Add To Cart

Section 25
Issues of Sensitivity and Specificity in Diagnosing PTSD in Children

Question 25 | Test | Table of Contents

The diagnosis of post-traumatic stress disorder (PTSD) in children, studied empirically for the last 20 years using the diagnostic criteria in the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association [APA], 1980, 1987, 1994), has been shown recently to be controversial and perhaps theoretically inconsistent regarding its original conceptual basis (Cohen, 1998; Yehuda & McFarlane, 1995). Diagnostic constructs used in work with child psychopathology often stem, historically, from an understanding of how the illness develops in and affects adults. The presentation of, and indeed the basis for psychiatric illnesses in children, however, often do differ significantly from the adult version of the illness, raising questions about the conceptual definitions of the illnesses. Differentiating between the child and adult versions of a diagnostic formulation is crucial to the relevance and usefulness of the diagnosis. In addition, controversies that exist for the diagnosis in adults, such as the recent evidence that PTSD is actually a relatively rare disorder with clear biological predisposing factors, also affect the clarity of the diagnosis in children (Yehuda & McFarlane, 1995). Questions regarding how severely traumatized children may meet the diagnostic criteria for an accurate diagnosis of PTSD, the comorbidity of PTSD with many other psychiatric illnesses, and the possibility that PTSD is not a valid diagnostic formulation as it has been applied to children demonstrate that the concept of PTSD may not be operationally sound.

Controversies: Sensitivity Versus Specificity: The current major controversies included in the literature regarding PTSD in children involve the construct validity of the formulation or, more particularly, the sensitivity and the specificity of the diagnosis. If the diagnosis is appropriately sensitive, it will accurately represent all those children suffering from PTSD. This implies a certain level of diagnostic responsiveness to signs and symptoms. Without this level of sensitivity, the diagnosis may be seen as too narrowly conceptualized. When this is the case, children who are suffering from PTSD and who would benefit from treatment may never be diagnosed. On the other hand, if the diagnosis is appropriately specific, it will capture only those individuals actually suffering from PTSD. If the diagnostic formulation is not specific enough or the conceptualization too diffuse, it may overlap with other psychiatric illnesses and become too broadly applied. Diluting the formulation will bring into question the precision of the diagnosis and its capability of singling out those truly in need of treatment. Lack of specificity renders PTSD into a chameleon of a diagnosis, mimicking many other psychiatric illnesses and making it difficult to identify its own specific characteristics. A diagnostic formulation, therefore, must be both adequately sensitive and specific in order to accomplish its goals.

Sensitivity: Controversies involving the sensitivity of the diagnosis of PTSD include questions regarding the number of symptoms from each category (reexperiencing, avoidance, and increased arousal) that must be present in order to diagnose PTSD (Cohen, 1998; Saigh, 1988). More specifically, do PTSD symptoms in children occur on a continuum of severity and frequency, or is PTSD a discrete diagnosis occurring only when the child reaches a particular threshold of symptoms (Terr, 1991)? Following from this question, it is noted that children may not be well served by the adult criteria for PTSD, as their reaction to extreme stress appears to be significantly different from that of adults (Schwartz & Kowalski, 1991). A developmental stage-specific set of diagnostic criteria may be needed to properly identify children with PTSD, or perhaps a separation of severe PTSD stressors into single-episode and chronic categories could be employed (Drell, Siegel, & Gaensbauer, 1993; Eth & Pynoos, 1985; Terr, 1991). Children may experience long periods of avoidance and numbing after experiencing trauma, which precludes the symptom of reexperiencing during that time, and thereby prompts underdiagnosis of PTSD because the child does not meet the classic diagnostic criteria (Arroyo & Eth, 1995). These questions imply that the core features of PTSD as they are understood for children may be too narrow in scope to capture adequately all those suffering the disorder. In other words, PTSD as a conceptualization is not sensitive enough to reliably identify all children suffering the illness. Therefore, the very definition of PTSD, as it applies to children, is called into question.

Specificity: PTSD also is problematic in relation to its specificity. Children may mask the presentation of PTSD with fear and rage, leading to a diagnosis of conduct disorder if they externalize aggression, or of borderline personality disorder if they injure themselves (Goodwin, 1985). Is PTSD, experienced at an early age, a precursor of a personality disorder? The comorbidity of major depressive disorder, dysthymic disorder, substance abuse, and attention deficit hyperactivity disorder (ADHD) with PTSD calls into question the breadth of the PTSD diagnosis, and suggests that the core features of PTSD have been conceptualized too broadly and not specifically enough. The conceptual basis of PTSD in children may be narrower than presently understood (Cohen, 1998). In addition, recent work regarding the neurobiology of PTSD in adults suggests that PTSD is not necessarily a result of a normative, adaptive stress response. Instead, PTSD may be the product of a progressive sensitization of biological systems that render the individual hyperresponsive to traumatic stimuli (kindling), further narrowing the formulation of PTSD from today's standpoint (Yehuda & McFarlane, 1995).  If the core features of PTSD are wider than currently understood, are children being diagnosed with other serious psychiatric diagnoses and not being treated correctly or effectively for their illness? If the core features are narrower than presently employed, is the desire to see PTSD as a normal response to extreme conditions-the stance that initially brought the diagnosis to the fore-no longer empirically relevant or useful? Do predisposing factors and physiological abnormalities that characterize those with PTSD exist instead, implying a threshold level of psychiatric damage in the victim, rather than normal adaptation (Yehuda & McFarlane, 1995)? Does this imply that PTSD, as a mental illness, is really something other than it is currently understood to be?

Significance of the Problem: The U.S. Department of Health and Human Services, Administration for Children and Families (Sedlack & Broadhurst, 1996) found the incidence of physical and sexual abuse of children doubled from 1.4 million to 2.8 million between 1986 and 1993, and that the number of children seriously injured by maltreatment increased almost fourfold, from 143,000 to 570,000, during that same period. War and ethnic cleansing in eastern Europe, ethnic cleansing and AIDS deaths in epidemic proportions in Africa, violence in urban and rural communities in the United States and elsewhere, and hosts of natural disasters worldwide from hurricanes to earthquakes have occurred recently and will continue to occur, traumatically affecting children.  With scientific advances, children undergo increasingly invasive and traumatic medical procedures, such as bone marrow transplants and intensive chemotherapy, to cure them of disease. In all these situations, children are subject to extreme trauma. Clearly, PTSD has been a useful formulation through which children, as victims in the midst of any extremely traumatic context, can be identified as at risk for or suffering due to psychological harm, and treated accordingly. It is important, therefore, that because the diagnosis is undergoing a conceptual metamorphosis, we explore the direction PTSD is taking, as a malleable and evolving conceptual framework, and how that applies to the diagnosis and care of children experiencing extreme stressors. Finally, if social and ethical reasons continue to exist to support the current popular formulation of PTSD in children, is it the responsibility of the disciplines employing the conceptual framework to continue to practice through this imperative?

Framework Issues: Sensitivity of the Diagnosis: When children are exposed to a severe discrete trauma, studies have found that relatively few of them actually fulfill the criteria for PTSD, while many of the children do meet the criteria partially. Using the Hurricane Hugo disaster in 1989 as a basis for a large-scale survey, Shannon, Lonigan, Finch, and Taylor (1994) interviewed 5,687 school-age children who had been exposed to the natural disaster, asking them to give a self-report about their experiences and reactions related to the storm. The authors note that one of their intents was to explore the hypothesis that children experience PTSD symptoms in the face of a natural disaster, despite the fact that prior studies suggested disasters did not produce severe post-traumatic reactions in children at all (Earls, Smith, Reich, & Jung, 1988). The Shannon et al. study found that approximately 5.42% of children and adolescents who were victims of the disaster developed post-traumatic reactions severe enough to be classified as exhibiting PTSD. There was also a large group of children and adolescents, however, found to be experiencing reactions strong enough to create a functional impairment but who did not meet the DSM-III-R (APA, 1987) classification for PTSD. More than 15% of the children in the study met at least two of the criteria for PTSD, and another 25% met at least one criterion.  The study was particularly strong in collecting data on a relatively large sample of children and adolescents within 3 months of the disaster's occurrence, lending considerable credibility to participants' self-reports. The measure used, the Frederick Reaction Index for Children (Frederick, 1985), is a self-report questionnaire with proven validity and reliability with the population tested. The self-report format, however, raises the question of a possible level of exaggeration in the responses. Teachers read the questionnaire to students in their homerooms, where they could easily interact with others in the classroom while answering. No attempt was made by the investigators to standardize the delivery of the questionnaires to the students.  If nearly half the children in the Hurricane Hugo study (Shannon et al., 1994) suffered some kind of functional impairment in school after the disaster, but only 5.42% of them actually met PTSD criteria, it may be questioned how useful the diagnosis is to capture or identify those in need of intervention. The relatively low incidence of PTSD after a discrete disaster is not unusual in the literature; Laor et al. (1997) found only 7.8% of the children who experienced missile attacks had sufficiently serious symptoms to warrant a diagnosis of PTSD. This rate is similar to that found after a school bus accident (Milgram, Toubiana, Klingman, Raviv, & Goldstein, 1988), and higher than that reported after a nuclear power plant accident (Handford et al., 1986).  By contrast, children exposed to chronic, long-term trauma, such as sexual abuse, have been found to have a much higher probability of meeting the full criteria for PTSD. McLeer, Deblinger, Henry, and Orvaschel (1992), found that 43.9% of the 92 sexually abused children they studied in a diagnostic treatment center met the full criteria for PTSD. The criteria, in addition, were met irrespective of the time elapsed since the last abusive episode. The finding that PTSD is common among clinically referred children who have been sexually abused may not be surprising; the subjects reflected a unique subpopulation, most of whom had experienced particularly serious abuse by a close and trusted family member and who had been self-selected for treatment in a clinic known for dealing with challenging cases. Of this group of children, however, 86.5% met the criteria for one or more reexperiencing behaviors, 52.4% met the criteria for three or more avoidant behaviors, and 72.0% met the criteria for two or more symptoms of autonomic hyperreactivity, indicating that most, if not all, of the children in this group either fully or partially met the criteria for PTSD (McLeer et al.). The subjects were assessed for PTSD in structured interviews, employing a PTSD symptom checklist, and were further assessed with several measures for determining evidence of depression, ADHD, and other psychiatric diagnoses.

Though the results of the McLeer et al. study (1992) may give the impression that the PTSD criteria are capturing an indiscriminate, inordinately large percentage of children in this group to be useful, there is another way to view these results. Is it inconceivable that all the children in this chronically sexually abused group might have the symptoms of PTSD? McLeer et al. question whether PTSD may be a much more prevalent disorder of childhood than was once thought. If so, is it not morally and ethically imperative that these children be identified and treated? If no more than 18 years ago the prevailing thought was that children did not suffer from PTSD symptoms at any time, it is not so far-fetched to posit that science has further ground to break regarding the disorder?  That the criteria for PTSD, as presented in the DSM-IV (APA, 1994), do not apply to very young children, was the focus of a pilot study conducted by Scheeringa, Zeanah, Drell, and Larrieu (1995).

Retrospectively reviewing 20 published cases of severely traumatized infants and children younger than 48 months, the authors found that none of the cases met the criteria for a PTSD diagnosis. Problematic with this group of subjects, vis-a-vis meeting the criteria, is that subjective experience is impossible to ascertain in children who are preverbal or barely verbal, making thoughts and feelings difficult for the clinician to infer. The subjects, however, did manifest persistent signs and symptoms that were recognized as impairments to their functioning lasting more than 1 month. The authors proposed that the criteria for PTSD should be revised to be more behaviorally anchored and developmentally sensitive to this age group.  The preceding discussion is limited by the small numbers of studies reviewed; however, several points are clear. The PTSD criteria as they are written do not clearly, reliably, and concisely capture the manifestation of the disorder in children. The formulation, therefore, may be seen as not sensitive enough, and at other times extremely sensitive in identifying those children suffering from PTSD. Each study uses a different measure or set of measures to determine who meets the criteria, and measures are employed by self-report and by structured interview, making comparison difficult between results and generalization of results. At present, there is no gold standard instrument to diagnose PTSD or to monitor its symptom course (Cohen, 1998). It appears from the studies involving natural or man-made disaster types of traumas that the formulation of PTSD may be too narrow and not sensitive enough. It leaves out what intuitively appeals to the concerned and responsible clinician as a good proportion of children who are suffering from PTSD. Though these children might meet, instead, the criteria for an adjustment disorder with anxiety or anxiety disorder not otherwise specified, there is a question of the relevance of treatment options guided by the diagnosis. If the PTSD formulation is not sensitive enough, are children being ignored who might benefit from treatment guided by PTSD principles?

It is clear that PTSD, as it is popularly understood as a conceptual framework, is not well delineated and is in the process of changing. In the case of PTSD in children, the current stress adaptation formulation appears to have been predicated on well-intentioned efforts to highlight the suffering of children in the context of extreme trauma, using the stress paradigm. The need to protect and care humanely for children who are victims of trauma continues to inform studies such as Steiner et al. (1997) that look at adolescents who were abused as children, and who are now incarcerated for their own abusive and antisocial acts. The effort to use PTSD to see these individuals, not as hopeless, hardened criminals but as human beings who can be helped, is to be lauded. The ability to use PTSD to explain the behavior of sexually and physically abused children is particularly useful in that it offers a treatment plan that may be used early in the abused child's life and soon after the trauma, thus avoiding further complications as the child matures. Certainly, the framework is seen as a way to help the trauma victim avoid social stigmatization or misunderstanding, to help others understand, cope with, and manage the often bewildering behaviors and symptoms that may appear in children after a trauma, and to highlight the impact of trauma in a way that is socially and politically cogent. As has been exemplified by the brief overview of literature, the PTSD formulation, however it is used for ethical and moral purposes, leaves much to be desired empirically.
- Tierney, June A.; Post-Traumatic Stress Disorder in Children: Controversies and Unresolved Issues; Journal of Child & Adolescent Psychiatric Nursing, Oct-Dec2000, Vol. 13 Issue 4, p147

Personal Reflection Exercise #11
The preceding section contained information regarding issues of sensitivity and specificity in diagnosing PTSD in children.  Write three case study examples regarding how you might use the content of this section in your practice.
Reviewed 2023

Update
Mental health screening and assessment tools for
forcibly displaced children: a systematic review

- Verhagen, I. L., Noom, M. J., Lindauer, R. J. L., Daams, J. G., & Hein, I. M. (2022). Mental health screening and assessment tools for forcibly displaced children: a systematic review. European journal of psychotraumatology, 13(2), 2126468. https://doi.org/10.1080/20008066.2022.2126468


Peer-Reviewed Journal Article References:
Grolnick, W. S., Schonfeld, D. J., Schreiber, M., Cohen, J., Cole, V., Jaycox, L., Lochman, J., Pfefferbaum, B., Ruggiero, K., Wells, K., Wong, M., & Zatzick, D. (2018). Improving adjustment and resilience in children following a disaster: Addressing research challenges. American Psychologist, 73(3), 215–229.

Hansel, T., Osofsky, H., Speier, A., & Osofsky, J. (2019). Postdisaster recovery and resilience: The mediating influences of mental health and environmental quality of life. Traumatology. Advance online publication.

Kilmer, R. P., Gil-Rivas, V., & Roof, K. A. (2020). Associations between children’s self-system functioning and depressive and posttraumatic stress symptoms following disaster. American Journal of Orthopsychiatry. Advance online publication.

QUESTION 25
In Shannon’s study of children affected by Hurricane Hugo, what conflicting data emerged? To select and enter your answer go to Test.

Test
Section 26
Table of Contents
Top