What Happens After Intensive Treatment? Post-Discharge Skill Use and Affect as Predictors of Depression Outcomes

What Happens After Intensive Treatment? Post-Discharge Skill Use and Affect as Predictors of Depression Outcomes | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article What Happens After Intensive Treatment? Post-Discharge Skill Use and Affect as Predictors of Depression Outcomes Shaan F. McGhie, Marie Forgeard, Kailyn Fan, Richard J. McNally, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7614725/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 12 Mar, 2026 Read the published version in Cognitive Therapy and Research → Version 1 posted 9 You are reading this latest preprint version Abstract Background. Patients may be vulnerable to relapse in the period immediately following discharge from intensive psychiatric care. Additional attention to factors that support continued recovery after discharge is therefore warranted. Methods. This study examined whether patterns of therapeutic skills use and affect during the acute post-discharge period predicted continued improvement in depression symptoms of patients who completed a partial hospital program. Using ecological momentary assessment and a predictive modeling approach, we assessed how individual differences in therapeutic skills use and their affective responses to skills use influenced recovery. Results. The final model retained five predictors and achieved acceptable discriminative ability (AUC = .74). Total use of behavioral activation predicted improvement (OR = 1.24). A within-person increase in positive affect following CBT skill use predicted non-improvement (OR = .84). Conclusions. These findings suggest that behavioral activation is an effective skill for sustaining depression symptom improvement in the weeks after discharging from a partial hospital program. Cognitive behavioral therapy relapse prevention prediction ecological momentary assessment depression Figures Figure 1 Introduction Mood and anxiety disorders are often chronic, and patients frequently reexperience symptoms after initial remission (Bruce et al., 2005; Burcusa & Iacono, 2007; Levy et al., 2021; Richards, 2011; Yonkers et al., 1996, 2003). Patients whose symptoms resurface will likely require additional treatment, further consuming scarce clinical resources and prolonging the distress of the individual. Thus, clinicians would benefit from better understanding the factors that predict stable recovery from these disorders following treatment. Novel statistical methods have been deployed to achieve this goal, including attempts to predict which treatment is better suited to a particular individual (Cohen et al., 2020; DeRubeis et al., 2014; Huibers et al., 2015). Machine learning can identify relations among variables in a dataset and then use those learned relations to predict patterns in new data (Coutanche & Hallion, 2020). This emphasis on prediction rather than explanation is well-suited to address the needs of mental health outcome research. This is what clinicians do every day as they treat their patients: they draw upon their knowledge of the literature and past patients (data) and based on patterns they have observed, they decide what treatment will likely be the most beneficial for specific patients (prediction). However, clinicians are constrained by a limited set of data (i.e., only their own caseloads and experience), and the accuracy of their predictions is not quantified. Accordingly, researchers have used machine learning to predict treatment assignment that offers the best fit for individuals (DeRubeis, 2019). For example, studies have examined the relative benefit for a particular depressed patient, based on pre-treatment characteristics, of cognitive behavioral therapy (CBT) versus a selective serotonin reuptake inhibitor (SSRI; DeRubeis et al., 2014), CBT versus psychodynamic therapy (Cohen et al., 2020), and CBT versus interpersonal therapy (Huibers et al., 2015), among others. The above studies predicted outcomes based on assignment to different treatments altogether. However, clinicians working in naturalistic settings often draw from multiple protocols to teach specific skills to the same patient (Cook et al., 2010), such as mindfulness (Michalak et al., 2020), dialectical behavior therapy (DBT) skills (DiGiorgio et al., 2010; Linehan, 2018), and cognitive behavior therapy skills, among others. The factors that influence a clinician’s decision to emphasize some skills over others vary (DiGiorgio et al., 2010; Michalak et al., 2020), and there is currently no data-driven way to decide what skills would optimize outcome for specific patients. To our knowledge, only one study has used prediction methods to identify which skills are best suited to an individual (Webb et al., 2021). The authors of that study used patient characteristics to predict which skills, either from CBT or DBT, were most closely tied to positive affect in the same hours they were used. However, short-term increases in positive affect do not necessarily signify whether these skills are beneficial overall. Indeed, many maladaptive coping strategies, such as drinking alcohol, avoidance, or drug use, boost higher positive affect in the moment, yet lead to negative long-term consequences. Conversely, increased positive affect after using a skill effectively might reinforce continued use of that skill, increasing the likelihood that patients implement behavioral change that is helpful for them. Thus, it is important to test whether changes in affect after using therapeutic skills are related to symptom improvement. As is typically done in outcome research, all of the predictive work reviewed thus far has focused on improvement from the beginning to the end of treatment. However, as discussed earlier, the time after discharging from treatment could be a fruitful period to ascertain factors that could predict stable recovery. This period is critical for patients to implement the skills they have learned in treatment. This may be especially true for those discharging from partial hospitalization or other similar intensive programs, whose lives have been suddenly interrupted by acute difficulties and intensive treatment delivered in a setting drastically different from that of their daily lives. Yet, little is known about which skills learned in intensive treatment offer the most benefit to patients after they leave. For example, particular skills acquired during treatment (e.g., cognitive reappraisal) may bolster recovery after treatment completion, especially for shorter-term and more acute levels of care. These may differ from the skills that are most useful to focus on during treatment. For instance, certain skills might be most useful for stabilization in acute distress (e.g., distress tolerance, emotion regulation) while others might be better employed when not in acute distress and when there is more time to master the skill (e.g., cognitive restructuring, exposure). Thus, outcome prediction studies should identify which skills are most useful for a patient’s continued improvement after treatment ends. The current project aims to address this question by identifying which therapeutic skills offer the most benefit for continued symptom improvement after discharge from a partial hospitalization program (PHP). The current study synthesizes the unanswered questions above to ask: (a) which individual skills are helpful for continued improvement after treatment, and (b) does how one feels shortly after using a skill indicate how helpful that skill is for symptom improvement? To answer these questions, we collected ecological momentary assessment (EMA) data (i.e., data collected multiple times throughout the day as patients go about their daily lives), during the two weeks post-discharge from a partial hospitalization program. EMA offers the opportunity to examine how these skills are being used in an ecologically valid manner that minimizes recall bias by having participants report skill use every few hours. Conducting EMA after discharge may illuminate how patients use skills without the structure and oversight of a treatment program or therapist. This intensive data collection also provides an opportunity to determine how skills and affect might covary over time within an individual. In addition to which skills they used, patients reported how they were feeling in the moment they completed each survey. We then calculated within-person correlations between therapeutic skills use (CBT skills or DBT skills) and affect, both measured every few hours over the course of the two weeks for each participant individually (as further described below). We examined the predictive validity of this assocatiation by predicting which patients maintained their symptom reduction two weeks after their discharge date. To do so, we separated patients into two outcome groups. Participants were included in the improvement group if they reported clinically-significant reduction in symptoms two weeks after discharge or if they discharged with few symptoms and retained this level, otherwise participants were in the non-improvement group. We determined the strength of this prediction model for clinical decision-making by using k-fold cross-validation and examining the resulting specificity and sensitivity. We hypothesized that stronger within-person associations between therapeutic skills and positive affect (cognitive behavior therapy skills-positive affect correlation; dialectical behavior skills - positive affect correlation) would predict decreased depression symptoms two weeks after discharge. We also hypothesized that the cross-fold validated prediction of continued improvement versus non-improvement from post-discharge skills use and the within-person association between positive affect and skills use would result in at least 80% accuracy. We also explored the added value of additional predictors, including individual mean and standard deviation of positive and negative affect, which have been previously linked to mental health (Heininga & Kuppens, 2021). Methods Participants Participants were 83 adults who had recently discharged from a partial hospitalization program (PHP) designed to stabilize patients’ symptoms so that they can transition to outpatient care. Data were collected in two rounds: 55 participants were enrolled as part of another study with different aims (Forgeard et al., 2021 ; Webb et al., 2021 ), and 28 were enrolled to increase the sample size for this study as well as to examine within-person processes, the analysis of which is not part of the current study. The two samples were combined for this study. Patients were required to own a smartphone to participate. Those with active mania/psychosis symptoms were excluded from participation. Participants who did not endorse using at least one BT skill and at least one DBT skill during the EMA period were dropped ( n = 4). As noted in the procedure section below, those in the first round of recruitment completed EMA for 14 days whereas those in the second round completed EMA for 16 days. If participants did not have a depressions score (the PHQ-9, as described below) at discharge and at least one PHQ-9 daily measurement in the last three days of the study (i.e., days 12 through 14 in the first sample, and days 14 through 16 in the second sample) they were removed ( n = 12). Finally, participants that did not complete another symptom measure(the BASIS-24, as described below) at discharge were removed ( n = 3), resulting in a final sample size of 60. Participants who completed EMA data but were dropped for the reasons above did not differ from the final sample in age, gender, education, or student status (all p s > .05). Participants were asked to report their gender (Female, Male, Not listed, please describe); 60% identified as female. Participants were 36 years old on average ( SD = 14, range = 18–70). The average length of stay at the PHP was 15 days ( SD = 3, range = 10–26). Demographics and diagnostic information for the full sample appears in Table 1 . The two outcome groups (improvement vs. non-improvement) did not differ by gender ( X 2 (1, N = 57) = .00, p = 1), age ( t (55) = − .06, p = 1), number of patients who were currently students ( X 2 (1, N = 60) = .09, p = 1), or by level of education X 2 (2, N = 58) = 4, p = 0.1). Due to low numbers in some categories, we could not compare the groups on sexual orientation or race. Table 1 Demographic and current diagnostic characteristics of the full sample n % Gender Female 36 60 Male 21 35 Genderfluid 1 1.7 Unreported 2 3.3 Race Asian 7 11.7 Black 2 3.3 Mixed 1 1.7 White (non-Hispanic/Latino) 50 83.3 Sexual Orientation Heterosexual 43 71.7 Bisexual 7 11.7 Gay/Lesbian 5 8.3 Queer 3 5 Something else 2 3.3 Education High school/GED 2 3.3 Some college, Associates, trade school 21 35 College graduate 19 31.7 Post-college education 18 30 Employment Not employed 24 40 Part time 11 18 Full time 25 42 Student 16 27 Current Diagnoses (sample one, n = 45) MDD current MDE 31 52 Bipolar I 6 10 Bipolar II 0 0 Panic Disorder 9 15 Agoraphobia 10 16.7 Social Anxiety Disorder 16 26.7 Generalized Anxiety Disorder 30 50 Obsessive Compulsive Disorder 13 21.7 Posttraumatic Stress Disorder 5 8.3 Alcohol Use Disorder 5 8.3 Measures At discharge, patients completed the Patient Health Questionnaire – 9 item (PHQ-9), a well-established and reliable measure of depressive symptoms (Spitzer, 1999 ). The PHQ-9 shows excellent internal consistency ( a = .86-.89) and test-retest reliability ( r = .84) in primary care patients (Kroenke et al., 2001 ), as well strong predictive validity for major depression (area under the curve = .95). In a psychiatric sample similar to the current study, internal consistency was good ( a = .87), as was sensitivity (.83) and specificity (.72) for detecting a current depressive episode (Beard et al., 2016 ). Participants also completed a modified version of the PHQ-9 daily during the EMA period. The modified version assessed symptoms in the past 24 hours, on a scale of 0 (not at all) to 3 (almost all of the time) and separated items that assessed different phenomena into individual questions (i.e., trouble falling asleep, trouble staying asleep, and sleeping too much were separated; Feeling down, depressed was separated from feeling hopeless; poor appetite was separated from overeating; moving or speaking slowly was separated from being fidgety and restless). To enable comparison between the two PHQ-9 versions, we reaggregated these items to match the original scale by using the highest ratings of the split items in our analyses. Within each EMA survey, patients rated how they felt immediately prior to the notification with regard to six positive affect items (excited, energized, activated, calm, peaceful, relaxed) and six negative affect items (nervous, frustrated, angry, bored, sad, tired). These items were chosen from prior studies’ momentary affect assessments (e.g., Peeters et al., 2010 ; Watson et al., 1988 ), modified to reduce the number of items to reduce participant burden and to ensure an equal number of high activation (i.e., excited, energized, active, nervous, frustrated, angry) versus low activation words (i.e., calm, peaceful, relaxed, bored, sad, tired) for both positive and negative affect (Forgeard et al., 2021 ). In sample one, participants rated affect on a Likert scale from 1 (not at all) to 7 (extremely), in sample two participants rated affect on a 0-100 slider scale from 0 (not at all) to (very much). Sample one’s affect ratings were rescaled to 0-100 prior to analyses. Also, within each survey, patients reported which skills they had used since the last survey administration, from a list of skills they had learned in treatment. These included cognitive behavior therapy (CBT) skills: behavioral activation, behavioral scheduling, exposure, and cognitive restructuring, and dialectical behavior therapy (DBT) skills: mindfulness, distress tolerance, emotion regulation, and interpersonal effectiveness/communication. To obtain diagnostic information, we had doctoral interns and clinical practicum students, trained by a postdoctoral fellow, administer the Mini-International Neuropsychiatric Interview (MINI; Sheehan et al., 1998 ). The MINI has strong reliability and validity relative to the Structured Clinical Interview for DSM-IV (Sheehan et al., 1998 ), and within this PHP, MINI diagnoses from interns and students correlate with those made by program psychiatrists (Kertz et al., 2012 ). Clinic practices changed between the collection of the first and second samples such that clinicians only completed MINI modules that were relevant to their patient, and not all clinicians were able to complete the MINI. Thus, diagnoses are reported for the first sample only. Procedure Upon discharge, patients were given information about the study, provided written informed consent, and then were trained on how to use the app to respond to surveys. Then, patients completed EMA surveys within the naturalistic setting of their lives. The first sample completed the surveys four times a day for 14 days, at semirandom intervals between 10 a.m. and 8 p.m., separated by at least two hours, with a maximum of 56 surveys. Because the second sample was recruited in part to examine within-person processes, we aimed to collect more surveys per person. Thus, the second sample completed the surveys 6 times a day (i.e., two more times) and for two extra days for a maximum of 96 data points for each participant. Participants in the second sample chose when the first survey of the day would be based on their typical waking time, and each subsequent survey was delivered at 2.5 hour increments after that. Participants were paid $ 20 for each week they participated in the study (participants in the second sample received $ 5 extra for the extra two days they completed), plus a bonus of $ 30 per week if they completed 80% of surveys (total possible $ 100 or $ 105). All procedures were approved by the [BLINDED] Human Research Committee Institutional Review Board. Analysis Analyses were preregistered on OSF ( https://osf.io/uems5 ). As noted above, sample one’s affect ratings were rescaled to 0-100 to match sample two. End of study PHQ-9 was calculated using the average of up to the last three days of daily PHQ-9 surveys, with missing measurements omitted such that if participants only completed one survey in the last three days, that survey was used as the end of study PHQ-9 score, and if they completed two of three, then the mean of those two measurements was used as the end of study PHQ-9 score. Explanatory Model We conducted a multilevel linear regression to examine whether within-person correlations between positive affect, CBT skills and DBT skills were associated with changes in PHQ-9 post-discharge. We entered PHQ-9 aggregate score as the dependent variable and person-specific CBT-PA correlation, person-specific DBT-PA correlation, and the time (discharge and end of study) as predictors, with interaction effects for each of the correlations with time. Subject ID was associated with a random intercept. We used the Lme4 package in R (Bates et al., 2015 ) with the following syntax: PHQ9 ~ TIME + CBT-PAcorr + DBT-PAcorr + CBT-PAcorr:TIME + DBT-PAcorr:TIME + (1|ID) Predictive Model To examine the predictive potential of skills use and affect for determining which participants would continue to improve after treatment vs which would deteriorate or remain unchanged, we conducted a regularized logistic regression predicting outcome group. Outcome group was determined as follows: patients were categorized as improving if they reported a clinically significant decrease of at least 5 points in PHQ-9 score from discharge to the end of the study or their PHQ-9 score remained < 5 at both time points; otherwise, patients were categorized as unchanged/deteriorating. In addition to the correlations between skills use (CBT/DBT) and positive affect, preregistered predictors included: the number of times over the two weeks that patients used each of the individual skills (i.e., psychoeducation, self-assessment, behavioral activation, behavioral scheduling, exposure, mindfulness, distress tolerance, emotion regulation, cognitive restructuring, and interpersonal effectiveness) and the total number of times a patient used any skill. The analyses for this study were originally preregistered with only positive affect items included, as a previous study with a subset of this sample found that on average, negative affect was not associated with therapeutic skills use (Forgeard et al., 2021 ). However, given that the correlations are calculated within-person, this may not preclude individual differences in this correlation corresponding to outcome. Additionally, given that prediction models benefit from additional predictors and that our regularization process, described below, drops non-important predictors from the model, we opted to include additional predictors. Thus, we also ran a model with the correlations between negative affect and CBT and DBT skills and an expanded set of other predictors, described below. We have included the preregistered model with fewer predictors in the supplementary materials. The additional predictors included BASIS-24 score at discharge to control for general severity; summaries of affect, including the within-individual mean and standard deviation of negative and positive affect, respectively, over the whole EMA period; total number of stressful events reported by the participant; and the proportion of stressful events where a skill was used, which we calculated by counting the number of times a stressful event was reported where any skill was used within the time period and dividing that number by the number of total reported stressful events. We used elastic net regularization to perform the prediction, which combines two regularization strategies, least absolute selection and shrinkage operator (LASSO) and ridge regression, to drop predictors from the model whose coefficients are likely spurious. This type of regularization is especially useful in cases where the predictors may be correlated, such as in the current study, wherein individuals may use skills simultaneously. To reduce the risk of overfitting, we used 10-fold cross-validation to select the lambda for the model. In this method, 1/10th of the data is a hold-out sample for testing, and 9/10ths as training, such that each tested hyperparameter predicted unseen data. We selected the minimum lambda that provided the best fit to the data. Elastic net regularization also requires a hyperparameter alpha, which determines the mix of LASSO and ridge regression; we selected alpha = .05 to balance each regularization method evenly. We report the accuracy, sensitivity, and specificity of the model by comparing predicted and observed group assignment. We used the R package glmnet to conduct the cross-validated elastic net regression (Friedman et al., 2010 ). Results Descriptives On average, participants completed 51.10 surveys ( SD = 17, range = 25–91). At discharge (i.e., before the EMA period), the improving group and the deteriorating/unchanged group did not significantly differ in their PHQ-9 severity ( t (30) = 1, p = .20) or total BASIS-24 severity ( t (32) = .30, p = .80). Descriptive information about each group’s affect and skills use over the course of the EMA period and their symptoms at discharge and after the EMA period are included in Table 2 . Table 2 Affect, skills use, and symptom descriptive information for each outcome group Variable Improving Non-improvement M ( SD ) Range M ( SD ) Range Within-person affect PA i M 29.77 (14.98) 10.16–79.38 26.87 (14.13) 1.3–57.93 NA i M 25.07 (19.49) 2.78–71.16 25.1 (12.07) 6.78–58.97 PA i SD 11.57 (4.56) 5.28–26.45 11.02 (3.29) 2.58–20.56 NA i SD 11.27 (6.04) 3.16–30.66 11.18 (3.5) 5.4–18.82 PA– CBT i r 0.03 (0.18) -0.38–0.36 0.14 (0.2) -0.19–0.63 PA-DBT i r 0.05 (0.2) -0.46–0.39 0.03 (0.18) -0.51–0.32 NA-CBT i r -0.04 (0.15) -0.33–0.37 -0.1 (0.21) -0.43–0.3 NA-DBT i r 0 (0.2) -0.22–0.5 0.04 (0.22) -0.28–0.61 Skills Use Total CBT skills 35.65 (27.33) 2 -138 25.26 (16.33) 4–72 Total DBT skills 32.12 (29.3) 2–121 20.26 (17.77) 4–88 Total skills 67.77 (49.14) 11–259 45.53 (29.91) 12–160 Total stressful events 6.92 (6.14) 0–22 7.09 (6.95) 0–31 Prop. events w skills 0.73 (0.36) 0–1 0.64 (0.32) 0–1 Self-Assessment 2.69 (3.72) 0–11 3.21 (5.96) 0–26 Behavioral Scheduling 6.08 (4.95) 0–17 5.29 (5.62) 0–21 Behavioral Activation 17.81 (15.32) 0–69 10.12 (8.61) 0–32 Exposure 6.12 (7.84) 0–29 4.74 (5.89) 0–18 Cognitive Restructuring 5.65 (6.01) 0–23 5.12 (7.18) 0–40 Mindfulness 11.35 (10.86) 0–49 7.88 (6.16) 0–21 Distress Tolerance 3.73 (4.83) 0–14 3.26 (4.65) 0 - 20 Emotion Regulation 9.27 (12.69) 0–56 4.21 (6.16) 0–25 Interpersonal Effectiveness communication 7.77 (10.01) 0–41 4.91 (5.55) 0–25 Proportion DBT 0.44 (0.22) 0.09–0.98 0.44 (0.19) 0.07–0.87 Symptoms PHQ-9 discharge 11.92 (8.1) 0–27 9.85 (2.78) 6–16 PHQ-9 end (last 3 days) 5.08 (3.96) 0–14.5 9.77 (4.16) 2–23 PHQ-9 change 6.84 (6.07) -3–24.67 0.08 (2.71) -7–4 BASIS-24 discharge 1.3 (0.79) 0.11–2.84 1.25 (0.34) 0.59–2.17 Depression Functioning 1.78 (1.13) 0.09–3.65 1.57 (0.48) 0.61–2.92 Relationships 0.99 (0.62) 0–2.12 1.26 (0.7) 0.28–3.16 Self-Harm 0.76 (0.93) 0–2.84 0.53 (0.76) 0–2.58 Emotional Lability 1.36 (1.06) 0–3.89 1.6 (0.75) 0–3 Psychosis 0.21 (0.38) 0–1.52 0.22 (0.35) 0–1.22 Substance Abuse 0.3 (0.46) 0–1.38 0.46 (0.62) 0–2.18 Simplified explanatory model In the explanatory model, the only significant predictor of PHQ-9 score was time (pre versus time post), B = 3.16, SE = .76, t (60) = 4.17, p .05). Prediction Model The results of the preregistered model are reported in the supplementary materials (Table S1 ). Differences between those models and the model reported below are minimal. Below, we describe the results for the full set of predictors described in the Method section. The 10-fold cross validation resulted in a minimum lambda of .15 selected for the final model. The accuracy of the model’s predictions was 62% (95% CI [.48, .74]), meaning that the model accurately categorized 62% of patients in the sample into the correct group. The area under the curve was acceptable (.74, Fig. 1), however the sensitivity was low (.23) while specificity was high (.91), indicating that the model was not as accurate at predicting improvement as it was predicting non-improvement. After regularization, five predictors were retained (Table 3 ). Total use of behavioral activation (OR = 1.23) predicted improvement. More use of emotion regulation was a weak predictor of improvement (OR = 1.09). Total use of DBT skills and total use of all skills were retained in the model as predictors of improvement but their predictive power was negligible. BASIS score was not retained in the model. Within-person concurrent positive affect and use of any CBT skill (OR = .84) predicted non-improvement. Table 3 Coefficients of the Predictors Retained in the Elastic Net Regression Model with Minimum Lambda. Predictor Log Odds Odds Ratio (Intercept) -0.27 0.76 PA-CBT i r -0.18 0.84 Total DBT 0.01 1.01 Total skills use 0.01 1.01 Behavioral Activation 0.21 1.24 Emotion Regulation 0.09 1.09 Note. PA = positive affect; CBT = Cognitive Behavior Therapy skills (second-wave); DBT = Dialectical Behavior Therapy skills. The following variables were not retained as predictors after regularization: PA-DBT i r , NA-BT i r , total CBT skills, number of stressful events, self-assessment, behavioral scheduling, cognitive restructuring, mindfulness, interpersonal effectiveness/communication, proportion of stressful events where a skill was used, exposure, distress tolerance, PA i M , PA i SD , NA i M , NA i SD , BASIS-24. Discussion This study examined the predictive accuracy of momentary affect, therapeutic skills use, and their association, for distinguishing between patients who continued to improve after discharge from a partial hospital program and patients who did not. To the authors’ knowledge, this is the first study to use a within-person correlation between affect and skills use, and one of few studies that focuses on treatment skills drawn from various modalities rather than entire treatment protocols, to predict follow-up outcome. The two outcome groups did not significantly differ in severity upon discharge from the treatment program, suggesting that the follow-up outcome is not a simple reflection of post-treatment severity. Thus, other post-discharge factors, including those examined in this study, are likely better predictors of continued improvement. This underscores the importance of better understanding the vulnerable period after discharge and which behaviors may support continued recovery. Within-person associations between affect and skills use were not associated with dimensional change on the PHQ-9 from discharge to follow up, however these variables in addition to total use of a range of DBT and CBT skills were acceptable predictors of clinically significant improvement (> 5 points reduction on PHQ-9) or stable retention of treatment gains (retained PHQ-9 below 5). One explanation for this discrepancy is that the model that included raw PHQ-9 scores assumes linear change, which may not hold. For instance, we would encounter floor effects when patients improve, and while the linear model would not differentiate between someone who stayed the same with zero symptoms from someone who stayed the same with moderate symptoms, the logistic model would group them separately, as reflects their real-world clinical disposition. Thus, the logistic model better reflects real-world clinical scenarios compared to the linear model. Of course, the logistic model was optimized for accurate prediction rather than significant proportion of variance explained. As such, additional predictors were included in an exploratory fashion compared to the simplified linear model; retained predictors performed over and above those dropped from the model and together were accurate in determining improving vs non-improving patients. In the regularized logistic regression, few predictors were retained in the model, but a clear pattern emerged: behavioral skills were stronger predictors of improvement than were CBT skills. The individual DBT skills most strongly predictive of improvement in this study were emotion regulation and communication/interpersonal effectiveness. These findings makes sense in the context of the higher-acuity treatment setting. For patients who have just stepped down from an inpatient unit or who have needed to step up to a higher level of care than outpatient treatment, DBT skills may be more appropriate, as several directly target acute emotional difficulties. These skills might be the most useful for limiting the detrimental effect of mental illness on patients’ functioning, such as managing their relationships or returning to work. This kind of improvement is the goal of the partial hospitalization program – to return patients to a level of functioning where they can go back to everyday life. Thus, skills that directly limit detrimental effects on functioning are likely to be the most helpful in this short period after discharge. This pattern is supported by some evidence from prior studies that DBT skills may be more effective than CBT skills in acute settings. One study of adolescents in an inpatient unit found that those assigned to DBT intervention experienced better behavioral functioning evidenced by fewer suicide attempts, less self-injury, fewer uses of restraints by treatment staff, and fewer days hospitalized post discharge, compared to those assigned to a CBT-based treatment as usual (Tebbett-Mock et al., 2020 ). Similarly, a study conducted in a small non-psychiatric sample of individuals who recently experienced myocardial infarction revealed preliminary evidence that those assigned to the DBT intervention experienced better emotion-focused coping compared to those in the cognitive therapy intervention (Nourisaeed et al., 2021 ). Another aspect of the skills that strongly predicted improvement is their focus on behavior. Importantly, behavioral activation was the only substantial predictor of improvement. This comports with previous findings showing that behavioral activation outperforms cognitive therapy in more severely depressed samples (Dimidjian et al., 2006 ). A prior study conducted within the same partial hospital program as the current study found that in depressed patients, behavioral activation significantly predicted next-day improvement in depression symptoms during treatment, while cognitive restructuring did not (Webb et al., 2016 ). Additionally in that sample, DBT skills use significantly predicted next-day improvement in anxiety symptoms. Thus, the current study suggests that the pattern of behavioral activation being associated with better improvement than cognitive restructuring extends to the use of these skills post-discharge as well. A unique aspect of this study is the use of within-person associations between skill use and affect to predict improvement. These associations are generated within the individual, but compared across individuals, to determine whether a certain tendency between individuals to experience more or less positive affect in conjunction with certain skills is associated with symptom improvement. Indeed, higher within-person associations between CBT skills use and positive affect predict non-improvement. In other words, patients who do not improve two weeks after discharge typically experience positive affect after completing a CBT skill. Considering that behavioral activation, which is expected to increase positive affect, is one of the CBT skills included in the composite associated with positive affect, it is surprising that the PA-CBT association is predictive of non-improvement while behavioral activation by itself was predictive of improvement. Perhaps other CBT skills included in this correlation, namely cognitive restructuring, behavioral scheduling, and self-assessment, interfered with this effect. If so, perhaps using these other skills to produce short-term increases in positive affect is ultimately unhelpful for long term improvement. Perhaps patients are replacing negative thoughts with unrealistically positive alternatives - for example, in Pollack et al. (2021), adolescents who imagined more unrealistic positive events exhibited a stronger association between defeat/entrapment and suicidal ideation than those with less positive thinking. Though increasing their short-term relief, ultimately patients may be disappointed when their unrealistic expectations do not materialize. Given that these patients are using these skills on their own, another possibility is that they are implementing them incorrectly. For instance, cognitive flexibility and behavioral scheduling are often easier to implement than behavioral activation, and therefore has the potential to be used maladaptively if it is used to avoid choosing to use other skills that are more warranted in the situation; e.g., choosing to reframe how one is thinking about a recent argument when a more effective strategy might be communicating effectively with the person with whom one was quarrelling. In this situation, the reframing may ultimately serve an avoidance function preventing the individual from solving problems effectively. Another possibility is that some patients may schedule activities but fail to carry them out. Finally, though behavioral activation is designed to improve mood, it may be unpleasant in the short term, so immediate positive affect while engaging in behavioral activation might indicate that the patient is not choosing the most challenging/rewarding activities. Although this study is a fruitful investigation into potential predictors of continued outcome, these findings would be strengthened by testing this model on a new sample. Additionally, these findings are merely correlational, and the absence of random assignment precludes confident claims about causality. However, a strength of this study is its ecological validity, given that it is common for patients in the real-world to be exposed to many different types of skills that could come from different treatment protocols (Cook et al., 2010 ). Indeed, this study benefits from the fact that the PHP that the participants in this completed exposes patients to a wide variety of skills, allowing for analyses such as this. Our study is also limited in its generalizability to non-white and non-heterosexual individuals, given the lack of diversity in our sample on these demographic variables. Additionally, improvement in this study was determined by the PHQ-9 which is not a comprehensive measure of recovery, though it is a common measure of depression in hospital settings, and has been shown to be sensitive to changes in depression symptoms (Löwe et al., 2004 ). For instance, the PHQ-9 would not capture aggression or anxiety symptoms that may have improved but may not covary strongly with PHQ-9 symptoms. We therefore cannot say whether these findings would generalize to improvement in other domains beyond depression symptoms. Additionally, the end of study PHQ-9 score was an average of three daily measurements. The validated version of the PHQ-9 assessed over two weeks might generalize better to findings in the real world. Future studies should aim to include more comprehensive assessments that are personalized to each patient’s diagnoses and treatment goals. Improvement beyond symptom reduction should also be explored in future studies, for instance by examining quality of life and other positive outcomes. Summary This study examined whether patients’ individual patterns of therapeutic skills use and affect in the vulnerable period directly after discharge from a partial hospital program would predict which of these patients continue to improve during the two weeks after they leave treatment. Behavioral activation predicted continued improvement, and a within-person trend of increased positive affect after using a CBT skill predicted non-improvement. These findings suggest that in the acute period directly after discharge from a partial hospital program, patients should focus on using behavioral activation to improve depression symptoms. Declarations Author Contribution SF and CB wrote the manuscript. SF, MF, RM and CB contributed to study design and aims. SF and KF collected the data. SF analyzed the data and prepared figures. All authors reviewed the manuscript and provided substantive revisions. Acknowledgement Thank you to Joshua Cetron and the Institute for Quantitative Social Sciences for guidance on the preregistration and analysis design for this project. Data Availability Code, data, and materials are available on OSF: https://osf.io/54fu6 References Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software , 67 , 1–48. https://doi.org/10.18637/jss.v067.i01 Beard, C., Hsu, K. J., Rifkin, L. S., Busch, A. B., & Björgvinsson, T. (2016). Validation of the PHQ-9 in a psychiatric sample. Journal of Affective Disorders , 193 , 267–273. https://doi.org/10.1016/j.jad.2015.12.075 Bruce, S. E., Yonkers, K. A., Otto, M. W., Eisen, J. L., Weisberg, R. B., Pagano, M., Shea, M. T., & Keller, M. B. (2005). Influence of Psychiatric Comorbidity on Recovery and Recurrence in Generalized Anxiety Disorder, Social Phobia, and Panic Disorder: A 12-Year Prospective Study. American Journal of Psychiatry , 162 (6), 1179–1187. https://doi.org/10.1176/appi.ajp.162.6.1179 Burcusa, S. L., & Iacono, W. G. (2007). Risk for recurrence in depression. Clinical Psychology Review , 27 (8), 959–985. https://doi.org/10.1016/j.cpr.2007.02.005 Cohen, Z. D., Kim, T. T., Van, H. L., Dekker, J. J. M., & Driessen, E. (2020). A demonstration of a multi-method variable selection approach for treatment selection: Recommending cognitive–behavioral versus psychodynamic therapy for mild to moderate adult depression. Psychotherapy Research , 30 (2), 137–150. https://doi.org/10.1080/10503307.2018.1563312 Cook, J. M., Biyanova, T., Elhai, J., Schnurr, P. P., & Coyne, J. C. (2010). What do psychotherapists really do in practice? An Internet study of over 2,000 practitioners. Psychotherapy: Theory, Research, Practice, Training , 47 (2), 260–267. https://doi.org/10.1037/a0019788 Coutanche, M. N., & Hallion, L. S. (2020). Machine learning for clinical psychology and clinical neuroscience. In The Cambridge handbook of research methods in clinical psychology (pp. 467–482). Cambridge University Press. https://doi.org/10.1017/9781316995808.041 DeRubeis, R. J. (2019). The history, current status, and possible future of precision mental health. Behaviour Research and Therapy , 123 , 103506. https://doi.org/10.1016/j.brat.2019.103506 DeRubeis, R. J., Cohen, Z. D., Forand, N. R., Fournier, J. C., Gelfand, L. A., & Lorenzo-Luaces, L. (2014). The Personalized Advantage Index: Translating Research on Prediction into Individualized Treatment Recommendations. A Demonstration. PLOS ONE , 9 (1), e83875. https://doi.org/10.1371/journal.pone.0083875 DiGiorgio, K. E., Glass, C. R., & Arnkoff, D. B. (2010). Therapists’ Use of DBT: A Survey Study of Clinical Practice. Cognitive and Behavioral Practice , 17 (2), 213–221. https://doi.org/10.1016/j.cbpra.2009.06.003 Dimidjian, S., Hollon, S. D., Dobson, K. S., Schmaling, K. B., Kohlenberg, R. J., Addis, M. E., Gallop, R., McGlinchey, J. B., Markley, D. K., Gollan, J. K., Atkins, D. C., Dunner, D. L., & Jacobson, N. S. (2006). Randomized trial of behavioral activation, cognitive therapy, and antidepressant medication in the acute treatment of adults with major depression. Journal of Consulting and Clinical Psychology , 74 (4), 658–670. https://doi.org/10.1037/0022-006X.74.4.658 Forgeard, M., Beard, C., Shayani, D., Silverman, A. L., Tsukayama, E., & Björgvinsson, T. (2021). Predictors of affect following discharge from partial hospitalization: A two-week ecological momentary assessment study. Psychological Medicine , 51 (7), 1157–1165. https://doi.org/10.1017/S0033291719004057 Friedman, J. H., Hastie, T., & Tibshirani, R. (2010). Regularization Paths for Generalized Linear Models via Coordinate Descent. Journal of Statistical Software , 33 , 1–22. https://doi.org/10.18637/jss.v033.i01 Heininga, V. E., & Kuppens, P. (2021). Psychopathology and positive emotions in daily life. Current Opinion in Behavioral Sciences , 39 , 10–18. https://doi.org/10.1016/j.cobeha.2020.11.005 Huibers, M. J. H., Cohen, Z. D., Lemmens, L. H. J. M., Arntz, A., Peeters, F. P. M. L., Cuijpers, P., & DeRubeis, R. J. (2015). Predicting Optimal Outcomes in Cognitive Therapy or Interpersonal Psychotherapy for Depressed Individuals Using the Personalized Advantage Index Approach. PLOS ONE , 10 (11), e0140771. https://doi.org/10.1371/journal.pone.0140771 Kertz, S. J., Bigda-Peyton, J. S., Rosmarin, D. H., & Björgvinsson, T. (2012). The importance of worry across diagnostic presentations: Prevalence, severity and associated symptoms in a partial hospital setting. Journal of Anxiety Disorders , 26 (1), 126–133. https://doi.org/10.1016/j.janxdis.2011.10.005 Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2001). The PHQ-9. Journal of General Internal Medicine , 16 (9), 606–613. https://doi.org/10.1046/j.1525-1497.2001.016009606.x Levy, H. C., O’Bryan, E. M., & Tolin, D. F. (2021). A meta-analysis of relapse rates in cognitive-behavioral therapy for anxiety disorders. Journal of Anxiety Disorders , 81 , 102407. https://doi.org/10.1016/j.janxdis.2021.102407 Linehan, M. M. (2018). Cognitive-Behavioral Treatment of Borderline Personality Disorder . Guilford Publications. Löwe, B., Kroenke, K., Herzog, W., & Gräfe, K. (2004). Measuring depression outcome with a brief self-report instrument: Sensitivity to change of the Patient Health Questionnaire (PHQ-9). Journal of Affective Disorders , 81 (1), 61–66. https://doi.org/10.1016/S0165-0327(03)00198-8 Michalak, J., Steinhaus, K., & Heidenreich, T. (2020). (How) Do Therapists Use Mindfulness in Their Clinical Work? A Study on the Implementation of Mindfulness Interventions. Mindfulness , 11 (2), 401–410. https://doi.org/10.1007/s12671-018-0929-9 Nourisaeed, A., Ghorban-Shiroudi, S., & Salari, A. (2021). Comparison of the effect of cognitive-behavioral therapy and dialectical behavioral therapy on perceived stress and coping skills in patients after myocardial infarction. ARYA Atherosclerosis , 17 (2), 1–9. https://doi.org/10.22122/arya.v17i0.2188 Peeters, F., Berkhof, J., Rottenberg, J., & Nicolson, N. A. (2010). Ambulatory emotional reactivity to negative daily life events predicts remission from major depressive disorder. Behaviour Research and Therapy , 48 (8), 754–760. https://doi.org/10.1016/j.brat.2010.04.008 Pollak, O. H., Guzmán, E. M., Shin, K. E., & Cha, C. B. (2021). Defeat, Entrapment, and Positive Future Thinking: Examining Key Theoretical Predictors of Suicidal Ideation Among Adolescents. Frontiers in Psychology , 12 , 590388. https://doi.org/10.3389/fpsyg.2021.590388 Richards, D. (2011). Prevalence and clinical course of depression: A review. Clinical Psychology Review , 31 (7), 1117–1125. https://doi.org/10.1016/j.cpr.2011.07.004 Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., Hergueta, T., Baker, R., & Dunbar, G. C. (1998). The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. The Journal of Clinical Psychiatry , 59 Suppl 20 , 22-33;quiz 34-57. Spitzer, R. L. (1999). Patient health questionnaire: PHQ . [New York] : [New York State Psychiatric Institute], [1999] ©1999. https://search.library.wisc.edu/catalog/999907635102121 Tebbett-Mock, A. A., Saito, E., McGee, M., Woloszyn, P., & Venuti, M. (2020). Efficacy of Dialectical Behavior Therapy Versus Treatment as Usual for Acute-Care Inpatient Adolescents. Journal of the American Academy of Child & Adolescent Psychiatry , 59 (1), 149–156. https://doi.org/10.1016/j.jaac.2019.01.020 Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology , 54 (6), 1063–1070. https://doi.org/10.1037/0022-3514.54.6.1063 Webb, C. A., Beard, C., Kertz, S. J., Hsu, K. J., & Björgvinsson, T. (2016). Differential role of CBT skills, DBT skills and psychological flexibility in predicting depressive versus anxiety symptom improvement. Behaviour Research and Therapy , 81 , 12–20. https://doi.org/10.1016/j.brat.2016.03.006 Webb, C. A., Forgeard, M., Israel, E. S., Lovell-Smith, N., Beard, C., & Björgvinsson, T. (2021). Personalized prescriptions of therapeutic skills from patient characteristics: An ecological momentary assessment approach. Journal of Consulting and Clinical Psychology . https://doi.org/10.1037/ccp0000555 Yonkers, K. A., Bruce, S. E., Dyck, I. R., & Keller, M. B. (2003). Chronicity, relapse, and illness—course of panic disorder, social phobia, and generalized anxiety disorder: Findings in men and women from 8 years of follow-up. Depression and Anxiety , 17 (3), 173–179. https://doi.org/10.1002/da.10106 Yonkers, K. A., Warshaw, M. G., Massion, A. O., & Keller, M. B. (1996). Phenomenology and Course of Generalised Anxiety Disorder. The British Journal of Psychiatry , 168 (3), 308–313. http://dx.doi.org/10.1192/bjp.168.3.308 Additional Declarations No competing interests reported. Supplementary Files supplementarymaterials.docx Cite Share Download PDF Status: Published Journal Publication published 12 Mar, 2026 Read the published version in Cognitive Therapy and Research → Version 1 posted Editorial decision: Revision requested 18 Jan, 2026 Reviews received at journal 17 Jan, 2026 Reviews received at journal 15 Jan, 2026 Reviewers agreed at journal 06 Jan, 2026 Reviewers agreed at journal 29 Dec, 2025 Reviewers invited by journal 15 Sep, 2025 Editor assigned by journal 15 Sep, 2025 Submission checks completed at journal 15 Sep, 2025 First submitted to journal 14 Sep, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7614725","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":517209879,"identity":"f0e5efa3-1c69-48ff-a31f-10451654d7d1","order_by":0,"name":"Shaan F. McGhie","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA9ElEQVRIiWNgGAWjYJACZgYDBjkGBh4o9wADgwQxWoxJ1cLAkNhAtBb5aYcfvy4osEmf78B78HNlm10+3wHmg7d58GgxuJ1mZj3DIC134wG+ZMmzbcmWMw+wJVvj1SKdw2bMY3A4d2MDj4FkwxlmA4MDPGbS+LTIzwZr+Z9u2MBj/LPhTD1QC/83vFoYbucwP+YxOJAgz8BjJtlQcRhkCxteLSC/MPMYJBtuYOYxs2yoOG4geZjN2HIOXoclP/7M88dOXr69x/hmg0G1Ad/x5oc33uBzGAMDGzgWDA7D+Mz4lYOVfABb10BY5SgYBaNgFIxQAAADGkYtw6uqBQAAAABJRU5ErkJggg==","orcid":"","institution":"Harvard University","correspondingAuthor":true,"prefix":"","firstName":"Shaan","middleName":"F.","lastName":"McGhie","suffix":""},{"id":517209880,"identity":"9b8f50d5-22cb-47d2-80f4-9b7262c072ca","order_by":1,"name":"Marie Forgeard","email":"","orcid":"","institution":"William James College","correspondingAuthor":false,"prefix":"","firstName":"Marie","middleName":"","lastName":"Forgeard","suffix":""},{"id":517209882,"identity":"9e131103-7f55-43d1-8502-217e35d1832e","order_by":2,"name":"Kailyn Fan","email":"","orcid":"","institution":"University of Denver","correspondingAuthor":false,"prefix":"","firstName":"Kailyn","middleName":"","lastName":"Fan","suffix":""},{"id":517209884,"identity":"d6c060a7-8738-459e-a119-a88a86bd0114","order_by":3,"name":"Richard J. McNally","email":"","orcid":"","institution":"Harvard University","correspondingAuthor":false,"prefix":"","firstName":"Richard","middleName":"J.","lastName":"McNally","suffix":""},{"id":517209886,"identity":"d216d52e-ecee-4ad3-9b4f-e2e83654f4fc","order_by":4,"name":"Courtney Beard","email":"","orcid":"","institution":"McLean Hospital","correspondingAuthor":false,"prefix":"","firstName":"Courtney","middleName":"","lastName":"Beard","suffix":""}],"badges":[],"createdAt":"2025-09-14 21:23:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7614725/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7614725/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s10608-026-10727-0","type":"published","date":"2026-03-12T15:58:24+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":91960710,"identity":"4cafb553-2af4-4972-ba07-1a069133e755","added_by":"auto","created_at":"2025-09-23 07:48:03","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":306179,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/ccc4de29e8273b7d380002c5.docx"},{"id":91960709,"identity":"1a25c6da-766d-40e2-bc18-ff08cd58dbbe","added_by":"auto","created_at":"2025-09-23 07:48:03","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6419,"visible":true,"origin":"","legend":"","description":"","filename":"b2e6cb59e2644c79a5ec31fbdc5106c2.json","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/086dff608df2a50aceb678f3.json"},{"id":91960712,"identity":"26d0941f-742d-4ea8-803b-f2900e1abd43","added_by":"auto","created_at":"2025-09-23 07:48:03","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40286,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/b7ffffae3fdf980a01861a96.docx"},{"id":91960713,"identity":"7629f553-eb56-418b-9a8b-a820b1a0593f","added_by":"auto","created_at":"2025-09-23 07:48:03","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":130612,"visible":true,"origin":"","legend":"","description":"","filename":"b2e6cb59e2644c79a5ec31fbdc5106c21enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/b553fdcf99ac26e43b1baafe.xml"},{"id":91962624,"identity":"0eaf235a-cf74-4fc1-9365-9af02d1d2ec0","added_by":"auto","created_at":"2025-09-23 07:56:04","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":205478,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/3b8c581347efbddf33bb40a3.jpeg"},{"id":91960715,"identity":"7252a3ef-e17a-496b-a5f1-65e2973f6d98","added_by":"auto","created_at":"2025-09-23 07:48:04","extension":"xml","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":127913,"visible":true,"origin":"","legend":"","description":"","filename":"b2e6cb59e2644c79a5ec31fbdc5106c21structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/dfcbc44e5f0240ac6f99c535.xml"},{"id":91960718,"identity":"33a70e8f-3dbc-40f2-b7ec-68cd203fbf21","added_by":"auto","created_at":"2025-09-23 07:48:08","extension":"html","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":136538,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/08d37d3774dc1ea61facff3e.html"},{"id":91960711,"identity":"df60256a-3102-40eb-897d-730832595526","added_by":"auto","created_at":"2025-09-23 07:48:03","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":27689,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eSensitivity and Specificity of the Selected Model\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/6db0db307febd26038801bbc.png"},{"id":104739417,"identity":"2c1b793d-6c6f-4c20-a455-c83126f5cd6d","added_by":"auto","created_at":"2026-03-16 16:06:42","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":857344,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/1f4e4158-8c01-4bc7-9d1d-dd040e93f099.pdf"},{"id":91960717,"identity":"2a292a6e-838d-4541-ab30-c61a95d26bb8","added_by":"auto","created_at":"2025-09-23 07:48:06","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":40286,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-7614725/v1/b57773dd25a4586ac3720cb8.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"What Happens After Intensive Treatment? Post-Discharge Skill Use and Affect as Predictors of Depression Outcomes","fulltext":[{"header":"Introduction","content":"\u003cp\u003eMood and anxiety disorders are often chronic, and patients frequently reexperience symptoms after initial remission (Bruce et al., 2005; Burcusa \u0026amp; Iacono, 2007; Levy et al., 2021; Richards, 2011; Yonkers et al., 1996, 2003). Patients whose symptoms resurface will likely require additional treatment, further consuming scarce clinical resources and prolonging the distress of the individual. Thus, clinicians would benefit from better understanding the factors that predict stable recovery from these disorders following treatment. Novel statistical methods have been deployed to achieve this goal, including attempts to predict which treatment is better suited to a particular individual (Cohen et al., 2020; DeRubeis et al., 2014; Huibers et al., 2015).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMachine learning can identify relations among variables in a dataset and then use those learned relations to \u003cem\u003epredict\u003c/em\u003e patterns in new data (Coutanche \u0026amp; Hallion, 2020). This emphasis on prediction rather than explanation is well-suited to address the needs of mental health outcome research. This is what clinicians do every day as they treat their patients: they draw upon their knowledge of the literature and past patients (data) and based on patterns they have observed, they decide what treatment will likely be the most beneficial for specific patients (prediction). However, clinicians are constrained by a limited set of data (i.e., only their own caseloads and experience), and the accuracy of their predictions is not quantified. Accordingly, researchers \u0026nbsp;have used machine learning to predict treatment assignment that offers the best fit for individuals (DeRubeis, 2019). For example, studies have examined the relative benefit for a particular depressed patient, based on pre-treatment characteristics, of cognitive behavioral therapy (CBT) versus a selective serotonin reuptake inhibitor (SSRI; DeRubeis et al., 2014), CBT versus psychodynamic therapy (Cohen et al., 2020), and CBT versus interpersonal therapy (Huibers et al., 2015), among others.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe above studies predicted outcomes based on assignment to different treatments altogether. However, clinicians working in naturalistic settings often draw from multiple protocols to teach specific skills to the same patient (Cook et al., 2010), such as mindfulness (Michalak et al., 2020), dialectical behavior therapy (DBT) skills (DiGiorgio et al., 2010; Linehan, 2018), and cognitive behavior therapy skills, among others. The factors that influence a clinician’s decision to emphasize some skills over others vary (DiGiorgio et al., 2010; Michalak et al., 2020), and there is currently no data-driven way to decide what skills would optimize outcome for specific patients. To our knowledge, only one study has used prediction methods to identify which \u003cem\u003eskills\u003c/em\u003e are best suited to an individual (Webb et al., 2021). The authors of that study used patient characteristics to predict which skills, either from CBT or DBT, were most closely tied to positive affect in the same hours they were used. However, short-term increases in positive affect do not necessarily signify whether these skills are beneficial overall. Indeed, many maladaptive coping strategies, such as drinking alcohol, avoidance, or drug use, boost higher positive affect in the moment, yet lead to negative long-term consequences. Conversely, increased positive affect after using a skill effectively might reinforce continued use of that skill, increasing the likelihood that patients implement behavioral change that is helpful for them. Thus, it is important to test whether changes in affect after using therapeutic skills are related to symptom improvement.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAs is typically done in outcome research, all of the predictive work reviewed thus far has focused on improvement from the beginning to the end of treatment. However, as discussed earlier, the time after discharging from treatment could be a fruitful period to ascertain factors that could predict stable recovery. This period is critical for patients to implement the skills they have learned in treatment. This may be especially true for those discharging from partial hospitalization or other similar intensive programs, whose lives have been suddenly interrupted by acute difficulties and intensive treatment delivered in a setting drastically different from that of their daily lives. Yet, little is known about which skills learned in intensive treatment offer the most benefit to patients after they leave. For example, particular skills acquired during treatment (e.g., cognitive reappraisal) may bolster recovery after treatment completion, especially for shorter-term and more acute levels of care. These may differ from the skills that are most useful to focus on during treatment. For instance, certain skills might be most useful for stabilization in acute distress (e.g., distress tolerance, emotion regulation) while others might be better employed when not in acute distress and when there is more time to master the skill (e.g., cognitive restructuring, exposure). Thus, outcome prediction studies should identify which skills are most useful for a patient’s continued improvement after treatment ends. The current project aims to address this question by identifying which therapeutic skills offer the most benefit for continued symptom improvement after discharge from a partial hospitalization program (PHP).\u003c/p\u003e\n\u003cp\u003eThe current study synthesizes the unanswered questions above to ask: (a) which \u003cem\u003eindividual\u003c/em\u003e \u003cem\u003eskills\u003c/em\u003e are helpful for continued improvement \u003cem\u003eafter\u003c/em\u003e treatment, and (b) does how one feels shortly after using a skill indicate how helpful that skill is for symptom improvement? To answer these questions, we collected ecological momentary assessment (EMA) data (i.e., data collected multiple times throughout the day as patients go about their daily lives), during the two weeks post-discharge from a partial hospitalization program. EMA offers the opportunity to examine how these skills are being used in an ecologically valid manner that minimizes recall bias by having participants report skill use every few hours. Conducting EMA after discharge may illuminate how patients use skills without the structure and oversight of a treatment program or therapist.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis intensive data collection also provides an opportunity to determine how skills and affect might covary over time within an individual. In addition to which skills they used, patients reported how they were feeling in the moment they completed each survey. We then calculated within-person correlations between therapeutic skills use (CBT skills or DBT skills) and affect, both measured every few hours over the course of the two weeks for each participant individually (as further described below). We examined the predictive validity of this assocatiation by predicting which patients maintained their symptom reduction two weeks after their discharge date. To do so, we separated patients into two outcome groups. Participants were included in the improvement group if they reported clinically-significant reduction in symptoms two weeks after discharge or if they discharged with few symptoms and retained this level, otherwise participants were in the non-improvement group. \u0026nbsp;We determined the strength of this prediction model for clinical decision-making by using k-fold cross-validation and examining the resulting specificity and sensitivity.\u003c/p\u003e\n\u003cp\u003eWe hypothesized that stronger within-person associations between therapeutic skills and positive affect (cognitive behavior therapy skills-positive affect correlation; dialectical behavior skills - positive affect correlation) would predict decreased depression symptoms two weeks after discharge. We also hypothesized that the cross-fold validated prediction of continued improvement versus non-improvement from post-discharge skills use and the within-person association between positive affect and skills use would result in at least 80% accuracy. We also explored the added value of additional predictors, including individual mean and standard deviation of positive and negative affect, which have been previously linked to mental health (Heininga \u0026amp; Kuppens, 2021).\u0026nbsp;\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e\u003ch2\u003eParticipants\u003c/h2\u003e\u003cp\u003eParticipants were 83 adults who had recently discharged from a partial hospitalization program (PHP) designed to stabilize patients\u0026rsquo; symptoms so that they can transition to outpatient care. Data were collected in two rounds: 55 participants were enrolled as part of another study with different aims (Forgeard et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Webb et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), and 28 were enrolled to increase the sample size for this study as well as to examine within-person processes, the analysis of which is not part of the current study. The two samples were combined for this study. Patients were required to own a smartphone to participate. Those with active mania/psychosis symptoms were excluded from participation.\u003c/p\u003e\u003cp\u003eParticipants who did not endorse using at least one BT skill and at least one DBT skill during the EMA period were dropped (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;4). As noted in the procedure section below, those in the first round of recruitment completed EMA for 14 days whereas those in the second round completed EMA for 16 days. If participants did not have a depressions score (the PHQ-9, as described below) at discharge and at least one PHQ-9 daily measurement in the last three days of the study (i.e., days 12 through 14 in the first sample, and days 14 through 16 in the second sample) they were removed (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;12). Finally, participants that did not complete another symptom measure(the BASIS-24, as described below) at discharge were removed (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3), resulting in a final sample size of 60. Participants who completed EMA data but were dropped for the reasons above did not differ from the final sample in age, gender, education, or student status (all \u003cem\u003ep\u003c/em\u003es\u0026thinsp;\u0026gt;\u0026thinsp;.05).\u003c/p\u003e\u003cp\u003eParticipants were asked to report their gender (Female, Male, Not listed, please describe); 60% identified as female. Participants were 36 years old on average (\u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;14, range\u0026thinsp;=\u0026thinsp;18\u0026ndash;70). The average length of stay at the PHP was 15 days (\u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3, range\u0026thinsp;=\u0026thinsp;10\u0026ndash;26). Demographics and diagnostic information for the full sample appears in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The two outcome groups (improvement vs. non-improvement) did not differ by gender (\u003cem\u003eX\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e (1, \u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;57)\u0026thinsp;=\u0026thinsp;.00, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1), age (\u003cem\u003et\u003c/em\u003e (55)\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;.06, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1), number of patients who were currently students (\u003cem\u003eX\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e (1, \u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;60)\u0026thinsp;=\u0026thinsp;.09, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1), or by level of education \u003cem\u003eX\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e(2, \u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;58)\u0026thinsp;=\u0026thinsp;4, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.1). Due to low numbers in some categories, we could not compare the groups on sexual orientation or race.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cem\u003eDemographic and current diagnostic characteristics of the full sample\u003c/em\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003en\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e%\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eGender\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e36\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e60\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGenderfluid\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUnreported\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eRace\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAsian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e11.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBlack\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMixed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWhite (non-Hispanic/Latino)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e50\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e83.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eSexual Orientation\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHeterosexual\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e43\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e71.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBisexual\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e11.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGay/Lesbian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e8.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eQueer\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSomething else\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eEducation\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHigh school/GED\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSome college, Associates, trade school\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCollege graduate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e19\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePost-college education\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e30\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eEmployment\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNot employed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e40\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePart time\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e18\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFull time\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e42\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eStudent\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eCurrent Diagnoses (sample one, n\u0026thinsp;=\u0026thinsp;45)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMDD current MDE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e31\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e52\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBipolar I\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBipolar II\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePanic Disorder\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAgoraphobia\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e16.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSocial Anxiety Disorder\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e26.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGeneralized Anxiety Disorder\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e50\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eObsessive Compulsive Disorder\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e21.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePosttraumatic Stress Disorder\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e8.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAlcohol Use Disorder\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e8.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eMeasures\u003c/h3\u003e\n\u003cp\u003eAt discharge, patients completed the Patient Health Questionnaire \u0026ndash; 9 item (PHQ-9), a well-established and reliable measure of depressive symptoms (Spitzer, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e1999\u003c/span\u003e). The PHQ-9 shows excellent internal consistency (\u003cem\u003ea\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.86-.89) and test-retest reliability (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.84) in primary care patients (Kroenke et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2001\u003c/span\u003e), as well strong predictive validity for major depression (area under the curve\u0026thinsp;=\u0026thinsp;.95). In a psychiatric sample similar to the current study, internal consistency was good (\u003cem\u003ea\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.87), as was sensitivity (.83) and specificity (.72) for detecting a current depressive episode (Beard et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Participants also completed a modified version of the PHQ-9 daily during the EMA period. The modified version assessed symptoms in the past 24 hours, on a scale of 0 (not at all) to 3 (almost all of the time) and separated items that assessed different phenomena into individual questions (i.e., trouble falling asleep, trouble staying asleep, and sleeping too much were separated; Feeling down, depressed was separated from feeling hopeless; poor appetite was separated from overeating; moving or speaking slowly was separated from being fidgety and restless). To enable comparison between the two PHQ-9 versions, we reaggregated these items to match the original scale by using the highest ratings of the split items in our analyses.\u003c/p\u003e\u003cp\u003eWithin each EMA survey, patients rated how they felt immediately prior to the notification with regard to six positive affect items (excited, energized, activated, calm, peaceful, relaxed) and six negative affect items (nervous, frustrated, angry, bored, sad, tired). These items were chosen from prior studies\u0026rsquo; momentary affect assessments (e.g., Peeters et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Watson et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e1988\u003c/span\u003e), modified to reduce the number of items to reduce participant burden and to ensure an equal number of high activation (i.e., excited, energized, active, nervous, frustrated, angry) versus low activation words (i.e., calm, peaceful, relaxed, bored, sad, tired) for both positive and negative affect (Forgeard et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn sample one, participants rated affect on a Likert scale from 1 (not at all) to 7 (extremely), in sample two participants rated affect on a 0-100 slider scale from 0 (not at all) to (very much). Sample one\u0026rsquo;s affect ratings were rescaled to 0-100 prior to analyses. Also, within each survey, patients reported which skills they had used since the last survey administration, from a list of skills they had learned in treatment. These included cognitive behavior therapy (CBT) skills: behavioral activation, behavioral scheduling, exposure, and cognitive restructuring, and dialectical behavior therapy (DBT) skills: mindfulness, distress tolerance, emotion regulation, and interpersonal effectiveness/communication.\u003c/p\u003e\u003cp\u003eTo obtain diagnostic information, we had doctoral interns and clinical practicum students, trained by a postdoctoral fellow, administer the Mini-International Neuropsychiatric Interview (MINI; Sheehan et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1998\u003c/span\u003e). The MINI has strong reliability and validity relative to the Structured Clinical Interview for DSM-IV (Sheehan et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1998\u003c/span\u003e), and within this PHP, MINI diagnoses from interns and students correlate with those made by program psychiatrists (Kertz et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Clinic practices changed between the collection of the first and second samples such that clinicians only completed MINI modules that were relevant to their patient, and not all clinicians were able to complete the MINI. Thus, diagnoses are reported for the first sample only.\u003c/p\u003e\n\u003ch3\u003eProcedure\u003c/h3\u003e\n\u003cp\u003eUpon discharge, patients were given information about the study, provided written informed consent, and then were trained on how to use the app to respond to surveys. Then, patients completed EMA surveys within the naturalistic setting of their lives. The first sample completed the surveys four times a day for 14 days, at semirandom intervals between 10 a.m. and 8 p.m., separated by at least two hours, with a maximum of 56 surveys. Because the second sample was recruited in part to examine within-person processes, we aimed to collect more surveys per person. Thus, the second sample completed the surveys 6 times a day (i.e., two more times) and for two extra days for a maximum of 96 data points for each participant. Participants in the second sample chose when the first survey of the day would be based on their typical waking time, and each subsequent survey was delivered at 2.5 hour increments after that.\u003c/p\u003e\u003cp\u003eParticipants were paid \u003cspan\u003e$\u003c/span\u003e20 for each week they participated in the study (participants in the second sample received \u003cspan\u003e$\u003c/span\u003e5 extra for the extra two days they completed), plus a bonus of \u003cspan\u003e$\u003c/span\u003e30 per week if they completed 80% of surveys (total possible \u003cspan\u003e$\u003c/span\u003e100 or \u003cspan\u003e$\u003c/span\u003e105). All procedures were approved by the [BLINDED] Human Research Committee Institutional Review Board.\u003c/p\u003e\n\u003ch3\u003eAnalysis\u003c/h3\u003e\n\u003cp\u003eAnalyses were preregistered on OSF (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://osf.io/uems5\u003c/span\u003e\u003cspan address=\"https://osf.io/uems5\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). As noted above, sample one\u0026rsquo;s affect ratings were rescaled to 0-100 to match sample two. End of study PHQ-9 was calculated using the average of up to the last three days of daily PHQ-9 surveys, with missing measurements omitted such that if participants only completed one survey in the last three days, that survey was used as the end of study PHQ-9 score, and if they completed two of three, then the mean of those two measurements was used as the end of study PHQ-9 score.\u003c/p\u003e\n\u003ch3\u003eExplanatory Model\u003c/h3\u003e\n\u003cp\u003eWe conducted a multilevel linear regression to examine whether within-person correlations between positive affect, CBT skills and DBT skills were associated with changes in PHQ-9 post-discharge. We entered PHQ-9 aggregate score as the dependent variable and person-specific CBT-PA correlation, person-specific DBT-PA correlation, and the time (discharge and end of study) as predictors, with interaction effects for each of the correlations with time. Subject ID was associated with a random intercept. We used the \u003cem\u003eLme4\u003c/em\u003e package in \u003cem\u003eR\u003c/em\u003e (Bates et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) with the following syntax:\u003c/p\u003e\u003cp\u003ePHQ9\u0026thinsp;~\u0026thinsp;TIME\u0026thinsp;+\u0026thinsp;CBT-PAcorr\u0026thinsp;+\u0026thinsp;DBT-PAcorr\u0026thinsp;+\u0026thinsp;CBT-PAcorr:TIME\u0026thinsp;+\u0026thinsp;DBT-PAcorr:TIME + (1|ID)\u003c/p\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003ePredictive Model\u003c/h2\u003e\u003cp\u003eTo examine the predictive potential of skills use and affect for determining which participants would continue to improve after treatment vs which would deteriorate or remain unchanged, we conducted a regularized logistic regression predicting outcome group. Outcome group was determined as follows: patients were categorized as improving if they reported a clinically significant decrease of at least 5 points in PHQ-9 score from discharge to the end of the study \u003cem\u003eor\u003c/em\u003e their PHQ-9 score remained\u0026thinsp;\u0026lt;\u0026thinsp;5 at both time points; otherwise, patients were categorized as unchanged/deteriorating.\u003c/p\u003e\u003cp\u003eIn addition to the correlations between skills use (CBT/DBT) and positive affect, preregistered predictors included: the number of times over the two weeks that patients used each of the individual skills (i.e., psychoeducation, self-assessment, behavioral activation, behavioral scheduling, exposure, mindfulness, distress tolerance, emotion regulation, cognitive restructuring, and interpersonal effectiveness) and the total number of times a patient used any skill.\u003c/p\u003e\u003cp\u003eThe analyses for this study were originally preregistered with only positive affect items included, as a previous study with a subset of this sample found that on average, negative affect was not associated with therapeutic skills use (Forgeard et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). However, given that the correlations are calculated within-person, this may not preclude individual differences in this correlation corresponding to outcome. Additionally, given that prediction models benefit from additional predictors and that our regularization process, described below, drops non-important predictors from the model, we opted to include additional predictors. Thus, we also ran a model with the correlations between negative affect and CBT and DBT skills and an expanded set of other predictors, described below. We have included the preregistered model with fewer predictors in the supplementary materials.\u003c/p\u003e\u003cp\u003eThe additional predictors included BASIS-24 score at discharge to control for general severity; summaries of affect, including the within-individual mean and standard deviation of negative and positive affect, respectively, over the whole EMA period; total number of stressful events reported by the participant; and the proportion of stressful events where a skill was used, which we calculated by counting the number of times a stressful event was reported where any skill was used within the time period and dividing that number by the number of total reported stressful events.\u003c/p\u003e\u003cp\u003eWe used elastic net regularization to perform the prediction, which combines two regularization strategies, least absolute selection and shrinkage operator (LASSO) and ridge regression, to drop predictors from the model whose coefficients are likely spurious. This type of regularization is especially useful in cases where the predictors may be correlated, such as in the current study, wherein individuals may use skills simultaneously. To reduce the risk of overfitting, we used 10-fold cross-validation to select the lambda for the model. In this method, 1/10th of the data is a hold-out sample for testing, and 9/10ths as training, such that each tested hyperparameter predicted unseen data. We selected the minimum lambda that provided the best fit to the data. Elastic net regularization also requires a hyperparameter alpha, which determines the mix of LASSO and ridge regression; we selected alpha\u0026thinsp;=\u0026thinsp;.05 to balance each regularization method evenly. We report the accuracy, sensitivity, and specificity of the model by comparing predicted and observed group assignment. We used the R package \u003cem\u003eglmnet\u003c/em\u003e to conduct the cross-validated elastic net regression (Friedman et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2010\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003eDescriptives\u003c/h2\u003e\u003cp\u003eOn average, participants completed 51.10 surveys (\u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;17, range\u0026thinsp;=\u0026thinsp;25\u0026ndash;91). At discharge (i.e., before the EMA period), the improving group and the deteriorating/unchanged group did not significantly differ in their PHQ-9 severity (\u003cem\u003et\u003c/em\u003e(30)\u0026thinsp;=\u0026thinsp;1, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.20) or total BASIS-24 severity (\u003cem\u003et\u003c/em\u003e(32)\u0026thinsp;=\u0026thinsp;.30, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.80). Descriptive information about each group\u0026rsquo;s affect and skills use over the course of the EMA period and their symptoms at discharge and after the EMA period are included in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cem\u003eAffect, skills use, and symptom descriptive information for each outcome group\u003c/em\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003eImproving\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eNon-improvement\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cem\u003eM\u003c/em\u003e (\u003cem\u003eSD\u003c/em\u003e)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRange\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cem\u003eM\u003c/em\u003e (\u003cem\u003eSD\u003c/em\u003e)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eRange\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eWithin-person affect\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA i\u003cem\u003eM\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e29.77 (14.98)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e10.16\u0026ndash;79.38\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e26.87 (14.13)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.3\u0026ndash;57.93\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNA i\u003cem\u003eM\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e25.07 (19.49)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.78\u0026ndash;71.16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e25.1 (12.07)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e6.78\u0026ndash;58.97\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA i\u003cem\u003eSD\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e11.57 (4.56)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.28\u0026ndash;26.45\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e11.02 (3.29)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.58\u0026ndash;20.56\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNA i\u003cem\u003eSD\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e11.27 (6.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.16\u0026ndash;30.66\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e11.18 (3.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.4\u0026ndash;18.82\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA\u0026ndash; CBT i\u003cem\u003er\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.03 (0.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.38\u0026ndash;0.36\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.14 (0.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.19\u0026ndash;0.63\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA-DBT i\u003cem\u003er\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.05 (0.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.46\u0026ndash;0.39\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.03 (0.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.51\u0026ndash;0.32\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNA-CBT i\u003cem\u003er\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.04 (0.15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.33\u0026ndash;0.37\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.1 (0.21)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.43\u0026ndash;0.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNA-DBT i\u003cem\u003er\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0 (0.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.22\u0026ndash;0.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.04 (0.22)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.28\u0026ndash;0.61\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eSkills Use\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal CBT skills\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e35.65 (27.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2 -138\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e25.26 (16.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e4\u0026ndash;72\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal DBT skills\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e32.12 (29.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2\u0026ndash;121\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e20.26 (17.77)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e4\u0026ndash;88\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal skills\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e67.77 (49.14)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e11\u0026ndash;259\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e45.53 (29.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e12\u0026ndash;160\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal stressful events\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.92 (6.14)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e7.09 (6.95)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;31\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eProp. events w skills\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.73 (0.36)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.64 (0.32)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;1\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSelf-Assessment\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.69 (3.72)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.21 (5.96)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;26\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBehavioral Scheduling\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.08 (4.95)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.29 (5.62)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;21\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBehavioral Activation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e17.81 (15.32)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;69\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e10.12 (8.61)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;32\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eExposure\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.12 (7.84)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;29\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.74 (5.89)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;18\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCognitive Restructuring\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.65 (6.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;23\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.12 (7.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;40\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMindfulness\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e11.35 (10.86)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;49\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e7.88 (6.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;21\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDistress Tolerance\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3.73 (4.83)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;14\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.26 (4.65)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0 - 20\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEmotion Regulation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e9.27 (12.69)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;56\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.21 (6.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;25\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInterpersonal Effectiveness\u003c/p\u003e\u003cp\u003ecommunication\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e7.77 (10.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;41\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.91 (5.55)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;25\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eProportion DBT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.44 (0.22)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.09\u0026ndash;0.98\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.44 (0.19)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.07\u0026ndash;0.87\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eSymptoms\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePHQ-9 discharge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e11.92 (8.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e9.85 (2.78)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e6\u0026ndash;16\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePHQ-9 end (last 3 days)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.08 (3.96)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;14.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e9.77 (4.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2\u0026ndash;23\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePHQ-9 change\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.84 (6.07)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-3\u0026ndash;24.67\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.08 (2.71)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-7\u0026ndash;4\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBASIS-24 discharge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.3 (0.79)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.11\u0026ndash;2.84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.25 (0.34)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.59\u0026ndash;2.17\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDepression Functioning\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.78 (1.13)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.09\u0026ndash;3.65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.57 (0.48)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.61\u0026ndash;2.92\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRelationships\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.99 (0.62)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;2.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.26 (0.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.28\u0026ndash;3.16\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSelf-Harm\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.76 (0.93)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;2.84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.53 (0.76)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;2.58\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEmotional Lability\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.36 (1.06)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;3.89\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.6 (0.75)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePsychosis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.21 (0.38)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;1.52\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.22 (0.35)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;1.22\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSubstance Abuse\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.3 (0.46)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u0026ndash;1.38\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.46 (0.62)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0\u0026ndash;2.18\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003eSimplified explanatory model\u003c/h2\u003e\u003cp\u003eIn the explanatory model, the only significant predictor of PHQ-9 score was time (pre versus time post), \u003cem\u003eB\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.16, SE\u0026thinsp;=\u0026thinsp;.76, \u003cem\u003et\u003c/em\u003e(60)\u0026thinsp;\u003cem\u003e=\u003c/em\u003e\u0026thinsp;4.17, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001). All remaining coefficients were nonsignificant (all \u003cem\u003ep\u003c/em\u003es\u0026thinsp;\u0026gt;\u0026thinsp;.05).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003ePrediction Model\u003c/h2\u003e\u003cp\u003eThe results of the preregistered model are reported in the supplementary materials (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Differences between those models and the model reported below are minimal. Below, we describe the results for the full set of predictors described in the Method section.\u003c/p\u003e\u003cp\u003eThe 10-fold cross validation resulted in a minimum lambda of .15 selected for the final model. The accuracy of the model\u0026rsquo;s predictions was 62% (95% CI [.48, .74]), meaning that the model accurately categorized 62% of patients in the sample into the correct group. The area under the curve was acceptable (.74, Fig.\u0026nbsp;1), however the sensitivity was low (.23) while specificity was high (.91), indicating that the model was not as accurate at predicting improvement as it was predicting non-improvement.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAfter regularization, five predictors were retained (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Total use of behavioral activation (OR\u0026thinsp;=\u0026thinsp;1.23) predicted improvement. More use of emotion regulation was a weak predictor of improvement (OR\u0026thinsp;=\u0026thinsp;1.09). Total use of DBT skills and total use of all skills were retained in the model as predictors of improvement but their predictive power was negligible. BASIS score was not retained in the model. Within-person concurrent positive affect and use of any CBT skill (OR\u0026thinsp;=\u0026thinsp;.84) predicted non-improvement.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cem\u003eCoefficients of the Predictors Retained in the Elastic Net Regression Model with Minimum Lambda.\u003c/em\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePredictor\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLog Odds\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eOdds Ratio\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e(Intercept)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e-0.27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.76\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA-CBT i\u003cem\u003er\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e-0.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.84\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal DBT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.01\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal skills use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.01\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBehavioral Activation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.24\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEmotion Regulation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.09\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.09\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"3\"\u003e\u003cem\u003eNote.\u003c/em\u003e PA\u0026thinsp;=\u0026thinsp;positive affect; CBT\u0026thinsp;=\u0026thinsp;Cognitive Behavior Therapy skills (second-wave); DBT\u0026thinsp;=\u0026thinsp;Dialectical Behavior Therapy skills. The following variables were not retained as predictors after regularization: PA-DBT i\u003cem\u003er\u003c/em\u003e, NA-BT i\u003cem\u003er\u003c/em\u003e, total CBT skills, number of stressful events, self-assessment, behavioral scheduling, cognitive restructuring, mindfulness, interpersonal effectiveness/communication, proportion of stressful events where a skill was used, exposure, distress tolerance, PA i\u003cem\u003eM\u003c/em\u003e, PA i\u003cem\u003eSD\u003c/em\u003e, NA i\u003cem\u003eM\u003c/em\u003e, NA i\u003cem\u003eSD\u003c/em\u003e, BASIS-24.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study examined the predictive accuracy of momentary affect, therapeutic skills use, and their association, for distinguishing between patients who continued to improve after discharge from a partial hospital program and patients who did not. To the authors\u0026rsquo; knowledge, this is the first study to use a within-person correlation between affect and skills use, and one of few studies that focuses on treatment skills drawn from various modalities rather than entire treatment protocols, to predict follow-up outcome.\u003c/p\u003e\u003cp\u003eThe two outcome groups did not significantly differ in severity upon discharge from the treatment program, suggesting that the follow-up outcome is not a simple reflection of post-treatment severity. Thus, other post-discharge factors, including those examined in this study, are likely better predictors of continued improvement. This underscores the importance of better understanding the vulnerable period after discharge and which behaviors may support continued recovery.\u003c/p\u003e\u003cp\u003eWithin-person associations between affect and skills use were not associated with dimensional change on the PHQ-9 from discharge to follow up, however these variables in addition to total use of a range of DBT and CBT skills were acceptable predictors of clinically significant improvement (\u0026gt;\u0026thinsp;5 points reduction on PHQ-9) or stable retention of treatment gains (retained PHQ-9 below 5). One explanation for this discrepancy is that the model that included raw PHQ-9 scores assumes linear change, which may not hold. For instance, we would encounter floor effects when patients improve, and while the linear model would not differentiate between someone who stayed the same with zero symptoms from someone who stayed the same with moderate symptoms, the logistic model would group them separately, as reflects their real-world clinical disposition. Thus, the logistic model better reflects real-world clinical scenarios compared to the linear model. Of course, the logistic model was optimized for accurate prediction rather than significant proportion of variance explained. As such, additional predictors were included in an exploratory fashion compared to the simplified linear model; retained predictors performed over and above those dropped from the model and together were accurate in determining improving vs non-improving patients.\u003c/p\u003e\u003cp\u003eIn the regularized logistic regression, few predictors were retained in the model, but a clear pattern emerged: behavioral skills were stronger predictors of improvement than were CBT skills. The individual DBT skills most strongly predictive of improvement in this study were emotion regulation and communication/interpersonal effectiveness. These findings makes sense in the context of the higher-acuity treatment setting. For patients who have just stepped down from an inpatient unit or who have needed to step up to a higher level of care than outpatient treatment, DBT skills may be more appropriate, as several directly target acute emotional difficulties. These skills might be the most useful for limiting the detrimental effect of mental illness on patients\u0026rsquo; functioning, such as managing their relationships or returning to work. This kind of improvement is the goal of the partial hospitalization program \u0026ndash; to return patients to a level of functioning where they can go back to everyday life. Thus, skills that directly limit detrimental effects on functioning are likely to be the most helpful in this short period after discharge.\u003c/p\u003e\u003cp\u003eThis pattern is supported by some evidence from prior studies that DBT skills may be more effective than CBT skills in acute settings. One study of adolescents in an inpatient unit found that those assigned to DBT intervention experienced better behavioral functioning evidenced by fewer suicide attempts, less self-injury, fewer uses of restraints by treatment staff, and fewer days hospitalized post discharge, compared to those assigned to a CBT-based treatment as usual (Tebbett-Mock et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Similarly, a study conducted in a small non-psychiatric sample of individuals who recently experienced myocardial infarction revealed preliminary evidence that those assigned to the DBT intervention experienced better emotion-focused coping compared to those in the cognitive therapy intervention (Nourisaeed et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAnother aspect of the skills that strongly predicted improvement is their focus on behavior. Importantly, behavioral activation was the only substantial predictor of improvement. This comports with previous findings showing that behavioral activation outperforms cognitive therapy in more severely depressed samples (Dimidjian et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). A prior study conducted within the same partial hospital program as the current study found that in depressed patients, behavioral activation significantly predicted next-day improvement in depression symptoms during treatment, while cognitive restructuring did not (Webb et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Additionally in that sample, DBT skills use significantly predicted next-day improvement in anxiety symptoms. Thus, the current study suggests that the pattern of behavioral activation being associated with better improvement than cognitive restructuring extends to the use of these skills post-discharge as well.\u003c/p\u003e\u003cp\u003eA unique aspect of this study is the use of within-person associations between skill use and affect to predict improvement. These associations are generated within the individual, but compared across individuals, to determine whether a certain tendency between individuals to experience more or less positive affect in conjunction with certain skills is associated with symptom improvement. Indeed, higher within-person associations between CBT skills use and positive affect predict non-improvement. In other words, patients who do not improve two weeks after discharge typically experience positive affect after completing a CBT skill.\u003c/p\u003e\u003cp\u003eConsidering that behavioral activation, which is expected to increase positive affect, is one of the CBT skills included in the composite associated with positive affect, it is surprising that the PA-CBT association is predictive of non-improvement while behavioral activation by itself was predictive of improvement. Perhaps other CBT skills included in this correlation, namely cognitive restructuring, behavioral scheduling, and self-assessment, interfered with this effect. If so, perhaps using these other skills to produce short-term increases in positive affect is ultimately unhelpful for long term improvement. Perhaps patients are replacing negative thoughts with unrealistically positive alternatives - for example, in Pollack et al. (2021), adolescents who imagined more unrealistic positive events exhibited a stronger association between defeat/entrapment and suicidal ideation than those with less positive thinking. Though increasing their short-term relief, ultimately patients may be disappointed when their unrealistic expectations do not materialize. Given that these patients are using these skills on their own, another possibility is that they are implementing them incorrectly. For instance, cognitive flexibility and behavioral scheduling are often easier to implement than behavioral activation, and therefore has the potential to be used maladaptively if it is used to avoid choosing to use other skills that are more warranted in the situation; e.g., choosing to reframe how one is thinking about a recent argument when a more effective strategy might be communicating effectively with the person with whom one was quarrelling. In this situation, the reframing may ultimately serve an avoidance function preventing the individual from solving problems effectively. Another possibility is that some patients may schedule activities but fail to carry them out. Finally, though behavioral activation is designed to improve mood, it may be unpleasant in the short term, so immediate positive affect while engaging in behavioral activation might indicate that the patient is not choosing the most challenging/rewarding activities.\u003c/p\u003e\u003cp\u003eAlthough this study is a fruitful investigation into potential predictors of continued outcome, these findings would be strengthened by testing this model on a new sample. Additionally, these findings are merely correlational, and the absence of random assignment precludes confident claims about causality. However, a strength of this study is its ecological validity, given that it is common for patients in the real-world to be exposed to many different types of skills that could come from different treatment protocols (Cook et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Indeed, this study benefits from the fact that the PHP that the participants in this completed exposes patients to a wide variety of skills, allowing for analyses such as this.\u003c/p\u003e\u003cp\u003eOur study is also limited in its generalizability to non-white and non-heterosexual individuals, given the lack of diversity in our sample on these demographic variables. Additionally, improvement in this study was determined by the PHQ-9 which is not a comprehensive measure of recovery, though it is a common measure of depression in hospital settings, and has been shown to be sensitive to changes in depression symptoms (L\u0026ouml;we et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). For instance, the PHQ-9 would not capture aggression or anxiety symptoms that may have improved but may not covary strongly with PHQ-9 symptoms. We therefore cannot say whether these findings would generalize to improvement in other domains beyond depression symptoms. Additionally, the end of study PHQ-9 score was an average of three daily measurements. The validated version of the PHQ-9 assessed over two weeks might generalize better to findings in the real world. Future studies should aim to include more comprehensive assessments that are personalized to each patient\u0026rsquo;s diagnoses and treatment goals. Improvement beyond symptom reduction should also be explored in future studies, for instance by examining quality of life and other positive outcomes.\u003c/p\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eSummary\u003c/h2\u003e\u003cp\u003eThis study examined whether patients\u0026rsquo; individual patterns of therapeutic skills use and affect in the vulnerable period directly after discharge from a partial hospital program would predict which of these patients continue to improve during the two weeks after they leave treatment. Behavioral activation predicted continued improvement, and a within-person trend of increased positive affect after using a CBT skill predicted non-improvement. These findings suggest that in the acute period directly after discharge from a partial hospital program, patients should focus on using behavioral activation to improve depression symptoms.\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eSF and CB wrote the manuscript. SF, MF, RM and CB contributed to study design and aims. SF and KF collected the data. SF analyzed the data and prepared figures. All authors reviewed the manuscript and provided substantive revisions.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThank you to Joshua Cetron and the Institute for Quantitative Social Sciences for guidance on the preregistration and analysis design for this project.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eCode, data, and materials are available on OSF: https://osf.io/54fu6\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBates, D., M\u0026auml;chler, M., Bolker, B., \u0026amp; Walker, S. (2015). Fitting Linear Mixed-Effects Models Using lme4. \u003cem\u003eJournal of Statistical Software\u003c/em\u003e, \u003cem\u003e67\u003c/em\u003e, 1\u0026ndash;48. https://doi.org/10.18637/jss.v067.i01\u003c/li\u003e\n\u003cli\u003eBeard, C., Hsu, K. J., Rifkin, L. S., Busch, A. B., \u0026amp; Bj\u0026ouml;rgvinsson, T. (2016). Validation of the PHQ-9 in a psychiatric sample. \u003cem\u003eJournal of Affective Disorders\u003c/em\u003e, \u003cem\u003e193\u003c/em\u003e, 267\u0026ndash;273. https://doi.org/10.1016/j.jad.2015.12.075\u003c/li\u003e\n\u003cli\u003eBruce, S. E., Yonkers, K. A., Otto, M. W., Eisen, J. L., Weisberg, R. B., Pagano, M., Shea, M. T., \u0026amp; Keller, M. B. (2005). Influence of Psychiatric Comorbidity on Recovery and Recurrence in Generalized Anxiety Disorder, Social Phobia, and Panic Disorder: A 12-Year Prospective Study. \u003cem\u003eAmerican Journal of Psychiatry\u003c/em\u003e, \u003cem\u003e162\u003c/em\u003e(6), 1179\u0026ndash;1187. https://doi.org/10.1176/appi.ajp.162.6.1179\u003c/li\u003e\n\u003cli\u003eBurcusa, S. L., \u0026amp; Iacono, W. G. (2007). Risk for recurrence in depression. \u003cem\u003eClinical Psychology Review\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e(8), 959\u0026ndash;985. https://doi.org/10.1016/j.cpr.2007.02.005\u003c/li\u003e\n\u003cli\u003eCohen, Z. D., Kim, T. T., Van, H. L., Dekker, J. J. M., \u0026amp; Driessen, E. (2020). A demonstration of a multi-method variable selection approach for treatment selection: Recommending cognitive\u0026ndash;behavioral versus psychodynamic therapy for mild to moderate adult depression. \u003cem\u003ePsychotherapy Research\u003c/em\u003e, \u003cem\u003e30\u003c/em\u003e(2), 137\u0026ndash;150. https://doi.org/10.1080/10503307.2018.1563312\u003c/li\u003e\n\u003cli\u003eCook, J. M., Biyanova, T., Elhai, J., Schnurr, P. P., \u0026amp; Coyne, J. C. (2010). What do psychotherapists really do in practice? An Internet study of over 2,000 practitioners. \u003cem\u003ePsychotherapy: Theory, Research, Practice, Training\u003c/em\u003e, \u003cem\u003e47\u003c/em\u003e(2), 260\u0026ndash;267. https://doi.org/10.1037/a0019788\u003c/li\u003e\n\u003cli\u003eCoutanche, M. N., \u0026amp; Hallion, L. S. (2020). Machine learning for clinical psychology and clinical neuroscience. In \u003cem\u003eThe Cambridge handbook of research methods in clinical psychology\u003c/em\u003e (pp. 467\u0026ndash;482). Cambridge University Press. https://doi.org/10.1017/9781316995808.041\u003c/li\u003e\n\u003cli\u003eDeRubeis, R. J. (2019). The history, current status, and possible future of precision mental health. \u003cem\u003eBehaviour Research and Therapy\u003c/em\u003e, \u003cem\u003e123\u003c/em\u003e, 103506. https://doi.org/10.1016/j.brat.2019.103506\u003c/li\u003e\n\u003cli\u003eDeRubeis, R. J., Cohen, Z. D., Forand, N. R., Fournier, J. C., Gelfand, L. A., \u0026amp; Lorenzo-Luaces, L. (2014). The Personalized Advantage Index: Translating Research on Prediction into Individualized Treatment Recommendations. A Demonstration. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(1), e83875. https://doi.org/10.1371/journal.pone.0083875\u003c/li\u003e\n\u003cli\u003eDiGiorgio, K. E., Glass, C. R., \u0026amp; Arnkoff, D. B. (2010). Therapists\u0026rsquo; Use of DBT: A Survey Study of Clinical Practice. \u003cem\u003eCognitive and Behavioral Practice\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(2), 213\u0026ndash;221. https://doi.org/10.1016/j.cbpra.2009.06.003\u003c/li\u003e\n\u003cli\u003eDimidjian, S., Hollon, S. D., Dobson, K. S., Schmaling, K. B., Kohlenberg, R. J., Addis, M. E., Gallop, R., McGlinchey, J. B., Markley, D. K., Gollan, J. K., Atkins, D. C., Dunner, D. L., \u0026amp; Jacobson, N. S. (2006). Randomized trial of behavioral activation, cognitive therapy, and antidepressant medication in the acute treatment of adults with major depression. \u003cem\u003eJournal of Consulting and Clinical Psychology\u003c/em\u003e, \u003cem\u003e74\u003c/em\u003e(4), 658\u0026ndash;670. https://doi.org/10.1037/0022-006X.74.4.658\u003c/li\u003e\n\u003cli\u003eForgeard, M., Beard, C., Shayani, D., Silverman, A. L., Tsukayama, E., \u0026amp; Bj\u0026ouml;rgvinsson, T. (2021). Predictors of affect following discharge from partial hospitalization: A two-week ecological momentary assessment study. \u003cem\u003ePsychological Medicine\u003c/em\u003e, \u003cem\u003e51\u003c/em\u003e(7), 1157\u0026ndash;1165. https://doi.org/10.1017/S0033291719004057\u003c/li\u003e\n\u003cli\u003eFriedman, J. H., Hastie, T., \u0026amp; Tibshirani, R. (2010). Regularization Paths for Generalized Linear Models via Coordinate Descent. \u003cem\u003eJournal of Statistical Software\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e, 1\u0026ndash;22. https://doi.org/10.18637/jss.v033.i01\u003c/li\u003e\n\u003cli\u003eHeininga, V. E., \u0026amp; Kuppens, P. (2021). Psychopathology and positive emotions in daily life. \u003cem\u003eCurrent Opinion in Behavioral Sciences\u003c/em\u003e, \u003cem\u003e39\u003c/em\u003e, 10\u0026ndash;18. https://doi.org/10.1016/j.cobeha.2020.11.005\u003c/li\u003e\n\u003cli\u003eHuibers, M. J. H., Cohen, Z. D., Lemmens, L. H. J. M., Arntz, A., Peeters, F. P. M. L., Cuijpers, P., \u0026amp; DeRubeis, R. J. (2015). Predicting Optimal Outcomes in Cognitive Therapy or Interpersonal Psychotherapy for Depressed Individuals Using the Personalized Advantage Index Approach. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(11), e0140771. https://doi.org/10.1371/journal.pone.0140771\u003c/li\u003e\n\u003cli\u003eKertz, S. J., Bigda-Peyton, J. S., Rosmarin, D. H., \u0026amp; Bj\u0026ouml;rgvinsson, T. (2012). The importance of worry across diagnostic presentations: Prevalence, severity and associated symptoms in a partial hospital setting. \u003cem\u003eJournal of Anxiety Disorders\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e(1), 126\u0026ndash;133. https://doi.org/10.1016/j.janxdis.2011.10.005\u003c/li\u003e\n\u003cli\u003eKroenke, K., Spitzer, R. L., \u0026amp; Williams, J. B. W. (2001). The PHQ-9. \u003cem\u003eJournal of General Internal Medicine\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(9), 606\u0026ndash;613. https://doi.org/10.1046/j.1525-1497.2001.016009606.x\u003c/li\u003e\n\u003cli\u003eLevy, H. C., O\u0026rsquo;Bryan, E. M., \u0026amp; Tolin, D. F. (2021). A meta-analysis of relapse rates in cognitive-behavioral therapy for anxiety disorders. \u003cem\u003eJournal of Anxiety Disorders\u003c/em\u003e, \u003cem\u003e81\u003c/em\u003e, 102407. https://doi.org/10.1016/j.janxdis.2021.102407\u003c/li\u003e\n\u003cli\u003eLinehan, M. M. (2018). \u003cem\u003eCognitive-Behavioral Treatment of Borderline Personality Disorder\u003c/em\u003e. Guilford Publications.\u003c/li\u003e\n\u003cli\u003eL\u0026ouml;we, B., Kroenke, K., Herzog, W., \u0026amp; Gr\u0026auml;fe, K. (2004). Measuring depression outcome with a brief self-report instrument: Sensitivity to change of the Patient Health Questionnaire (PHQ-9). \u003cem\u003eJournal of Affective Disorders\u003c/em\u003e, \u003cem\u003e81\u003c/em\u003e(1), 61\u0026ndash;66. https://doi.org/10.1016/S0165-0327(03)00198-8\u003c/li\u003e\n\u003cli\u003eMichalak, J., Steinhaus, K., \u0026amp; Heidenreich, T. (2020). (How) Do Therapists Use Mindfulness in Their Clinical Work? A Study on the Implementation of Mindfulness Interventions. \u003cem\u003eMindfulness\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(2), 401\u0026ndash;410. https://doi.org/10.1007/s12671-018-0929-9\u003c/li\u003e\n\u003cli\u003eNourisaeed, A., Ghorban-Shiroudi, S., \u0026amp; Salari, A. (2021). Comparison of the effect of cognitive-behavioral therapy and dialectical behavioral therapy on perceived stress and coping skills in patients after myocardial infarction. \u003cem\u003eARYA Atherosclerosis\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(2), 1\u0026ndash;9. https://doi.org/10.22122/arya.v17i0.2188\u003c/li\u003e\n\u003cli\u003ePeeters, F., Berkhof, J., Rottenberg, J., \u0026amp; Nicolson, N. A. (2010). Ambulatory emotional reactivity to negative daily life events predicts remission from major depressive disorder. \u003cem\u003eBehaviour Research and Therapy\u003c/em\u003e, \u003cem\u003e48\u003c/em\u003e(8), 754\u0026ndash;760. https://doi.org/10.1016/j.brat.2010.04.008\u003c/li\u003e\n\u003cli\u003ePollak, O. H., Guzm\u0026aacute;n, E. M., Shin, K. E., \u0026amp; Cha, C. B. (2021). Defeat, Entrapment, and Positive Future Thinking: Examining Key Theoretical Predictors of Suicidal Ideation Among Adolescents. \u003cem\u003eFrontiers in Psychology\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e, 590388. https://doi.org/10.3389/fpsyg.2021.590388\u003c/li\u003e\n\u003cli\u003eRichards, D. (2011). Prevalence and clinical course of depression: A review. \u003cem\u003eClinical Psychology Review\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e(7), 1117\u0026ndash;1125. https://doi.org/10.1016/j.cpr.2011.07.004\u003c/li\u003e\n\u003cli\u003eSheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., Hergueta, T., Baker, R., \u0026amp; Dunbar, G. C. (1998). The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. \u003cem\u003eThe Journal of Clinical Psychiatry\u003c/em\u003e, \u003cem\u003e59 Suppl 20\u003c/em\u003e, 22-33;quiz 34-57.\u003c/li\u003e\n\u003cli\u003eSpitzer, R. L. (1999). \u003cem\u003ePatient health questionnaire: PHQ\u003c/em\u003e. [New York] : [New York State Psychiatric Institute], [1999] \u0026copy;1999. https://search.library.wisc.edu/catalog/999907635102121\u003c/li\u003e\n\u003cli\u003eTebbett-Mock, A. A., Saito, E., McGee, M., Woloszyn, P., \u0026amp; Venuti, M. (2020). Efficacy of Dialectical Behavior Therapy Versus Treatment as Usual for Acute-Care Inpatient Adolescents. \u003cem\u003eJournal of the American Academy of Child \u0026amp; Adolescent Psychiatry\u003c/em\u003e, \u003cem\u003e59\u003c/em\u003e(1), 149\u0026ndash;156. https://doi.org/10.1016/j.jaac.2019.01.020\u003c/li\u003e\n\u003cli\u003eWatson, D., Clark, L. A., \u0026amp; Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e, \u003cem\u003e54\u003c/em\u003e(6), 1063\u0026ndash;1070. https://doi.org/10.1037/0022-3514.54.6.1063\u003c/li\u003e\n\u003cli\u003eWebb, C. A., Beard, C., Kertz, S. J., Hsu, K. J., \u0026amp; Bj\u0026ouml;rgvinsson, T. (2016). Differential role of CBT skills, DBT skills and psychological flexibility in predicting depressive versus anxiety symptom improvement. \u003cem\u003eBehaviour Research and Therapy\u003c/em\u003e, \u003cem\u003e81\u003c/em\u003e, 12\u0026ndash;20. https://doi.org/10.1016/j.brat.2016.03.006\u003c/li\u003e\n\u003cli\u003eWebb, C. A., Forgeard, M., Israel, E. S., Lovell-Smith, N., Beard, C., \u0026amp; Bj\u0026ouml;rgvinsson, T. (2021). Personalized prescriptions of therapeutic skills from patient characteristics: An ecological momentary assessment approach. \u003cem\u003eJournal of Consulting and Clinical Psychology\u003c/em\u003e. https://doi.org/10.1037/ccp0000555\u003c/li\u003e\n\u003cli\u003eYonkers, K. A., Bruce, S. E., Dyck, I. R., \u0026amp; Keller, M. B. (2003). Chronicity, relapse, and illness\u0026mdash;course of panic disorder, social phobia, and generalized anxiety disorder: Findings in men and women from 8 years of follow-up. \u003cem\u003eDepression and Anxiety\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(3), 173\u0026ndash;179. https://doi.org/10.1002/da.10106\u003c/li\u003e\n\u003cli\u003eYonkers, K. A., Warshaw, M. G., Massion, A. O., \u0026amp; Keller, M. B. (1996). Phenomenology and Course of Generalised Anxiety Disorder. \u003cem\u003eThe British Journal of Psychiatry\u003c/em\u003e, \u003cem\u003e168\u003c/em\u003e(3), 308\u0026ndash;313. http://dx.doi.org/10.1192/bjp.168.3.308\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"cognitive-therapy-and-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"cotr","sideBox":"Learn more about [Cognitive Therapy and Research](http://link.springer.com/journal/10608)","snPcode":"10608","submissionUrl":"https://www.editorialmanager.com/cotr/default.aspx","title":"Cognitive Therapy and Research","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Cognitive behavioral therapy, relapse prevention, prediction, ecological momentary assessment, depression","lastPublishedDoi":"10.21203/rs.3.rs-7614725/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7614725/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground. \u003c/strong\u003ePatients may be vulnerable to relapse in the period immediately following discharge from intensive psychiatric care. Additional attention to factors that support continued recovery after discharge is therefore warranted.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods.\u003c/strong\u003e This study examined whether patterns of therapeutic skills use and affect during the acute post-discharge period predicted continued improvement in depression symptoms of patients who completed a partial hospital program. Using ecological momentary assessment and a predictive modeling approach, we assessed how individual differences in therapeutic skills use and their affective responses to skills use influenced recovery.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults.\u003c/strong\u003eThe final model retained five predictors and achieved acceptable discriminative ability (AUC = .74). Total use of behavioral activation predicted improvement (OR = 1.24). A within-person increase in positive affect following CBT skill use predicted non-improvement (OR = .84).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions. \u003c/strong\u003eThese findings suggest that behavioral activation is an effective skill for sustaining depression symptom improvement in the weeks after discharging from a partial hospital program.\u003c/p\u003e","manuscriptTitle":"What Happens After Intensive Treatment? Post-Discharge Skill Use and Affect as Predictors of Depression Outcomes","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-23 07:47:59","doi":"10.21203/rs.3.rs-7614725/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-01-18T07:08:03+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-17T20:59:03+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-15T12:07:25+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"223878731152305693747189291454844117152","date":"2026-01-06T11:28:02+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"146986999407517736860758178406898798715","date":"2025-12-29T18:17:41+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-09-15T07:05:53+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-09-15T05:36:51+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-09-15T05:35:45+00:00","index":"","fulltext":""},{"type":"submitted","content":"Cognitive Therapy and Research","date":"2025-09-14T21:21:19+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"cognitive-therapy-and-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"cotr","sideBox":"Learn more about [Cognitive Therapy and Research](http://link.springer.com/journal/10608)","snPcode":"10608","submissionUrl":"https://www.editorialmanager.com/cotr/default.aspx","title":"Cognitive Therapy and Research","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"e2fe9cca-58ab-492c-b3f0-042ffe3799ce","owner":[],"postedDate":"September 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-03-16T16:03:05+00:00","versionOfRecord":{"articleIdentity":"rs-7614725","link":"https://doi.org/10.1007/s10608-026-10727-0","journal":{"identity":"cognitive-therapy-and-research","isVorOnly":false,"title":"Cognitive Therapy and Research"},"publishedOn":"2026-03-12 15:58:24","publishedOnDateReadable":"March 12th, 2026"},"versionCreatedAt":"2025-09-23 07:47:59","video":"","vorDoi":"10.1007/s10608-026-10727-0","vorDoiUrl":"https://doi.org/10.1007/s10608-026-10727-0","workflowStages":[]},"version":"v1","identity":"rs-7614725","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7614725","identity":"rs-7614725","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}