Integrated, exposure-based treatment for PTSD and comorbid substance use disorders: Predictors of treatment dropout

Highlights

• To date, one study has examined dropout from treatment targeting PTSD/SUD simultaneously.

• 43% of participants dropped out of treatment.

• The majority of dropout occurred in the later stages of treatment (session 9 and 10).

• Greater baseline PTSD symptom severity was associated with dropout.

• Procedural changes regarding assessment during treatment are suggested.

Abstract

High rates of comorbid posttraumatic stress disorder (PTSD) and substance use disorders (SUD) have been noted in veteran populations. Fortunately, there are a number of evidence-based psychotherapies designed to address comorbid PTSD and SUD. However, treatments targeting PTSD and SUD simultaneously often report high dropout rates. To date, only one study has examined predictors of dropout from PTSD/SUD treatment. To address this gap in the literature, this study aimed to 1) examine when in the course of treatment dropout occurred, and 2) identify predictors of dropout from a concurrent treatment for PTSD and SUD. Participants were 51 male and female veterans diagnosed with current PTSD and SUD. All participants completed at least one session of a cognitive-behavioral treatment (COPE) designed to simultaneously address PTSD and SUD symptoms. Of the 51 participants, 22 (43.1%) dropped out of treatment prior to completing the full 12 session COPE protocol. Results indicated that the majority of dropout (55%) occurred after session 6, with the largest amount of dropout occurring between sessions 9 and 10. Results also indicated a marginally significant relationship between greater baseline PTSD symptom severity and premature dropout. These findings highlight inconsistencies related to timing and predictors of dropout, as well as the dearth of information noted about treatment dropout within PTSD and SUD literature. Suggestions for procedural changes, such as implementing continual symptom assessments during treatment and increasing dialog between provider and patient about dropout were made with the hopes of increasing consistency of findings and eventually reducing treatment dropout.

Introduction

A recent meta-analysis estimated that 23% of Operation Enduring Freedom (OEF) and Operation Iraqi Freedom (OIF) veterans meet criteria for posttraumatic stress disorder (PTSD; Fulton et al., 2015). PTSD is a chronic disorder characterized by physiological hyperarousal, avoidant behaviors and maladaptive cognitions resulting from directly experiencing or witnessing a life threatening event(s) (American Psychiatric Association, 2013). Although some PTSD symptoms (e.g., hypervigilance) may be considered adaptive in the combat theatre given threat of physical harm, these symptoms can become debilitating once the veteran returns home (Breslau, Lucia, & Davis, 2004). Research has linked PTSD to deficits in occupational performance (Taylor, Wald, & Asmundson, 2006), social functioning (Frueh, Turner, Beidel, & Cahill, 2001), and physical health (Jakupcak, Luterek, Hunt, Conybeare, & McFall, 2008). PTSD is also associated with decreased quality of life (Gill et al., 2014), elevated rates of suicidal ideation (Jakeupcak et al., 2009), and substance use disorders (SUD; Carlson et al., 2010). In fact, veterans with PTSD are roughly three times more likely than veterans without PTSD to develop an SUD than civilians (Petrakis, Rosenheck, & Desai, 2011).

Research on the etiology of comorbid PTSD and SUD suggest that, for most individuals, the onset of the PTSD precedes the onset of the SUD (Hien et al., 2000, Jacobsen et al., 2001), and that alcohol and drugs are commonly used to self-medicate distressing PTSD symptoms, such as nightmares, intrusive memories, and hyperarousal (Back et al., 2012, Back et al., 2014, Back et al., 2014, Brady et al., 2004, Khantzian, 1985, Possemato et al., 2015). Substances such as alcohol reduce physiological arousal, increase disinhibition and provide temporary cognitive distractions (Grosso et al., 2014). Although substance use may result in short-term relief of symptoms, the long-term consequences of addiction can be severe. For instance, veterans diagnosed with PTSD and comorbid SUD tend to display higher rates of homelessness and are more likely to receive VA disability than individuals with PTSD alone (Bowe & Rosenheck, 2015). Veterans with PTSD and SUD also tend to display greater levels of intolerance for uncertainty, lower tolerance for distress (Banducci, Bujarski, Bonn-Miller, Patel, & Connolly, 2016) and have poorer treatment outcomes (Norman, Tate, Anderson, & Brown, 2007).

Fortunately, a variety of treatments have been developed to address PTSD symptoms. For instance, exposure (e.g., Prolonged Exposure; PE), cognitive (e.g., Cognitive Processing Therapy-Cognitive Therapy; CPT-C) and socially focused treatments (e.g., Interpersonal Psychotherapy; IPT) have shown to be effective in reducing PTSD symptoms (Foa et al., 2013, Markowitz et al., 2015, Walter et al., 2014). More recently, a number of evidence-based psychotherapies have been developed to address comorbid PTSD and SUD simultaneously (Carrico et al., 2015, McCauley et al., 2012). These treatments seamlessly integrate components for both disorders and have been found to provide significant reduction in symptoms of both conditions. However, treatment dropout is a significant problem in both PTSD and SUD treatment populations. A recent meta-analysis found that the average dropout rate for trauma-focused PTSD treatments was 36% (Imel, Laska, Jakupcak, & Simpson, 2013). Similar dropout rates (30–40%) have been reported in SUD treatments (Kelly and Moos, 2003, Rabinowitz and Marjefsky, 1998). Individuals that dropout of treatment tend to report less symptom reduction and increased future service utilization for both PTSD (Tuerk et al., 2013) and SUD (Moos and Moos, 2003, Moos et al., 1995). Among individuals seeking treatment for comorbid PTSD and SUD, dropout rates as high as 61% have been reported (Brady et al., 2001, Mills et al., 2012, Zandberg et al., 2016). Given this, investigations of predictors of dropout would help inform treatment design and retain patients in treatment longer, which is associated with increased benefits and more favorable treatment outcomes (Tuerk et al., 2013).

Findings from previous studies indicate that the vast majority of subjects dropout from PTSD treatment (Gutner et al., 2016, Szafranski et al., 2015) and SUD treatment (McKellar et al., 2006, McMahon et al., 1999) early on, and before mid-treatment. Studies examining dropout from PTSD treatment show that demographic and clinical predictors of dropout include younger age (Gros, Yoder, Tuerk, Lozano, & Acierno, 2011), male gender (van Minnen Arntz, & Keijsers, 2002), African American race (Lester, Artz, Resick, & Young-Xu, 2010), lower levels of education (Rizvi, Vogt, & Resick, 2009), higher military rank (Szafranski et al., 2016), greater concurrent drug use (Szafranski, Gros, Menefee, Wanner, & Norton, 2014), lower income (Galovski, Blain, Mott, Elwood, & Houle, 2012), greater disability status and lower social support (Gros, Price, Yuen, & Acierno, 2013), and higher pretreatment symptom severity (Garcia, Kelley, Rentz, & Lee, 2011). Similarly, studies examining dropout from SUD treatment include younger age, lower income, being unemployed (Mertens & Weisner, 2000), African American race (Milligan, Nich, & Carroll, 2004), lower education level (Mammo & Weinbaum, 1993), female gender (Arfken, Klein, di Menza, & Schuster, 2001) and more frequent drug use (McKellar, Kelly, Harris, & Moos, 2006).

Studies have almost exclusively examined dropout from PTSD and SUD treatments separately. This is surprising given the recent evidence that integrated treatments targeting PTSD and SUD simultaneously are effective in reducing symptoms of both disorders (Foa et al., 2013, McGovern et al., 2015, Roberts et al., 2015). In fact, only one study to date has examined dropout from concurrent treatments for PTSD and alcohol use disorder (Zandberg, Rosenfield, Alpert, McLean, & Foa, 2016). Zandberg, Rosenfield, Alpert, McLean, and Foa (2016) found that accidents and “other” types of trauma were associated with the highest rates of dropout, whereas physical assault was associated with the lowest amount of dropout. Interestingly, both fast and slow rates of PTSD symptom improvement predicted treatment attrition. To address current gaps in the literature, this study aimed to 1) examine when in the course of treatment dropout occurred, and 2) identify predictors of dropout from a concurrent treatment for PTSD and SUD. Consistent with previous studies (Gutner et al., 2016, McKellar et al., 2006, McMahon et al., 1999, Szafranski et al., 2015), it was hypothesized that the majority of dropout would occur prior to mid-treatment. Younger age (Gros et al., 2011, Mertens and Weisner, 2000), African American race (Lester et al., 2010, Milligan et al., 2004), lower education (Mammo and Weinbaum, 1993, Rizvi et al., 2009), greater concurrent drug use (McKellar et al., 2006, Szafranski et al., 2014), lower income (Galovski, Blain, Mott, Elwood, & Houle, 2012) and higher pretreatment symptom severity (Garcia et al., 2011, McKellar et al., 2006) were expected to predict dropout from treatment.