“Listen to your patient; he is telling you the diagnosis.”
–Sir William Osler
“You can see a lot just by observing.”
–Yogi Berra
“Are you saying, doctor, that your diagnosis is based on what the individual told you?”
–Typical question for the expert
Lawyers attack the objectivity of data, or perceived lack of it, to impeach expert witnesses whose findings lack psychometric verification and quantitation. A valid theory of knowledge, according to this tactic, would exclude what an evaluee tells us (as self-serving) and how we interpret it (as biased). In my world, the client is paying for experience, the value-added component of the assessment beyond mechanical algorithms. I'm worried that subjectivity has been canceled.
Perhaps expert witnesses and attorneys mean different things by objectivity. We usually mean that our data, however derived, have not been influenced by our personal sentiments, that they are impartial. For expert witnesses, the ability to confront explicit and implicit bias must be incorporated into practice and embedded into our attitudes, predispositions, and behavior.
A perennial concern in forensic psychiatry, the objectivity problem was observed by Dr. Isaac Ray1in his advice to witnesses in 1851:
[The expert] is, in form at least, the party's witness, engaged by him, and by him made acquainted with all that he knows respecting the merits of the case. The consequence of such a relation is that he can scarcely help testifying under a bias (Ref. 1, p 57).
A century later, Dr. Bernard Diamond2 admonished that the impartial expert does not exist; the idea is a “fallacy.” Undertaking to interpret information in an adversarial way and presenting it to the justice system as the truth runs afoul of our shared values. Embracing bias, simply because we are proxy combatants, cannot be condoned. Ethics requires us to recognize it.
This journal has alerted readers repeatedly to potential sources of compromised objectivity. Concerns have included expert witnesses' responses to pressure from retaining counsel, assessing trauma, role ambiguities, and assessing professional negligence. Being objective, in the sense of unbiased, is an unrelenting challenge for our profession, more so when biases embedded in our education and technique are not constantly monitored. The harder problem is repairing attitudinal coloring of our opinions. A recent journal editorial3 observed that the longstanding AAPL Ethics Guidelines, an aspirational document, is a bit long in the tooth. While objectivity is demanded, there may be a scotoma in contextualization. As Martinez and Candilis3 point out consistently, technical skills in forensic psychiatry must be paired with contextual understandings of culture, individual differences, and systemic inequities. Performative truth-telling that ignores the bigger frame of justice and our potential role in it, misses the mark. Agreed.
There is another obstacle to truth: whether the data used to form our opinions are objective, as if we could detach ourselves from the assessment. Perhaps “detach” is too strong a word, since the adversarial system keeps experts alert to their partisan roles. Nevertheless, removing our sensory impressions from clinical data has never been required and would weaken the mental status examination. The following example illustrates the translocation of objectivity from the witness's distortions to the theory of the expert's knowledge.
A Jarring Encounter
In a work-injury case, I testified at a discovery deposition on behalf of the plaintiff. The physical trauma was not to the brain, but the employee was left with significant psychic residue. My methodology was standard and straightforward: record review, collation of sources, and mental status examination informed my opinion. There was little pushback on my qualifications on voir dire. The direct examination conducted by the defense attorney, however, turned out to be more than the usual fishing trip. My initial responses noted my examination of the plaintiff, all sources of information, and my conclusion that a life-threatening incident had left the plaintiff with posttraumatic stress disorder (PTSD). I was expecting questions about the basis of the diagnosis and how I concluded the plaintiff was not malingering, but the attorney treaded lightly, presumably because the underlying incident was documented (objective). Anyone could relate to the folk-psychological dynamic of physical trauma causing psychic injury. So far, so good, I thought.
What happened next was odd and unsettling. Mind you, this was not my first rodeo, but I was given a new horse to ride. The questions pivoted to objectivity, though not the objectivity of removing bias; rather, whether I had objective findings to assert at all. It was beyond the typical questions about ruling out exaggerated symptoms and the degree to which I relied on self-reported symptoms of PTSD, matters I had been encountering throughout my career. Instead, the next questions attempted to paint me as lacking objective findings and attitude, as if I had nothing to bring to the table. I knew that, by rule, expert testimony was required when medical matters were salient. It felt like an evidentiary hearing on the admissibility of the psychiatric examination itself, where I was being accused of peddling snake oil. If my opinions were based, even in part, on what the examinee told me and how I regarded it, my work was to be nullified as lacking objectivity. An existential crisis had erupted.
An Unexpected Philosophical Excursion
I found myself in an anxiety dream sequence, sitting for an exam on an unknown topic. During the cross-examination, I greeted Descartes and Kant and lightly tripped over solipsism (roughly, the idea that only the self can be known, not the mind of others) and epistemology (how we know what we know). I found myself having to justify standard forensic interviewing as a valid source of knowledge. The colloquy continued (here in a less nightmarish, more condensed version):
Q. Is it fair to say that your diagnosis of PTSD was based on the plaintiff's self-report?
A. In part. The rest was all the documents reviewed, including photos and videos of the accident, official reports…
Q. Yes, but what objective evidence did you have?
A. All of the reports of the accident, notes of other clinicians, and observations of the man afterward…
To this point, I thought I was just being questioned by a skeptic on the thoroughness of my examination and my reliance on sources other than self-report. But it was more: a conversation about how I did not extract objective data, whatever that meant, beyond self-report. It became a philosophical exercise in the meaning of “objective,” in the sense of data untainted by my mind. With no philosophers by my side, I had to respond to questions while wrangling with how to defend my methods. I started to have flashbacks of college philosophy, the time and the place for confronting all those brave souls who took on the big questions of how we know anything and whether there is an objective world beyond our own perceptions.
Philosopher Barry Stroud,4 commenting on Kant and epistemological skeptics, said, “At every point in the attempted justification of a knowledge claim, the skeptic will always have another question yet to be answered, another relevant possibility yet to be dismissed, and so he can't be answered directly” (Ref. 4, p 242). I have long known, and taught, that the courts do not tolerate gray areas, but this was a whole new ballgame, and it wasn't what I'd signed up for. I was now defending my very profession. Imagine if I had retorted something about philosophy and psychiatry, that I was describing mental phenomena in the natural world, the nosology of natural kinds (reflecting Nature rather than human categories), universal truth untainted by human diagnostic constructs. But alas, there was no time for metaphysics, and I was at the mercy of my questioner.
In the conventional view of testimony (the Federal Rule of Evidence 703, for example), a doctor can report findings without evidentiary scrutiny: “An expert may base an opinion on facts or data in the case that the expert has been made aware of or personally observed.” That's what I'm talking about. But here I was in deeper Stygian water trying to get past a gatekeeper barking about the reliability of the psychiatric examination as a source of information. As much as I may enjoy the intellectual exercise, I was not prepared for this philosophical inquisition under oath.
The examination continued:
Q. How do you differentiate subjective from objective?
A. Subjective, in medicine, is what is reported by patients in their own words. Objective is what we, the examiners, observe. I take what the person says, along with what I perceive, and regard it by standard psychiatric principles.
Q. How do you make sure you remain objective?
A. That is an ethical matter. We need to be mindful not to color the information or infuse it with our own values; not to be biased.
Welcome to Square One. I was being examined by Lewis Carroll, now sure that the attorney and I were not using “objective” in the same way. As Humpty Dumpty said to Alice, “When I use a word, it means just what I choose it to mean—neither more nor less” (Ref. 5, p 132, emphasis in original). Very well, as long as you explain it, and it doesn't undermine time-tested bases of medical procedure and knowledge, or so I thought until recently. Well, there was no explanation or consensual definition of objectivity, just questions. It was not so much a challenge about bias as it was about whether I could take myself out of the equation, leaving pristine data. Even to say “Clinical data,” data derived by examination, seemed dubious. Perhaps that memo had gone right into my spam folder.
From SOAP to SUDs
My medical education took place in the early years of the problem-oriented medical record. The purpose of it, according to its originator, Dr. Larry Weed, was to clean up the clutter and chaos of existing records and to provide a roadmap, a means of communication among providers.6 Each medical problem was to be recorded in four domains: Subjective, Objective, Assessment, and Plan (SOAP). If we were not being cynical, we could see a clear distinction between Subjective (what the patient said, the chief complaint, symptoms) and Objective (what we took in with our own senses, examination findings, signs). I had been practicing in the Oslerian tradition of using sensory data and knowledge or experience to make a diagnosis. So, to me, when asked for my clinical objective findings, I always equated them with what I, the objective (unbiased) clinician, observed. Was I wrong to regard the medical examination as sacrosanct, immune from the falsifiability domain of the Daubert7 inquiry? My responses did not satisfy the attorney in this case, though this was not a Daubert challenge (yet).
In the weeds, my pulse quickened as I felt the methodology of forensic psychiatry threatened. SOAP had slipped out of my hands and, to borrow a behavior therapy term, my Subjective Units of Distress (SUDs) were spilling onto the witness stand. I resonated with criminal suspects facing incredulous detectives during interrogation (Fig. 1).
The author, objectively distressed and experiencing subjective units of distress (SUDs), during mock polygraphy using antique instrumentation (2014). Assisted by Yan Xuan, MD (behind). From the author's collection. Photo by Evi Numen and equipment from George Grigonis, Mütter Museum of the College of Physicians of Philadelphia.
The colloquy continued, proving Stroud's4 point about skeptics:
Q. Well, can't you test the person against some standard? Why didn't you use psychological testing?
A. That is usually the domain of psychologists, who prefer to use structured instruments and numbers. In my experience, both roads can lead to valid results.
But the gold standard for psychiatrists is the examination, in which the person's narrative is obtained, we ask questions, and we make observations. It's how we do it in psychiatry. There's no presumption that we sacrifice objectivity by using a less structured or quantified approach. Both psychiatrists and psychologists use the same diagnostic criteria.
Q. If you're controlling the interview, then how can it be verified for accuracy?
A. I need to know things other than what the person reports. For example, that the accident occurred, the details of it, and how a person might react to the situation. Then we want to know what other clinicians said, people who were not connected to the litigation. There has to be congruence with the agreed-upon facts of the case and the individual's clinical presentation. It's not simply a matter of my personal impressions; rather, how all the pieces fit together.
Wittgenstein on the Witness Stand
The remainder of the examination was about how the injured person met diagnostic criteria for PTSD. Is it not self-evident that traumatic events are associated with a degree of psychic distress, or am I expected to carve Nature at the joints, an impossible task: ultimate truth-finding?8 Today, Plato's standard of a universal truth would far exceed “reasonable scientific certainty.” We psychiatrists appreciate self-report, which would be an element of many structured and unstructured protocols. If I had used a structured instrument for PTSD (for example, the PTSD Checklist for DSM-5 [PCL-5] or the Clinician Administered PTSD Scale [CAPS-5]), it would have been another version of self-report. And if I had used a broader inventory of personality, there was a risk of extraneous interpretations muddying the water (traits, for example). Besides, I could neither interpret nor vouch for the accuracy of such instruments, despite their enduring popularity. No, I had to go it alone, unrepentant, using crude but established methodology. The attorney unceremoniously moved on from the philosophical underpinnings of expert testimony to practical matters of degree of injury and prognosis, the money part. This interlude of the grilling, while not comfortable, was at least terra cognita. These were the mundane issues of the A and P of Weed's charting schema, Assessment and Plan.
Back to my crisis of applied epistemology. There may be a shift in how objectivity is defined in forensic settings: from preventing bias to taking oneself entirely out of the equation. In a conversation with a psychiatric resident, I mentioned the importance of separating subjective (experienced) from objective (observed) data. Without missing a beat, she replied, “Yes, but you're making the observations, so doesn't that make it subjective on your part?” She was conveying the subjectivity of contemporary medical thinking, that clinical impressions blend the patient's experience, my observations, and my experience. But in my heart of hearts, I believe that is, in fact, what doctors do and the reason we trained for so long. The resident's question delivered me to a safer place, the essential reason persons with specialized knowledge are called to court.
Medical experts are expected to use experience and learning to move beyond the data into an interpretation. I must articulate my findings as they bear on the question before the court; they do not speak for themselves. Opinions without bases are inadmissible. It is a convention embedded in the rules of evidence; for example, that knowledge of one's specialty counts for something and is a prerequisite for saying anything. The attorney and I were each using the English language from different points of reference: mine from that of a person educating the court, albeit in an adversarial environment, and the attorney's from that of a partisan arguing to win the case. If we let ourselves get tied up in knots over it, we land in Wittgenstein's Philosophical In-vestigations,9 where the impossibility of interspecies communication is exposed. Narrative and quantitation are not mutually exclusive. When a psychiatrist testifies, however, there are rules that permit the results of an assessment to stand as evidence. After that, the weight of the testimony, not its epistemology, becomes a question for the fact finder.
As a Matter of Fact, It's a Matter of Opinion
In the above colloquy, then, I could not simply assert, ipse dixit, that my conclusions were objective in the sense of their being unbiased. My results required independent existence beyond my brain and perhaps, in the opposing attorney's view, only written or quantitated protocols could breathe truth into my statements. I was not abusing my role in the case. I wasn't Prometheus stealing fire from the gods, only using a medical model in a legal setting. Far be it from me to transition arrogantly from opinion to truth, though it has been argued that such liberties are possible under decisions like Kumho Tire Co. v. Carmichael,10 which provided some leeway in the “forensic identification sciences” (fingerprinting, etc.).11 I took comfort from the fact that courts relied on people like me to make diagnoses and prognostications and on judges to decide whether my opinions had more probative than prejudicial influence. Still, there was a nagging question about my resistance to relying on social science techniques to reify, objectify, or validate my medical opinion. I wondered whether it was time to graduate from old-school psychiatry and add a second layer of objectivity to my clinical method, even if it felt counterfeit.
Lawyers are urged to challenge mental health testimony that seems to be based on subjective (biased) statements rather than scientifically supportable evidence. Here, the shift in meaning that I experienced is exposed: reliable opinions must be grounded in something more than the psychiatric examination, even when filtered through specialized knowledge, expertise, and accepted methods. Faust12 has written:
In [some] instances, what is presented as a scientific assertion is merely an expert's subjective impression, with the expert doing little more than serving as a mouthpiece for one side of a case. This danger is especially pronounced in cases involving mental health testimony. For example, an expert may assert that Mr. Smith's mental status became much worse after a purported head injury, an impression based almost entirely on Mr. Smith's description of a change in status following an event. The expert's appraisal may not be a scientific one, but rather a subjective conclusion that Mr. Smith's assertions are accurate. In this way, the expert is merely mirroring what Mr. Smith is saying, but doing so under the guise of scientific appraisal or knowledge. For such reasons as these Daubert assigned the judge the gatekeeper role in order to avoid circumstances in which jurors may be overly influenced by the appearance of science rather than the true probative value of information (Ref. 12, p 44, emphasis in original).
Faust's concern, and likely that of the attorney deposing me, is that psychiatric methodology is dependent, to a necessary degree, on our mind interpreting what the evaluee reports to us paired with what we observe. I agree that an opinion based on a “subjective impression” would carry no weight without what I have called the “objective” elements of the data. They are synergistic. Imagine if I decided to circumvent the problem of self-report by declining to examine the plaintiff on the grounds that I did not want to be influenced by a highly motivated person and then contaminate the jury with bias. I would then be disqualified for failing to employ the standard techniques and ethics of my profession. Impressionistic opinions are redolent of bias, as Sadoff13 recalled in a hospital malpractice case:
When asked, on cross-examination, how he knew the patient was schizophrenic, he stated, “I'm a psychiatrist. It's a gut feeling. I know.” He said this without having examined the patient and from his review of the records. That type of idiosyncratic behavior would not be allowed in the courtroom today (Ref. 13, p 123).
Indeed, such an expert is likely to be eviscerated; testifying explicitly on gut feelings is unacceptably subjective. Retrospective testamentary capacity assessments require great circumspection, especially because the parties' emotions are involved. The well-known and related problem is the treating clinician serving as an expert, with its attendant loss of objectivity.14
Old Dogs and Their Day (in Court)
If some of this sounds familiar, I refer the reader to Dr. Charles Scott's AAPL presidential address in 2012.15 He talked persuasively about upgrading forensic psychiatry's objectivity, in the service of truth, by relying less on subjective assessments. Recalling Dr. Paul Appelbaum's theory of ethics, he said:
[F]irst, the forensic psychiatrist has an obligation for both subjective and objective truth-telling. In other words, not only must the psychiatrist present information that he believes to be true, but the information presented also must accurately reflect the current scientific evidence and consensus of the field (Ref. 15, p 19).
Further reflecting on the need to upgrade forensic sciences (including ours), Scott cited the National Academy of Science's concern over the less-than-100 percent accuracy of fingerprint analysis. If forensic psychiatry cannot hold up under scrutiny like Faust's, we risk the impression of modern-day phrenology: certitude sans science. Scott discussed the literature on risk assessment and detection of malingering, concluding that a structured approach had more literature support than the unstructured. Accordingly, forensic training programs have an obligation to provide education on validated tools, he argued.
Speaking as an old dog contemplating new tricks, I do not dispute that, if structured assessments that provide a common vocabulary become a standard of practice, we should expect trainees to learn about them. The evolving standard for suicide risk, for example, may involve a blended approach.16 Sex offender assessments are more problematic, with tension between static (historical, objective) data and dynamic information gleaned largely from self-report. In any event, it would not be desirable for manualized approaches to choke off open-ended ones simply because they are amenable to measurement. That would be the moral equivalent of refusal to teach the arts because outcome measures are too soft. There must also be vigilance around the possibility that structured assessments contain racial or cultural bias, which gets us back to Martinez and Candilis'3 point.
At the end of the day, credibly explaining one's technique, staying in one's lane, not shading one's opinion through conscious bias, and stating it with hubris-free confidence remain serviceable qualities in the expert witness. Ray17 located our role concisely 150 years ago:
To promote the ends of justice, it sometimes becomes necessary to obtain the opinion of the witness rather than any facts he may possess, because what is wanted is, not merely the naked facts, but the inferences derived from them,—their significance with reference to the points in issue (Ref. 17, p 409).
Acknowledgments
The author is grateful to Michelle T. Joy, MD for her insights.
Footnotes
Disclosures of financial or other potential conflicts of interest: None.
- © 2021 American Academy of Psychiatry and the Law
