The Methodological Problem of Identifying Criterion A Traumatic Events During the COVID‐19 Era: A Commentary on Karatzias et al. (2020)

Abstract

A recent study published in the Journal of Traumatic Stress demonstrated that posttraumatic stress disorder (PTSD) rates in Ireland are as high as 17.7% and that this could be related to the COVID‐19 pandemic (Karatzias et al., 2020). However, this number is probably skewed, as the fundamental requirement for a PTSD diagnosis—namely, a life‐threatening or severely stressful event—was not fulfilled. In this comment, the consideration of COVID‐19–related PTSD to represent a diagnosis is questioned based on the definitions of PTSD in the ICD‐11 and DSM‐5.

A recent study by Karatzias et al. (2020) investigated posttraumatic stress symptoms (PTSS) in Ireland during the COVID‐19 pandemic. It showed that 17.7% of the sample have possible posttraumatic stress disorder (PTSD) related to the COVID‐19 pandemic. Such a number is staggering considering that in a United States–based sample, Cloitre and colleagues (2019) found a rate of only 3.4% when using the same PTSD scale. However, in the study by Karatzias et al. (2020), there are several methodological issues related to how PTSD was measured.

Following exposure to a traumatic event, PTSD can be diagnosed after an individual has experienced 1 month of specific symptoms, such as the reexperiencing of the event or the feeling of being threatened (American Psychiatric Association [APA], 2014; World Health Organization [WHO], 2018). It is important to note that integral to this diagnosis is the identification of the traumatic event to which these symptoms are related. Thus, the event must always be determined, and according to both the 11th revision of the International Classification of Diseases (ICD‐11) and the Diagnostic and Statistical Manual of Mental Disorders (DSM‐5), this event must constitute a life‐threatening or extremely stressful event (APA, 2014; WHO, 2018). Examples of such events are sexual assault, violence, or the sudden or unexpected death of a loved one (Cloitre et al., 2018).

However, in the study by Karatzias et al., (2020), the questions used to determine the presence of PTSD were based on the participant's COVID‐19 experience. Thus, the question the study really asked of its participants was: “What is your experience with COVID‐19?” Such experiences vary greatly, with very few people outside of the medical profession experiencing direct contact with COVID‐19–infected individuals. The “COVID‐19 experience” can mean anything, from nonthreatening experiences to COVID‐19 infection. The vagueness of this concept becomes clear when it is compared to concrete events, such as sexual assault or violent robbery. People who are in quarantine might have more exposure to domestic abuse, child abuse, or sexual abuse, all possible consequences of quarantine as suggested by many commentators. However, when one's symptoms are considered relative to something as broad and abstract as the “COVID‐19 experience,” this information on the specific traumatic event is lost.

Thus, we do not know whether the event the participants experienced was life‐threatening or extremely stressful. Without such an event, the requirements for PTSD are not fulfilled. The danger of applying PTSD criteria without considering the event is that overestimations might occur, which was previously noted by Durodié and Wainwright (2019) in their review of studies on terrorism that use PTSD symptom scales.

Karatzias et al. (2020) argue that there is a rationale for using such a broad and abstract event as the “COVID‐19 experience” as the basis for a study on the association between COVID‐19 and PTSD. The authors state that:

Data derived from previous outbreaks of respiratory infections, such as severe acute respiratory syndrome (SARS), demonstrate that being infected or the threat of being infected can be a potentially traumatic event and increase the risk of developing PTSD… The COVID‐19 pandemic represents a threatening and potentially traumatic event, as it can lead to hospitalization and even death (p. 365).

However, the studies Karatzias et al. (2020) cite do not support this argument. Studies by Mak et al. (2009) and Chen et al. (2006) examined survivors of SARS, whereas Wu et al. (2009) studied individuals working in hospitals who were exposed to SARS. Such events can be considered life‐threatening because participants were directly exposed to a potentially fatal disease; in other words, these events fulfill the basic requirement of a PTSD diagnosis, namely the experience of a life‐threatening or extremely stressful event. Thus, the sources cited by Karatzias et al. (2020) show that COVID‐19 is a potentially traumatic event if you are directly exposed to the virus. However, in their own study, it is not clear at all whether participants were directly exposed to life‐threatening danger caused by COVID‐19.

To know whether participants answer items on PTSD assessments using a traumatic event as a point of reference, a checklist or scale of life events is often used in such studies. The authors refer to Cloitre et al. (2018) for the use of their PTSD scale, yet Cloitre et al. (2018) also used the Life Events Checklist to determine whether someone fulfilled the first requirement for PTSD. A consideration of the nature of the event decreases the number of people who can be classified as suffering from PTSD, thus causing a large difference between a study that uses an assessment of life events finding 3.4% of the sample to have possible PTSD and a rate of 17.7% in a study that did not use a similar assessment of life events (Karatzias, et al., 2020).

Naturally, the problems raised herein do not mean that COVID‐19 did not affect public mental health. Population studies on mental health during COVID‐19 are needed and will be useful for policymakers and therapists. The problem with the study by Karatzias, et al. (2020) is that there is no clear indication of what caused participants’ mental distress. The authors briefly acknowledge the limitations of their study, but these limitations call into question whether there is something such as “COVID‐19–related PTSD” and whether the high percentage of 17.7% is accurate (Karatzias et al., 2020).

When trying to study the impact of the “COVID‐19 experience,” we should perhaps start to look at other diagnoses; for example, an adjustment disorder might be more appropriate. As there is no concrete event to which all symptoms can be tied, PTSD is likely not suitable for the goal studies such as that by Karatzias et al. (2020) aim to achieve.