Rationale: Posttraumatic stress disorder (PTSD) is a serious public health condition and prevalence is much higher in veterans than in the general population. Although evidence-based PTSD treatments significantly reduce PTSD symptoms, veterans face several barriers to access care and many drop out of therapy prematurely. Approximately 36% of veterans dropout of PTSD treatments; however, dropout rates vary greatly across studies and service settings. Premature dropout prevents veterans from receiving an adequate dose of treatment. Novel treatment delivery modalities, such as videoconferencing and home-based care, have been widely implemented in the Veterans Affairs Healthcare System to overcome barriers to care, which may thereby reduce dropout rates. However, the current literature has revealed mixed findings about which factors contribute to veterans dropping out of PTSD treatment. Further, little research has examined differences in dropout rates between delivery modalities and if veterans’ reasons for dropping out of treatment differ between delivery modalities. The current study has three aims: 1) to determine if there are significant differences in dropout rates from PTSD Prolonged Exposure therapy (PE) between three modalities of care including in-home, in-person therapy (IHIP), home-based telehealth (HBT), and office-based telehealth (OBT); 2) to identify baseline and process factors, including demographic, baseline PTSD and depression symptoms, working alliance, attitudes and beliefs about mental health, and perceived barriers to care, that may predict dropout from PE; and 3) to explore whether there are differences in predominant themes and factors related to Veteran’s reported reasons for dropout among the three modalities of care (IHIP, HBT, and OBT).

Design: This study was a QUANTITATIVE  qualitative explanatory sequential mixed methods study that examined data from an ongoing federally funded randomized controlled trial that evaluated the efficacy of variable length PE delivered via three delivery modalities: IHIP, HBT, and OBT. Participants were 159 veterans aged 18 years or older who were diagnosed with PTSD using the Clinician Assessed PTSD Scale for DSM-5, and who were randomized to one of the three delivery modalities to receive up to 15 sessions of PE. Data from all 159 veterans were included in the quantitative analyses. For the qualitative data analyses, approximately a third of the veterans who dropped out of PE (n = 22), and who was assigned to one of the three delivery modalities, participated in an individual interview about potential contextual and individual factors related to dropping out of PE. Veterans completed the Beck Depression Inventory-II, Working Alliance Inventory-SR, Credibility/Expectancy Questionnaire, a treatment delivery modality preference measure, a demographics questionnaire, and a modified version of the Stigma/Barriers to Care scale. For Aim 1, differences in dropout rates between delivery modality were tested. For Aim 2, a logistic regression was conducted to determine quantitative predictors of dropout. Individual semi-structured interviews were then conducted to explore veterans’ reasons for dropping out and to contextualize the quantitative findings. Team based coding was used to conduct open and focused coding. Qualitative and quantitative results were triangulated to identify which factors predicted veterans’ dropout from PE. For Aim 3, a constant comparison approach was used to identify differences in reasons for dropping out of treatment between the three modalities.

Results: Forty-three percent of veterans dropped out of PE (n = 69) but dropout rates varied by treatment modality; 60% of veterans dropped out from OBT (n = 31), 44% from HBT (n = 24), and 26% (n = 14) from IHIP. Veterans in the OBT condition were more likely to dropout from therapy than individuals in the IHIP condition β = 1.414 p < .01, OR = 4.112, 95% CI [1.083, 9.379]. Compared to veterans in IHIP, Veterans in the HBT condition were also more likely to drop out of PE, β = .801, p = .053, OR = 2.229, 95% CI [.998, 5.025]. Individuals in OBT were more likely to drop out of therapy compared to individuals in HBT, β = .613, p = .120, OR = 1.845, 95% CI [.853, 3.991] but the difference in drop out was not statistically significant. Lower perceived credibility of PE at baseline, greater perceived stigma at baseline, and OBT also predicted a higher likelihood of dropout. Qualitative interviews were conducted with almost a third of the veterans (n = 22) who dropped out of PE. The qualitative interviews revealed that practical barriers, psychological and emotional factors, and the therapeutic context contributed to veterans’ decisions to drop out from PE. There were also differences between modalities about how the modality itself impacted dropout. IHIP had the lowest impact on dropout; veterans in IHIP reported that the modality did not influence their decision to drop out of PE. Half of the veterans interviewed from the HBT condition reported that the modality had an impact, and the majority of veterans in OBT reported that the modality affected their decision to drop out. Veterans in OBT described greater logistical barriers that contributed to dropout. Some veterans in both HBT and OBT said that the telehealth modality was impersonal and contributed to dropping out from PE. There were more reported internet connectivity issues in HBT influencing dropout compared to OBT.

Conclusion: Veterans drop out of PE for several reasons and many veterans have multiple reasons for dropping out of therapy. IHIP can decrease dropout but also may be more costly compared to HBT and OBT. The VA Healthcare System should continue providing PE via telehealth technologies and providers should openly discuss treatment concerns with veterans while in PE in an effort to reduce dropout. Identifying which factors contribute to dropout could lead to the development of engagement strategies to increase retention and maximize clinical gains.