The Therapeutic Alliance in Digital Mental Health Interventions for Serious Mental Illnesses: Narrative Review
Open accessPublished version of a preprint
The DOI matches its Crossref record; 159 of the 159 references it cites were found in a registry or an index; 35 of the 169 sources its claims cite were checked for support.
Abstract
Background: Digital mental health interventions offer unique advantages, and research indicates that these interventions are effective for a range of mental health concerns. Although these interventions are less established for individuals with serious mental illnesses, they demonstrate significant promise. A central consideration in traditional face-to-face therapies is the therapeutic alliance, whereas the nature of a digital therapeutic alliance and its relationship with outcomes requires further attention, particularly for individuals with serious mental illnesses. Objective: This narrative review aims to encourage further consideration and critical evaluation of the therapeutic alliance in digital mental health, specifically for individuals with serious mental illnesses. Methods: A narrative review was conducted by combining 3 main areas of the literature: the first examining the evidence for digital mental health interventions for serious mental illnesses, the second illuminating the nature and role of the therapeutic alliance in digital interventions, and the third surrounding practical considerations to enhance a digital therapeutic alliance. Results: Results indicated that a therapeutic alliance can be cultivated in digital interventions for those with serious mental illnesses, but that it may have unique, yet-to-be-confirmed characteristics in digital contexts. In addition, a therapeutic alliance appears to be less directly associated with outcomes in digital interventions than with those in face-to-face therapies. One possibility is that the digital therapeutic alliance is associated with increased engagement and adherence to digital interventions, through which it appears to influence outcomes. A number of design and implementation considerations may enhance the digital therapeutic alliance, including human support and technological features. Conclusions: More research is required to further understand the nature and specific role of a therapeutic alliance in digital interventions for serious mental illnesses, particularly in informing their design. This review revealed several key research priorities to advance the therapeutic alliance in digital interventions.
Overview
CoChat Summary
TL;DR
This narrative review argues that a therapeutic alliance can be cultivated in digital mental health interventions for serious mental illnesses, but its nature and role may differ from face-to-face therapy. It concludes that the digital alliance may be less directly linked to outcomes and may instead influence outcomes through engagement and adherence, with human support and persuasive design features as potential enhancers.
Key findings
- A therapeutic alliance can be experienced in digital interventions, including in a small number of samples with serious mental illnesses, but it may have unique, yet-to-be-confirmed characteristics in digital contexts.
- Evidence for a direct alliance–outcome link in digital interventions is conflicting: a systematic review found an association only in 3 supported-intervention studies, while multiple supported and unsupported studies found no association; no evidence of such an association was found in unsupported digital interventions.
- The review suggests the digital therapeutic alliance may be associated with increased engagement and adherence, which in turn influence outcomes; across the few studies considering this, all reported an association between alliance and engagement/adherence.
- Human support appears beneficial for alliance, engagement, and adherence, but is not always feasible or desirable; design features such as tailoring/personalization, dialogue support, credibility support, and social support may help cultivate a digital therapeutic alliance.
Methods
- Narrative review integrating 3 literature areas: digital mental health interventions for serious mental illnesses, the therapeutic alliance in digital interventions, and practical considerations to enhance a digital therapeutic alliance.
- Focus on intervention studies with participants with psychotic spectrum or bipolar disorder or any author-defined serious mental illness, drawing on broader mental health literature where minimal SMI data exist; excluded contexts where a human delivers the majority of the intervention, web-delivered traditional face-to-face therapy, and blended therapy.
- No systematic search methods or formal bias/quality assessment were used; findings are presented as a broad integration of relevant literature to identify gaps.
Limitations
- Stated: no systematic search methods or formal bias and quality assessment, creating a significant risk of selection bias; included studies cannot be presumed to represent the available literature, and firm conclusions are not warranted.
- Stated: many included analyses were secondary analyses and many were underpowered due to sample size, inflating type I and type II error risk; conceptual issues may also affect results, such as how social withdrawal, mistrust, or paranoia could interact with engagement and alliance.
- Evident from the text: minimal data were available for serious mental illnesses on alliance–outcome and engagement/adherence relationships; alliance measurement was heterogeneous, measured at different times, and engagement/adherence were often used interchangeably or poorly defined, limiting conclusions.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures
Verification
Claims the paper makes
21 Valid · 7 Not valid · 97 Not checked · 44 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“However, developing digital interventions for individuals experiencing serious mental illnesses can introduce additional challenges because of illness-related factors, such as cognitive impairments or mistrust, thus requiring specific design considerations [12,16].”
§ The Therapeutic Alliance in Digital Mental Health Interventions for Serious Mental Illnesses: Narrative Review > Results > Specific Benefits of Digital Interventions for Serious Mental Illnesses
What the sources say
- Partly supportedcites [12] sourceExperimental — accuracy not yet measured
Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.
- Partly supportedcites [16] sourceExperimental — accuracy not yet measured
Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“3 meta-analyses demonstrated that supported digital interventions for a range of mental health concerns were comparable in efficacy (moderate effect sizes) with face-to-face psychological interventions [20-22].”
§ The Therapeutic Alliance in Digital Mental Health Interventions for Serious Mental Illnesses: Narrative Review > Results > Effectiveness of Digital Interventions
What the sources say
- Not checked yetcites [20] source
- Partly supportedcites [21] sourceExperimental — accuracy not yet measured
Our copy does carry the core comparative point for supported (therapist-guided) online interventions: the Abstract and the anxiety Results state effect sizes of 0.7–1.7 for therapist-assisted interventions with 'efficacy similar to face-to-face CBT', and the Discussion adds that they were 'at least as large as standard psychological treatment (0.31) in primary care'. However, three elements of the claim are not supported by this paper: (1) it is a systematic review, not a meta-analysis — the Discussion states that differences in study groups, interventions and periods made generating a pooled standardized mean difference impossible; (2) its scope is depression and anxiety disorders only (panic disorders and phobias explicitly excluded in the anxiety Methods/Results), not 'a range of mental health concerns'; (3) the effect sizes it reports for supported interventions are described as large/robust (0.6–1.9 for therapist-guided depression CBT; 0.7–1.7 for anxiety), with 'modest' reserved for stand-alone, unguided interventions (0.3–0.7). The explicit 'as effective as face-to-face psychotherapy' framing in the Introduction is credited to other works it reviews (Titov, 2011; Reger & Gahm, 2009), though this paper adopts and relies on the comparison in its own Discussion. (LLM, conf=0.72)
- Not checked yetcites [22] source
boosted- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Every source the paper cites for this claim was checked and supports it.
In the paper
“Similarly, Berry et al [55] adapted the ARM for digital contexts by consulting end users and mental health professionals.”
§ The Therapeutic Alliance in Digital Mental Health Interventions for Serious Mental Illnesses: Narrative Review > Results > Therapeutic Alliance
What the sources say
- Supportedcites [55] source
The Abstract Methods and Methods ('Stage 3: Development of the Mobile Agnew Relationship Measure (mARM)') state that the ARM was adapted for the mHealth/digital context using ratings and wording suggestions from service users and mental health staff, filtered through a decision-making algorithm on which items to retain, drop or reword. The Discussion ('Contextualizing Findings Within Existing Literature') likewise refers to 'our adaptions to the ARM items' and to what 'service users and staff in our study found hard to relate to'. Both parts of the claim — adaptation of the ARM for digital contexts and consultation of end users plus mental health professionals — are therefore stated in our copy. (LLM, conf=0.93)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“However, it should be noted that some studies involving direct comparisons of supported and unsupported interventions have failed to demonstrate significant differences in effectiveness and/or adherence [89-91]”
§ The Therapeutic Alliance in Digital Mental Health Interventions for Serious Mental Illnesses: Narrative Review > Results > Therapeutic Alliance
What the sources say
- Supportedcites [89] source
Titov et al. (2015) is precisely such a direct comparison: the Discussion (p. null) states 'A surprising finding from the present study was that no consistent differences were observed between the clinical outcomes of participants receiving weekly clinician contact and those who received treatment in a self-guided format,' and that these findings are 'inconsistent with results of reviews that have found that clinician-guided treatment is associated with better adherence and treatment outcomes than self-guided interventions.' Section 3.3.4 (Treatment completion and satisfaction rates) reports no difference in lessons completed between CG-CBT and SG-CBT (F1,288 = 3.59, p = .059) and no significant differences in satisfaction/recommendation, and the Discussion summary (p. null) notes 'Treatment completion rates were also similar across the groups.' A minor caveat is that the paper defines its self-guided arm as preceded by an initial therapist interview with monitoring, rather than fully automated unsupported treatment, but this does not alter that the trial found no significant differences in effectiveness or adherence between supported and unsupported delivery. (LLM, conf=0.88)
- Not checked yetcites [90] source
- Supportedcites [91] source
The Discussion 'Summary of main results' states that 'results also generally showed no difference in outcomes following therapist-supported ICBT versus unguided CBT at posttreatment', and Implications for research credits this to the directly comparing studies ('the studies included in our comparison of therapist-supported ICBT versus unguided CBT ... suggest no difference in treatment outcome between the two interventions'). Section 2 documents the five head-to-head trials (Titov 2008c; Furmark 2009a/b; Berger 2011; Greist 2012) with non-significant effect estimates (e.g. 2.4 quality of life SMD 0.07, 95% CI -0.37 to 0.50; Berger 2011 16/27 vs 15/27), and 'Quality of the evidence' adds 'generally quite similar rates of dropout across the interventions investigated (experimental and comparator)', covering the adherence limb of the 'and/or' claim. As a meta-analysis it reports rather than originates these study results, which the rubric accepts. (LLM, conf=0.85)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“The findings of multiple meta-analytic and systematic reviews suggest that digital interventions with support tend to have greater effect sizes than those without human support [21,79-83] for outcomes, including depression and anxiety symptoms, and general well-being.”
§ The Therapeutic Alliance in Digital Mental Health Interventions for Serious Mental Illnesses: Narrative Review > Results > Therapeutic Alliance
What the sources say
- Partly supportedcites [21] sourceExperimental — accuracy not yet measured
The Introduction reports that a prior meta-analysis found larger effect sizes for therapist-supported than unsupported interventions. In its sections on depression and anxiety interventions (page unavailable), this review reports larger effect sizes for guided depression CBT, but says the anxiety studies are too few to compare reliably. It does not establish the claimed difference for general well-being. (LLM, conf=0.87)
- Not checked yetcites [79] source
- Not checked yetcites [80] source
- Partly supportedcites [81] sourceExperimental — accuracy not yet measured
Section 3.2 reports a pooled symptom-severity effect favoring guided over unguided interventions, including trials involving depression and social phobia. The paper does not report general well-being as an outcome of that comparison, so it supports the symptom claim but not that example. (LLM, conf=0.86)
- Not checked yetcites [82] source
- Could not verifycites [83] sourceExperimental — accuracy not yet measured
The prompt contains no matching passage. Searches for therapist support, effect sizes, and well-being returned no full-text passages; the only matches were abstract or metadata text, which cannot serve as evidence under the stated rules. The claim therefore cannot be verified from the available passages of our copy. (LLM, conf=0.65)
hedgedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
88 more claims · 12 with their cited sources checked
Create a free account to see all 169 claims and to verify the 134 references still open.
- It may state two separate claims as one. 64
- Probably not a substantial claim. 21
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“meta-analyses indicate no impairment of the therapeutic alliance by digital mental health interventions of various types (Berger, 2017; Seuling et al., 2023; Tremain et al., 2020)”
From Virtual Reality exposure therapy in the treatment of public speaking anxiety and social anxiety disorder (2024) · cites this paper as Tremain et al., 2020
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Recent narrative reviews have highlighted the importance of combining digital health interventions with human support in a bid to cultivate digital alliances, thereby promoting engagement and effectiveness (Tremain et al., 2020).”
From Therapeutic games to reduce anxiety and depression in young people: A systematic review and exploratory meta‐analysis of their use and effectiveness. (2024) · cites this paper as Tremain et al., 2020
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ The Therapeutic Alliance in Digital Mental Health Interventions for Serious Mental Illnesses: Narrative Review > Results > Therapeutic Alliance
“In a further supported intervention, the WAI-Tech goal and bond subscales were positively correlated with the number of modules completed [53].”
§ The Therapeutic Alliance in Digital Mental Health Interventions for Serious Mental Illnesses: Narrative Review > Results > Therapeutic Alliance
“In another study, ratings on the WAI-Tech for a digital intervention were similar to those on the WAI in a face-to-face (treatment as usual) group but were not associated with the outcome: cocaine abstinence [53].”
Citations
Cites
The Elaboration Likelihood Model of Persuasion
Petty, R. E., & Cacioppo, J. T. (1986) · Advances in Experimental Social Psychology · doi:10.1016/s0065-2601(08)60214-2 · Cited by 8,822
The generalizability of the psychoanalytic concept of the working alliance.
Bordin, E. S. (1979) · Psychotherapy: Theory, Research & Practice · doi:10.1037/h0085885 · Cited by 5,238 · cited 4×
“How the definition by Bordin [39] on therapeutic alliance can apply to unsupported interventions is not yet clear, but it has been suggested that therapeutic goals and therapeutic task components (or analogs thereof) may be cultivated within interactions between users and the intervention, while the bond element may b…”
BackgroundMethodInfluentialDevelopment and validation of the working alliance inventory
Horvath, A. O., & Greenberg, L. S. (1989) · Journal of Counseling Psychology · doi:10.1037//0022-0167.36.2.223 · Cited by 3,520
Relation of the therapeutic alliance with outcome and other variables: a meta-analytic review
Martin, D. J., Garske, J. P., & Davis, M. K. (2000) · Journal of Consulting and Clinical Psychology · doi:10.1037//0022-006x.68.3.438 · Cited by 3,355
Client-centered therapy; its current practice, implications, and theory.
Miles, D. W. (1952) · Psychological Bulletin · doi:10.1037/h0053095 · Cited by 2,112
Alliance in individual psychotherapy
Horvath, A. O., Del Re, A. C., Flückiger, C., & Symonds, D. (2011) · Psychotherapy · doi:10.1037/a0022186 · Cited by 2,072
Yatham, L. N., Kennedy, S. H., Parikh, S. V., Schaffer, A., Bond, D. J., Frey, B. N., Sharma, V., Goldstein, B. I., Rej, S., Beaulieu, S., et al. (2018) · Bipolar Disorders · doi:10.1111/bdi.12609 · Cited by 1,852 · cited 1×
“The current Canadian Network for Mood and Anxiety Treatments and International Society for Bipolar Disorders treatment guidelines conclude that, to date, there is insufficient evidence to support the use of digital interventions as adjunctive therapies for BD [33].”
BackgroundThe alliance in adult psychotherapy: A meta-analytic synthesis
Flückiger, C., Del Re, A. C., Wampold, B. E., & Horvath, A. O. (2018) · Psychotherapy · doi:10.1037/pst0000172 · Cited by 1,763
Persuasive Systems Design: Key Issues, Process Model, and System Features
Oinas-Kukkonen, H., & Harjumaa, M. (2009) · Communications of the Association for Information Systems · doi:10.17705/1cais.02428 · Cited by 1,663
Internet-based cognitive behaviour therapy for symptoms of depression and anxiety: A meta-analysis
SPEK, V., CUIJPERS, P., NYKLÍČEK, I., RIPER, H., KEYZER, J., & POP, V. (2007) · Psychological Medicine · doi:10.1017/s0033291706008944 · Cited by 1,586 · cited 1×
“The findings of multiple meta-analytic and systematic reviews suggest that digital interventions with support tend to have greater effect sizes than those without human support [21,79-83] for outcomes, including depression and anxiety symptoms, and general well-being.”
Background
Research works printed but not linked: 1
- Kessler, RC, 2001, Health Serv Res, 36, 987
Cited by
Items 101–101 of 101
Bergin, A., Allison, A. M., & Hazell, C. M. (2023) · doi:10.2196/preprints.48537 · Cited by 3
Metadata
- Authors
- Tremain, H., McEnery, C., Fletcher, K., & Murray, G.
- Affiliations
- Swinburne University of Technology (AU)
- Orygen (AU)
- Corresponding author
- Tremain, Hailey
- Venue
- JMIR Mental Health
- Publisher
- JMIR Publications Inc.
- Published
- 2020-08-07
- Volume
- 7
- Issue
- 8
- Pages
- e17204
- ISSN
- 2368-7959 (ISSN-L)
- PMC
- PMC7442952
- OpenAlex
- W3033785558
- Cited by
- 193
- References
- 150
- Citations by year
- 2026: 37 · 2025: 45 · 2024: 25 · 2023: 37 · 2022: 23 · 2021: 24 · 2020: 1
- FWCI
- 3.34
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- JMIR Mental Health (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Minerva Access (University of Melbourne) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Swinburne figshare (Swinburne University of Technology) (submittedVersion)
- Topics
- Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
- Social Sciences › Psychology › Clinical Psychology › Psychotherapy Techniques and Applications
- Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Telemedicine and Telehealth Implementation
- Keywords
- therapeutic relationship, digital mental health interventions, severe mental illness, treatment engagement, intervention adherence, intervention design
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-04, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-04 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash and gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.
