Article Text

Download PDFPDF
Randomised controlled trial
Current evidence does not support the widespread use of collaborative care for PTSD, but it is limited by not evaluating context
  1. Bradley V Watts
  1. VA National Center for Patient Safety 215 N. Main Street, White River Junction, Vermont, USA
  1. Correspondence to: VA National Center for Patient Safety 215 N. Main Street, White River Junction, Vermont, USA; Bradley.v.watts{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Commentary on: OpenUrlCrossRefPubMed.


Posttraumatic stress disorder (PTSD) is a common disorder that affects many patients in the primary care (PC) setting. There are multiple evidence-based treatments for PTSD, and some of which could easily be delivered in a PC setting. Collaborative care (CC) models have been used to improve depression care in PC. In comparison to depression, PTSD is less commonly treated in PC. There have been several attempts to examine CC models for PTSD; however, the results have been inconclusive. This study seeks to examine the effectiveness of CC for PTSD in Federally Qualified Health Centers (FQHCs).


This study is a randomised, controlled, clinical trial comparing minimally enhanced usual care (MEU) with PTSD care management (PCM) in six FQHCs. The MEU consisted of education to PC providers about trauma, PTSD and evidence-based psychopharmacology. The PCM included those educational interventions plus coordinated care through direct communication with patients and providers to facilitate continuity of care and evidence-based treatment. The unit of randomisation in this study was the patient. The primary outcome was PTSD symptom severity based on the clinician-administered PTSD scale.


Four hundred and four patients with PTSD treated at one of the four FQHCs were randomised to either MEU or PCM. Those receiving PCM showed a decrease in PTSD symptom severity from 71.1 to 46.9, but those in the control group showed a similar decrease from 71.0 to 44.2 (p=0.33). Both groups had similar percentages of participants who no longer met the diagnostic criteria for PTSD after the trial, including 66.7% for PCM and 70.6% for MEU.


This study highlights several important findings when considered in the context of other relevant studies. In this study, CC for PTSD was associated with a substantial improvement in the PTSD symptoms. However, the control condition was also associated similar improvements. The conventional interpretation of these results would typically be that CC is not an effective intervention for PTSD. However, that analysis glosses over the fact that more than half of all patients in the study no longer had PTSD by the end of the study. The authors postulate, “patients had a more favourable prognosis” and “FQHC could have been better positioned to implement care improvements.” These explanations can be amplified through the consideration of alternative frameworks. Pawson and Tilley1 have advanced the idea of Realistic Evaluation which considers not just intervention and outcome, but also the context in which the intervention was used. This approach is especially germane in the case of CC for PTSD which is complex intervention which aims for changes in providers and patients. This framework might interpret the results of this study as basic education regarding the diagnosis and treatment of PTSD to staff of an FQHC (with little specialty mental healthcare access), conferring substantial improvement in patients who have little prior treatment history.

Three other studies of CC for PTSD have been completed. All were performed in either VA or Department of Defence clinical settings. RESPECT-PTSD used a similar intervention as this study.2 Like the current study, RESPECT-PTSD's intervention did not separate from the control group. However, unlike this study, RESPECT-PTSD had a very modest effect on PTSD symptoms. Two studies of CC for PTSD have shown effectiveness compared to control conditions, yet3 ,4 both involved a much more intensive intervention. Their effect on PTSD symptoms was intermediate between PCM and RESPECT-PTSD.

What are possible conclusions of these inconsistent results? It appears that in a clinical context lacking mental health resources, even basic education about PTSD and its effective treatments can have profound effects. In contrast, in settings like the VA where access to mental health is relatively good and knowledge about PTSD is common, models must include intensive interventions to effect even modest changes in symptoms. This is owing to the fact that providers are already more informed and some patients have already failed to respond to basic interventions.

This idea informs clinicians, but it can also map out future research efforts. CC research for treating depression has already embraced these conceptual models. Studies such as 3CM employed a cluster randomisation of clinics that allowed for analysis of context in addition to intervention and outcome.5 Future studies of CC and other complex interventions for PTSD should avail themselves of similar study designs.

Implications for practice

This study does not provide evidence of effectiveness that would support wide-spread use of CC for PTSD. Taken in the context of recent research with similar interventions, it suggests that subtle differences in the intervention and the clinical site may have a large influence on effectiveness.


View Abstract


  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.