In Sorensen, our supreme court recognized a duty of confidentiality between physicians and patients, and held that this duty prevents a physician from "disclosing information received through the physician-patient relationship." See id. ¶ 12. Specifically, the court held that this duty prohibits "ex parte communications between a plaintiff's treating physician" and defense counsel, who is acting as the attorney for the plaintiff's litigation opponent. Id. ¶ 21. In the event that defense counsel wishes to obtain information from a plaintiff's treating physician, that attorney must use "traditional forms of formal discovery," such as interrogatories and depositions. Id. ¶ 24. The supreme court listed two important reasons for this rule: furthering "patient expectations of physician-patient confidentiality," and providing sufficient "judicial monitoring" of information exchanges between treating physicians and defense attorneys. Id. ¶ 21. ¶61 Our supreme court elaborated on this standard in Wilson v. IHC Hosps., Inc., 2012 UT 43, 289 P.3d 369. In that case, the court clarified that once an improper ex parte communication has occurred, the "[s]election of an appropriate sanction requires fact-finding combined with the exercise of discretion" by the trial court. Id. ¶ 94. Thus, when a trial court becomes aware of an improper ex parte communication, it may exercise its power to levy sanctions, which "may include fines, attorney fees, exclusion of evidence, disqualification of counsel, or any other appropriate response," and which should be "appropriately related to the nature of the misconduct and the resulting prejudice, either actual or potential." Id. ¶62 In this case, Williams's counsel contacted Nurse ex parte on the evening before Nurse's testimony, and secured Nurse's agreement to testify against Lee, chiefly regarding some of Williams's general practice habits. Lee maintains that this conversation violated the principles our supreme court set forth in Sorensen and Wilson. At various points during this appeal, Defendants have posited two arguments supporting their assertion that this conversation was not improper under Sorensen. First, Defendants argued in their brief that Nurse was not a treating medical provider for Lee and therefore not subject to Sorensen's bar. Second, Defendants asserted at oral argument that counsel's ex parte communications with Nurse were not legally prohibited because those communications did not include any discussion of Williams's specific treatment of Lee, but instead were limited merely to a discussion of Williams's practice habits in general. Neither argument is convincing. ¶63 The first argument merits little discussion. Defendants' representation, in their brief, that Nurse "is not and was not [Lee's] treating medical provider" is plainly incorrect. After Defendants made that assertion in their brief, Lee attached to her reply brief some ten pages of medical records demonstrating that Nurse treated Lee during Lee's 2008 pregnancy, including treatment as early as October 2008, roughly the period when Lee should have received her first RhoGAM injection. There is no question that Nurse was one of Lee's treating providers during the pregnancy in question, and therefore Nurse is a health care provider with whom defense counsel should not have had ex parte conversations. ¶64 The second argument differs from the first one, in that the factual assertions Defendants make are correct, but the argument is ultimately no more persuasive. As a factual matter, the conversation between Nurse and Williams's attorney appears to have been limited to a discussion of some of Williams's general treatment habits. Specifically, Williams wanted to call Nurse as a witness so that she could explain to the jury that Williams was the type of doctor who typically explains medical injuries to his patients in a thorough manner. Lee's counsel was afforded the opportunity, at trial (and outside the presence of the jury), to ask Nurse about the contents of her conversation with Williams's attorney, and Nurse confirmed counsel's account that there was no specific discussion of Lee at all, and certainly no discussion about Williams's or Nurse's treatment of Lee. Although the conversation did not specifically include any discussion of Lee or Williams's care of Lee, we are persuaded that the conversation was nevertheless forbidden by the rule set forth in Sorensen. ¶65 One of the reasons our supreme court gave for forbidding ex parte communications between treating physicians and defense attorneys is the need to assure "patients that their honest and complete disclosures of symptoms and medical history to treating physicians will be kept confidential." Sorensen, 2008 UT 8, ¶ 22. The court noted that "judicial monitoring" of those conversations is necessary to further that goal, because otherwise "patients would lack adequate assurance that their candid responses to questions important to determining their appropriate medical treatment would remain confidential." Id. The court considered, but rejected, the notion of "placing the burden of determining relevancy" on either the attorney or the physician, noting as follows:
Placing the burden of determining relevancy on an attorney, who does not know the nature of the confidential disclosure about to be elicited, is risky. Asking the physician, untrained in the law, to assume this burden is a greater gamble and is unfair to the physician. We believe this determination is better made in a setting in which counsel for each party is present and the court is available to settle disputes.Id. ¶ 23 (citation omitted); see also Wilson, 2012 UT 43, ¶ 91 (noting that defense counsel "has a duty in the underlying lawsuit to neither instigate nor facilitate a treating physician's breach of the duty of confidentiality to his patient through an improper ex parte meeting"). ¶66 In both Sorensen and Wilson, the ex parte communications in question actually did concern the specific treatment of the plaintiffs. We are unaware of any Utah case discussing whether an ex parte conversation between a treating physician and a defense attorney that does not discuss the patient's care would violate the principles set forth in Sorensen. As a practical matter, we recognize that a conversation simply about setting a date for scheduling a deposition, or a friendly conversation in a grocery store about the weather or the local football team's fortunes, would not be problematic. See, e.g., Goin v. United States, No. 3:13-cv-564-NJR-DGW, 2015 WL 1577771, *2 (S.D. Ill. Apr. 2, 2015) (holding that a conversation between an attorney and a doctor that was limited to scheduling a date for a deposition did not violate Illinois's version of Sorensen). But a conversation about the merits or substance of the case, even if that conversation does not explicitly touch on the patient's care, is too close to the line, and in our view is still within the Sorensen prohibition. See, e.g., Burns v. Michelotti, 604 N.E.2d 1144, 1148-49 (Ill. App. Ct. 1992) (determining that an ex parte communication that concerned only details relevant to a change-of-venue motion, including whether the doctor's office was in a certain county and whether all of the doctor's treatment of the patient had been in a certain county, violated Illinois's version of Sorensen, even though the communications did not concern the specifics of the doctor's treatment of the patient); see also Smith v. Orthopedics Int'l Ltd., 244 P.3d 939, 945 (Wash. 2010) (holding that even transmitting publicly available documents to a patient's treating physician without asking any questions violates Washington's prohibition on ex parte communications). ¶67 Accordingly, we conclude that any ex parte communication between a defense attorney and a plaintiff's treating physician that is related to the merits or substance of the plaintiff's case in any respect violates the rule set forth in Sorensen, regardless of whether the confidential details of patient's care are in fact discussed and regardless of whether actual prejudice results. Accordingly, the conversation between Williams's counsel and Nurse was improper, even though it did not include any discussion of Williams's or Nurse's specific treatment of Lee. The trial court's conclusion to the contrary was incorrect. ¶68 The trial court, on remand, must determine what (if any) sanction is appropriate under the circumstances. When considering what, if any, sanction to impose, a trial court may take into account whether and to what extent the patient's care was actually discussed, and whether any confidential details were divulged. In this case, the fact that the conversation between counsel and Nurse included neither a discussion of Lee's medical treatment nor a discussion of any confidential information Lee may have shared with Nurse or Williams during the course of her treatment may counsel in favor of a light sanction. See, e.g., Burns, 604 N.E.2d at 1150-51 (imposing a very light "sanction" in view of the "minimal nature of the contact in question and its lack of relevance to the substance of the case"). But selection of that sanction is for the trial court to determine. See Wilson, 2012 UT 43, ¶ 94 (stating that "[s]election of an appropriate sanction" is "a decision [that] is best made in the first instance by a trial court," because it "requires fact finding combined with the exercise of discretion").
In Daniels v. Gamma West Brachytherapy, LLC, our supreme court stated that "the statute of limitations did not begin to run until [the plaintiff] discovered that the [d]efendants' treatment and care might have been negligent and thus might have caused his injuries." 2009 UT 66, ¶ 31, 221 P.3d 256 (emphasis added). We read this passage as referring to the possibility of negligence, not to the possibility of injury—that if the health care professionals had been negligent, then there might be a negligent cause for his already-known injury. On the facts of that case, there was no question that the plaintiff already knew of his injury. Id. ¶¶ 6-7.
While the standards for granting or denying motions to strike a juror "for cause" should remain the same for small-town trials as they are for big-city trials, we note that the issue of potential jurors being acquainted with parties, witnesses, or other jurors will, as a practical matter, arise far more often in rural areas than in counties with large populations. If jurors could be removed for cause simply because they are acquainted with a party or a witness or another juror, seating a jury in a small-town trial would be a difficult exercise indeed. To raise a presumption of bias, whether in a big city or in a small town, there must be more than a simple, neutral acquaintance; instead, there must be a "relationship of affection, respect, or esteem" (or its opposite) with a witness or party "that cannot be deemed disinterested, indifferent, or impartial." State v. Cox, 826 P.2d 656, 660 (Utah Ct. App. 1992) (citation and internal quotation marks omitted).
We can, of course, affirm a trial court's decision on any ground supported by the record. See Bailey v. Bayles, 2002 UT 58, ¶ 10, 52 P.3d 1158.
The court also vacated a previous decision of the Utah State Bar Ethics Advisory Opinion Committee that had authorized ex parte communications between defense lawyers and treating physicians. Sorensen v. Barbuto, 2008 UT 8, ¶ 27, 177 P.3d 614. In so doing, the court "instruct[ed] lawyers," as a matter of attorney ethics, "to confine their contact and communications with a physician or therapist who treated their adversary to formal discovery methods." Id.
Defendants make no argument that the duty of confidentiality articulated in Sorensen applies only to physicians and not to other health care professionals, such as nurses. Indeed, both sides appear to be operating under the assumption that the duty applies to nurses in the same way it applies to physicians. Because the parties do not raise the issue, we do not directly address it, other than to note, in passing, that courts in other states that have prohibited ex parte communications between plaintiff's treating physicians and defense counsel have extended the fiduciary duty of confidentiality to nurses as well. See, e.g., Roberson v. Liu, 555 N.E.2d 999, 1002 (Ill. App. Ct. 1990) (ex parte sworn statement provided to defense counsel by nurse who was present during plaintiff's surgery breached the fiduciary duty of confidentiality). --------
III. Harmless Error
¶69 Having determined that the trial court made three errors, we next consider whether those errors, taken together, merit reversal. "The court at every stage of the proceeding must disregard any error or defect in the proceeding which does not affect the substantial rights of the parties." Utah R. Civ. P. 61. Accordingly, when we determine that a trial court erred, we do not reverse unless "there is a reasonable likelihood that a different result would have been reached" absent the errors. Belden v. Dalbo, Inc., 752 P.2d 1317, 1321 (Utah Ct. App. 1988). The burden of showing that a party was harmed by the trial court's error rests on that party. State v. Lafferty, 2001 UT 19, ¶ 35, 20 P.3d 342. ¶70 Here, we have noted that the trial court erred in three instances, by: (1) adopting jury instructions (and a special verdict form) that misstated the legal standard for when a plaintiff discovers her injury, thus relaxing Defendants' burden of proof for their statute of limitations defense; (2) excluding evidence at trial that was relevant to the question of what Lee knew and when she knew it; and (3) denying outright Lee's motion regarding defense counsel's improper ex parte communication with Nurse. The first two of these errors cannot be categorized as harmless, and therefore a new trial is required. ¶71 In our view, the first issue is particularly significant. The jury was presented with an incorrect legal standard in both the jury instructions as well as in the special verdict form—a standard that effectively relaxed the burden of proof that Defendants had to meet for their affirmative defense. The jury was told that Lee would be considered to have discovered her injury as soon as she "knew or should have known she might have been injured." As we have explained, this is incorrect; the jury should have been instructed, in keeping with applicable case law and MUJI 2d CV325, that a plaintiff is not considered to have discovered her injury until she knows, or reasonably should know, that she has sustained an injury. After being given an incorrect legal standard, the jury may well have determined that Lee knew that she might be injured as soon as she discovered Williams's alleged negligence, regardless of whether she knew or should have known she was actually Rh-sensitized at that point. There is a reasonable likelihood that the jury may have reached a different result had they been given jury instructions and a verdict form that articulated the correct legal standard. ¶72 The second issue is also material. Lee correctly notes that the only witness to testify that Lee knew of her injury prior to the birth of her second child was Williams, who testified that he told Lee about it on December 31, 2008, the day after her first child was born. But Williams's own medical records do not independently support his testimony, and the jury was not permitted to see all of those records, including the records showing that no RhoGAM injection was ever ordered for Lee during her pregnancy coupled with the record containing an unsupported post-birth assertion to the contrary. As we conclude above, all of these records were relevant to the jury's determination during the first phase of the bifurcated trial, and should not have been kept from the jury. We are unable to say with any degree of certainty that the outcome of the trial would have been the same had the jury been given the opportunity to consider all of Williams's records. ¶73 In the end, we conclude that these two errors—especially taken together—may have made a difference in the outcome of the trial, and that "there is a reasonable likelihood that a different result would have been reached" absent the errors. Belden, 752 P.2d at 1321. Accordingly, the trial court's adoption of an improper jury instruction and exclusion of some of Williams's medical records were not harmless errors. ¶74 The third error we identify—the trial court's failure to identify the improper ex parte communication between Nurse and Williams's attorney as a violation of the Sorensen standard—may not have been an error that led, on its own, to a different outcome. However, because we remand the case for a new trial based on the harmful effect of the first two errors, we direct the district court, on remand, to consider an appropriate sanction (if any) for the improper ex parte communication.
CONCLUSION
¶75 We affirm the trial court's summary judgment decision, and conclude that the trial court appropriately narrowed the issues for trial. During the trial, however, the district court committed three errors, two of which are potentially harmful enough to warrant reversal. Accordingly, we affirm in part and reverse in part, and remand this case to the trial court for a new trial or for such other further proceedings as may be consistent with this opinion.