By Alex Stein
Most psychiatrists don’t know about it, but the switch from Frye to Daubert in the admission of expert testimony matters for them a lot. Psychiatrists treat patients with second-generation antipsychotics: Zyprexa, Risperdal, Clozaril, Seroquel, and similar drugs. A reputable, but still controversial, body of research links those drugs to tardive dyskinesia: a serious neurological disorder involving uncontrollable facial grimacing, repetitive tongue thrusting, and other untoward bodily movements. Under Frye, expert evidence can only be admitted upon showing that it received “standing and scientific recognition” from the relevant community of experts. Absence of a solid consensus disqualifies the evidence. Expert testimony linking tardive dyskinesia to antipsychotic drugs consequently would not be admissible under Frye. Under Daubert, however, it would go into evidence because its underlying research is grounded in scientific method and procedure that can be replicated, examined, and properly explained to the jury.
This is exactly what happened in a recent case decided by the United States District Court for the District of Columbia: Patteson v. Maloney— F.Supp.2d —-, 2013 WL 5133495 (D.D.C. 2013).
This case featured a patient who developed tardive dyskinesia (TD) following her insomnia treatment by an antipsychotic drug, Seroquel. After acknowledging that replicable and peer-reviewed studies linking TD to Seroquel have been published in reputable medical journals, the court decided that the patient’s causation expert passes the Daubert test. The court explained that the fact that medical community is not unanimous about these studies goes to the weight of the expert’s testimony and does not make it inadmissible. The court also underscored Daubert’s “liberal thrust” toward the admission of expert testimony in general.
Moreover, the court ruled that the expert can establish patient-specific causation by differential etiology. This method, in the court’s words, involves “the creation of a list of possible and/or most likely causes for a patient’s signs and symptoms, based on his/her medical history, examination findings, and ancillary testing. … The [expert] then eliminates options from the list until the most likely cause is found. Picture a whiteboard filled with possible medical culprits for a patient’s symptoms—familiar to fans of the medical television drama House—and then watch each being methodically crossed off the list through testing and deduction until a single diagnosis remains.”
Based on this method, plaintiffs who never experienced TD prior to taking the medication would normally be able to move their case to the jury. This ruling was also based on Daubert. The differential etiology method could hardly be admitted into evidence under Frye.
The consequent liability risk for psychiatrists is obvious, given the frequency of TD among psychiatric patients. Antipsychotic drug prescriptions follow the trial and error method. This method involves patient-specific observations, tradeoffs and adjustments that rely on the psychiatrist’s intuition. There are no hard-and-fast rules and protocols similar to those that provide legal “safe harbors” for other doctors (see here, at pp. 1208-16). As a result, plaintiffs’ experts would often be able to second-guess the psychiatrist’s drug prescription and plausibly describe it as negligent.
This new legal reality increases the liability risk for psychiatrists while limiting their ability to defeat malpractice suits summarily. Psychiatrists face this increased liability prospect in every state court that follows Daubert as well as in federal courts sitting in diversity or deciding malpractice suits under the Federal Tort Claims Act. State courts that follow Frye (as in New York and California) will protect psychiatrists against this risk.
A possible solution to this problem can be found in the rules of informed consent that deem psychiatric patients competent to consent to an antipsychotic drug prescription. Excluded from this presumption are patients proven to be “substantially incapable of applying an understanding of the advantages and disadvantages of particular medication to [their] mental illness.” See In the Matter of the Mental Commitment of Melanie L., 833 N.W.2d 607 (Wis. 2013). Based on this presumption, psychiatrists can have their patients sign an informed consent form specifying the TD risk among the anticipated side effects of the prescribed medication. The form should also explain the tradeoff recommended by the psychiatrist and agreed by the patient. Finally, the form should expressly state that the patient chooses this tradeoff over its alternatives and agrees to the drug adjustment process by trial and error. The patient will then be deemed to have assumed the risk of developing TD from taking the medication.
Admittedly, this is not a foolproof solution, and I can also think of psychiatrists who would not like it very much. This solution, however, minimizes the psychiatrists’ risk of liability for TD. Psychiatrists who feel uncomfortable about my proposed informed-consent form should read Patteson v. Maloney— F.Supp.2d —-, 2013 WL 5133495 (D.D.C. 2013) and think about it again.
Thank you for this interesting post, Alex. One thing that is under-recognized in regards to informed consent in the context of schizophrenia is the extent of cognitive impairment that many patients experience. If a patient cannot hold in working memory the information from the consent form long enough to process it, then it is not at all clear that the patient can give informed consent at all let alone understand adequately a risk like TD. We see these concerns about competency for consent most clearly in the context of Alzheimers disease, and it makes me doubt that an appeal to a signed consent form would often provide much of a defense where schizophrenia is concerned. That said, your approach may be much stronger for other uses of antipsychotics. Would love to hear your thoughts on this.
Thanks, Matthew!
I agree with you that psychiatric patients are often unable to give informed consent to treatment. The law, nonetheless, deems them capable of giving informed consent –- in many cases, counterfactually — unless and until their presumed capability is disproved by clear and convincing evidence. The recent Wisconsin case I cited illustrates this rule. The constitutional origin of this rule and the resulting misalignment between law and psychiatry is Addington v. Texas, 441 U.S. 418 (1979). Hence, my proposed solution does work. Admittedly, it is somewhat (or more than somewhat) cynical. For that reason, I wrote that I can think of psychiatrists who will refuse to follow it and risk liability instead. Importantly, in severe schizophrenia cases to which you refer, psychiatrists would happily obtain informed consent from the patient’s guardian (a family member or local social services). That’s an easy case for a psychiatrist. My proposed form should be used in these cases as well.
Thank you for this interesting post, Alex. One thing that is under-recognized in regards to informed consent in the context of schizophrenia is the extent of cognitive impairment that many patients experience. If a patient cannot hold in working memory the information from the consent form long enough to process it, then it is not at all clear that the patient can give informed consent at all let alone understand adequately a risk like TD. We see these concerns about competency for consent most clearly in the context of Alzheimers disease, and it makes me doubt that an appeal to a signed consent form would often provide much of a defense where schizophrenia is concerned. That said, your approach may be much stronger for other uses of antipsychotics. Would love to hear your thoughts on this.
Very interesting observations, Alex.
I did stumble on your comment that this is a “new legal reality that increases liability risk for psychiatrists.” This year marks the 20th anniversary of the Daubert decision, which rejected the “general acceptance” criterion for admissibility of expert testimony, and the Patterson decision seems like a pretty straightforward application to a new “reputable … body of research” about the causal relationship between a particular set of drugs and a particular set of adverse outcomes. I would argue that this is an old legal reality, just being applied to a new set of facts.
I agree completely that psychiatrists should notify their patients of this risk, as part of the “informed consent” process. (Seems like a material fact to me.) But if it is negligent to use the drug under these circumstances, then informed consent should not immunize doctors. That would be equivalent to a written liability waiver, which we don’t enforce in MedMal. Thus, I see informed consent as a second theory of liability, not as a defense.
Thank you so much for your input, Chris!
I respond to your observations in the order you made them:
I. Patteson as a New Reality for Psychiatrists
As a multifactor standard, Daubert indeed is 20 years old. However, this standard was designed to be transformed or translated into field-specific high-resolution rules: every new application of Daubert was supposed to – and in fact did – create a more specific rule for the expert knowledge it covered. And it is this field-specific rule – not the Daubert general standard – that actually determines the admissibility of the relevant expert knowledge.
This widely accepted account of how courts use general standards to develop specific rules over time (see, e.g., Louis Kaplow, Rules Versus Standards: An Economic Analysis, 42 Duke L.J. 557 (1992)) has been anticipated in connection with Daubert. See Federal Judicial Center’s Reference Manual that organizes courts’ applications of Daubert into specific fields of scientific and other specialized knowledge.
For the field of antipsychotic drug prescriptions, what Patteson did is pretty new. This new reality is still unfolding because there is no assurance that this decision will be followed. Indeed, I think that there are good reasons for not following this decision. I believe that courts should align their rulings on the TD causation issue with what they would have decided under Frye (as they did, for example, with the “Bendectin causes birth defects” theory in the Daubert litigation and with different speculative “event studies” in securities class actions). In particular, I question the differential etiology part. Generally, I am in favor of the differential etiology method (see here, on pages 1221-23). But in the case of TD associated with psychiatrists’ prescriptions of antipsychotics that follow the method of trial and error, I see a potentially devastating chilling effect. And so, unlike you, I don’t think that Patteson was a straightforward application of Daubert. Policy-wise, it was a questionable application; and I therefore believe that some courts will decide the issue differently.
II. My “Informed Consent” Solution of the Problem
You are right in saying that informed consent cannot vindicate a substandard treatment. A psychiatrist cannot give an antipsychotic medication to a patient for whom this medication is contraindicated (by current tests or medical history). But this plain negligence scenario is not what Patteson and my post are about. My post is about the standard and recurrent trial-and-error work that psychiatrists do with patients whose condition might actually be improved by the right antipsychotic drug. Absence of bright-line rules in this field gives psychiatrists a lot of discretion, but it also exposes them to suits, many of which – incidentally – raise the “lack of informed consent” allegation as well.
As I explain in my post, Patteson and its likes allow those suits to get to the jury, and hence the heightened liability risk for psychiatrists. I don’t believe that a psychiatrist is actually negligent in this type of case, but the jury might decide it differently. Consequently, the law should enable psychiatrists to defeat those suits summarily.
For this recurrent type of case – that might be decided both ways because of Patteson and its likes – my advice to psychiatrists is this: Choose any medically defensible trial-and-error method and get it preapproved by having your patient sign an elaborate informed-consent form. Importantly, this will open up the school-of-thought defense possibility as well.
Thanks again,
A.S.
Thanks for your followup, Alex. I think we agree in substance. In my view, however, courts do not and should not create “field specific rules” based on applications of Daubert. Such mid-level rules may have an ossifying effect that lags behind scientific progress. Instead, when presented with a Daubert challenge, they should apply the standard to the facts presented.
Even if scientific knowledge were unchanging, no single trial judge (and its Court of Appeals) can really purport to decide a scientific rule for an entire field. Instead, they must decide the evidentiary motion on the evidence presented, which will often involve credibility assessments of the particular expert witnesses presented, and a careful parsing of their words, as applied to the particularities of the case presented, with due regard for the burden on the proponent of the evidence. Just because one plaintiff does a lame job of getting his evidence admitted, shouldn’t prejudice others in the future on similar questions.
(I concede however that there are opinions where courts do purport to set such rules, and it is probably not so egregious when it is a ruling admitting evidence. And I concede that Daubert decisions in either direction will tend to persuade other courts to the extent that the other courts are faced with similar questions. Thus in practice, the Daubert decisions become quite rule-like. That’s why I think we may ultimately agree in substance.)
Thanks, Chris! We do have a disagreement, but not one that would be of interest to the Bill of Health readers. Your view of expert evidence doctrine aligns with that of a number of evidence scholars with whom I disagreed in my book, Foundations of Evidence Law, where I criticized the idea that evidence rules’ designers should care about scientific progress. Briefly, my argument is that evidence rules — all of them — should be geared toward one goal: minimizing the cost of error and error-avoidance as a total sum. For that reason, I am quite pleased to see courts utilizing economies of scale by developing mid-level rules that are cheap to apply. A.
Very interesting observations, Alex.
I did stumble on your comment that this is a “new legal reality that increases liability risk for psychiatrists.” This year marks the 20th anniversary of the Daubert decision, which rejected the “general acceptance” criterion for admissibility of expert testimony, and the Patterson decision seems like a pretty straightforward application to a new “reputable … body of research” about the causal relationship between a particular set of drugs and a particular set of adverse outcomes. I would argue that this is an old legal reality, just being applied to a new set of facts.
I agree completely that psychiatrists should notify their patients of this risk, as part of the “informed consent” process. (Seems like a material fact to me.) But if it is negligent to use the drug under these circumstances, then informed consent should not immunize doctors. That would be equivalent to a written liability waiver, which we don’t enforce in MedMal. Thus, I see informed consent as a second theory of liability, not as a defense.
Thank you so much for your input, Chris!
I respond to your observations in the order you made them:
I. Patteson as a New Reality for Psychiatrists
As a multifactor standard, Daubert indeed is 20 years old. However, this standard was designed to be transformed or translated into field-specific high-resolution rules: every new application of Daubert was supposed to – and in fact did – create a more specific rule for the expert knowledge it covered. And it is this field-specific rule – not the Daubert general standard – that actually determines the admissibility of the relevant expert knowledge.
This widely accepted account of how courts use general standards to develop specific rules over time (see, e.g., Louis Kaplow, Rules Versus Standards: An Economic Analysis, 42 Duke L.J. 557 (1992)) has been anticipated in connection with Daubert. See Federal Judicial Center’s Reference Manual that organizes courts’ applications of Daubert into specific fields of scientific and other specialized knowledge.
For the field of antipsychotic drug prescriptions, what Patteson did is pretty new. This new reality is still unfolding because there is no assurance that this decision will be followed. Indeed, I think that there are good reasons for not following this decision. I believe that courts should align their rulings on the TD causation issue with what they would have decided under Frye (as they did, for example, with the “Bendectin causes birth defects” theory in the Daubert litigation and with different speculative “event studies” in securities class actions). In particular, I question the differential etiology part. Generally, I am in favor of the differential etiology method (see here, on pages 1221-23). But in the case of TD associated with psychiatrists’ prescriptions of antipsychotics that follow the method of trial and error, I see a potentially devastating chilling effect. And so, unlike you, I don’t think that Patteson was a straightforward application of Daubert. Policy-wise, it was a questionable application; and I therefore believe that some courts will decide the issue differently.
II. My “Informed Consent” Solution of the Problem
You are right in saying that informed consent cannot vindicate a substandard treatment. A psychiatrist cannot give an antipsychotic medication to a patient for whom this medication is contraindicated (by current tests or medical history). But this plain negligence scenario is not what Patteson and my post are about. My post is about the standard and recurrent trial-and-error work that psychiatrists do with patients whose condition might actually be improved by the right antipsychotic drug. Absence of bright-line rules in this field gives psychiatrists a lot of discretion, but it also exposes them to suits, many of which – incidentally – raise the “lack of informed consent” allegation as well.
As I explain in my post, Patteson and its likes allow those suits to get to the jury, and hence the heightened liability risk for psychiatrists. I don’t believe that a psychiatrist is actually negligent in this type of case, but the jury might decide it differently. Consequently, the law should enable psychiatrists to defeat those suits summarily.
For this recurrent type of case – that might be decided both ways because of Patteson and its likes – my advice to psychiatrists is this: Choose any medically defensible trial-and-error method and get it preapproved by having your patient sign an elaborate informed-consent form. Importantly, this will open up the school-of-thought defense possibility as well.
Thanks again,
A.S.
Thanks for your followup, Alex. I think we agree in substance. In my view, however, courts do not and should not create “field specific rules” based on applications of Daubert. Such mid-level rules may have an ossifying effect that lags behind scientific progress. Instead, when presented with a Daubert challenge, they should apply the standard to the facts presented.
Even if scientific knowledge were unchanging, no single trial judge (and its Court of Appeals) can really purport to decide a scientific rule for an entire field. Instead, they must decide the evidentiary motion on the evidence presented, which will often involve credibility assessments of the particular expert witnesses presented, and a careful parsing of their words, as applied to the particularities of the case presented, with due regard for the burden on the proponent of the evidence. Just because one plaintiff does a lame job of getting his evidence admitted, shouldn’t prejudice others in the future on similar questions.
(I concede however that there are opinions where courts do purport to set such rules, and it is probably not so egregious when it is a ruling admitting evidence. And I concede that Daubert decisions in either direction will tend to persuade other courts to the extent that the other courts are faced with similar questions. Thus in practice, the Daubert decisions become quite rule-like. That’s why I think we may ultimately agree in substance.)
Thanks, Chris! We do have a disagreement, but not one that would be of interest to the Bill of Health readers. Your view of expert evidence doctrine aligns with that of a number of evidence scholars with whom I disagreed in my book, Foundations of Evidence Law, where I criticized the idea that evidence rules’ designers should care about scientific progress. Briefly, my argument is that evidence rules — all of them — should be geared toward one goal: minimizing the cost of error and error-avoidance as a total sum. For that reason, I am quite pleased to see courts utilizing economies of scale by developing mid-level rules that are cheap to apply. A.