Therapist-supported internet therapy for mental disorders – a health technology assessment
Health technology assessment
|Published
The Norwegian Institute of Public Health was commissioned to prepare a complete method health technology assessment on the use of guided internet treatment for mental disorders (eMeistring).
Key message
Mental disorders and substance abuse disorders are common in the Norwegian population. Only a few seek treatment for the disorders. Internet treatment opens for more people who need it to seek help.
We found that:
Therapist-supported internet therapy for anxiety in adults at the end of the study gave:
- Better effect, measured as clinically important improvement and reduction of anxiety-specific symptoms (low confidence), and better quality of life (medium confidence) than no treatment.
- No difference in effect (very low confidence) and slightly better quality of life (very low confidence) compared to non-therapist-supported internet treatment.
- No difference in effect (low confidence) and slightly better quality of life (low confidence) compared to face-to-face treatment.
- Patients were mainly satisfied with the treatment.
- Negative effects were inadequately reported.
Internet treatment with and without therapist support compared to no treatment at the end of the study gave:
- Better effect on symptom relief and functional level than no treatment for depression, anxiety or sleep disturbance in adults and mental disorders in children and adolescents. This was also the case for students who abused alcohol, but not in adults who abused alcohol. We have from medium to very low confidence in these effect estimates.
To put the results in a Norwegian context, we conducted a cost-minimization analysis comparing therapist-supported internet-based cognitive behavioural therapy (“eMestring”) with conventional cognitive behavioural therapy in the Western Norway Regional Health Authority. We found that:
- Direct costs associated with therapist-supported internet therapy are comparable to direct treatment costs associated with conventional face-to-face therapy.
- When patients’ travel costs are included in the calculation, therapist-supported internet therapy have a potential to generate cost savings compared with conventional face-to-face therapy.
The budget implications of introduction of therapist-supported internet therapy as routine treatment are uncertain. Increased use of therapist-supported internet therapy could lead to some savings in terms of reduced travel costs. It can also lead to an increase in the number of patients receiving treatment and thus an increase in total cost.
Background
Mental disorders and substance abuse disorders are common in the Norwegian population. The disorders contribute to significant health loss in the form of reduced workforce, increased sickness absence, increased risk of physical illness and early death. Only a few seek treatment for the disorders. Technological development has opened for more people in need of assistance, seeking help.
Objective
There are several recent systematic reviews on the effect of therapist-supported internet therapy for various diagnoses and the aim of this health technology assessment was to provide an overview of high quality systematic reviews summarizing the clinical effect of therapist-supported internet therapy for various mental disorders in children, adolescents and adults, as well as assessing certainty of effect estimates. In addition, we performed a health economic evaluation and described budgetary consequences of a therapist-supported internet program for mental disorders compared with conventional cognitive behavioural therapy in the Western Norway Regional Health Authority. The health technology assessment is prepared on behalf of the Commission Forum in the National System for Managed Introduction of New Health Technologies within the Specialist Health Service in Norway, i.e. New Methods (“Nye metoder”).
Method
Effect
We conducted systematic searches for literature in nine databases. At least two researchers went independently through title, abstracts and full-text articles. The following criteria were used for search strategy and article selection:
Population: Adults with primary diagnosis of mild to moderate depression and / or anxiety, alcohol dependence or insomnia, and psychical disorders in children and adolescents
Intervention:Therapist-supported internet-based therapy on all platforms (data machine, tablets, mobile phone)
Comparator: Face-to-face treatment, unguided internet therapy, waiting list control
Outcomes: Symptoms, functional level, workforce, quality of life, patient satisfaction, adverse events, waiting time for treatment
Study design: High quality systematic reviews based on Scientific Quality Assessment of Review, Cochrane EPOC group
Language: English, Scandinavian
Exclusion criteria: Systematic reviews with low methodological quality
Relevant systematic reviews within each diagnosis were quality-assessed using relevant checklist. The certainty in the documentation was assessed using Grading of Recommendations Assessment, Development and Evaluation (GRADE).
Health economy
To put the results in a Norwegian context, we performed a cost-minimization analysis in the payer’s perspective, comparing "eMeistring", a therapist-supported internet-based program for mental disorders available in the Western Norway Regional Health Authority, with conventional face- to- face treatment. In addition, we described the budget implications of introduction of this program as a routine treatment in the Western Norway Regional Health Authority.
Results
Effect
The systematic literature searches gave 3684 hits. We reviewed 107 publications in full text after excluding non-relevant publications by title and abstract review. We excluded another 90 publications after full text assessment and was left with 17 systematic reviews. These were quality assessed. For each of the relevant diagnoses, we included the systematic review with the latest search and highest methodological quality. Only one of the systematic reviews exclusively assessed therapist-supported internet therapy. The other systematic reviews included studies that assessed internet treatment with both with and without therapist support. In some of the systematic reviews, separate subgroup analyses were performed between therapist-supported and unguided treatment. In other systematic reviews, it was discussed and possibly qualitatively investigated if there were differences between therapist-supported internet therapy and unguided internet treatment.
Therapist-supported internet therapy
Anxiety:
Compared with minimal or no treatment, adults
Therapist-supported internet therapy with cognitive behavioral therapy gave better symptom relief (standardised mean difference (SMD) -1.06 [95% CI -1.25 to -0.82], low confidence), higher functional level (RR 3.75 [95% CI 2, 51 to 5.60], low confidence) and better quality of life (SMD 0.47 [95% CI 0.38 to 0.57], moderate confidence) than no or minimal treatment. The review did not assess any of the other relevant outcomes for our health technology assessment.
Compared with unguided internet therapy, adults
Therapist-supported internet therapy with cognitive behavioral therapy did not improve symptom relief (SMD -0.22 [95% CI -0.56 to 0.13], very low confidence) and quality of life (SMD 0.07 [95 % CI -0.37 to 0.50], very low confidence) compared to unguided internet therapy. Functional level estimates were not measurable. The review did not assess any of the other relevant outcomes for our health technology assessment.
Compared with face-to-face treatment, adults
Therapist-supported internet therapy with cognitive behavioral therapy did not improve symptom relief (SMD 0.06 [95% CI -0.25 to 0.37], low confidence) or function level (RR 1.09 [95% CI 0.89 to 1.34], low confidence) compared to face-to-face treatment. There was a small difference in quality of life in favour of therapist-supported internet therapy (SMD 0.26 [95% CI 0.06 to 0.45], low confidence). The review did not assess any of the other relevant outcomes for our health technology assessment.
Internet therapy with and without therapist support
Depression:
Compared with no treatment, adults
Internet therapy (with and without therapist-support) with cognitive behavioral therapy gave better symptom relief (SMD 0.74 [95% CI 0.62 to 0.86], moderate confidence) than no treatment. The review did not assess any of the other relevant outcomes for our health technology assessment.
Depression and anxiety (overall effect estimates):
Compared with minimal or no treatment, adults
Internet therapy (with and without therapist-support) with cognitive behavioral therapy gave larger reduction of mental symptoms (g = 0.90 [95% CI 0.74 to 1.00], low confidence) than minimal and no treatment. Participants in the internet therapy group were mostly satisfied with the treatment (median 86% satisfied). The review did not assess any of the other relevant outcomes for our health technology assessment.
Compared with face-to-face treatment
Internet therapy (with and without therapist-support) with cognitive behavioral therapy gave larger reduction of mental symptoms (g = 0.38 [95% CI 0.18 to 0.59], low confidence) than face-to-face treatment.
Insomnia:
Compared with no treatment, adults
Internet therapy (with and without therapist-support) with cognitive behavioral therapy gave better sleep efficiency (g = 0.58 [95% CI 0.36 to 0.81], very low confidence) and improvement in insomnia severity (g = 1.09 [95% CI 0.74 to 1.45], very low confidence) than no treatment. The review did not assess any of the other relevant outcomes for our health technology assessment.
Alcohol abuse:
Compared with no treatment, adults
Internet therapy (with and without therapist-support) with different forms of evidence-based treatment for alcohol abuse gave no difference in alcohol consumption (MD -25.0 grams / week [95% CI -51.9 to 1.9], low confidence) or proportion of participants meeting drinking limit guidelines (RR 1.22 [95% CI 0.79 to 1.89], very low confidence) in adults who misused alcohol as compared to no treatment. There was no difference between the groups in self-reported social problems (very low confidence). The review did not assess any of the other relevant outcomes for our health technology assessment.
Compared with no treatment, students
Internet therapy (with and without therapist-support) with different forms of evidence-based treatment for alcohol abuse lowered alcohol consumption (MD -11.7 grams / week [95% CI -19.3 to -4.1], moderate confidence) and the proportion of participants meeting drinking limit guidelines (RR 1.53 [95% CI 1.09 to 2.17], moderate confidence) in students who misused alcohol compared to no treatment. There were no difference in self-reported social problems (SMD 0 [95% CI -0.10 to 0.10], moderate confidence) between the groups.
Mental disorders in children and adolescents:
Compared with no treatment, children and adolescents
Internet therapy (with and without therapist-support) with cognitive behavioral therapy gave improvement in depressive symptoms (MD 1.68 [95% CI -3.11 to -0.25], very low confidence), anxiety (MD -1.47 [95 % CI -2.36 to -0.59], low confidence) and quality of life (MD -5.55 [95% CI -10.88 to -0.22], very low confidence) in children and adolescents compared to no treatment or waiting list control. More in the control group than in the intervention group left the study early (OR 1.31 [95% CI 1.08 to 1.58], moderate confidence). The review did not assess any of the other relevant outcomes for our health technology assessment.
Health economy
Direct treatment costs associated with therapist-supported internet-based treatment are NOK 17,600 for a treatment series. A treatment series delivered in conventional setting costs about NOK 17,100. When patients’ transportation costs are included in the calculation, internet-based therapy have a potential to generate cost savings compared with conventional face-to-face therapy.
The budget implications of introduction of internet-based therapy as routine treatment are uncertain. Increased use of therapist-supported internet-based therapy could lead to some savings in terms of reduced travel costs. It can also lead to an increase in the number of patients receiving treatment and thus an increase in total cost.
Discussion
Effect
The aim of this health technology assessment was to assess the effect of therapist-supported internet therapy by providing an overview of systematic reviews and assessing the effect estimates in these reviews. The review of systematic reviews in the different areas showed that many high methodological quality reviews have been performed. The assessment of the included randomized controlled studies performed by the authors of the systematic reviews showed that there was variation in the performance of the studies, and this resulted in that our confidence to the effect estimates varied from moderate to very low.
Only one of the included systematic reviews exclusively included studies with therapist-supported internet programs. In the other systematic reviews, the degree of therapist support varied between no support to a large extent of support (6.5 hours). There were consistency in results between therapist-supported internet therapy and aggregated results for therapist-supported and unguided internet therapy.
There seem to be a lack of systematic reviews comparing therapist-supported internet therapy with unguided internet therapy for adults with depression, insomnia or alcohol dependence and mental disorders for children and adolescents. It may also be useful to perform systematic reviews on the effectiveness of the intervention in clinical practice for different diagnoses.
Health economy
The results show that direct treatment costs associated with therapist-supported internet treatment are comparable to costs associated with conventional face-to-face treatment, despite of additional investment cost and platform operation costs. This may indicate that therapists can treat more patients with the help of internet-based programmes compared with standard treatment within the same time. When transportation costs are included in the calculation, therapist-supported internet treatment can generate cost savings.
Conclusion
Therapist-based internet therapy for anxiety in adults at the end of the study gave:
- Better effect, measured as clinically important improvement and reduction of anxiety-specific symptoms, (low confidence) and better quality of life (medium confidence) than no treatment.
- No difference in effect (very low confidence) and slightly better quality of life (very low confidence) compared to non-therapist-based internet treatment.
- No difference in effect (low confidence) and slightly better quality of life (low confidence) compared to face-to-face treatment.
- Patients were mainly satisfied with the treatment.
- Negative effects were inadequately reported.
Internet treatment with and without therapist support compared to no treatment at the end of the study gave:
- Better effect on symptom relief and functional level than no treatment for depression, anxiety or sleep disturbance in adults and mental disorders in children and adolescents. This was also the case for students who abused alcohol, but not in adults who abused alcohol. We have from medium to very low confidence in these effect estimates.
Budget impact
- Direct costs associated with therapist-supported internet therapy are comparable to direct treatment costs associated with conventional face-to-face therapy.
- When patients’ travel costs are included in the calculation, therapist-supported internet therapy have a potential to generate cost savings compared with conventional face-to-face therapy.
The budget implications of introduction of therapist-supported internet therapy as routine treatment are uncertain. Increased use of therapist-supported internet therapy could lead to some savings in terms of reduced travel costs. It can also lead to an increase in the number of patients receiving treatment and thus an increase in total cost.
