Dual treatments for dual diagnoses

You’re no doubt aware of the long-standing debate among therapists and researchers about the “correct” approach to treating clients who have both mental health and substance use disorders. Many cognitive therapy studies from the 90s excluded people with substance use disorders and many workshops recommended that clients first abstain from drug and alcohol use before receiving therapy for the mental health disorder. At the other extreme, I’ve heard some substance abuse counsellors argue to pay little attention to people’s substance use in favour of focusing on their mental health problems. I’ve heard of counsellors in high profile drug treatment centres providing trauma therapy to people who were homeless and using amphetamines most days. Clinical experience led me to often employ a dual agenda/therapy program because most of my clients were more motivated to improve their mental health than change their drug and alcohol use, however the drug and alcohol use clearly worsened their mental health and ability to manage it.  This study by Tripp and colleague (2020) provides just a little more evidence for the integrated approach.

They were conducting a randomized controlled trial of two 12-16 session individual therapies for 107 Veterans with PostTraumatic Stress Disorder (PTSD) and Substance Use Disorders (SUD): COPE and Safety Seeking. COPE combines prolonged exposure for PTSD with relapse prevention for SUD. Sessions 1-3 focus on goal setting, psychoeducation and CBT strategies. From session 4, recipients undertake an in vivo or imaginal exposure exercise of 30-45 minutes and the final 15 minutes of the session focuses on relapse prevention strategies for SUD. Safety Seeking does not involve discussions about traumatic events, but instead focuses on helping individuals to envision feeling safe and then teaches coping strategies to achieve that feeling. The manual covers 25 diverse topics from grounding strategies to self-nurturing. Tripp and colleagues measured self-reported alcohol use (via the Substance Use Inventory) and PTSD symptoms (via PCL-5) each week. They were interested in seeing which relationship was stronger - PTSD symptoms predicting subsequent drinking, or drinking predicting subsequent PTSD symptoms – and whether this differed between treatments.

Irrespective of which treatment Veterans received, there was a statistically significant reduction in drinking over time. Controlling for time, therapy and baseline drinking, greater PTSD severity predicted greater future alcohol use. The effects were fairly small. The range of the PCL-5 is 0-80. An increase in PCL-5 scores from 26 to 43 was associated with a 0.17 of a standard deviation increase in drinking, and an increase in PCL-5 scores from 26 to 59 was associated with a 0.39 of a standard deviation increase in alcohol. The relationship between PTSD severity and subsequent alcohol use was not influenced by which therapy participants received. Those who received COPE had lower alcohol use overall than those who received Safety Seeking.

Greater alcohol use during treatment predicted greater PTSD symptom severity the following week. Again, the effects were small. The difference between no drink and having one drink was associated with 1.75 more points on the PCL-5 the next week. The difference between 0 and 3.2 drinks was associated with 3.51 more points on the PCL-5 the next week (d  = 0.20). Again, the COPE condition had a greater effect on reducing PTSD symptoms than safety seeking.

The authors were somewhat surprised that alcohol use predicted subsequent PTSD severity, which suggests that there is a reciprocal relationship between PTSD and alcohol use whereby people drink to self-medicate trauma symptoms but alcohol in turn, worsens those symptoms. The authors suggested that integrated treatment of PTSD and alcohol use is preferable to treating each problem sequentially.

As interesting as these results are, there are limitations to the study design which means we have to treat the data cautiously and keep them in perspective. Retention declined in the latter third of the study. The study started only with those who had attended at least one session and by session 4 85% remained, but this dropped to 63% by session 8 and was only 28% at session 12. Although it would be tempting to interpret this as the participants having got what they needed from therapy ahead of time, looking at the parent study from which Tripp et al.’s data was taken, the mean PTSD scores measured by the CAPS at the end of treatment were estimated to be 25.8 [22.1, 29.6] for COPE and 32.9 [29.3, 36.6] for Safety Seeking (Norman et al., 2019), where the suggested threshold for recovery is a score < 12. The sample experienced significant pathology and although COPE was clearly a superior program to Safety Seeking, there remains plenty of room for improvement. For the purposes of understanding PTSD symptom-alcohol relations, the attrition wasn’t properly modelled: data was assumed to be missing at random and it certainly wouldn’t have been. It is more likely that those who stuck around to provide data were more cooperative and in better psychological health. The change in “concentration” of better functioning individuals over time may be both due to effective treatments working early AND the loss of those who don’t respond. Clinical experience would suggest that those with heavier drinking would be more likely to drop out earlier than lighter drinkers and abstainers. It’s hard to see how these factors would erase the relationships between PTSD symptoms and alcohol use, but the magnitude of their association is probably not well estimated in this sample.

Clinical Implications

• Alcohol use makes PTSD symptoms worse and PTSD symptoms lead to greater drinking. Wherever possible, address both problems together unless the alcohol use is so heavy that it interferes with psychotherapy (i.e., the person cannot attend an appointment sober).

• Although there is no substantial difference between some therapy protocols for certain high prevalence problems like depression, this cannot be extrapolated to all problems. For some psychological problems there is evidence that some approaches are more effective than others. It is our responsibility as clinicians to follow the literature and know when the approach we use makes a difference. In this study, Safety Seeking, which did not explicitly address PTSD or substance use but was hoped to be helpful with both, was less effective at reducing PTSD symptoms and less effective in reducing drinking among those with significant baseline alcohol use.

For the original article, go to:
https://psycnet.apa.org/record/2020-13730-001

REFERENCES
Norman, S.B., Trim, R., Haller, M., Davis, B.C., Myers, U.S., Colvonen, P.J., Blanes, E., Lyons, R., Siegel, E.Y., Angkaw, A.C., Norman, G.J., & Mayes, T. (2019). Efficacy of integrated exposure therapy vs integrated coping skills therapy for comorbid posttraumatic stress disorder and alcohol use: A randomized clinical trial. JAMA Psychiatry, 76, 791-799.

Tripp, J.C., Worley, M.J., Straus, E., Angkaw, A.C., Trim, R.S., & Norman, S.B. (2020). Bidirectional relationship of posttraumatic stress disorder (PTSD) symptom severity and alcohol use over the course of integrated treatment. Psychology of Addictive Behaviors, 34, 506-511.