Editorials

Post-traumatic stress disorder in intensive care survivors

Natalia Wege, Mary Princip

BMJ 2025; 389 doi: https://doi.org/10.1136/bmj.r841 (Published 07 May 2025)Cite this as: BMJ 2025;389:r841

A role for primary care

Advances in intensive care medicine have improved survival rates, yet post-traumatic stress disorder (PTSD) remains a prevalent and often underdiagnosed consequence among patients who are discharged from intensive care units.1 Approximately 20% of these individuals develop PTSD symptoms, which can lead to lasting impairments in quality of life, occupational functioning, and overall physical health.23 Contributing factors include exposure to life threatening conditions, invasive procedures, prolonged isolation, and a profound loss of control.45Early identification and targeted support of PTSD symptoms after ICU care are essential because systematic screening and tailored intervention can significantly reduce long term psychiatric impairment.6

Despite increasing awareness of PTSD after ICU care, access to effective treatments remains limited. Cognitive behavioural therapy and eye movement desensitisation and reprocessing are well established treatments for PTSD, yet their accessibility is limited due to long wait times and workforce shortages.7 Given these barriers, and considering that many of these affected patients initially reconnect with the healthcare system through general practitioners (GPs), primary care settings might have a role for feasible, scalable interventions. GPs frequently serve as first line providers following ICU discharge and are well positioned to identify early signs of mental health impairment and to deliver early stage mental health support. However, structured interventions in primary care contexts for PTSD have been largely absent.

A linked study addresses this gap. Gensichen and colleagues (doi: 10.1136/bmj-2024-082092) conducted a multicentre, observer blind, randomised controlled trial evaluating a novel, brief, GP-led narrative exposure therapy tailored for people discharged from the ICU.8

The study, involving 319 general practices in Germany, tested an intervention consisting of three structured GP consultations and eight follow-up nurse interactions, targeting post-traumatic stress symptoms. The primary outcome was the severity of post-traumatic stress symptoms at six months, measured using the post-traumatic diagnostic scale for DSM-5 (PDS-5), a validated 20 item patient reported outcome measure (range 0-80). The predefined minimal clinically important difference was six points.

The findings suggest that the intervention was feasible, with over 90% of participating GPs adhering to the structured protocol, despite known time constraints. Nearly a third of patients in the intervention group met a more than 50% reduction in PTSD symptoms compared with 12.6% in the control group. These results represent a clinically meaningful outcome considering the brief and low intensity nature of the intervention. Additionally, improvements in secondary outcomes such as depression, disability, and quality of life, indicated potential broader psychosocial benefits beyond reduction of post-traumatic stress symptoms.

However, the intervention did not meet the predefined minimal clinically important difference for the primary outcome. At six months, symptom reduction averaged 1.5 points in the control group and 6.2 points in the intervention group, resulting in a group difference of 4.7 points. At 12 months, symptom reduction was 2.5 points in the control group and 7.9 points in the intervention group, corresponding to a group difference of 5.4 points. In both cases, the differences did not reach the predefined threshold of six points. Moreover, the intervention did not impact core symptom clusters such as avoidance and hyperarousal. These findings suggest that narrative processing and biographical integration may alleviate certain cognitive affective dimensions of PTSD (eg, intrusions, mood, and distress), yet additional emotionally activating or exposure based components might be required to address the full spectrum of symptoms. Combining exposure based methods with cognitive restructuring could target these resistant symptom clusters more effectively.2

Another limitation concerns the inclusion criteria, which excluded patients with severe PTSS (PDS-5 score of >70) and people already receiving psychiatric care. While this approach enhances generalisability to typical primary care populations, it also limits the applicability of the findings for people with the highest clinical need. Moreover, the intervention was delivered by GPs following brief training, without ongoing supervision or structured case discussion, which may have constrained the therapeutic depth and adaptability of the intervention.

Although the observed treatment effects were moderate, their importance lies in the intervention’s potential for scalability and broad accessibility, especially in healthcare systems facing limited specialist resources. This represents an important step towards designing and evaluating trauma informed primary care interventions.

Integration of structured PTSD interventions and trauma informed principles into GP training, including brief screening tools and stepped care models, could help to address the growing mental health burden, particularly in underserved areas. Moreover, embedding trauma insights into routine care, as emphasised by McBain and Cordova,5 such as anticipatory guidance, validation of trauma related symptoms, and pacing of medical communication, may further enhance recovery and resilience.

That said, addressing PTSD in people who were in ICUs requires a broad approach, including strengthening interfaces between ICU and primary care, embedding trauma informed diagnostics earlier in the treatment pathway, and establishing preventive measures during ICU stays.

The PICTURE trial represents an important advance in trauma informed primary care interventions, bridging acute care and long term psychotherapeutic support. While not a replacement for specialised psychiatric treatment, such models offer a pragmatic strategy to reduce the psychological burden of critical illness. The ability to deliver structured, low risk psychotherapeutical support within a familiar, trusted setting is valuable; yet feasibility alone should not define the limits of evidence based care. Feasibility is a starting point, not the endpoint, for the development of high quality GP-led interventions for post-traumatic stress symptoms after ICU care. As research continues, the challenge will be to refine these early interventions without diluting their therapeutic effectiveness. Future research should focus on refining content of therapy, optimising delivery of care, and ensuring broad integration across healthcare systems.