Long-term Effectiveness and Predictors of Transdiagnostic Internet-Delivered Cognitive Behavioral Therapy for Emotional Disorders in Specialized Care: Secondary Analysis of a Randomized Controlled Trial

Background Transdiagnostic internet-delivered cognitive behavioral therapy (iCBT) for emotional disorders has been shown to be effective in specialized care in the short term. However, less is known about its long-term effects in this specific setting. In addition, predictors of long-term effectiveness may help to identify what treatments are more suitable for certain individuals. Objective This study aimed to analyze the long-term effectiveness of transdiagnostic iCBT compared with that of treatment as usual (TAU) in specialized care and explore predictors of long-term effectiveness. Methods Mixed models were performed to analyze the long-term effectiveness and predictors of transdiagnostic iCBT (n=99) versus TAU (n=101) in public specialized mental health care. Outcomes included symptoms of depression and anxiety, health-related quality of life (QoL), behavioral inhibition and behavioral activation, comorbidity, and diagnostic status (ie, loss of principal diagnosis) from baseline to 1-year follow-up. Sociodemographic characteristics (sex, age, and education) and clinical variables (principal diagnosis, comorbidity, and symptom severity at baseline) were selected as predictors of long-term changes. Results Compared with baseline, transdiagnostic iCBT was more effective than TAU in improving symptoms of depression (b=–4.16, SE 1.80, 95% CI –7.68 to –0.67), health-related QoL (b=7.63, SE 3.41, 95% CI 1.00-14.28), diagnostic status (b=–0.24, SE 0.09, 95% CI –1.00 to –0.15), and comorbidity at 1-year follow-up (b=–0.58, SE 0.22, 95% CI –1.00 to –0.15). From pretreatment assessment to follow-up, anxiety symptoms improved in both transdiagnostic iCBT and TAU groups, but no significant differences were found between the groups. Regarding the predictors of the long-term effectiveness of transdiagnostic iCBT compared with that of TAU, higher health-related QoL at follow-up was predicted by a baseline diagnosis of anxiety, male sex, and the use of psychiatric medication; fewer comorbid disorders at follow-up were predicted by older age and higher baseline scores on health-related QoL; and fewer depressive symptoms at follow-up were predicted by baseline diagnosis of depression. However, this pattern was not observed for baseline anxiety diagnoses and anxiety symptoms. Conclusions The results suggest that transdiagnostic iCBT is more effective than TAU to target depressive symptoms among patients with emotional disorders. Anxiety symptoms remained stable at 1-year follow-up, with no differences between the groups. Results on predictors suggest that some groups of patients may obtain specific gains after transdiagnostic iCBT. Specifically, and consistent with the literature, patients with baseline depression improved their depression scores at follow-up. However, this pattern was not found for baseline anxiety disorders. More studies on the predictor role of sociodemographic and clinical variables in long-term outcomes of transdiagnostic iCBT are warranted. Future studies should focus on studying the implementation of transdiagnostic iCBT in Spanish public specialized mental health care. Trial Registration ClinicalTrials.gov NCT02345668; https://clinicaltrials.gov/ct2/show/NCT02345668

INTRODUCTION 2a-i) Problem and the type of system/solution "Mixed models were performed to analyze the long-term effectiveness and predictors of transdiagnostic iCBT (EmotionRegulation) (n= 99) versus treatment as usual (n = 101) in public specialized mental health care" "To the best of our knowledge, studies on the long-term effectiveness of iCBT in public specialized mental health care are scarce in the literature. However, the high demand for mental health resources along with the lack of resources in this specific setting [24,25] highlight the need for evidence-based interventions that are also effective in the long term. In a previously published RCT, we examined the effectiveness of a transdiagnostic iCBT compared to treatment as usual (TAU) in public specialized mental health care (González-Robles et al., 2020). Transdiagnostic iCBT was found to be more effective than TAU on measures of anxiety (d= 0.35), depression (d= 0.41), and health-related QoL (d= -0.45) at post-treatment. However, the effects of this intervention in the long term (1-year follow-up) have not yet been analyzed. Therefore, the aim of the present investigation was twofold: a) to analyze the long-term outcomes of transdiagnostic iCBT for emotional disorders, compared to TAU; and b) to analyze potential predictors of the long-term effectiveness of transdiagnostic iCBT versus TAU." 2a-ii) Scientific background, rationale: What is known about the (type of) system "Among the range of iCBT programs, transdiagnostic iCBT for emotional disorders has shown its efficacy and effectiveness in a growing number of randomized controlled trials (RCT). Several meta-analyses have shown that transdiagnostic iCBT is effective in the short term, with pooled effect sizes (Hedges g) in the medium to large range for overall measures of anxiety (0.78-0.82), depression (0.79-0.84), and quality of life (QoL) (0.48-0.56) at posttreatment [13,14]. However, most of these available meta-analytic studies mainly report post-treatment outcomes, which suggests that more research is needed on long-term effects of transdiagnostic iCBT." Does your paper address CONSORT subitem 2b? "Therefore, the aim of the present investigation was twofold: a) to analyze the long-term outcomes of transdiagnostic iCBT for emotional disorders, compared to TAU; and b) to analyze potential predictors of the long-term effectiveness of transdiagnostic iCBT versus TAU" METHODS 3a) CONSORT: Description of trial design (such as parallel, factorial) including allocation ratio "The current study analyzes long-term data (1-year follow-up) on a previously published RCT that compared transdiagnostic iCBT to treatment as usual in public specialized mental healthcare services [26]." 3b) CONSORT: Important changes to methods after trial commencement (such as eligibility criteria), with reasons No changes in the methods have happened after trial commencement.

3b-i) Bug fixes, Downtimes, Content Changes
There are no unexpected events to report in this study. 4a) CONSORT: Eligibility criteria for participants "To participate in the study, patients had to meet the following eligibility criteria: (1) aged 18 years or older; (2) ability to understand and read Spanish; (3) access to the Internet at home and an email address; (4) fulfill Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV-TR) diagnostic criteria [28] for ED, including major depressive disorder, dysthymic disorder, depression not otherwise specified, panic disorder, agoraphobia, social anxiety disorder, generalized anxiety disorder, anxiety not otherwise specified, and obsessive-compulsive disorder; (5) provide written informed consent; (6) absence of schizophrenia, bipolar disorder, and alcohol and/or substance dependence disorder; (7) absence of high risk of suicide; (8) absence of a disabling medical disease that prevented the participant from carrying out the psychological treatment; and (9) not receiving another psychological treatment during the study (in the experimental group). Pharmacological treatment was allowed, but participants had to be taking the same dose during the two months prior to enrolling in the study. In addition, participants in the experimental group whose medication was increased or changed during the study period were excluded from the trial (decreases in pharmacological treatment were accepted)." 4a-i) Computer / Internet literacy This aspect is not included in the eligibility criteria. 4a-ii) Open vs. closed, web-based vs. face-to-face assessments: "Participants were adults who attended public mental health units in Spain to seek psychological and/or psychiatric help between July 2015 and June 2019. Potential participants were identified by the psychiatrists and psychologists in these centers and referred to the study researchers for eligibility assessment (see González-Robles et al., 2020 for a full description of the recruitment process)." 4a-iii) Information giving during recruitment Participants were briefed for recruitment. However, this information is not provided either in any of the previous published studies related to this paper. 4b) CONSORT: Settings and locations where the data were collected "Participants were adults who attended public mental health units in Spain to seek psychological and/or psychiatric help between July 2015 and June 2019. Potential participants were identified by the psychiatrists and psychologists in these centers and referred to the study researchers for eligibility assessment (see González-Robles et al., 2020 for a full description of the recruitment process)." 4b-i) Report if outcomes were (self-)assessed through online questionnaires This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 4b-ii) Report how institutional affiliations are displayed 5) CONSORT: Describe the interventions for each group with sufficient details to allow replication, including how and when they were actually administered 5-i) Mention names, credential, affiliations of the developers, sponsors, and owners

5-vii) Access
This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 5-viii) Mode of delivery, features/functionalities/components of the intervention and comparator, and the theoretical framework "All participants received a 12-module transdiagnostic iCBT protocol (EmotionRegulation) through the web platform a web platform design by our research group [36]. The core components of the treatment are based on the Unified Protocol [37, 38] and some treatment strategies from Dialectical Behavioral Therapy (e.g., what and how techniques) [39]. Participants are trained to learn and practice adaptive emotion regulation skills through the following components: a) present-focused emotional awareness (Modules 4 and 5); b) cognitive flexibility (Modules 6 and 7); c) identification and modification of emotional avoidance patterns and emotion-driven behaviors (Modules 8 and 9); and d) exposure (interoceptive and situational) (Modules 10 and 11). The treatment contains four additional modules: an introduction module (Module 1), a module to facilitate the patient's engagement with the therapy (Module 2), a module with psychoeducation on emotions (Module 3), and a relapse prevention module at the end of the treatment (Module 12). In addition, EmotionRegulation includes a Module 0 (Welcome module) with information and recommendations about how to use the protocol. The modules are presented sequentially to enable step-by-step movement through the program. All participants had access to the protocol for a maximum period of 18 weeks, and they were allowed to use the program any time they wanted to during the trial period (i.e., including the follow-up periods). Additional details about this treatment have been reported [26,27]. The treatment modules and their goals are depicted in Table 1."

5-ix) Describe use parameters
This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)"

5-x) Clarify the level of human involvement
This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)"

5-xi) Report any prompts/reminders used
This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)"

5-xii) Describe any co-interventions (incl. training/support)
This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 6a) CONSORT: Completely defined pre-specified primary and secondary outcome measures, including how and when they were assessed "Principal outcomes Beck Depression Inventory, Second Edition (BDI-II) [29]. The BDI-II is a self-report scale consisting of 21 items about the symptoms that characterize MDD. Scores on each item range from 0 to 3, and the maximum score is 63 points. The instrument has demonstrated internal consistency in both the original version (α=.76-.95) and the Spanish version (α=.87 to .89) [30]. Cronbach α for the BDI-II in this study at baseline was .90. Beck Anxiety Inventory (BAI) [31] is a 21-item self-report questionnaire that assesses anxiety, with a maximum score of 63 points. Each item has a 4-point severity scale (from 0: not at all to 3: severely) that addresses anxiety symptoms experienced during the previous week. Several validation studies have shown adequate psychometric properties, with good to excellent internal consistency (α between 0.85 and 0.94) and convergent and divergent validity. The Spanish version of the BAI has demonstrated high internal consistency (α=.93) [32]. Cronbach α for the BAI in this study at baseline was .92.

Secondary outcomes
Quality of Life EuroQoL-5D-3L Questionnaire (EQ-5D-3L) [33] is a generic instrument that measures health-related quality of life (QoL) and consists of two parts. Part 1 assesses self-reported problems in each of the following five domains: mobility, self-care, daily activities, pain/discomfort, and anxiety/depression. Part 2 records the subject's self-assessed health on a visual analog scale (VAS), a 10-cm vertical line on which the best and worst imaginable health states score 100 and 0, respectively. In this study, health-related QoL was assessed using the VAS. Behavioral Inhibition Scale and Behavioral Activation Scale (BIS/BAS) [34,35] contains 20 items rated from 1 to 4, with seven BIS subscale items that evaluate individuals' emotional responses to impending negative events and 13 BAS items that evaluate individuals' behavioral and emotional responses to potentially positive events. The BIS and BAS have shown good reliability in individuals with ED (α= .73-.92) and good convergent and discriminant validity as indicators of temperament. The internal consistency of the Spanish version ranges between 0.65 and 0.82. Cronbach α for the BIS and BAS subscales in this study at baseline were .61 and .80, respectively."

6a-i) Online questionnaires: describe if they were validated for online use and apply CHERRIES items to describe how the questionnaires were designed/deployed 6a-ii) Describe whether and how "use" (including intensity of use/dosage) was defined/measured/monitored
This aspect is not reported in the current paper. See González-Robles et al. (2021) for a full description of how use was measured and analyzed. 6a-iii) Describe whether, how, and when qualitative feedback from participants was obtained 6b) CONSORT: Any changes to trial outcomes after the trial commenced, with reasons "Participants were adults who attended public mental health units in Spain to seek psychological and/or psychiatric help between July 2015 and June 2019. Potential participants were identified by the psychiatrists and psychologists in these centers and referred to the study researchers for eligibility assessment (see González-Robles et al., 2020 for a full description of the recruitment process)." 7a) CONSORT: How sample size was determined 7a-i) Describe whether and how expected attrition was taken into account when calculating the sample size This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 7b) CONSORT: When applicable, explanation of any interim analyses and stopping guidelines "Principal outcomes Beck Depression Inventory, Second Edition (BDI-II) [29]. The BDI-II is a self-report scale consisting of 21 items about the symptoms that characterize MDD. Scores on each item range from 0 to 3, and the maximum score is 63 points. The instrument has demonstrated internal consistency in both the original version (α=.76-.95) and the Spanish version (α=.87 to .89) [30]. Cronbach α for the BDI-II in this study at baseline was .90. Beck Anxiety Inventory (BAI) [31] is a 21-item self-report questionnaire that assesses anxiety, with a maximum score of 63 points. Each item has a 4-point severity scale (from 0: not at all to 3: severely) that addresses anxiety symptoms experienced during the previous week. Several validation studies have shown adequate psychometric properties, with good to excellent internal consistency (α between 0.85 and 0.94) and convergent and divergent validity. The Spanish version of the BAI has demonstrated high internal consistency (α=.93) [32]. Cronbach α for the BAI in this study at baseline was .92.

Secondary outcomes
Quality of Life EuroQoL-5D-3L Questionnaire (EQ-5D-3L) [33] is a generic instrument that measures health-related quality of life (QoL) and consists of two parts. Part 1 assesses self-reported problems in each of the following five domains: mobility, self-care, daily activities, pain/discomfort, and anxiety/depression. Part 2 records the subject's self-assessed health on a visual analog scale (VAS), a 10-cm vertical line on which the best and worst imaginable health states score 100 and 0, respectively. In this study, health-related QoL was assessed using the VAS. Behavioral Inhibition Scale and Behavioral Activation Scale (BIS/BAS) [34,35] contains 20 items rated from 1 to 4, with seven BIS subscale items that evaluate individuals' emotional responses to impending negative events and 13 BAS items that evaluate individuals' behavioral and emotional responses to potentially positive events. The BIS and BAS have shown good reliability in individuals with ED (α= .73-.92) and good convergent and discriminant validity as indicators of temperament. The internal consistency of the Spanish version ranges between 0.65 and 0.82. Cronbach α for the BIS and BAS subscales in this study at baseline were .61 and .80, respectively." 8a) CONSORT: Method used to generate the random allocation sequence This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 8b) CONSORT: Type of randomisation; details of any restriction (such as blocking and block size) This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 9) CONSORT: Mechanism used to implement the random allocation sequence (such as sequentially numbered containers), describing any steps taken to conceal the sequence until interventions were assigned This aspect is not mentioned in the current paper because it is a secondary analysis but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 10) CONSORT: Who generated the random allocation sequence, who enrolled participants, and who assigned participants to interventions This aspect is not mentioned in the current paper because it is a secondary analysis but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 11a) CONSORT: Blinding -If done, who was blinded after assignment to interventions (for example, participants, care providers, those assessing outcomes) and how 11a-i) Specify who was blinded, and who wasn't This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 11a-ii) Discuss e.g., whether participants knew which intervention was the "intervention of interest" and which one was the "comparator" This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 11b) CONSORT: If relevant, description of the similarity of interventions Not applicable because the comparator was treatment as usual. 12a) CONSORT: Statistical methods used to compare groups for primary and secondary outcomes "Mixed-effects models were conducted to analyze the long-term effects and predictors of EmotionRegulation using the lmer function from the lme4 Rpackage [41], with R version 4.0.2 [42]. Analyses were conducted via Restricted Maximum Likelihood estimation (REML) [43,44]. In contrast to multiple imputation methods (i.e., to fill in missing data) or complete-case data through detection methods (which results in biased estimations), the REML method allows incomplete/unbalanced data to be modeled by finding parameters that maximize the likelihood using all the available data, providing a less-biased estimate of variance components with smaller sample sizes [43,45,46]. To compute the magnitude of between-group changes at one-year follow-up, effect sizes (Cohen's d) were calculated by dividing the differences in means by the pooled SD. Effect sizes were interpreted according to Cohen's convention: effect sizes of 0.20 are considered low, effect sizes of 0.50 are considered medium, and effect sizes of 0.80 and above are considered large [47]." 12a-i) Imputation techniques to deal with attrition / missing values "In contrast to multiple imputation methods (i.e., to fill in missing data) or complete-case data through detection methods (which results in biased estimations), the REML method allows incomplete/unbalanced data to be modeled by finding parameters that maximize the likelihood using all the available data, providing a less-biased estimate of variance components with smaller sample sizes [43,45,46]. " 12b) CONSORT: Methods for additional analyses, such as subgroup analyses and adjusted analyses We analyzed predictors of long-term effectivenes: "Mixed-effects models were conducted to analyze the long-term effects and predictors of EmotionRegulation using the lmer function from the lme4 R-package [41], with R version 4.0.2 [42]." "Furthermore, we selected different baseline variables as potential predictors of long-term outcomes, including demographic variables (i.e., sex [0=men; 1=women], age, and education [0=non-university (basic and medium studies); 1=university]), clinical status (i.e., medication [0=no; 1=yes], principal diagnosis [0=depression; 1=anxiety; 2=OCD], comorbidity (number of clinical diagnoses), and diagnostic status [0=does not meet diagnostic criteria; 1= meets diagnostic criteria]), dispositional traits (i.e., behavioral inhibition [BI] and behavioral activation [BA]), and symptomatology (anxiety [BAI], depression [BDI-II], and health-related QoL [EQ-5D-5L]). Given that we were interested in long-term changes (i.e., one-year follow-up relative to baseline), we focused on Group1*Time4 interactions (i.e., EmotionRegulation*1year follow-up) to analyze the long-term effects and predictors [48]." RESULTS 13a) CONSORT: For each group, the numbers of participants who were randomly assigned, received intended treatment, and were analysed for the primary outcome "The flowchart of participants from baseline to one-year follow-up can be seen in Figure 1. A total of 326 patients expressed interest in the study, 281 of whom were assessed for eligibility. Of these 281 participants, 67 were excluded from the study. A total of 214 participants were randomized to either EmotionRegulation (n=106) or TAU (n=108). In addition, seven patients in each condition withdrew from the study before the pretreatment assessment, and so they were excluded from the analyses. Therefore, the final sample at baseline comprised 99 participants in EmotionRegulation and 101 participants in TAU. One-year follow-up data were obtained from 46 participants (46%) in the EmotionRegulation condition and from 47 participants (47%) in the TAU condition. " 13b) CONSORT: For each group, losses and exclusions after randomisation, together with reasons This aspect is not mentioned in the current paper but the reader is referred to the main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 13b-i) Attrition diagram 14a) CONSORT: Dates defining the periods of recruitment and follow-up This aspect is not mentioned in the current paper but the reader is referred to the study protocol and main outcomes paper for a description of these aspects. "The study design of the RCT has been fully described elsewhere [27]." "(see González-Robles et al., 2020 for a full description of the recruitment process)" 14a-i) Indicate if critical "secular events" fell into the study period 14b) CONSORT: Why the trial ended or was stopped (early) This item is not applicable because the trial did not end or stopped early.

15) CONSORT: A table showing baseline demographic and clinical characteristics for each group
The manuscript includes a table that contains demographic and clinical characteristics of the sample in the results section (Baseline data) ( Table 2).

15-i) Report demographics associated with digital divide issues
The manuscript includes a table that contains these types of demographic data in the results section (Baseline data) ( Table 2).