JUSTICE BLAND delivered the opinion of the Court, in which Chief Justice Hecht, Justice Devine, Justice Blacklock, Justice Busby, Justice Huddle, and Justice Young joined. JUSTICE BUSBY filed a concurring opinion. JUSTICE BOYD filed a dissenting opinion, in which Justice Lehrmann joined. In this health-care liability case, we must determine the appropriate causation standard to apply. The facts are tragic. A teen committed suicide after seeking treatment for depression from her pediatric health-care providers. The expert testimony at trial established the medical providers’ negligence, but it did not establish that, but for the negligence, the teen would not have committed suicide. In affirming a verdict for the teen’s family, the court of appeals incorrectly omitted an analysis of but-for causation. Because the testimony does not establish but-for causation, and our precedent requires it in cases like this one, we reverse and render. I A Pediatrics Cool Care is a pediatric clinic supervised by Dr. Jose Salguero. On March 1, 2012, Ginger Thompson brought A.W., her thirteen-year-old daughter, to the practice, where A.W. had been a patient since 2010. A.W. told Jenelle Robinson, a physician assistant employed by the practice, that she was feeling “sad all the time” and “can’t control her feelings.” After a brief consultation, Robinson diagnosed A.W. with depression and prescribed Celexa, an anti- depressant. Robinson could not recall whether she warned A.W. and Thompson that Celexa could cause suicidal ideation and to watch A.W. closely, but she testified it was her usual practice to do so. Robinson also could not recall whether she had asked A.W. if she currently was experiencing any suicidal ideation or thoughts of self-harm, though she thought it was “likely” she had made that inquiry. Robinson did not ask to interview A.W. outside of Thompson’s presence. Nor did Robinson use diagnostic checklists the clinic employed for adolescents presenting with depression. Robinson said that she “strongly recommended” that A.W. seek counseling. The medical record reflects that Robinson provided a list of counselors to Thompson. Thompson, however, testified that the practice did not. Even so, Thompson said that A.W. had refused to go to counseling because “she wanted to try the medication and that she had a—she had a teacher at school she could talk to and she didn’t want to go and talk to a stranger.” Robinson testified that she instructed A.W. to follow up one week later, but Thompson denied that Robinson gave that instruction. Although Robinson had prescribed thirty days’ worth of Celexa, the medical assistant who transcribed the record indicated that A.W. was to receive three thirty-day refills. Robinson testified that the U.S. Drug Enforcement Agency limits Celexa prescriptions to ninety days. About six weeks later, on April 17, A.W. returned to the practice, complaining of migraine headaches. Nurse practitioner Allyn Kawalek examined A.W. According to the medical records, both A.W. and Thompson reported a positive change in A.W.’s mood. Almost two years after the visit, and after A.W.’s parents filed this suit, an unknown person altered the record of this visit to add the phrase “patient is to come back in 30 days for follow-up.” On July 31, Thompson called the practice, asking to refill A.W.’s Celexa prescription. Bernadette Aguillon, a medical assistant, took Thompson’s call. She initially told Thompson that Thompson could refill the Celexa over the phone. Aguillon later saw that A.W. was overdue to follow up, and she attempted to call Thompson back to schedule an appointment. After failing to reach Thompson, Aguillon approved the refill, despite lacking authorization from Dr. Salguero or any of the providers. Aguillon testified that she regularly wrote prescription refills on Dr. Salguero’s behalf without consulting him. After learning of A.W.’s suicide, Aguillon attempted to alter A.W.’s medical records to conceal her error. About two weeks later, on the evening of August 14, Thompson discovered her daughter’s body. A.W.’s cause of death was determined to be suicide by an overdose of Benadryl. Thompson testified that she had no idea that A.W. was suicidal. Neither Thompson nor her husband had noticed anything unusual about A.W.’s behavior leading up to her suicide. A.W. never revealed to either her mother or her father that she had any suicidal thoughts. None of A.W.’s friends reported to Thompson that they had suspected A.W. to be suicidal, or even that she was depressed. A.W.’s father, Brad Washington, testified that A.W. did not seem sad or depressed the last time they visited, about two weeks before her death. A.W. was fourteen years old at the time of her death. B Thompson and Washington sued Pediatrics Cool Care, Dr. Salguero,[1] Robinson, and Kawalek (collectively, the providers) for negligence and gross negligence. The jury heard expert testimony from Dr. Herschel Lessin, a pediatrician, on the deficiencies in the providers’ care, including: Robinson’s “[t]otally inadequate” workup, particularly her failure to interview A.W. outside the presence of her mother, failure to use a standardized depression-screening questionnaire, and failure to determine the scope and severity of A.W.’s depression; Robinson’s decision to prescribe Celexa after a single visit; The transcription error resulting in A.W.’s receiving three refills of Celexa; The providers’ failure to follow up with A.W.; Failures in the practice’s record-keeping, including insufficient documentation, alterations made after the fact, and the providers’ failure to review the records to catch errors; and Aguillon’s decision to refill the Celexa prescription without authorization. Dr. Lessin did not testify as to the cause of A.W.’s death. A.W.’s parents presented Dr. Fred Moss, a psychiatrist, to testify that the health-care providers’ negligence caused A.W.’s death. Though he testified that A.W. should not have been prescribed Celexa, Dr. Moss confirmed that the Celexa was not a cause of A.W.’s suicide: Q. Now, are you here to tell the jury that this Celexa caused her suicide? A. Oh, no.[2] Dr. Moss instead testified that, had Robinson asked the right questions when she examined A.W. for depression, A.W.’s answers would have “created pathways towards treatment options” that then would have prevented A.W. from committing suicide: Q. All right. Now, I want you to talk about the—can you— can you tell us the ways, sir, as you sit here today, that the treatment options that were available to [A.W.] to—that you believe, based upon reasonable medical probability, would have prevented her committing suicide on August 14th of 2012? A. Sure. So they’re really going to be defined in no small part for—based on the answers to the questions that weren’t asked in [A.W.]‘s case, unfortunately, but if would—they would have been asked would have created pathways towards treatment options that would then be made available and then would have prevented her from committing—from, unfortunately, committing suicide on August 14th, 2012. Based on “answers to the questions that weren’t asked,” Dr. Moss listed several “pathways” that A.W. and her parents could have explored, including counseling, nutritional counseling, group therapy, sports, exercise, meditation, and establishing relationships with teachers and advocates. Dr. Moss testified that, had Robinson interviewed A.W. outside her mother’s presence, and had A.W. disclosed any suicidal tendencies she was feeling to Robinson, then Robinson would have had “a great space to work from.” In Dr. Moss’s experience, exposing suicidal ideation results in “kids find[ing] a reason to live.” Dr. Moss testified that “an accumulation” of the providers’ errors resulted in a failure to create pathways and connections for further treatment: It’s a cluster of so many things, so many things, so many acts and maybe more so. So many omissions of all the things I’ve listed here plus some that would have created pathways, that could have created connections, that could have created—I don’t even know what it would have created had they been addressed initially. When pressed whether a particular path would have prevented A.W.’s suicide, Dr. Moss responded, “I cannot list a specific—one specific path that [Robinson] might have taken that would have prevented suicide reliably,” and repeated that “a cluster of a cumulative number of things” led to A.W.’s suicide. He further conceded that, even had the providers done everything correctly, A.W. still might have committed suicide. But, if the providers had taken an adequate history, “several cumulative factors would have showed up, maybe, just maybe more likely than not preventing her suicide.” Moss formed his opinions based on his extensive psychiatric experience treating pediatric and adult patients. He also relied on literature that connected the use of psychotherapy in addition to medication as leading to “better outcomes,” though not preventing suicide. When asked on cross-examination about the certainty of his conclusions, he explained that “[t]his work that we’re speaking of is not grounded in science.” The jury also heard expert testimony from Dr. Armando Correa, an assistant professor in the Department of Pediatrics at Baylor College of Medicine. Dr. Correa testified that “suicide in teenagers is usually impulsive. It’s unforeseeable. It’s just an action that they take without thinking of the consequences. And, sadly, most of the time it cannot be prevented.” Of the two thousand teenage suicides per year, Dr. Correa testified that “the majority of those are impulsive.” Dr. Correa asserted with “a reasonable degree of medical probability” that A.W.’s suicide “was an impulsive, unpreventable act.” However, he conceded that “an ordinary prudent physician can foresee that if you—if you don’t properly treat a 14 year old with depression that suicide can occur.” The trial court asked the jury whether each of the providers proximately caused A.W.’s death. The charge defined proximate cause as “a cause that was a substantial factor in bringing about an occurrence, and without which cause, such occurrence would not have occurred.”[3] Additionally, “the act or omission complained of must be such that a [provider][4] using ordinary care would have foreseen that the occurrence, or some similar occurrence, might reasonably result therefrom. There may be more than one proximate cause of an occurrence.” The jury found that Dr. Salguero and Robinson proximately caused A.W.’s death, but found Kawalek—the last provider to see A.W.—not liable. The jury rendered a multi-million-dollar verdict, which the trial court reduced to $1.285 million, plus interest, on final judgment. The providers who were found liable appealed, challenging the sufficiency of the liability and causation evidence supporting the verdict, the admission of Dr. Moss’s testimony, and the trial court’s calculation of prejudgment interest. The court of appeals affirmed.[5] Although the trial court had submitted a but-for, or cause-in-fact, causation standard as part of the jury’s charge on proximate cause, the court of appeals omitted any analysis of it, citing our opinion in Bustamante v. Ponte.[6] The court instead confined its analysis to substantial-factor causation. Using this relaxed causation standard, the court of appeals held that Dr. Moss’s testimony was sufficient evidence that Dr. Salguero’s and Robinson’s negligence caused A.W.’s death.[7] The court of appeals further concluded that Dr. Moss’s testimony was factually grounded and reliable, based upon his clinical experience and training as a psychiatrist.[8] Finally, the court of appeals upheld the interest calculation. We granted the providers’ petition for review. II Ordinarily, to recover for medical malpractice, a plaintiff must prove “to a reasonable medical probability that the injuries complained of were proximately caused by the negligence of a defendant.”[9] The two elements of proximate cause are cause-in-fact and foreseeability.[10] A defendant’s negligence is the cause-in-fact of a plaintiff’s injury if “(1) the negligence was a substantial factor in causing the injury, and (2) without the act or omission, the harm would not have occurred.”[11] Courts refer to these two components as “substantial factor” causation and “but for” causation. In this Court, the medical providers do not challenge the jury’s negligence findings. Rather, the providers contend that the court of appeals erred in its legal-sufficiency review when it declined to conduct any but-for causation analysis. Had the court of appeals employed the correct causation standard, the providers argue, it would have concluded that Dr. Moss’s testimony was legally insufficient to support a finding that Dr. Salguero and Robinson proximately caused A.W.’s suicide. In answering the providers’ causation challenge, we must first determine the appropriate causation standard, and then evaluate the evidence for legal sufficiency under that standard.[12] A In a typical medical malpractice case, the plaintiff must adduce evidence that the defendant’s negligence was a substantial factor in causing the injury and that, but for the defendant’s negligence, the plaintiff would not have been injured.[13] Our Court has applied this causation standard in two other medical malpractice cases involving a patient’s suicide. In Providence Health Center v. Dowell, we rendered judgment for a hospital, holding that the medical providers’ conduct was too attenuated to be a cause of their patient’s suicide.[14] The patient had expressed suicidal intentions, but the hospital released him after he refused inpatient treatment and promised to seek care at a mental health clinic.[15] Thirty-three hours later, he hanged himself.[16] Our Court held that the hospital’s decision to release the patient was not a proximate cause of his death.[17] In reaching that conclusion, we observed that there was no evidence the patient would have consented to treatment, the expert never testified that hospitalization would have prevented the patient’s suicide, and the decision to discharge the patient was “too remote from his death in terms of time and circumstances.”[18] For those reasons, we concluded that “the defendants’ negligence was too attenuated from the suicide to have been a substantial factor in bringing it about.”[19] In Rodriguez-Escobar v. Goss, we similarly rendered judgment for the defendant physician because no evidence supported a but-for causal link between the physician’s treatment and the patient’s suicide.[20] In that case, the police brought the patient to the hospital after she had discharged a shotgun inside her bedroom and expressed suicidal intent.[21] The patient’s son obtained a Mental Health Warrant for Emergency Detention to have her involuntarily admitted to a state mental health hospital.[22] The physician who conducted triage at the state hospital concluded that the patient did not meet the criteria for involuntary hospitalization and discharged her.[23] Three days later, the patient committed suicide.[24] We considered whether the evidence was “legally sufficient to support the finding that absent the negligence of [the physician]—but for his negligence—[the patient] would not have committed suicide.”[25] The expert in Rodriguez-Escobar testified that “if [the patient] had been in the hospital, I don’t think that she would have been able to kill herself, at least not shoot herself. And hopefully if a plan had been in place, then her chances of having a better life would have been there.”[26] The expert conceded “I don’t know long term what her prognosis would have been. It would have depended upon a lot of things.”[27] Following our decision in Providence, our Court held that the expert’s testimony did not establish that the physician’s negligence proximately caused the patient’s death.[28] Citing our decision in Bustamante v. Ponte, the court of appeals disregarded the cause-in-fact analysis applied in Providence and Rodriguez-Escobar, instead announcing that it would “apply the substantial factor test,” to the exclusion of requiring but-for causation.[29] In Bustamante, we rejected “a stringent but-for causation test” for an individual actor when the evidence demonstrates that concurrent acts of negligence combined to cause the injury.[30] Bustamante, however, did not eliminate but-for causation for medical malpractice cases involving multiple negligent actors. Rather, when the facts establish that concurrent causation exists for multiple negligent actors—each whose negligence is itself a but-for cause of the injury in question absent the others’ concurrent negligence—then the but-for requirement shifts from the individual level to the aggregate level of defendant tortfeasors.[31] Bustamante concerned a premature infant who had a 90% to 100% chance of developing retinopathy of prematurity, an abnormal blood-vessel growth pattern that can cause diminished vision or blindness.[32] Experts testified that a properly screened and diagnosed infant would have received a laser therapy that was “successful in over 75% of ‘all comers’” and, when timely, prevented retinal detachment in almost 90% of eyes studied.[33] The failure to timely diagnose and treat the infant’s retinopathy lay equally on two physicians, Dr. Ponte and Dr. Llamas, and a jury found both responsible.[34] A divided court of appeals reversed, holding that there was no evidence that either physician’s negligence was a but-for cause of the infant’s injuries because the other physician’s negligence also contributed to causing the injury.[35] “Specifically, the court of appeals criticized [the expert] for testifying that it was ‘more likely than not’ [the infant] would have a sighted life if not for Dr. Ponte’s and Dr. Llamas’s combined negligence, rather than quantifying the negative impact of each negligent act.”[36] Our Court held that the court of appeals erred in applying “a stringent but-for causation requirement in a case that should have been resolved under the substantial-factor test.”[37] Because both physicians had failed to diagnose and treat the retinopathy, it was impossible to say that, but for the actions of either physician, the infant would have a sighted life. The evidence of but-for causation was nonetheless present at the aggregate level—but for the combined negligence of Dr. Ponte and Dr. Llamas, the infant more likely than not would have a sighted life. And, had either physician acted alone, his negligence in failing to diagnose retinopathy would have been a cause-in-fact of the injury. This case also presents multiple defendants. If the negligent acts of each provider are so concurrent that they cannot be examined in isolation, the correct approach is to consider whether each provider’s individual negligence was a substantial factor in A.W.’s death and whether the providers’ combined negligence was a but-for cause of A.W.’s death.[38] The court of appeals erred in eliminating a but-for causation requirement. The parents further argue that our decision in Windrum v. Kareh suggests that Bustamante eliminated but-for causation in medical negligence cases involving multiple actors.[39] In Windrum, we rejected the court of appeals’ reliance on the physician’s failure to diagnose and treat not being the “immediate” cause of death to demonstrate that substantial-factor causation was lacking.[40] Instead, we held, “the proof required is that the negligence be a substantial factor, not that it be the ‘immediate cause.’”[41] Contrary to the parents’ suggestion, however, our Court required but-for causation: “The ultimate question, then, ‘is whether, by a preponderance of the evidence, the negligent act or omission is shown to be a substantial factor in bringing about the harm and without which the harm would not have occurred.’”[42] We have not eliminated but-for causation; we do not do so today. B In assessing the legal sufficiency of the evidence supporting the finding that the providers’ combined negligence was a but-for cause of A.W.’s death, we consider the evidence in a light favorable to the verdict.[43] In reviewing the legal sufficiency of the evidence, we evaluate “whether the evidence at trial would enable reasonable and fair-minded people to reach the verdict under review.”[44] To prove that medical negligence proximately caused an injury or death requires expert testimony. An expert’s scientific testimony must be “grounded ‘in the methods and procedures of science.’”[45] “Otherwise, the testimony is ‘no more than subjective belief or unsupported speculation.’”[46] Thus, an expert’s bare assertions about causation do not suffice.[47] Nor can the expert rely on “magic language” to establish that the testimony is based on reasonable medical probability instead of possibility, speculation, or surmise.[48] “[I]f the record contains no evidence supporting an expert’s material factual assumptions, or if such assumptions are contrary to conclusively proven facts, opinion testimony founded on those assumptions is not competent evidence.”[49] Finally, when the evidence demonstrates other plausible causes of an injury, the expert must exclude those other causes with reasonable certainty.[50] Dr. Moss’s testimony fails to do more than speculate that, but for the providers’ negligent care, A.W. would not have committed suicide. His conclusions are not based on facts in this record. He testified that if the providers had questioned A.W. outside her mother’s presence, as they should have, and if A.W. then had divulged information about any current suicidal ideation, and if A.W. and her parents then had availed themselves of resources that should have been provided in response to that information, then “based upon reasonable medical probability” A.W. would still be alive: Q: . . . [C]an you tell us the ways, sir, as you sit here today, that the treatment options that were available to [A.W.] to—that you believe, based upon reasonable medical probability, would have prevented her committing suicide on August 14th of 2012? A: Sure. So they’re really going to be defined in no small part for—based on the answers to the questions that weren’t asked in [A.W.]‘s case, unfortunately, but if would—they would have been asked would have created pathways towards treatment options that would then be made available and then would have prevented her from committing—from, unfortunately, committing suicide on August 14th, 2012. Dr. Moss’s conclusion relies on a series of assumptions, beginning with the assumption that, had Robinson properly conducted A.W.’s intake assessment, A.W. would have disclosed that she was suicidal, which further assumes that she had experienced such thoughts at that point. Dr. Moss then assumes that A.W. would have availed herself of one or more “pathways towards treatment.” Finally, he assumes that A.W. and her parents’ engagement along these pathways would have prevented A.W. from committing suicide. To properly conclude that this attenuated chain of events proximately caused A.W.’s suicide requires an evidentiary basis. The first assumption—that A.W. would have disclosed that she was suicidal—is contradicted by the uncontroverted testimony that A.W. had never disclosed any suicidal ideation to anyone. There is no evidence to suggest that A.W. was, in fact, suicidal at the time of her intake assessment or at any other point before August 14.[51] Dr. Moss confirmed that he could not know how A.W. would have responded to a proper diagnostic evaluation: Q. The questions and inquiries you believe should have been made that you’ve listed, the different paths to go down that—those inquiries, questions, you don’t know what the answers to those questions would have been, do you? A. I do not. Q. So you don’t know if the answer to those questions, then, because you don’t know what the answers would have been, you don’t know whether the answers to those questions would be something that would prevent [A.W.] from committing suicide? A. I don’t know that, correct. Dr. Moss’s second assumption—that A.W. and her parents would have consented to any recommended pathways toward treatment—is belied by Thompson’s testimony that A.W. had refused to go to counseling. This assumption, which underpins his conclusion, is not only not supported but is contradicted by the evidence presented. Dr. Moss’s third assumption—that additional treatments would have prevented A.W.’s suicide—is also not reliably supported. Of the pathways Dr. Moss suggested, he could not identify a particular treatment or combination of treatments that “would have prevented suicide reliably.” Pressed about whether the cumulative effect of all the proposed treatments would have reliably prevented suicide, Dr. Moss testified that “upon pursuing all the pathways that I brought up and more, several cumulative factors would have showed up, maybe, just maybe more likely than not preventing her suicide.” Even if the providers had done everything perfectly, Dr. Moss agreed that A.W. “might have still committed suicide,” though he thought it was “more likely than not” that she would not have.[52] This A. I certainly cannot conclude that one. assertion was based on his experience treating thousands of minors for depression, not one of whom had committed suicide. Dr. Moss did not, however, provide a reliable basis to differentiate between patients who, with ordinary psychiatric care, would not commit suicide, and those who would, even with proper care. His conclusion was “not grounded in science.” Dr. Moss testified that unspecified literature discussed “coupling psychotherapy with—with medication in a more appropriate form and a form of treatment that leads to better outcomes,” but he did not opine that those treatments prevent suicide. While Dr. Moss appropriately relied on his experience in treating patients with depression, his connection to A.W.’s case relies on facts not borne out by the record. His first two assumptions about what A.W. might have done are speculative in light of Thompson’s testimony that A.W. never expressed suicidal ideation to anyone and had refused counseling. Thus, his experience with psychiatric patients who accept treatment, in general, having better outcomes presents no basis for concluding that A.W. would have been such a patient. Dr. Moss asserted that his patients were “just like” A.W., but he did not connect that assertion with evidence of why A.W. was a treatable patient and not a patient who would have committed suicide despite treatment. We do not require certainty to the extent the dissent suggests.[53] Expert opinions, however, must be based on the facts in the record and not controverted by them, even when relying on experience and training as a basis for a medical opinion. Dr. Moss presented no factual, verifiable basis for concluding what A.W. would have done had the medical providers asked questions outside her mother’s presence, followed up, or provided a list of counselors or other treatment options. Dr. Moss’s testimony also did not exclude the alternative possibility proposed by Dr. Correa: that A.W.’s suicide was a spontaneous, impulsive—and thus, unpreventable—act. There is no evidence in the record at all that the jury could rely on to exclude this possibility. “[W]hen the facts support several possible conclusions, only some of which establish that the defendant’s negligence caused the plaintiff’s injury, the expert must explain to the fact finder why those conclusions are superior based on verifiable medical evidence, not simply the expert’s opinion.”[54] Our precedent confirms the infirmity of the causation evidence here. Dr. Moss’s testimony is like the expert testimony in Rodriguez- Escobar, which similarly conditioned better outcomes on the patient’s availing herself of additional treatment: “hopefully if a plan had been in place, then her chances of having a better life would have been there.”[55] It is also like the testimony in Providence that our Court rejected as speculative: the plaintiffs’ expert, “when asked directly about whether hospitalization would have prevented [the] suicide . . . , answered only that [the patient] ‘would have improved’ and been at a ‘lower risk’ of suicide when he left.”[56] In Providence, moreover, the patient and his loved ones had reported his suicidal thoughts and his attempts to take his life to the health-care providers.[57] The evidence in Rodriguez- Escobar and Providence that the patients would have lived is stronger than that present here. Both cases presented more evidence of the patients’ mental health, and the denied treatment in both cases had a closer temporal connection to the suicide. We do not hold that medical malpractice could never be the cause of a suicide. Nor, by our holding, do we countenance the providers’ conduct. Negligence, however, cannot substitute for legally sufficient evidence of causation. * * * There is no evidence that the providers’ care proximately caused A.W.’s suicide. We therefore reverse the court of appeals’ judgment and render judgment for the providers. Jane N. Bland Justice OPINION DELIVERED: May 13, 2022
