Improving the Assessment of COVID-19–Associated Posttraumatic Stress Disorder—Reply

In Reply We thank Marx and colleagues for their thoughtful comment on our recent article investigating the prevalence of posttraumatic stress disorder (PTSD) in postacute COVID-19 illness.¹ We fully agree with the authors on the importance of assessing COVID-19–associated PTSD through the criterion-standard Clinician Administered PTSD Scale for DSM-5 (CAPS-5).² Based on the CAPS-5, we found a 30.2% prevalence of PTSD in patients after acute COVID-19. Our results partly match the pooled PTSD prevalence of 21.94% (95% CI, 9.37-43.31) found in the meta-analysis by Cénat et al,³ which Marx and colleagues cited. However, this meta-analysis³ pooled actual patients with individuals from the general population. Hence, the higher PTSD prevalence found in our study in patients who experienced acute COVID-19 illness is not surprising and in line with the rate of 32.2% (95% CI, 23.7-42.0) reported by Rogers et al⁴ in postacute patients after previous coronavirus epidemics. Because the field is rapidly expanding, it is likely that many studies will be added and used to meta-analyze and synthesize published evidence.

Regarding the methodological concerns raised by Marx and colleagues on criterion A, all patients in the study sought and received treatment at the emergency department, where their clinical conditions were considered severe. Consistently, 81% of participants were hospitalized. Medical events and injuries related to COVID-19 were assessed during the interview according to the CAPS-5² and the Life Events Checklist for DSM-5 (LEC-5).⁵ Patients were asked to describe their traumatic experience in their own words, then the interviewer asked for more details regarding the event. All participants with PTSD reported to have directly experienced life-threatening medical symptoms and/or personally witnessed the death of other patients at the emergency department or during hospitalization.

We understand Marx and colleagues’ surprise that in our study the only acute COVID-19 characteristic associated with PTSD diagnosis was delirium or agitation. This result may be explained in light of the fact that COVID-19 is a life-threating illness per se and was recently reported to be the leading cause of death in the US.⁶ Accordingly, our data may be related to the unpredictable severity of the illness, which can quickly escalate to serious and possibly life-threatening. In the PTSD group, we found more persistent medical symptoms specifically related to COVID-19 (ie, fatigue, dyspnea, ageusia, and dysgeusia), compared with the non-PTSD group. These results seem to suggest that the characteristics of both the acute phase and postacute phase of COVID-19 might contribute to patients reexperiencing trauma associated with the illness.

Regarding the severity of symptoms, the mean (SD) CAPS-5 score differed significantly for patients with PTSD (38.80 [10.32]) and those without PTSD (4.96 [5.11]) (P < .001). In patients reporting PTSD, mean symptom severity was within the moderate-severity range, which often requires therapeutic intervention. Nevertheless, the CAPS-5 total score in the PTSD group ranged from 14 to 57, covering a wide range of degrees of severity. Future studies could stratify patients according to PTSD symptom severity to better identify factors associated with PTSD, as well as protective measures, and to help design therapeutic and prevention strategies.

Corresponding Author: Delfina Janiri, MD, Department of Psychiatry, Fondazione Policlinico Universitario Agostino Gemelli IRCCS, Largo Francesco Vito 1, 00168 Rome, Italy (delfina.janiri@uniroma1.it).

Published Online: June 2, 2021. doi:10.1001/jamapsychiatry.2021.1126

Conflict of Interest Disclosures: Dr Sani reports personal fees from Janssen, Angelini Spa, and Lundbeck. No other disclosures were reported.