Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Impact of an acceptance facilitating intervention on psychotherapists’ acceptance of blended therapy

  • Harald Baumeister ,

    Roles Conceptualization, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing

    harald.baumeister@uni-ulm.de

    Affiliation Department of Clinical Psychology and Psychotherapy, Institute of Psychology and Education, Ulm University, Ulm, Germany

  • Yannik Terhorst,

    Roles Formal analysis, Methodology, Writing – review & editing

    Affiliation Department of Clinical Psychology and Psychotherapy, Institute of Psychology and Education, Ulm University, Ulm, Germany

  • Cora Grässle,

    Roles Investigation, Methodology, Project administration, Writing – review & editing

    Affiliation Department of Clinical Psychology and Psychotherapy, Institute of Psychology and Education, Ulm University, Ulm, Germany

  • Maren Freudenstein,

    Roles Investigation, Methodology, Project administration, Writing – review & editing

    Affiliation Department of Clinical Psychology and Psychotherapy, Institute of Psychology and Education, Ulm University, Ulm, Germany

  • Rüdiger Nübling,

    Roles Conceptualization, Resources, Supervision, Validation, Writing – review & editing

    Affiliation Chamber of Psychotherapists Baden-Württemberg, Stuttgart, Germany

  • David Daniel Ebert

    Roles Conceptualization, Supervision, Validation, Writing – review & editing

    Affiliation Faculty of Behavioural and Movement Sciences, Clinical, Neuro- & Developmental Psychology, VU Amsterdam, Amsterdam, The Netherlands

Abstract

Blended therapy is a new approach combining advantages of face-to-face psychotherapy and Internet- and mobile-based interventions. Acceptance is a fundamental precondition for its implementation. The aim of this study was to assess 1) the acceptance of psychotherapists towards blended therapy, 2) the effectiveness of an acceptance facilitating intervention (AFI) on psychotherapists’ acceptance towards blended therapy and 3) to identify potential effect moderators. Psychotherapists (N = 284) were randomly assigned to a control (CG) or an intervention group (IG). The IG received a short video showing an example of blended therapy, the CG an attention placebo video. Both groups received a reliable online questionnaire assessing acceptance, effort expectancy, performance expectancy, facilitating conditions, social influence and internet anxiety. Between group differences were examined using t-tests and Mann-Whitney tests. Exploratory analysis was conducted to identify moderators. Psychotherapists in CG showed mixed baseline acceptance towards blended therapy (low = 40%, moderate = 33%, high = 27%). IG showed significantly higher acceptance compared to CG (d = .27, pone-sided = .029; low = 24%, moderate = 47%, high = 30%). Bootstrapped confidence intervals were overlapping. Performance expectancy (d = .35), effort expectancy (d = .44) and facilitating conditions (d = .28) were significantly increased (p < .05). No effects on social influence and internet anxiety were found (p>.05). Exploratory analysis indicated psychodynamic oriented psychotherapists profiting particularly from the AFI. Blended therapy is a promising approach to improve healthcare. Psychotherapists show mixed acceptance, which might be improvable by AFIs, particularly in subpopulations of initially rather skeptical psychotherapists. Forthcoming studies should extend the present study by shifting focus from attitudes to the impact of different forms of AFIs on uptake.

Introduction

Mental disorders are globally highly prevalent and affect people in all regions worldwide [14], accounting for 32.4% of years lived with disability and 13% of disability adjusted life years [5]. Moreover, the pooled relative risk (RR = 2.22, 95%-CI: 2.12–2.33) of mortality among persons with mental disorders is increased by over 120% compared to persons without mental disorders [6].

Several psychotherapeutic and psychopharmacological interventions are effective in the treatment of mental health conditions [7]. However, low perceived need for help and obstructive attitudes towards mental health treatments limit treatment seeking behavior and staying in treatment [8]. Thus, measures to improve peoples´ attitudes towards mental health care are needed to improve peoples´ intention to use mental health services and ultimately make use of the available evidence-based treatment approaches.

This general call for active dissemination of mental health treatments is particularly true for the evidence-based treatment and prevention approach of internet- and mobile-based interventions (IMIs) [9]. IMIs have been shown to be effective and cost-effective across several mental health conditions [1014]. They frequently have been suggested as one option to increase the dissemination of mental health care in light of advantages over conventional face to face therapies such as being time efficient, being spatial and timely independently usable and allowing a higher degree of anonymity for those who perceive psychotherapy as stigmatizing [1518]. On the other hand, IMIs require more self-regulation, self-reflection and self-management competencies, which might lead to time pressure or frustration [19] and might come along with communication problems given the lack of visual cues in the therapeutic process [20].

A newer approach called “blended therapy” integrates the advantages of both IMIs and conventional psychotherapy, aiming to combine the best of two worlds [21]. Blended therapy might save clinicians´ time compared to traditional psychotherapy, increase the effectiveness of current state-of the art treatment, or might on the other side lead to lower dropout rates in IMIs [21]. As such, blended therapy is suggested as a promising innovation for the psychotherapeutic setting [22].

A fundamental precondition to implement blended therapy in routine care would be that both patients and psychotherapists are willing to use blended therapy [2327]. In the field of stand-alone IMIs several studies showed that the baseline acceptance rate of patients towards IMIs is low amongst different clinical target populations [2832]. Informational material (acceptance facilitating interventions [AFIs]) such as short informational videos, aiming to provide trustworthy information, reduce apprehensions and misconceptions, proofed to be capable of improving patients´ acceptance of IMIs [28,29,31,33,34]. Moreover, patients seem to prefer psychotherapist assisted e-mental health services over stand-alone IMIs [25], suggesting blended therapy as a way to go [35]. Still, improving patients´ acceptance and offering blended therapy approaches to patients might not be sufficient, when psychotherapists´ attitudes towards digital supported therapies are not positive. Previous studies focusing on stand-alone IMIs showed advocating attitudes in psychotherapists [3638]. However, only a few studies evaluated attitudes towards blended therapy [27,39]. Schuster and colleagues [27] reported evidence for the general acceptance of blended therapy with no preference of psychotherapists’ attitudes towards web-based or blended therapy compared to face-to-face therapy [27]. Also Becker and Jensen-Doss [39] reported positive average attitudes towards blended therapy. Psychotherapists are important gate keepers of patients’ treatment choice [40]. Hence, increasing the acceptance of psychotherapists towards blended therapy is of utmost importance. Given their effectiveness in patients AFIs might facilitate psychotherapists´ acceptance of digital supported therapies [2729,31,33,34]. Currently, there is no randomized controlled trail evaluating the effectiveness of an AFI on the acceptance of psychotherapists’ towards blend therapy.

Hence, the present studies aimed (1) to investigate the degree of psychotherapists’ acceptance towards blended therapy and (2) to examine the effectiveness of an AFI on psychotherapists’ acceptance towards blended therapy. In the context of psychotherapists’ attitudes towards stand-alone IMIs theoretical orientation (e.g. cognitive behavioral or psychodynamic) is argued to influence the acceptance [3638]. Thus, the effect of an AFI might be moderated by the theoretical orientation. To further examine this thesis as well as to identify other potential moderators (3) exploratory analyses were conducted.

Materials and methods

This study was an experimental study with a balanced (1:1) randomization scheme. In cooperation with five German psychotherapy chambers (Landespsychotherapeutenkammern Baden-Württemberg, Schleswig-Holstein, Bayern, Hessen, Hamburg) and one medical association (Landesärztekammer Hessen) psychotherapists were recruited from November 2016 till February 2017 via e-mail, internet websites and postal mail. Data were collected via unipark (https://www.unipark.com/ [last accessed on 18.02.2020]) and unipark’s randomization feature was used to allocate survey participants randomly to either the intervention group (IG) or control group (CG). To be included in the study, persons had to be a licensed psychotherapist or a psychotherapist in training. The IG received an AFI video and the CG received a placebo video. Detailed information on the intervention and control video are provided the section “experimental conditions” below. The study design was presented to the ethics committee of Ulm University which deemed this study as ethically uncritical.

About 13,740 psychotherapists were contacted primarily via the email distribution lists of the aforementioned psychotherapy chambers. Moreover, the study was advertised via chambers´ homepages. A total of 513 psychotherapists visited the online survey, 284 started the survey, and 233 were included in per-protocol analyses. A post-hoc power analysis (one-sided Wilcoxon–Mann–Whitney test for independent groups) revealed a power of 1-ß = .64 to detect the present effect of d = 0.27 on the acceptance score between IG and CG at an α-level of 5%.

Experimental conditions

Intervention group—Acceptance facilitating intervention (AFI): The AFI was a 5-minute video presenting information about blended therapy. The video discusses potential psychotherapists’ worries about the use of IMIs [41] and facets of the Unified Theory of Acceptance and Use of Technology (UTAUT) [42] to influence the attitude towards blended therapy positively. In this way, dysfunctional beliefs and worries were challenged and advantages of blended therapy was emphasized. The video was framed by scenes from a F2F-psychotherapy session, showing a role-play scene between a psychotherapist and a patient, with actors acting the parts. The video showed an exemplarily integration of internet-based interventions into psychotherapy. After a short psychotherapy scene, an expert in IMIs (HB) presented various ways in which IMIs could be integrated in psychotherapy. For example, exercises of an internet-based intervention between F2F-sessions or the use of internet-interventions for comorbid disorders were shown. This was accompanied by further information about patients’ empowerment, increased self-efficacy and autonomy, efficient use of F2F-sessions, and improvement of healthcare. To further illustrate benefits of blended therapy as well as to reduce worries, an example patient reported about her positive experience with blended therapy (e.g. experienced flexibility, additional support, simplicity and usability of internet-interventions or data security). The video was developed in cooperation with the School of Advanced Professional Studies, University Ulm and elements of a real internet-intervention against panic attacks (https://www.geton-training.de/Panik.php) were used for illustrative purposes. For further information about the AFI the video script (German) is presented in the supporting information.

Control group—Attention placebo video: The attention placebo video was a video of four minutes with a psychotherapist talking about work load and work burden of psychotherapists. Thus, the video was relevant for psychotherapists, however, without an expected specific impact on psychotherapists attitude towards blended therapy. The placebo video is available under: http://www.kbv.de/html/22421.php [last accessed on 18.02.2020].

Measures

Primary outcome.

Acceptance was operationalized based on the Unified Theory of Acceptance and Use of Technology (UTAUT) [42], which emerged from eight different acceptance models: Theory of Reasoned Action (TRA; [43]), Technology Acceptance Model (TAM, [44]), Motivational Model (MM; [45]), Theory of Planned Behavior (TPB; [46]), Combined TAM and TPB (C-TAM-TPB; [47]), Model of PC Utilization (MPCU; [48]), Innovation Diffusion Theory (IDT; [49] and Social Cognitive Theory (SCT; [50]). While the model was initially developed and validated in the work context [42], the questionnaire has been successfully transferred to the medical field in prior studies (eg. [28,29,31]). As in the original questionnaire, all items were rated on 5-point scales with response options ranging from “does not apply at all (1)” to “applies completely (5)”. Four items assessed acceptance: 1. Generally, I would consider to test blended therapy, 2. I would use blended therapy regularly, if I had the possibility, 3. I would recommend blended therapy to colleagues and 4. I would NOT use blended therapy (inverted item). Items are summed for a total acceptance score (range: 5–20, mean = 12.5). For the original items of the UTAUT questionnaire see [42]. Reliability was excellent (ωtotal = .94). Furthermore, the construct validity of the adopted UTAUT questionnaire was confirmed in a validation study: Confirmatory factor analysis and structural equation modelling for the proposed UTAUT model yielded an excellent fit [51]. Acceptance was assessed after participants watched the intervention (IG) or attention placebo video (CG).

Secondary outcomes.

Based on UTAUT four key predictors (performance expectancy, effort expectancy, social influence, facilitating conditions) were operationalized. Items were based on the UTAUT model [42]. Similar, to the acceptance items, the original UTAUT items were adopted to the medical setting. In addition, internet anxiety was added as a dimension based on previous studies (eg. [28,29,31]). Furthermore, inverted items were created and added in this study to reduce biasing effects caused by the assessment methodology. As outline above the model fit of the adopted UTAUT questionnaire and model (including internet anxiety) is excellent [51].

Performance expectancy was measured by 8 items (reliability, ωtotal = .93), effort expectancy by 6 items (ωtotal = .86), social influence by 3 items (ωtotal = .76), facilitating conditions by 7 items (ωtotal = .80), and Internet anxiety by 3 items (ωtotal = .83). (see Table 1 for items). Items are summed for a total score for each predictor. All five predictor scales were assessed after participants watched the intervention (IG) or attention placebo video (CG).

Sociodemographic data and other variables.

In addition to age and sex, type of psychotherapy license (psychological psychotherapist, child and adolescents psychotherapist, psychiatrist, psychosomatic practitioner, child and adolescents practitioner, other), therapeutic background (behavior psychotherapy, depth psychotherapy, psychodynamic psychotherapy, other), work setting (practice, counseling center, outpatient clinic, inpatient clinic, other), employment (fulltime, part-time, unemployed, other), technology access at work and home (yes/no), frequency of technology use at work and home (5-point scale), expertise using PCs or internet (5-point scales), prior knowledge of blended therapy (5-point scale), experience with blended therapy (5-point scale) were assessed.

Data analysis

All outcomes were analyzed on a per-protocol basis (PP). Individuals were included in per-protocol analyses (= IG/CG watched the video), if the automated system check, whether the video was played completely, and the self-report check “I watched the video” were positive.

For descriptive purposes, the acceptance scale was split in three categories. Cut-off values were defined by the authors as: low acceptance (acceptance sum score: 5–9), medium acceptance (sum score 10–15) and high acceptance (sum score 15–20). Percentages were calculated for each category in total and for both groups. Differences in the frequencies were assessed by Chi-square test.

To assess whether acceptance of intervention group differs significantly from the control group’s mean acceptance, one-sided t-test was used for mean differences with alpha level set to 5%. The acceptance score in the IG was compared against the acceptance score of the CG. Acceptance was asses after participants watched the intervention and control video, respectively. In presence of non-normally distributed data Mann-Whitney test and bootstrapping were used. The bootstrapped 2.5% and 97.5% quantile of the distribution of the resampled group means were used to identify a group difference. If the groups’ quantiles were overlapping no differences were assumed. The normal distribution assumption was tested via Shapiro-Wilk test. Similar to acceptance Mann-Whitney test and bootstrap was used for secondary outcomes (performance expectancy, effort expectancy, social influence, internet anxiety and facilitating conditions), if scales were not normally distributed. Two-sided tests were used for all secondary outcomes.

A linear regression model was used as explorative analysis to identify moderators on the effect of AFI. The variables age, gender, therapeutic background (behavioral psychotherapy, depth psychotherapy, psychodynamic psychotherapy, other), type of license (psychological psychotherapist, child and adolescents psychotherapist, other) and their interactions with group were inserted in an initial model (all variables were effect coded or z-standardized). Group was dummy-coded (1 = IG). The initial model included a total of 20 predictors, inclusive the intercept and all interactions. In a stepwise procedure non-significant predictors were removed from the model until a final model was achieved. Ordinary Least Square (OLS) estimator was employed.

Missingness

A total of 31 items were used to measure acceptance, effort expectancy, performance expectancy, internet anxiety, social influence and facilitating conditions. Missingness for acceptance score was 2.2%. For effort expectancy, performance expectancy, internet anxiety and facilitating conditions missingness was 4.7%, respectively. Dropout is assumed to be independent of the included variables and missing values. A missing completely at random mechanism (MCAR) [52] was assumed for all six variables. Analyses based on the original data using list-wise exclusion yield similar results as analyses using multiple imputations based on predictive mean matching (m = 20). Since no differences occurred, only results from analyses based on the original data using list-wise exclusion are reported in the present study.

Results

Of 284 participants 140 were randomly assigned to the intervention group and 144 to the control group. Mean age was 48.6 (SD = 11.7) and 59% of all participants were female. Based on the system check 21.4% of the participants of the intervention group and 7.8% in the control group did not watch the video completely. Accordingly, for the per protocol analyses 107 participants remain as IG and 126 as CG (see Fig 1). Further demographics of the analyzed sample are summarized in Table 2.

Level of acceptance

Acceptance measured in CG was low to moderate (M = 11.4, SD = 4.8; low = 39.8%, moderate = 33.3%, high = 26.8%). Acceptance in IG was moderate to high (M = 12.7, SD = 4.5; low = 23.8%, medium = 46.7%, high = 29.5%) (Fig 2). Chi-square test revealed significant differences in the frequencies of acceptance categories (χ2 (2, N = 228) = 7.18, p = 0.028).

thumbnail
Fig 2. Acceptance of psychotherapists towards blended therapy in relation to the experimental conditions.

https://doi.org/10.1371/journal.pone.0236995.g002

Between-group effect on acceptance

Acceptance was not normally distributed (Shapiro-Wilk test: p < .001). One-sided Mann-Whitney test showed significant differences between groups (pone-sided = .026). IG showed higher acceptance (MCG = 11.5 [95%-CI: 10.6–12.3], MIG = 12.7 [95%-CI: 11.9–13.6]). The difference between CG and IG in standard deviations is d = 0.27 (95% CI: .01-.53).

Between-group effects on secondary outcomes

Performance expectancy was not normally distributed (Shapiro-Wilk test: p < .001). Mann-Whitney test showed significant differences in location parameters (p = .011). The effect in standard deviations is d = 0.34 (95%-CI: .08–.60) favoring intervention group. Bootstrap resulted in overlapping quantiles (MCG = 24.0 [95%-CI: 22.6–25.3], MIG = 26.5 [95%-CI: 25.1–27.9]). Effort expectancy was normally distributed (Shapiro-Wilk test: p = .166). A t-test revealed significant higher values for IG (MCG = 17.9, MIG = 20.2, t(219.93) = -3.51, p < .001). The effect in standard deviations is d = 0.46 (95%-CI: .20–.71) favoring the IG. Facilitating conditions were normally distributed (Shapiro-Wilk test: p = .130). T-test was significant (t(216,58) = -2.00, p = .046). The mean of IG was MIG = 23.2 and the mean of CG MCG = 22.0 and the effect in standard deviations was d = 0.27 (95%-CI: .01–.53) favoring intervention group. Social influence was not normally distributed (Shapiro-Wilk test: p < .001). Mann-Whitney test showed no significant differences in location parameters (p = .301) and quantiles of bootstrapped means were overlapping (MCG = 7.4 [95%-CI: 7.0–7.9], MIG = 7.87 [95%-CI: 7.41 – 8.34]). Internet anxiety was not normally distributed (Shapiro-Wilk test: p < .001). Mann-Whitney test showed no significant differences in location parameters (p = .759) and quantiles of bootstrapped means were overlapping (MCG = 5.5 [95%-CI: 5.1–6.0], MIG = 5.3 [95%-CI: 4.9–5.8]). Effects on secondary outcomes are summarized in Table 3.

Exploratory analysis of effect moderating variables

Results of the exploratory analysis are summarized in Table 4.

thumbnail
Table 4. Exploratory regression results for z-standardized acceptance scores.

https://doi.org/10.1371/journal.pone.0236995.t004

Psychotherapists using depth psychology and other approaches showed less acceptance towards blended therapy compared to psychotherapists with other type of therapeutic background (see Table 1). Individuals within the IG had an increased acceptance by 0.28 standard deviations compared to average. This effect was increased by 0.26 standard deviations for psychodynamic psychotherapists (see Table 1 and Fig 3).

thumbnail
Fig 3. Interaction effect between effect coded group and effect coded psychodynamic therapist.

https://doi.org/10.1371/journal.pone.0236995.g003

Discussion

This is the first study examining the effectiveness of an acceptance facilitating intervention [AFI] on the acceptance of psychotherapists towards blended therapy. The effect on the primary outcome acceptance was small to medium (d = 0.27). Further small to medium effects on performance expectancy, effort expectancy and on facilitating conditions were found, but no effects were observed on social influence and internet anxiety. Overall, most psychotherapists showed a moderate to high acceptance in the CG, with an average acceptance substantially above the scale mean. Given the placebo video had no effect on the acceptance, we assume that the acceptance in the CG represents the acceptance in the general psychotherapist population. Explorative analysis revealed that AFI effect on acceptance is almost doubled for psychodynamic psychotherapists. Further, explorative analysis showed that the subpopulation of psychotherapist from a depth psychology background had lower than average acceptance.

Blended therapy is a new approach, which combines advantages of both face-to-face psychotherapy and IMIs [21,26] and seems to be a promising approach to improve the current mental health care situation. Based on the above average baseline acceptance found in this study the fundamental precondition for the implementation of online elements into on-site psychotherapy, namely psychotherapists´ acceptance towards blended therapy, seems to be given, which is in line with previous findings [27]. However, the overall acceptance towards blended therapy should be interpreted carefully, as the present study sample is not representative with only a minimal percentage of the psychotherapists following the study invitation, most probably with a bias towards digitally open psychotherapists. The AFI used in this study showed positive effects on psychotherapists´ acceptance towards blended therapy. This result is in line with other studies examining the effects of AFIs on patients’ acceptance (e.g. [28,29,31]). However, prior studies found higher effect sizes (e.g. d = 0.71 [31]), while showing lower baseline acceptance (e.g. 93.7% of all participants reported a low to moderate acceptance [31]). Based on these differences, one could assume that AFIs are especially effective in populations with low acceptance. The present study design is not able to test this assumption, since a pre-test in the intervention group would have been necessary to estimate an interaction between baseline acceptance and AFI. However, such a design adaptation should be considered carefully, as a pre-post-test design might not be adequate for examining an ultra-short AFI delivered in under 5 minutes [28,29,31]. Furthermore, it has to be highlighted that the present study was not powered to identify a small to medium effect size of d 0.27. Thus, future studies with confirmatory design should replicate the present findings.

Another aim of the study was to identify potential effect moderators. The explorative analysis revealed a subpopulation, which showed lower than average acceptance (psychotherapist using depth psychology). Further, being a psychodynamic psychotherapist was found to be a meaningful moderator, with psychodynamic psychotherapists profiting to a greater extend from the AFI compared to psychotherapists with another background. This finding suggests that short informational videos might be particularly useful as a first step for those who are skeptical or uninformed regarding blended therapy, whereas short AFIs might not be sufficient to further increase acceptance of psychotherapists who are already to some degree open to this approach.

When interpreting the present findings some limitations need to be taken into account. First, some data within this study was not normally distributed. Typically used parametric methods would have led to biased results. Non-parametric tests and bootstrap were used in the analyses to obtain robust results [5355]. Thus, differences between CG and IG could be detected and further exploratory analysis was feasible with all assumptions met. For the latter the explorative character should be highlighted. No a-priori assumptions about which variables moderate the effect of the AFI were made, except for psychotherapeutic orientation. Moreover, the study was neither designed nor powered to detected moderator effects. In addition, group sizes for main effects in the explorative analyses were highly unequal (e.g. ratio for other approaches was roughly 1:6), which means the present analyses was also highly underpowered to detect main effects (e.g. post-hoc power analysis to detect a main effect of d = 0.3 for other approaches yield a power of 33%). Thus, the generalizability of the exploratory analysis should be interpreted carefully and further studies validating the present findings are needed.

Second, in all analyses missing cases were excluded. This procedure leads to unbiased estimates, if missingness is missing completely at random (MCAR) [52]. Yet, this assumption cannot be statistically verified. In the present study, missingness was rather low (< 5%) and a replication with multiple imputations yielded no different results. Hence, results are expected to be at low risk of bias.

Third, there might be a baseline imbalance between the experimental conditions regarding gender, which could have biased the present findings. As the more important it seems to replicate the present findings in further experimental studies in order to substantiate the present evidence.

Fourth, only 284 from potentially over 13,000 psychotherapists took part in the present survey and only 233 were included in analyses. Moreover, participants were recruited mainly via online ways (e.g. emails or website) and the survey was also online. This may have led to a selective “internet friendly/familiar” sample, which is not representative for all psychotherapists. This could also be an explanation for the rather high level of acceptance compared to prior studies in representative patient samples [28,29,31]. At the same time, this argues for a higher impact of AFIs in the whole target group, given the findings, that AFIs were seemingly more effective in initially more skeptical participants.

Fifth, in this study the UTAUT model [42] was used and the AFI and the scales were developed based on this model, extended by internet anxiety as a predictor. All outcomes were measured reliably according to typical cut-off values for internal consistency [56]. Reliability scores were calculated using McDonald’s Omega, which is argued to be a better estimator than the often used Cronbach’s alpha [57,58]. Although the UTAUT model was used in this study, the UTAUT model itself and the legitimacy of the introduced predictor internet anxiety was not evaluated. Future studies should test whether the UTAUT model and its extension applies for blended therapy.

Finally, a different design including a pre-test (with substantial time between pre- and post-assessment to avoid a retest/recall bias) to test whether AFIs are indeed more effective in low-acceptance population should be applied. Thereby, different AFI designs (e.g. information paper, presentation format, testing of an example online-component, targeting special population characteristics such as therapeutic background) might facilitate acceptance in different ways and should be tested accordingly in order to examine the most efficient way of increasing participants´ acceptance. For optimization purposes a multiphase optimization strategy (MOST) using fractional factorial designs as recommended by Collins and colleagues could be used [59,60].

Conclusion

Currently psychotherapists in Germany show a mixed acceptance towards blended therapy, which will likely be in a similar vein found in other countries that are not yet much familiar with digital approaches for treating mental disorders. The AFI within this study had a significant small to moderate overall effect on psychotherapists´ acceptance. Thus, AFIs might be an easy to distribute way of facilitating psychotherapists´ attitudes towards blended therapy. Given that not all participants watched the video, implementation strategies should be developed which ensure that psychotherapists´ actually do watch the video (e.g. as controlled CME training). Finally, forthcoming studies need to go beyond acceptance as the outcome and examine psychotherapists´ actual use of blended therapies, which might shift the focus from attitudes to practical, legal and monetary aspect of implementing blended therapies in our daily psychotherapeutic work. Aiming at increasing actual use of blended therapy, we also expect that AFIs need to be more complex than a five-minute video, using a longitudinal approach based on interventions ranging from information over initial workshops to continuous training on the job.

References

  1. 1. Baumeister H, Härter M. Prevalence of mental disorders based on general population surveys. Soc Psychiatry Psychiatr Epidemiol. 2007;42: 537–546. pmid:17516013
  2. 2. Polanczyk G V., Salum GA, Sugaya LS, Caye A, Rohde LA. Annual research review: A meta-analysis of the worldwide prevalence of mental disorders in children and adolescents. J Child Psychol Psychiatry Allied Discip. 2015;56: 345–365. pmid:25649325
  3. 3. Steel Z, Marnane C, Iranpour C, Chey T, Jackson JW, Patel V, et al. The global prevalence of common mental disorders: A systematic review and meta-analysis 1980–2013. Int J Epidemiol. 2014;43: 476–493. pmid:24648481
  4. 4. Wittchen HU, Jacobi F, Rehm J, Gustavsson A, Svensson M, Jönsson B, et al. The size and burden of mental disorders and other disorders of the brain in Europe 2010. Eur Neuropsychopharmacol. 2011;21: 655–679. pmid:21896369
  5. 5. Vigo D, Thornicroft G, Atun R. Estimating the true global burden of mental illness. The Lancet Psychiatry. 2016;3: 171–178. pmid:26851330
  6. 6. Walker ER, McGee RE, Druss BG. Mortality in Mental Disorders and Global Disease Burden Implications. JAMA Psychiatry. 2015;72: 334. pmid:25671328
  7. 7. Nathan PE, Gorman JM. A Guide to Treatments That Work. 4th ed. Oxford University Press; 2015. https://doi.org/10.3233/WOR-152185
  8. 8. Andrade LH, Alonso J, Mneimneh Z, Wells JE, Al-Hamzawi A, Borges G, et al. Barriers to Mental Health Treatment: Results from the WHO World Mental Health (WMH) Surveys. Psychol Med. 2014;44: 1303–1317. pmid:23931656
  9. 9. Ebert DD, Van Daele T, Nordgreen T, Karekla M, Compare A, Zarbo C, et al. Internet- and Mobile-Based Psychological Interventions: Applications, Efficacy, and Potential for Improving Mental Health. Eur Psychol. 2018;23: 167–187.
  10. 10. Andersson G, Cuijpers P, Carlbring P, Riper H, Hedman E. Guided Internet-based vs. face-to-face cognitive behavior therapy for psychiatric and somatic disorders: A systematic review and meta-analysis. World Psychiatry. 2014;13: 288–295. pmid:25273302
  11. 11. Ebert DD, Zarski A-C, Christensen H, Stikkelbroek Y, Cuijpers P, Berking M, et al. Internet and computer-based cognitive behavioral therapy for anxiety and depression in youth: a meta-analysis of randomized controlled outcome trials. PLoS One. 2015;10: e0119895. pmid:25786025
  12. 12. Königbauer J, Letsch J, Doebler P, Ebert DD, Baumeister H. Internet- and mobile-based depression interventions for people with diagnosed depression: A systematic review and meta-analysis. J Affect Disord. 2017;223: 28–40. pmid:28715726
  13. 13. Paganini S, Teigelkötter W, Buntrock C, Baumeister H. Economic evaluations of internet- and mobile-based interventions for the treatment and prevention of depression: A systematic review. J Affect Disord. 2018;225: 733–755. pmid:28922737
  14. 14. Sander L, Rausch L, Baumeister H. Effectiveness of internet-based interventions for the prevention of mental disorders: a systematic review. JMIR Ment Heal. 2016;3: e38.
  15. 15. Andersson G, Titov N. Advantages and limitations of Internet-based interventions for common mental disorders. World Psychiatry. 2014;13: 4–11. pmid:24497236
  16. 16. Baumeister H, Lin J, Ebert DD. Internet-und mobilebasierte Ansätze Psychosoziale Diagnostik und Behandlung in der medizinischen Rehabilitation. Leitthema Bundesgesundheitsbl. 2017;60: 436–444. pmid:28224187
  17. 17. Cuijpers P, Van Straten A, Andersson G. Internet-administered cognitive behavior therapy for health problems: A systematic review. J Behav Med. 2008;31: 169–177. pmid:18165893
  18. 18. Ebert DD, Berking M, Thiart H, Riper H, Laferton JAC, Cuijpers P, et al. Restoring depleted resources: Efficacy and mechanisms of change of an internet-based unguided recovery training for better sleep and psychological detachment from work. Heal Psychol. 2015;34: 1240–1251. pmid:26651465
  19. 19. Rozental A, Boettcher J, Andersson G, Schmidt B, Carlbring P. Negative Effects of Internet Interventions: A Qualitative Content Analysis of Patients’ Experiences with Treatments Delivered Online. Cogn Behav Ther. 2015;44: 223–236. pmid:25705924
  20. 20. Moritz S, Schröder J, Meyer B, Hauschildt M. The more it is needed, the less it is wanted: Attitudes toward face-to-face intervention among depressed patients undergoing online treatment. Depress Anxiety. 2013;30: 157–167. pmid:22930656
  21. 21. Erbe D, Eichert H-C, Riper H, Ebert DD. Blending Face-to-Face and Internet-Based Interventions for the Treatment of Mental Disorders in Adults: Systematic Review. J Med Internet Res. 2017;19: e306. pmid:28916506
  22. 22. Mathiasen K, Andersen TE, Riper H, Kleiboer AAM, Roessler KK. Blended CBT versus face-to-face CBT: a randomised non-inferiority trial. BMC Psychiatry. 2016;16: 432. pmid:27919234
  23. 23. Crutzen R, de Nooijer J, Brouwer W, Oenema A, Brug J, de Vries NK. Strategies to Facilitate Exposure to Internet-Delivered Health Behavior Change Interventions Aimed at Adolescents or Young Adults: A Systematic Review. Heal Educ Behav. 2011;38: 49–62. pmid:21189422
  24. 24. Donkin L, Glozier N. Motivators and motivations to persist with online psychological interventions: a qualitative study of treatment completers. J Med Internet Res. 2012;14: e91. pmid:22743581
  25. 25. Topooco N, Riper H, Araya R, Berking M, Brunn M, Chevreul K, et al. Attitudes towards digital treatment for depression: A European stakeholder survey. Internet Interv. 2017;8: 1–9. pmid:30135823
  26. 26. Wentzel J, van der Vaart R, Bohlmeijer ET, van Gemert-Pijnen JEWC. Mixing Online and Face-to-Face Therapy: How to Benefit From Blended Care in Mental Health Care. JMIR Ment Heal. 2016;3: e9. pmid:26860537
  27. 27. Schuster R, Pokorny R, Berger T, Topooco N, Laireiter A-R. The Advantages and Disadvantages of Online and Blended Therapy: Survey Study Amongst Licensed Psychotherapists in Austria. J Med Internet Res. 2018;20: e11007. pmid:30563817
  28. 28. Baumeister H, Nowoczin L, Lin J, Seifferth H, Seufert J, Laubner K, et al. Impact of an acceptance facilitating intervention on diabetes patients’ acceptance of Internet-based interventions for depression: A randomized controlled trial. Diabetes Res Clin Pract. 2014;105: 30–39. pmid:24862240
  29. 29. Baumeister H, Seifferth H, Lin J, Nowoczin L, Lüking M, Ebert DD. Impact of an Acceptance Facilitating Intervention on Patients’ Acceptance of Internet-based Pain Interventions—A Randomised Controlled Trial. Clin J Pain. 2014;31: 528–535. pmid:24866854
  30. 30. Casey LM, Joy A, Clough BA. The impact of information on attitudes toward e-mental health services. Cyberpsychol Behav Soc Netw. 2013;16: 593–598. pmid:23679567
  31. 31. Ebert DD, Berking M, Cuijpers P, Lehr D, Pörtner M, Baumeister H. Increasing the acceptance of internet-based mental health interventions in primary care patients with depressive symptoms. A randomized controlled trial. J Affect Disord. 2015;176: 9–17. pmid:25682378
  32. 32. Mayer G, Gronewold N, Alvarez S, Bruns B, Hilbel T, Schultz J-H. Acceptance and expectations of medical experts, students and patients towards e-mental health apps (Preprint). JMIR Ment Heal. 2019;6: e14018. pmid:31763990
  33. 33. Brouwer W, Oenema A, Crutzen R, Nooijer J de, Vries NK de, Brug J. What makes people decide to visit and use an internet-delivered behavior-change intervention?: A qualitative study among adults. Health Education. 2009.
  34. 34. Cranen K, Drossaert CHC, Brinkman ES, Braakman-Jansen ALM, Ijzerman MJ, Vollenbroek-Hutten MMR. An exploration of chronic pain patients’ perceptions of home telerehabilitation services. Heal Expect. 2012;15: 339–350. pmid:21348905
  35. 35. Apolinário-Hagen J, Kemper J, Stürmer C. Public Acceptability of E-Mental Health Treatment Services for Psychological Problems: A Scoping Review. JMIR Ment Heal. 2017;4: e10. pmid:28373153
  36. 36. Schröder J, Berger T, Meyer B, Lutz W, Hautzinger M, Späth C, et al. Attitudes Towards Internet Interventions Among Psychotherapists and Individuals with Mild to Moderate Depression Symptoms. Cognit Ther Res. 2017;41: 745–756.
  37. 37. Vigerland S, Ljótsson B, Bergdahl Gustafsson F, Hagert S, Thulin U, Andersson G, et al. Attitudes towards the use of computerized cognitive behavior therapy (cCBT) with children and adolescents: A survey among Swedish mental health professionals. Internet Interv. 2014;1: 111–117.
  38. 38. Wangberg SC, Gammon D, Spitznogle K. In the eyes of the beholder: Exploring psychologists’ attitudes towards and use of e-therapy in Norway. Cyberpsychology Behav. 2007;10: 418–423. pmid:17594266
  39. 39. Becker EM, Jensen-Doss A. Computer-assisted therapies: Examination of therapist-level barriers to their use. Behav Ther. 2013;44: 614–624. pmid:24094787
  40. 40. Du E, Quayle E, Macleod H. Service providers’ perceptions on the uptake of computerised cognitive behavioural therapy (CCBT). PsychNology J. 2013;11: 213–233.
  41. 41. Schröder J. Psychometrische Messung von Einstellungen gegenüber psychologischen Online- Interventionen bei depressiven Personen und Psychotherapeuten (doctoral dissertation). University Hamburg. 2015.
  42. 42. Venkatesh V, Morris MG, Davis GB, Davis FD. User Acceptance of Information Technology: Toward a Unified View. MIS Q. 2003;27: 425–478. Available: http://www.jstor.org/stable/30036540
  43. 43. Fishbein M, Ajzen I. Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley; 1975.
  44. 44. Davis FD. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Q Manag Inf Syst. 1989;13: 319–339.
  45. 45. Davis FD, Bagozzi RP, Warshaw PR. Extrinsic and Intrinsic Motivation to Use Computers in the Workplace. J Appl Soc Psychol. 1992;22: 1111–1132.
  46. 46. Ajzen I. From intentions to actions: A theory of planned behavior. Action control: From cognition to behavior. New York: Springer; 1985. pp. 11–39.
  47. 47. Taylor S, Todd P. Assessing IT usage: The role of prior experience. MIS Q Manag Inf Syst. 1995;19: 561–568.
  48. 48. Thompson RL, Higgins CA, Howell JM. Personal computing: Toward a conceptual model of utilization. MIS Q Manag Inf Syst. 1991;15: 125–142.
  49. 49. Moore GC, Benbasat I. Development of an instrument to measure the perceptions of adopting an information technology innovation. Inf Syst Res. 1991;2: 192–222.
  50. 50. Bandura A. Social foundations of thought and action: A social cognitive theory. Englewood Cliffs: Prentice Hall; 1986.
  51. 51. Terhorst Y, Philippi P, Ebert DD, Lin J, Messner E-M, Baumeister H. Unified Theory of Acceptance and Use of Technology–A Model Validation in the Context of Digital Health. Prep. 2020.
  52. 52. Enders CK. Applied Missing Data Analysis. Library. 2010.
  53. 53. Efron B. Nonparametric estimates of standard error: The jackknife, the bootstrap and other methods. Biomtlrika. 1981;68: 589–9.
  54. 54. Leonhart R. Lehrbuch Statistik: Einstieg und Vertiefung. Huber; 2013.
  55. 55. Mooney CZ, Duval RD. Bootstrapping: A nonparametric approach to statistical inference. Sage; 1993.
  56. 56. George D, Mallery P. SPSS for Windows step by step: a simple guide and reference, 11.0 update. Allyn and Bacon; 2003.
  57. 57. Dunn TJ, Baguley T, Brunsden V. From alpha to omega: A practical solution to the pervasive problem of internal consistency estimation. Br J Psychol. 2014;105: 399–412. pmid:24844115
  58. 58. Revelle W, Zinbarg R. Coefficients Alpha, Beta, Omega and GLB: Comments on Sijtsma. Psychometrika. 2009;74: 145–154. Available: http://personality-project.org/revelle/publications/revelle.zinbarg.08.pdf
  59. 59. Collins LM, Chakraborty B, Murphy SA, Strecher VJ. Comparison of a phased experimental approach and a single randomized clinical trial for developing multicomponent behavioral interventions. Clin Trials. 2009;6: 5–15. pmid:19254929
  60. 60. Collins LM, Murphy SA, Nair VN, Strecher VJ. A strategy for optimizing and evaluating behavioral interventions. Ann Behav Med. 2005;30: 65–73. pmid:16097907