Abstract
Warren and her colleagues’ timely exploration of the difficulties and uncertainties in diagnosing PTSD and personality disorders in the female inmate population raises fundamental questions for clinical as well as forensic analysis. Questions of under-reporting, over-reporting, and comorbidity in this population point to serious inadequacies in the scheme of categorical, context-independent diagnosis in the Diagnostic and Statistical Manual of Mental Disorders. The interplay of predisposition, life history, and current setting and circumstances can best be captured by a progressive refining of probability estimates. Such a diagnostic process calls for psychodynamically informed clinical and forensic interviewing. Additional recommendations are made for the purpose of achieving multidimensional, context-sensitive diagnosis and forensic evaluation.
The exploration of diagnostic conundrums in the female inmate population by Warren and her colleagues1 raises important questions that should be asked in the clinical and forensic analysis of the posttraumatic stress disorder (PTSD)/personality disorder spectrum. These questions, in turn, point to fundamental limitations, both conceptual and practical, in the scheme of diagnostic categorization in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR).2
As Warren and colleagues make clear, posttraumatic stress disorder is likely to be both over- and under-reported in female inmate populations. One reason for this imprecision is that this particular diagnosis relies heavily on self-reports of symptoms. In the study reported, the researchers used criteria for classifying the inmates into those with and without PTSD based on the inmates’ self-reports of experiences and their detailed recollections of how they reacted to them at the time. This method is essentially one that is used to diagnose posttraumatic spectrum disorders clinically. Warren and colleagues present an excellent example of how easily PTSD symptom reporting can be manipulated by interested parties, leading to over-reporting.
At the same time, people who actually have PTSD spectrum traits may under-report their past or even current symptoms as a way of avoiding retraumatization. Some of those who have untreated PTSD have learned to minimize the memory of the traumatic event, to maintain relatively normal functioning. This suppression of memory can be especially common in correctional settings, where day-to-day interactions with officers and other inmates can trigger PTSD symptoms.
The diagnostic picture is further complicated by the comorbidity and potential confusion of PTSD spectrum and personality disorder spectrum traits. Borderline personality spectrum traits, some of which may have their origins in genetics as well as experience,3 can predispose a person to PTSD when exposed to a trauma. A history of intense and unstable relationships, affective dysregulation, behavioral dyscontrol, extreme sensitivity to rejection, fear of abandonment, and failures of mentalization are all part of the borderline personality spectrum. A person who is hypersensitive to her environment in the manner described by this symptom cluster is likely to experience environmental changes and challenges in a more vivid and (when relevant) more threatening way than less emotionally reactive individuals. A potentially traumatic event is likely to have a greater psychological impact on such a person. Because some people with borderline personality traits simultaneously seek and reject intimate relationships, they often lack sufficient social support or are unable to trust and rely on the social support they do have, making it difficult to form stabilizing connections within criminal justice settings.
Additional variables come into play when history and predisposition (biological or experiential) meet current setting and circumstances. Incarceration can be traumatic in itself, especially in view of the prevalence of sexual assaults perpetrated by both inmates and guards4 and the risk of retaliation for reporting an assault.5 The traumatic effects of this environment can be exacerbated by the reluctance of some individuals with PTSD to seek psychiatric treatment or, as noted, discuss their trauma histories. Nonetheless, Warren and colleagues1 report that some women in their sample felt safer and better taken care of in prison than in the community. The interplay between the setting and circumstances and the individual's psychiatric history raises important questions stimulated by their paper. For example, does a history of repeated traumas in the family and community set a person up for more severe traumas in prison, or does it immunize her against whatever she may face there? The most reliable answers to questions such as these lie not in diagnostic taxonomy but in in-depth history-taking in any given case and in a greater body of research on the effects of incarceration on persons with trauma histories, as well as the association between trauma history and involvement in the criminal justice system.
As difficult as it is to tease out the interplay of PTSD symptomatology and the myriad contributing factors in a jail or prison, forensic evaluation of persons who present with PTSD/personality disorder spectrum traits raises additional complexities. Two cases from the forensic practice of one of the authors provide examples for consideration.
Case Examples
Case 1
An older woman claimed to have accidentally shot her husband. Initial police interview tapes showed clear evidence of dissociation. A few years earlier, the woman's daughter had committed suicide. The deceased had allegedly committed incest with his daughter when the latter was a child. Crime-scene evidence supported the defendant's claim that she was in fear of her life at the time of the shooting. She also had a documented history of multiple medications with potential interaction effects, medications she was taking at the time of the shooting. During the forensic psychiatric interview, as the defendant gave a history of the fatal incident, she was observed to exhibit a variety of mental and physiological changes consistent with dissociation.
Case 2
A woman faced criminal charges related to the death of her young child at the hands of her husband. Initially, a defense of battered-woman syndrome was raised. However, there was no reliable corroborative history and no evidence of diminished capacity in a police interview videotaped shortly after the fatal incident. Instead, the defendant's self-reports of symptoms began subsequent to placement in a battered-women's group in prison. In the course of extended forensic psychiatric interviews, the examiner observed no major mental and physiological changes associated primarily with recollection of the child's death or of earlier alleged traumas. What physiological changes were observed were related to the defendant's repeated claims concerning the alleged unfairness of the legal process.
Analysis
Warren and colleagues aptly cite the interactional model by Hacking6 of the classification of human experiences to reflect personal experience combined with social and cultural needs. From Hacking, it is not a long step to a Bayesian diagnostic approach in which biological predisposition sets up prior probability estimates that are then modified, in turn, by life history, current life circumstances and setting (e.g., prison), and examination data.7 Of course, the sequence in which these data are considered may vary, depending on which data are made available first; for example, one may start with current observations or life history and then obtain physiological or genetic data through laboratory testing. As part of this process, the importance of thoughtful and psychodynamically informed clinical and forensic interviewing cannot be overemphasized. Such interviewing, together with review and analysis of corroborative data, allows one to ask first what kind of character the patient or examinee is, which then allows one to ask the questions: “What am I hearing? What am I observing? What does it mean?” Such progressive refining of probability estimates is at variance with the “snapshot” approach of the DSM.7
The timely study by Warren and colleagues,1 interpreted against a background of forensic psychiatric experience, prompts the following observations and recommendations, some of which are more fully discussed elsewhere.8
First, the DSM-IV-TR2 contains useful cautions regarding the need for clinical judgment when applying diagnostic categories to patients, with special cautions concerning the use of DSM categories in forensic settings. Much of this information, however, is confined to the introduction and therefore is ignored or never seen by many users of the manual. It would be very helpful to weave the caveats into the presentation of material in the text.
Second, the validity of applying the DSM diagnostic criteria independent of setting is called into question by, inter alia, the self-selection of the incarcerated population for antisocial behavior (a diagnostic criterion for borderline as well as antisocial personality disorder). At the same time, the prison population may show a self-selection for avoidance (a diagnostic criterion for PTSD). This questioning of the validity of setting-independent diagnosis extends to the clinical as well as the forensic setting, and therefore to clinical trials that rely on DSM categories and DSM-based practice guidelines.
Third, self-reports may be an insufficient basis for reliable clinical diagnosis or forensic evaluation and opinion formulation. Especially in the forensic context, reliable history-taking requires corroborative data. This concern is especially relevant to PTSD, where the diagnostic formulation is based largely on self-reported events and reactions that may be products of malingering or misattribution in the legal arena.
Fourth, personality disorder diagnoses as currently conceived may obscure salient characterological data that lie on a continuum. For example, the notions of borderline and narcissistic personality disorders are cleansed of the often accompanying elements of meanness, ruthlessness, and vindictiveness captured by older diagnostic terminology such as the pseudopatient who is a sadist or an imposter.
Fifth, the ICD-10 diagnosis of enduring personality change after catastrophic experience (ICD-10 F62.0) provides a useful alternative and corrective to the DSM diagnosis of PTSD in that it incorporates a temporal process in which a patient's condition may evolve in a changing life context. The ICD-10 approach is more favorable to a dimensional understanding in which traits and disorders are seen as lying on a spectrum of variability, as opposed to the all-or-none formulation (as noted by Warren and colleagues) of PTSD in the DSM. In particular, the spectrum model allows for recognition and identification of personality traits that do not necessarily rise to the level of personality disorders.
Sixth, clinical and forensic interviewing are becoming lost arts in psychiatry. All too many training programs in psychiatry are neglecting psychodynamically informed interviewing and clinical reasoning skills. Recently, the American Board of Psychiatry and Neurology eliminated oral interviewing as a requirement for board certification.9 The rationale has been given that oral interviewing will be evaluated during residency, when problems can be remedied, rather than belatedly at the completion of training. That this rationale would not apply to other skills being taught is curious. One question that should be studied is whether this change is a consequence of a shift in psychiatric training and practice related to pharmaceutical industry ties.10,11 Whatever the current gaps in general psychiatric education, Warren and colleagues’ study is a wake-up call that forensic psychiatry must teach and practice interviewing and opinion formulation that are informed by both a psychodynamic perspective12 and an awareness of decision-making heuristics.13,14
As Warren and colleagues1 remind us with respect to the deep understanding of PTSD,15 there is a continuing need for psychodynamically informed (though not psychodynamically dominated) forensic psychiatry. To paraphrase Cicero, those who forget history will forever be children, and those who forget an individual's history and the setting in which that history is being taken will forever present naïve views.
Acknowledgments
The authors gratefully acknowledge the very helpful comments and suggestions of Debra Pinals, Ruth Greenberg, Angela Hegarty, Irene Coletsos, Michael Williams, Archie Brodsky, Jason Huffman, and members of the Program in Psychiatry and the Law, Beth Israel Deaconess Medical Center, Harvard Medical School.
- American Academy of Psychiatry and the Law
