- Email: [email protected]
Clinical and Cost-Effectiveness of Therapist-Guided Internet-Delivered Cognitive Behavior Therapy for Older Adults With Symptoms of Depression: A Randomized Controlled Trial
Available online at www.sciencedirect.com
ScienceDirect Behavior Therapy 46 (2015) 193 – 205 www.elsevier.com/locate/bt
Clinical and Cost-Effectiveness of Therapist-Guided Internet-Delivered Cognitive Behavior Therapy for Older Adults With Symptoms of Depression: A Randomized Controlled Trial Nickolai Titov Blake F. Dear Macquarie University Shehzad Ali University of York Judy B. Zou Carolyn N. Lorian Luke Johnston Matthew D. Terides Rony Kayrouz Macquarie University Britt Klein Federation University Australia, The Australian National University, and Swinburne University of Technology Milena Gandy Vincent J. Fogliati Macquarie University
The authors gratefully acknowledge the participants for their involvement and helpful comments. This research was enabled by a National Priority Driven Research Program Grant from beyondblue. The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication. BFD is supported by a National Health and Medical Research Council (NHMRC) Australian Public Health Fellowship. BFD, MDT, and SA had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Address correspondence to Blake F. Dear, Ph.D., eCentreClinic, Department of Psychology, Macquarie University, New South Wales, Australia; e-mail: [email protected]. 0005-7894/© 2014 Association for Behavioral and Cognitive Therapies. Published by Elsevier Ltd. All rights reserved.
Depression is a common and significant health problem among older adults. Unfortunately, while effective psychological treatments exist, few older adults access treatment. The aim of the present randomized controlled trial (RCT) was to examine the efficacy, long-term outcomes, and cost-effectiveness of a therapist-guided internet-delivered cognitive behavior therapy (iCBT) intervention for Australian adults over 60 years of age with symptoms of depression. Participants were randomly allocated to either a treatment group (n = 29) or a delayed-treatment waitlist control group (n = 25). Twenty-seven treatment group participants started the iCBT treatment and 70% completed the treatment within the 8-week course, with 85% of participants providing data at posttreatment. Treatment comprised an online 5-lesson iCBT
194
titov et al.
course with brief weekly contact with a clinical psychologist, delivered over 8 weeks. The primary outcome measure was the Patient Health Questionnaire-9 Item (PHQ-9), a measure of symptoms and severity of depression. Significantly lower scores on the PHQ-9 (Cohen’s d = 2.08; 95% CI: 1.38 – 2.72) and on a measure of anxiety (Generalized Anxiety Disorder-7 Item) (Cohen’s d = 1.22; 95% CI: 0.61 – 1.79) were observed in the treatment group compared to the control group at posttreatment. The treatment group maintained these lower scores at the 3-month and 12-month follow-up time points and the iCBT treatment was rated as acceptable by participants. The treatment group had slightly higher Quality-Adjusted Life-Years (QALYs) than the control group at posttreatment (estimate: 0.012; 95% CI: 0.004 to 0.020) and, while being a higher cost (estimate $52.9 l 95% CI: −23.8 to 128.2), the intervention was cost-effective according to commonly used willingness-to-pay thresholds in Australia. The results support the potential efficacy and cost-effectiveness of therapist-guided iCBT as a treatment for older adults with symptoms of depression. Australian and New Zealand Clinical Trials Registry: ACTRN12611000927921; https://www.anzctr.org.au/ Trial/Registration/TrialReview.aspx?id=343384. Keywords: depression; older adults; internet; cognitive behavior therapy (CBT); cost-effectiveness
DEPRESSION IS A MAJOR HEALTH ISSUE among older adults. Research indicates that 8% to 10% of people over 60 years of age experience clinically significant symptoms (Pirkis et al., 2009) but that depression is often underdiagnosed in older adults (Byers, Yaffe, Covinsky, Friedman, & Bruce, 2010; Gum, KingKallimanis, & Kohn, 2009). This is significant because depression in older adults is associated with poorer physical health (Brenes et al., 2008; Cockayne et al., 2011) and an increased risk of suicide (Grek, 2007). Unfortunately, despite the availability of effective treatments (Ayers, Sorrell, Thorp, & Wetherell, 2007; Gould, Coulson, & Howard, 2012; Nordhus & Pallesen, 2003; Scogin, Welsh, Hanson, Stump, & Coates, 2005), research indicates that the number of older adults seeking and receiving evidence-based treatment is low (Beekman et al., 2002; Wetherell, Lenze & Stanley, 2005). There are numerous barriers to traditional face-to-face psychological treatments and the most significant to older adults include stigma, mobility limitations, costs of treatment, and the limited number of trained therapists. Thus, there is considerable need for treatment approaches that overcome barriers and increase access to evidencebased treatment for older adults. Internet-delivered cognitive behavior therapy (iCBT) is one approach that may reduce barriers to psychological treatments (Andersson & Cuijpers,
2009; Andrews, Cuijpers, Craske, McEvoy, & Titov, 2010; Cuijpers et al., 2009). iCBT treatments provide the same information and teach similar skills as traditional face-to-face CBT treatments, but do so via the internet using structured materials. iCBT may be administered as a self-guided intervention, or may include therapist support provided via email and telephone (Titov, 2011). There is now considerable meta-analytic data supporting the efficacy of iCBT in treating adults with anxiety and depression (Andersson & Cuijpers, 2009; Andrews et al., 2010; Cuijpers et al., 2009). Evidence is also emerging indicating that the results of iCBT are comparable to those obtained in traditional face-to-face CBT (Cuijpers Donker, van Straten, Li, & Andersson, 2010; Kiropoulos et al., 2008) and that iCBT is cost-effective (Hedman, Ljotsson, & Lindfors, 2012). Clinical trials of computer-delivered CBT and iCBT have, however, focussed on adults aged 20 to 60 years to date (Crabb et al., 2012). However, the promising results of this research and the increasing use of the internet by older adults in many countries, including Australia (Ewing & Thomas, 2012), has triggered targeted research focusing on the efficacy of iCBT with older adults (e.g., adults 60 years and older) (Dear et al., 2013; Spek et al., 2007; Zou et al., 2012). For example, a randomized controlled trial (RCT; n = 301) targeting adults over the age of 50 (M = 55; SD = 4.6) with subclinical symptoms of depression revealed an iCBT intervention was at least as effective as group CBT in reducing symptoms of depression, and that both treatments were superior to a waitlist control group (Spek et al., 2007). More recently, a feasibility open trial evaluated an iCBT intervention for adults over the age of 60 (M = 63.4; SD = 5.08) with depression and found a large within-group effect size (Cohen’s d N 1.0) (Dear et al., 2013). While the results of this open trial are encouraging, a controlled research trial is needed to explore the efficacy and cost-effectiveness of iCBT for older adults with symptoms of depression. The present study was designed to examine the efficacy, longer-term outcomes, and cost-effectiveness of a new iCBT intervention, the Managing Your Mood Course, for older adults with symptoms of depression, using an RCT design comparing a treatment and a waitlist control group. The Managing Your Mood Course is a five-lesson intervention delivered over 8 weeks with telephone and email support from a therapist. It was hypothesized that: (a) the treatment group would report significantly reduced symptoms of depression at posttreatment compared with the waitlist-control group, (b) symptom reductions of the treatment group would be sustained at 3-month and 12-month follow-up, and (c) the iCBT treatment would be cost-effective based
internet cbt for older adults with depression
FIGURE 1
195
Participant flow chart.
on Quality-Adjusted Life-Years (QALYs) and commonly employed willingness-to-pay thresholds.
Method participants A total of 111 individuals applied to participate in the study via the eCentreClinic website (www. ecentreclinic.org) and 54 met the criteria for inclusion into the study. The inclusion criteria were as follows:
(a) resident of Australia; (b) 60 years of age and over; (c) reported that they have been assessed by a general practitioner or medical specialist to rule out a reversible physical cause for their depression. An initial additional criterion that applicants had a total score ≥ 10 on the Patient Health Questionnaire-9 Item (PHQ-9; Kroenke, Spitzer, & Williams, 2001) was removed during the trial due to the significant number of applicants who did not meet this criterion,
196
titov et al.
Table 1
Demographic Characteristics of the Treatment and Control Groups Variable
Gender Male Female Age Mean Range Marital Status Single/Never Married Married/ De Facto Separated/Divorced/Widowed Education None High School Trade / Technical Certificate Diploma / Degree Employment Full-time Part-time / Student Unemployed, retired or disabled Previous Mental Health Treatment Taking Medication Hours Per Week Using Internet Mean Range Confidence using the Internet Mean (rating 1 to 5) Range
Treatment Group
Control Group
n
%
n
%
5 22
18.5% 81.5%
9 16
36% 64%
64.52 (SD: 2.58) 61 to 69
66.16 (SD: 3.80) 62 to 76
Significance Statistics
χ 2 (1) = 2.02, p = .16
F = .011, p = .93
1 14 12
3.7% 51.9% 44.4%
4 9 12
16.0% 36.0% 48.0%
χ 2 (2) = 2.81, p = .25
1 6 3 17
3.7% 22.2% 11.1% 63.0%
2 9 6 8
8.0% 36.0% 24.0% 32.0%
χ 2(3) = 5.10, p = .16
2 7 18 18 13
7.4% 25.9% 66.7% 66.7% 48.1%
0 4 21 13 12
0.0% 16.0% 84.0% 52.0% 48.0%
χ 2 (2) = 2.98, p = .23
χ 2(1) = 1.16, p = .28 χ 2 (1) = 0.00, p = .99
15.26 (14.98) 2 to 60
16.28 (11.96) 3 to 50
F = .076, p = .78
2.89 (1.05) 1 to 5
2.88 (1.42 1 to 5
F b .001, p = .98
but who still reported that they felt depressed. Thereafter, all applicants were required to verbally confirm during the telephone interview that they were experiencing difficulties with depression and that they wanted treatment for these symptoms. The exclusion criteria were as follows: (a) current participation in CBT; (b) use of illicit drugs or consumption of more than three standard drinks/day; (c) current diagnosis of schizophrenia or bipolar disorder; (d) severe symptoms of depression (defined as a total score N 19 or responding N 2 to Question 9 (suicidal ideation) on the PHQ-9; (e) if taking medication for anxiety or depression, not having been on a stable dose for at least a month. Details of participant flow are shown in Figure 1. Successful applicants were contacted by the researchers who confirmed the applicant met the inclusion criteria and administered a structured diagnostic interview, the Mini International Neuropsychiatric Interview Version 5 (MINI), via telephone. Group allocation preceded the telephone interview. A total of 29 participants were randomized to the treatment group and 25 were randomized to the waitlist control group. A permuted block randomiza-
tion sequence was generated by an independent researcher at another institution using a random number generator (www.randomizer.org) and each allocation was kept in a sealed envelope until allocation, which was opened after an applicant met the inclusion criteria. The demographic characteristics of the sample are shown in Table 1. The study was approved by the Human Research Ethics Committee (HREC) of Macquarie University, Sydney, Australia, and the trial was registered on the Australian and New Zealand Clinical Trials Registry (ANZCTR) as ACTRN12611000927921.
design and measures The design comprised a CONSORT-revised compliant RCT comparing a treatment group with a delayed-treatment waitlist control group. Questionnaires were administered to the treatment group immediately prior to the treatment, immediately after the treatment, and again 3 months and 12 months posttreatment. All questionnaires were administered online. The MINI diagnostic assessments were administered to the treatment group at application and again at 3 months posttreatment. The waitlist
internet cbt for older adults with depression control group commenced treatment immediately after the treatment group completed the treatment and were administered a self-guided version of the course, that is, without therapist contact, as a part of another trial (Dear et al., submitted). The waitlist control group were administered questionnaires and diagnostic assessments at the same time points as the treatment group.
measures Patient Health Questionnaire-9 Item (PHQ-9; Kroenke et al., 2001) The PHQ-9 comprises 9 items measuring symptoms of major depressive disorder based on the DSM-IV criteria for depression. A total score ≥ 10 has been identified as an important threshold for identifying DSM-IV congruent depression with increasing scores indicating greater symptom severity (Kroenke et al., 2001). The PHQ-9 has been found to have good psychometric properties generally (Kroenke, Spitzer, Williams, & Löwe, 2010; Titov, Dear, McMillan, et al., 2011) and in older adults (Phelan et al., 2010). Cronbach’s alpha for the PHQ-9 in the current study was .84. Generalized Anxiety Disorder 7-Item Scale (GAD-7; Spitzer, Kroenke, Williams, & Löwe, 2006) The GAD-7 comprises 7 items measuring symptoms of anxiety (Löwe et al., 2008). A total score ≥ 8 has been shown to be the optimum sensitive and specific threshold for the possible presence of an anxiety disorder (Kroenke, Spitzer, Williams, Monahan, & Löwe, 2007). The GAD-7 has been found to exhibit good psychometric properties (Dear, Titov, Sunderland, et al., 2011; Spitzer et al., 2006). However, one recent study has found a lower cutoff score of 5 may be required when used with older adults (Wild et al., in press). Cronbach’s alpha for the GAD-7 in the current study was .86. EuroQol - 5 Dimensions - 5 Levels (EQ-5D-5L; Herdman et al., 2011) The EQ-5D-5L is a standard and widely used measure of health-related quality of life (Brazier, 2007; Herdman et al., 2011). It comprises 5 items that inquire about five dimensions of health: mobility, self-care, usual activities, pain and discomfort, and anxiety and depression. Responses to the five dimensions are weighted and summed to create a total score between 0 (i.e., indicating deceased) and 1 (i.e., indicating full-health). The EQ-5D-5L has been shown to have good convergent validity with the World Health Organization 5-item Well Being Questionnaire (Janssen et al., 2013). EQ-5D scores were converted into utility values using an algorithm for the Australian population (Norman et al., 2013).
197
Then, QALYs were calculated using the timeweighted average of the utility scores (Glick, 2007). Mini International Neuropsychiatric Interview Version 5.0.0 (MINI; Sheehan et al., 1998) The MINI is a brief diagnostic interview developed to determine the presence of current Axis-I disorders using DSM-IV diagnostic criteria. It has excellent interrater reliability (k = 0.88 to 1.00) and adequate concurrent validity with the Composite International Diagnostic Interview (World Health Organization, 1990).
intervention The Managing Your Mood Course is a new 8-week treatment intervention based on the general structure and core psychological principles included in other evidence-based CBT interventions developed by the authors (Dear, Titov, Schwencke, et al., 2011; Dear et al., 2013; Titov, Dear, Schwencke, et al., 2011; Titov et al., 2013; Titov et al., 2014; Zou et al., 2012). The course consists of five online lessons, five lesson summaries and homework assignments, regular automated reminder and notification emails, secure internal email-type messaging facility with a therapist, and case stories detailing the experiences of older adults recovering from depression (see Table S1). Each lesson is presented in a slide show format that includes a combination of didactic material (i.e., text-based instructions and information) and case-enhanced learning material (i.e., educational stories which demonstrate the application of skills and problem resolution). To facilitate gradual learning and mastery of skills, an additional week is provided for the more complex lessons (i.e., thought challenging and graded exposure), and participants are prevented from skipping lessons or progressing too quickly through the materials. Participants are instructed to read the lessons over 8 weeks according to a timetable. Homework assignments are provided for each lesson. Additional resources are provided during the 8 weeks that include materials about assertiveness communication skills, problem solving, managing worry, challenging beliefs, and sleep hygiene. therapists Two registered and experienced clinical psychologists (CL and JBZ) with doctoral qualifications provided all clinical contact with participants via weekly telephone calls or a secure email-type private messaging system. Both therapists were trained in the administration of MINI diagnostic assessments and in the treatment of depression. Two senior therapists, authors BFD and NT, provided scheduled weekly supervision sessions during which all cases were reviewed. Supervision was provided at other times as required. The therapist
198
titov et al.
attempted to contact participants each week during the course and tried to limit contact to 10 minutes, although more time could be used if clinically indicated. A research assistant provided administrative support to collate data and provide technical support. All contact with participants was recorded as was the total therapist time spent per participant. Therapists were not blinded to group status when administering the MINI diagnostic assessments either at assessment or at the 3-month follow-up. No audio recordings of therapist contact with participants were made.
statistical analyses All analyses were conducted using SPSS Version 22. Participants who did not start treatment were not included in analyses. Group differences in demographic data and pretreatment measures were analysed with one-way analyses of variance (ANOVAs) or chi-square tests (shown in Table 1). Mixed-models analyses employing an autoregressive covariance structure, maximum likelihood estimation and pretreatment scores as a covariate, were used to analyze changes on the symptom measures. These mixed models analyses were run separately for the overall sample and for the clinical subsample, with the latter comprising those participants scoring ≥ 10 on the PHQ-9 at pretreatment. The mixed-models approach is consistent with intention-to-treat analytic approaches under the assumption that data are missing at random. Effect sizes (Cohen’s d) and 95% confidence intervals were calculated for both within-group and between-group effects based on the observed means. The following criteria of clinical significance were used. First, reliable improvement was calculated using standard procedures for determining reliable change (Jacobson & Truax, 1991), which take into account the reliability and measurement error associated with the questionnaire being employed. Specifically, a participant was deemed to have made a reliable improvement if they scored above the total cutoff at pretreatment (i.e., ≥ 8 on the GAD-7 or ≥ 10 on the PHQ-9) and their symptoms improved by a reliable amount; that is, more than 5.20 or 3.53 on the PHQ-9 and the GAD-7, respectively (Gyani, Shafran, Layard, & Clark, 2013). Second, reliable recovery was determined to have occurred if a participant scored above the clinical cutoff at pretreatment, made a reliable improvement, and scored below the clinical cutoff at the posttreatment or follow-up time point of interest (Gyani et al., 2013). Importantly, these analyses were calculated separately for the PHQ-9 the GAD-7 and baseline observations were carried forward in these clinical significance analyses as a
conservative approach for controlling for missing data.
economic analysis The economic analysis was conducted from the perspective of the national health provider in Australia. The key outcome of the economic analysis is the incremental cost-effectiveness ratio (ICER), which is a ratio of the difference in costs (incremental costs) and difference in QALYs (incremental QALYs) between the treatment and control groups. Data on health-related resource use and health-related quality of life (HRQoL) were collected at baseline and Week 8 (i.e., end of treatment) and a resource use questionnaire recorded health-care visits over the last 8 weeks. Individual patient-level costs were calculated as the product of health-care resource use and relevant unit costs for: (a) primary and secondary health care consultations and admissions, (b) use of antidepressant and antianxiolytic medications, and (c) resource use associated with the iCBT treatment provided. The resource use associated with the iCBT treatment included the cost of the therapist’s time required, supervisor’s time, internet access as well as computer and telephone use, but did not include the cost of creating the iCBT intervention or the clinic’s software platform or website. Resource use was valued using costs from the Medicare Benefits Schedule (Australian Government Department of Health, 2013a) and the Pharmaceutical Benefits Scheme (Australian Government Department of Health, 2013b). The measure of health benefit used was the QALY. The health state descriptor measure was the EQ-5D-5L (Brazier, 2007; Herdman et al., 2011). EQ-5D scores were converted into utility values using an algorithm for the Australian population (Norman et al., 2013). Then QALYs were calculated using the time-weighted average of the utility scores (Glick, 2007). The incremental difference in costs and QALYs between groups was estimated using seemingly unrelated regression model, which takes account of the correlation between costs and QALYs. The analysis controlled for baseline covariates, including baseline utility, cost, and previous history of mental health treatment. The regression coefficient on the treatment variable in the cost and QALY equations represents the incremental difference in costs and QALYs, respectively. The ratio of these coefficients provided the ICER. The confidence interval around the ICER was estimated using 10,000 nonparametric bootstrap replicates, each representing a potential outcome of the economic analysis (Brazier, 2007). In order to evaluate cost-effectiveness of the treatment over a range of WTP thresholds, while taking account of uncertainty in incremental cost and QALY
199
internet cbt for older adults with depression estimates, we used the conventional cost-effectiveness acceptability curve (CEAC). The CEAC uses the bootstrap replicates to represent the probability of the intervention being cost-effective over a range of willingness to pay thresholds per QALY (Fenwick et al., 2004).
p = .001). Pairwise comparisons revealed the treatment group had significantly lower PHQ-9 and GAD-7 scores than the control group (p b .001) at posttreatment. The treatment group’s PHQ-9 and GAD-7 scores were significantly reduced at posttreatment, 3-month follow-up, and 12-month follow-up, compared to the pretreatment scores (p b .001). There were no significant differences from posttreatment to 3-month follow-up or from 3-month to 12-month follow-up on the PHQ-9 or the GAD-7 (ps N .05).
Results preliminary tests No differences were found between the treatment and control groups in demographic characteristics or in the proportions meeting diagnostic criteria (ps N .05). No differences were found between participants who did and did not complete posttreatment questionnaires (ps N .05). The numbers of treatment group participants meeting diagnostic criteria at pretreatment and at 3-month follow-up are shown in Table 2.
symptom outcomes for clinical sample Means and standard deviations for the clinical sample (PHQ-9 total scores ≥ 10) on the PHQ-9 and the GAD-7 are shown in Table 3. The mixedmodels analyses revealed a significant main effect for Time (F3,64 = 22.53, p b .001) and a significant Time × Group interaction, (F1,88 = 40.03, p b .001) on the PHQ-9. Similarly, a significant main effect for Time (F3, 47 = 14.39, p b .001) and a significant Time × Group interaction (F1, 75 =10.61, p = .002) were found on the GAD-7. Pairwise comparisons revealed the treatment group had significantly lower PHQ-9 and GAD-7 scores than the waitlist control group (p b .001) at posttreatment. The PHQ-9 and the GAD-7 scores of the treatment group at posttreatment, 3-month, and 12-month follow-up were all significantly lower than their pretreatment scores
symptom outcomes for overall sample Means and standard deviations for the PHQ-9 and the GAD-7 for the treatment and waitlist control groups are shown in Table 3. The mixed models analyses examining PHQ-9 scores revealed a significant effect for Time (F3, 90 = 17.35, p b .001) and Time × Group interaction (F1, 127 = 35.51, p b .001). Similarly, the mixed-models analyses examining GAD-7 scores revealed significant effects for Time (F3, 80 = 14.33, p b .001) and the Time × Group interaction (F1, 127 = 12.18,
Table 2
Frequency Data of Participants’ Diagnoses At time of application Treatment Group
MINI-Diagnosis MDE PAN/AG SP OCD PTSD GAD Number of diagnoses 0 1 2 3 4 5
3-Month Follow-upa
Control Group
Total
Treatment Group
n
%
n
%
n
%
n
%
16 4 5 2 2 10
59 14 18 7 7 37
19 1 4 0 3 7
79 4 16 4 12 29
35 5 9 3 5 17
68 9 17 5 9 33
7 2 2 1 1 7
25 7 7 3 3 25
7 9 7 1 2 1
25 33 25 3 7 3
4 13 2 3 2 0
16 54 8 12 8 0
11 22 9 4 4 1
21 43 17 7 7 2
17 5 3 0 1 1
62 18 11 0 3 3
Note. Baseline observations were carried forward for missing data. Abbreviations: MDE, Major Depressive Episode; PAN/AG, panic disorder/Agoraphobia; SP, social phobia; OCD, Obsessive Compulsive Disorder; PTSD, Post Traumatic Stress Disorder; GAD, Generalised Anxiety Disorder.
200
Table 3
Means, Standard Deviations and Effect Sizes (Cohen’s d) for the Observed and Estimated Marginal Means n
Observed Means (Completer Data) Pre
Post
Estimated Means (Mixed Models Data) 3-Month Follow-up
12-Month Pre Follow-up
Post
Effect sizes (Based On Observed Means) 3-Month Follow-up
12-Month Within Group Follow-up pre to post
Between Group Within Group Within Group post treatment pre to 3-month pre to 12-month
OVERALL SAMPLE PHQ-9 Treatment
11.21 4.29 (3.16) 5.55 (3.36) 5.57 (2.92) (3.41) 11.43 11.50 (3.14) – – (2.93)
1.63 (0.99 – 2.22) − 0.12 (− 0.67– 0.44)
2.08 (1.38 – 2.72)
1.25 (0.65 – 1.82) –
1.25 (0.65 – 1.82) –
GAD-7 Treatment
27 7.44 (5.15)
2.96 (2.87)
Control
25 8.28 (4.90)
7.64 (4.68)
7.57 (2.37) 7.82 (2.37)
3.47 (2.75) 3.43 (2.82) – –
1.07 1.22 (0.49 – 1.63) (0.61 – 1.79) 0.13 (− 0.42 – 0.69)
1.00 (0.42 – 1.55) –
1.12 (0.54 – 1.68) –
3.05 (3.44) 2.84 (2.63) – –
3.18 (2.57) 6.87 (2.54)
CLINICAL SAMPLE PHQ-9 (PHQ9 ≥ 10) Treatment 16 14.56 (4.39) Control 18 14.89 (3.04) GAD-7 (PHQ9 ≥ 10) Treatment 16 9.63 (5.54) Control 18 10.44 (3.83)
4.79 (2.32) 14.76 (4.26)
5.85 (4.52) –
5.64 (4.47) –
14.62 (2.86) 14.34 (.2.87)
5.16 (3.06) 14.09 (2.96)
6.57 (3.16) –
6.60 (3.36) –
2.91 2.79 (1.76 – 3.68) (1.85 – 3.81) 0.04 (− 0.62 – 0.69)
1.96 (1.08 – 2.75) –
2.02 (1.12 – 2.81) –
3.43 (3.34) 8.88 (4.51)
3.92 (3.92) –
3.45 (2.73) –
9.54 (2.58) 9.84 (2.59)
3.82 (2.76) 8.34 (2.66)
4.57 (2.87) –
4.10 (3.11) –
1.35 1.36 (0.55 – 2.08) (0.59 – 2.07) 0.37 (− 0.29 – 1.02)
1.19 (0.41 – 1.91) –
1.42 (0.61 – 2.15) –
Note. GAD-7; Generalised Anxiety Disorder 7-item. PHQ-9; Patient Health Questionnaire 9-item.
titov et al.
4.90 (4.05) 4.68 (4.47) 25 12.04 (5.42) 12.68 (5.48) – –
Control
27 11.04 (5.62) 3.96 (2.48)
201
internet cbt for older adults with depression (ps b .001). There were no significant changes in the treatment group from posttreatment to 3-month follow-up or from the 3-month to the 12-month follow-up (ps N .05), indicating that initial symptom reductions from pretreatment to posttreatment were maintained at follow-up.
reported that they were somewhat dissatisfied. No participants reported that they were very dissatisfied with the course. All responding participants (n = 20/20; 100%) indicated they would recommend the course to a friend and (n = 19/19; 100%) that the course was worth their time.
clinical contact Forty minutes of therapist time in total was required per participant for administrative purposes and for the administration of diagnostic interviews. The mean total therapist time per participant in the treatment group during treatment was 45.07 minutes (SD = 32.51). This time was spent sending and reading secure email messages and telephoning participants.
economic analysis After adjustment for baseline covariates in the regression analysis, the mean cost per participant in the treatment group was $52.0 more over the 8-week period (95% bias corrected CI: $-23.80 to $128.20) compared with the waitlist control group (Table 5). Moreover, after adjustment, individuals in the treatment group had, on average, slightly higher QALYs than individuals in the control group (difference in QALYs of 0.012; 95% CI: 0.004 to 0.020), but incurred higher costs than those participants allocated to the control group. The incremental cost-effectiveness ratio was $4,392 (95% bias-corrected CI: b 0 to $63,962) which means that, on average, the intervention will produce one additional QALY for additional cost of $4,392. Uncertainty in the cost-effectiveness estimates is represented in Figure S1. The cost-effectiveness plane shows that treatment is very likely to be more costly but more effective (with small probability of being more effective and less costly). The CEAC shows the probability of the treatment being costeffective at different WTP thresholds. There is 50% probability of treatment being cost-effective if the decision-maker is only willing to pay $4,392 per QALY (which is also the mean value of ICER) and
clinical significance The proportions of participants reporting reliable improvement and reliable recovery are shown in Table 4. Chi-square analyses indicated greater rates of reliable improvement and reliable recovery on the PHQ-9 and the GAD-7 in the treatment group compared to the control group (ps b .007). More than 65% of treatment group participants reported clinically reliable improvement and recovery in symptoms of depression at posttreatment. treatment satisfaction Of the 20 participants who provided feedback about the Course, 80% (n = 16) reported that they were satisfied or very satisfied with the course, 15% (n = 3) gave a rating of neutral and 5% (n = 1)
Table 4
Proportions Reporting Reliable Improvement and Reliable Recovery CLINICAL SAMPLE
n
PHQ-9 (PHQ-9 ≥ 10) Treatment 16 Control GAD-7 (GAD-7 ≥ 8) Treatment Control
17
16 17
Posttreatment
3- Month Follow-Up
12- Month Follow-Up
Reliable Improvement
χ2
Reliable Recovery
χ2
11 / 16 (68.7%) 1 / 17 (5.8%)
χ 2 = 14.81, p b .001
11 / 16 (68.7%) 0 / 17 (0%)
χ 2 = 18.29, p b .001
9 / 16 (56.2%) –
8 / 16 (50.0%) –
8 / 16 (50.0%) –
8 / 16 (50.0%) –
8 / 16 (50.0%) 3 / 17 (17.6%)
χ 2 = 8.76, p = .005
7 / 16 (43.7%) 2 / 17 (11.7%)
χ 2 = 8.36, p = .007
6 / 16 (37.5%) –
4/ 16 (25.0%) –
5 / 16 (31.2%) –
4/ 16 (25.0%) –
Reliable Reliable Reliable Reliable Improvement Recovery Improvement Recovery
Note. Diagnostic data was not available for one control group participant. Baseline observations were carried forward for missing data. A person was deemed to have made a reliable improvement if they scored above the total cut-off at pretreatment and their symptoms improved by a reliable amount. A person was deemed to have reliably recovered if they scored above the clinical cut-off at pretreatment, made a Reliable Improvement, and scored below the clinical cut-off at the posttreatment or follow-up time point of interest. GAD-7; Generalised Anxiety Disorder 7-item. PHQ-9; Patient Health Questionnaire 9-item.
202
titov et al.
Table 5
Adjusted Costs and QALYs for the Treatment and Control Groups at Posttreatment
Treatment Control Difference
Mean costs ($)
95% Bias Corrected Confidence Interval ($)
Mean QALYs
95% Bias Corrected Confidence Interval
198.6 146.6 52.0
155.5 - 292.3 80.8 - 250.6 − 23.8 - 128.2
0.114 0.102 0.012
0.002 - 0.150 − 0.009 - 0.140 0.004 - 0.020
Note. Adjusted for baseline utility, baseline cost (in the cost equation) and history of mental health problems. QALYs; quality-adjusted life-years. 1AUD = 0.937 USD.
it increases as decision maker’s WTP increases. The CEAC shows that treatment is very likely to be cost-effective at the commonly used willingness-topay threshold of $50,000/QALY in Australia (probability of being cost-effective is N 95%) (Fenwick et al., 2004; McCabe, Claxton, & Culyer, 2008; Figure S1).
Discussion This RCT examined a new iCBT intervention, the Managing Your Mood Course, for older adults experiencing symptoms of depression. It was hypothesized that (a) the treatment group would report significantly reduced symptoms of depression at posttreatment compared with the waitlist control group, (b) symptom reductions of the treatment group would be sustained at 3-month and 12-month follow-up, and (c) the iCBT treatment would be cost-effective based on QALYs and commonly employed willingness-to-pay thresholds. These hypotheses were supported. The treatment group reported lower symptoms of depression at posttreatment and these improvements were maintained at 3-month and 12-month follow-up. Moreover, the treatment group reported slightly higher QALYs and costs at posttreatment and the cost of the incremental cost-effectiveness ratio showed that the intervention is highly likely to be cost-effective at the commonly used willingness-to-pay threshold of $50,000/QALY in Australia. The findings of the present RCT are consistent with and extend the findings of a feasibility open trial which examined iCBT for older adults with low mood and depression (Dear et al., 2013). The present findings also extend the results of an earlier RCT (n = 301), which targeted a younger cohort of older adults (N 50 years of age) with subclinical symptoms of depression (Spek et al., 2007). These results also support the finding that psychological treatments including CBT are effective in older adults (Ayers et al., 2007; Gould et al., 2012; Nordhus & Pallesen, 2003; Scogin et al., 2005) and extend the positive results of therapist-guided iCBT for depression (Andersson & Cuijpers, 2009; Andrews et al., 2010) to older adults with depression. Thus, the present results support the potential of iCBT as an approach
for the dissemination of evidence-based psychological treatments to older adults with depression. Several findings from the present study deserve brief comment. First, consistent with the broader iCBT literature, relatively little therapist time (M = 45.07 minutes; SD = 32.51 minutes) was required per person during treatment compared with traditional face-to-face treatments. This suggests that the treatment model reported here may represent a cost-effective model for dissemination of psychological interventions in health services. Second, while the present study initially sought to recruit only participants with clinical level symptoms of depression as measured by the primary outcome measure (i.e., total score ≥ 10 on the PHQ-9; Kroenke et al., 2001), this criterion was abandoned due to concerns about false negatives. In particular, it was noted that applicants who were not meeting the cutoff were still describing significant and distressing symptoms of depression, raising questions about the relevant cutoffs for the PHQ-9 in older adults. However, while the removal of this cutoff criterion means some caution is required when interpreting the present results, it is notable that large clinical improvements were also found in the subgroup analyses that included only participants who met the initial cutoff criterion. Finally, while the intervention targeted symptoms of depression, significant clinical improvements were also observed on the measure of anxiety (Cohen’s d N 1.0), indicating that the benefits generalized to other symptom domains. This raises the possibility of modifying the existing intervention in order to intentionally create a transdiagnostic intervention for older adults that targets common and underlying symptoms of depressive and anxiety disorders (Barlow, Allen, & Choate, 2004; Titov, Dear, Johnston, & Terides, 2012). Some limitations of the current RCT are important to acknowledge. First, the use of a waitlist control group limits the conclusions that can be made about the exact cause of the observed treatment effect and the long-term differences in clinical outcomes. Second, although the present study provides encouraging support regarding iCBT for older adults with symptoms of depression, the participants in the trial were a younger cohort of older adults (mean age =
internet cbt for older adults with depression 65.31 years; SD = 3.30 years; range 61 to 76 years). It is therefore unclear how well the results of the present study will generalize to older cohorts. Third, the structured diagnostic interviewers were not blinded, which may have unintentionally biased the diagnostic process. Fourth, while there is evidence for their comparability (Hedman et al., 2010; Rohde, Lewinsohn, & Seeley, 1997), it is important to note that all questionnaires were administered via the internet and all diagnostic interviews were conducted via telephone rather than face-to-face. Fifth, although supportive, the cost-effectiveness analyses conducted were preliminary in nature and require replication with larger samples and longer follow-up periods. Finally, further studies involving a broader range and larger number of older adults are also needed in order to determine the characteristics of older adults who may or may not benefit from the iCBT approach. This is essential if iCBT is to be offered as a part of mental health services for older adults. In conclusion, large clinical improvements in symptoms of both depression and anxiety were observed following treatment with the Managing Your Mood iCBT intervention and were sustained at 3-month and 12-month follow-up. The treatment was associated with slightly higher QALYs and the preliminary cost-effectiveness analyses were broadly supportive of the iCBT intervention. Thus, the present study supports the potential utility of iCBT as an innovative and cost-effective approach to increasing access to treatment for older adults with symptoms of depression. Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.beth.2014.09. 008. Conflict of Interest Statement N. Titov and B. Dear are authors and developers of the Managing Your Mood Course, but derive no personal or financial benefit from it. N. Titov, B. Dear, and B. Klein are funded by the Australian Government to develop and provide a free national internet and telephone-delivered treatment service, the MindSpot Clinic (www.mindspot.org.au), for people with anxiety and depression.
References Andersson, G., & Cuijpers, P. (2009). Internet-based and other computerized psychological treatments for adult depression: A meta-analysis. Cognitive Behaviour Therapy, 38, 196–205. Andrews, G., Cuijpers, P., Craske, M. G., McEvoy, P., & Titov, N. (2010). Computer therapy for the anxiety and depressive disorders is effective, acceptable and practical health care: A meta-analysis. PLoS One, 5, e13196. Australian Government Department of Health. (2003a). Medicare benefits schedule book. ISBN: 978-1-74241-825-4. Australian Government Department of Health. (2003b). Pharmaceutical benefits schedule. Updated December 1, 2013. Retrieved December 3, 2013, from http://www.pbs. gov.au/pbs/home
203
Ayers, C. R., Sorrell, J. T., Thorp, S. R., & Wetherell, J. L. (2007). Evidence-based psychological treatments for late-life anxiety. Psychology and Aging, 22, 8–17. Barlow, D., Allen, L., & Choate, M. (2004). Toward a unified treatment for emotional disorders. Behavior Therapy, 35, 205–230. Beekman, A. T., Geerlings, S. W., Deeg, D. J., Smit, J. H., Schoevers, R. S., de Beurs, E., . . . van Tilburg, W. (2002). The natural history of late-life depression: A 6-year prospective study in the community. Archives of General Psychiatry, 59, 605–611. Brazier, J. (2007). Measuring and valuing health benefits for economic evaluation. New York: Oxford University Press. Brenes, G. A., Penninx, B. W., Judd, P. H., Rockwell, E., Sewell, D. D., & Wetherell, J. L. (2008). Anxiety, depression and disability across the lifespan. Aging and Mental Health, 12, 158–163. Byers, A. L., Yaffe, K., Covinsky, K. E., Friedman, M. B., & Bruce, M. L. (2010). High occurrence of mood and anxiety disorders among older adults: The National Comorbidity Survey Replication. Archives of General Psychiatry, 67, 489–496. Crabb, R. M., Cavanagh, K., Proudfoot, J., Learmonth, D., Rafie, S., & Weingradt, K. R. (2012). Is computerized cognitive-behavioural therapy a treatment option for depression in late-life? A systematic review. British Journal of Clinical Psychology, 51, 459–464. Cockayne, N. L., Glozier, N., Naismith, S. L., Christensen, H., Neal, B., & Hickie, I. B. (2011). Internet-based treatment for older adults with depression and co-morbid cardiovascular disease: Protocol for a randomised, double-blind, placebo controlled trial. BMC Psychiatry, 11, 10. Cuijpers, P., Donker, T., van Straten, A., Li, J., & Andersson, G. (2010). Is guided self-help as effective as face-to-face psychotherapy for depression and anxiety disorders? A systematic review and meta-analysis of comparative outcome studies. Psychological Medicine, 40, 1943–1957. Cuijpers, P., Marks, I. M., van Straten, A., Cavanagh, K., Gega, L., & Andersson, G. (2009). Computer-aided psychotherapy for anxiety disorders: A meta-analytic review. Cognitive Behaviour Therapy, 38, 66–82. Dear, B. F., Titov, N., Schwencke, G., Andrews, G., Johnston, L., Craske, M. G., & McEvoy, P. (2011). An open trial of a brief transdiagnostic internet treatment for anxiety and depression. Behaviour Research and Therapy, 49, 830–837. Dear, B. F., Titov, N., Sunderland, M., McMillan, D., Anderson, T., Lorian, C., & Robinson, E. (2011). Psychometric comparison of the Generalized Anxiety Disorder Scale-7 and the Penn State Worry Questionnaire for measuring response during treatment of generalised anxiety disorder. Cognitive Behaviour Therapy, 40, 216–227. Dear, B. F., Zou, J., Titov, N., Lorian, C., Johnston, L., Spence, J., . . . Knight, R. G. (2013). Internet-delivered cognitive behavioural therapy for depression: a feasibility open trial for older adults. Australian and New Zealand Journal of Psychiatry, 47, 169–176. Dear, B. F., Zou, J., Ali, S., Lorian, C., Johnston, L., Terides, M. D., … Titov, N. (submitted). Examining self-guided internet-delivered cognitive behaviour therapy for older adults with anxiety and depression: Two feasibility open trials. Manuscript submitted for publication. Ewing, S., & Thomas, J. (2012). CCi Digital Futures 2012: The Internet in Australia. Kelvin Grove, QLD: ARC Centre of Excellence for Creative Industries and Innovation. Fenwick, E., O'Brien, B., & Briggs, A. (2004). Cost-effectiveness acceptability curves— Facts, fallacies and frequently asked questions. Health Economics, 13, 11.
204
titov et al.
Glick, H. (2007). Economic evaluation in clinical trials. Oxford: Oxford University Press. Gould, R. L., Coulson, M. C., & Howard, R. J. (2012). Efficacy of cognitive behavioral therapy for anxiety disorders in older people: A meta-analysis and meta-regression of randomized controlled trials. Journal of the American Geriatrics Society, 60, 218–229. Grek, A. (2007). Clinical management of suicidality in the elderly: An opportunity for involvement in the lives of older patients. Canadian Journal of Psychiatry, 52(Suppl. 1), 47S–57S. Gum, A. M., King-Kallimanis, B., & Kohn, R. (2009). Prevalence of mood, anxiety, and substance-abuse disorders for older Americans in the National Comorbidity Survey-Replication. American Journal of Geriatric Psychiatry, 17, 769–781. Gyani, A., Shafran, R., Layard, R., & Clark, D. M. (2013). Enhancing recovery rates: Lessons from one year of IAPT. Behavior Research and Therapy, 51, 597–606. Hedman, E., Ljotsson, B., & Lindfors, N. (2012). Cognitive behaviour therapy via the internet: A systematic review of applications, clinical efficacy and cost-effectiveness. Expert Review of Pharmacoeconics and Outcomes Research, 12, 745–764. Hedman, E., Ljotsson, B., Ruck, C., Furmark, T., Carlbring, P., Lindefors, N., & Andersson, G. (2010). Internet administration of self-report measures commonly used in research on social anxiety disorder: A psychometric evaluation. Computers in Human Behaviour, 26, 736–740. Herdman, M., Gudex, C., Lloyd, A., Janssen, M., Kind, P., Parkin, D., Bonsel, G., & Badia, X. (2011). Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Quality of Life Research, 20, 1727–1736. Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach to defining meaningful change in psychotherapy research. Psychology, 59, 12–19. Janssen, M. F., Pickard, A.,S., Golicki, D., Gudex, C., Niewada, M., Scalone, L., Swinburn, P., & Busschbach, J. (2013). Measurement properties of the EQ-5D-5L compared with the EQ-5D-3L across eight patient groups: A multi-country study. Quality of Life Research, 22, 1717–1727. Kiropoulos, L., Klein, B., Austin, D. W., Gilson, K., Pier, C., Mitchell, J., & Ciechomski, L. (2008). Is internet-based CBT for panic disorder and agoraphobia as effective as face-toface? Journal of Anxiety Disorders, 22, 1273–1284. Kroenke, K., Spitzer, R. L., & Williams, J. B. (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16, 606–613. Kroenke, K., Spitzer, R. L., Williams, J. B., & Löwe, B. (2010). The Patient Health Questionnaire Somatic, Anxiety, and Depressive Symptom Scales: A systematic review. General Hospital Psychiatry, 32, 345–359. Kroenke, K., Spitzer, R. L., Williams, J. B., Monahan, P. O., & Löwe, B. (2007). Anxiety disorders in primary care: Prevalence, impairment, comorbidity, and detection. Annals of Internal Medicine, 146(5), 317–325. Löwe, B., Decker, O., Müller, S., Brähler, E., Schellberg, D., Herzog, W., & Herzberg, P. Y. (2008). Validation and standardization of the Generalized Anxiety Disorder Screener (GAD-7) in the general population. Medical Care, 46, 266–274. McCabe, C., Claxton, K., & Culyer, A. J. (2008). The NICE Cost-Effectiveness Threshold: What it is and what that means. PharmacoEconomics, 26(9), 12. Nordhus, I. H., & Pallesen, S. (2003). Psychological treatment of late-life anxiety: An empirical review. Journal of Consulting and Clinical Psychology, 71, 643–651.
Norman, R., Cronin, P. A., & Viney, R. C. (2013). A pilot discrete choice experiment to explore preferences for EQ-5D-5L health states. Applied Health Economics and Health Policy, 11, 287–298. Phelan, E., Williams, B., Meeker, K., Bonn, K., Frederick, J., Logerfo, J., & Snowden, M. (2010). A study of the diagnostic accuracy of the PHQ-9 in primary care elderly. BMC Family Practice, 11, 63. Pirkis, J., Pfaff, J., Williamson, M., Tyson, O., Stocks, N., Goldney, R., . . . Almeida, O. P. (2009). The community prevalence of depression in older Australians. Journal of Affective Disorders, 115, 54–61. Rohde, P., Lewinsohn, P. M., & Seeley, J. R. (1997). Comparability of telephone and face-to-face interviews in assessing axis I and II disorders. American Journal of Psychiatry, 154, 1593–1598. Scogin, F., Welsh, D., Hanson, A., Stump, J., & Coates, A. (2005). Evidence-based psychotherapies for depression in older adults. Clinical Psychology: Science and Practice, 12, 222–237. Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., . . . Dunbar, G. C. (1998). The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59 (Suppl. 20), 22–33. Spek, V., Nyklicek, I., Smits, N., Cuijpers, P., Riper, H., & Keyzer, J. (2007). Internet-based cognitive behaviour therapy for subthreshold depression in people over 50 years old: A randomiszed controlled trial. Psychological Medicine, 37, 1797–1806. Spitzer, R., Kroenke, K., Williams, J. B., & Löwe, B. (2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Archives of Internal Medicine, 166, 1092–1097. Titov, N. (2011). Internet-delivered psychotherapy for depression in adults. Current Opinion in Psychiatry, 24, 18–23. Titov, N., Dear, B. F., Johnston, L., Lorian, C., Zou, J., Wootton, B., McEvoy, P. M., Rapee, R. M. (2013). Improving adherence and clinical outcomes in self-guided internet treatment for anxiety and depression: randomised controlled trial. PLoS One, 262873. Titov, N., Dear, B. F., Johnston, L., McEvoy, P. M., Wootton, B., Terides, M. D., . . . Rapee, R. M. (2014). Improving adherence and clinical outcomes in self-guided internet treatment for anxiety and depression: A 12 month follow-up of a randomised controlled trial. PLoS One, e89591. Titov, N., Dear, B. F., Johnston, L., & Terides, M. D. (2012). Transdiagnostic internet treatment for anxiety and depression. Spanish Journal of Clinical Psychology, 17, 237–260. Titov, N., Dear, B. F., McMillan, D., Anderson, T., Zou, J., & Sunderland, M. (2011). Psychometric comparison of the PHQ-9 and BDI-II for measuring response during treatment of depression. Cognitive Behaviour Therapy, 40, 126–136. Titov, N., Dear, B. F., Schwencke, G., Andews, G., Johnston, L., Craske, M. G., & McEvoy, P. (2011). Transdiagnostic treatment for anxiety and depression: A randomised Controlled trial. Behaviour Research and Therapy, 49, 331–452. Wetherell, J. L., Lenze, E. J., & Stanley, M. A. (2005). Evidence-based treatment of geriatric anxiety disorders. Psychiatric Clinics of North America, 28, 871–896. Wild, B., Eckl, A., Herzog, W., Niehoff, D., Lechner, S., Maatouk, I., … Löwe, B. (in press). Assessing generalized anxiety disorder in elderly people using the GAD-7 and
internet cbt for older adults with depression GAD-2 scales: Results of a validation study. American Journal of Geriatric Psychiatry. World Health Organization. (1990). Composite International Diagnostic Interview, Version 1.0. Geneva: World Health Organization. Zou, J. B., Dear, B. F., Titov, N., Lorian, C. N., Johnston, L., Spence, J., . . . Sachdev, P. (2012). Brief internet-delivered
205
cognitive behavioural therapy for anxiety in older adults: A feasibility trial. The Journal of Affective Disorders, 26, 650–655. R E C E I V E D : March 5, 2014 A C C E P T E D : September 16, 2014 Available online 22 October 2014
ScienceDirect Behavior Therapy 46 (2015) 193 – 205 www.elsevier.com/locate/bt
Clinical and Cost-Effectiveness of Therapist-Guided Internet-Delivered Cognitive Behavior Therapy for Older Adults With Symptoms of Depression: A Randomized Controlled Trial Nickolai Titov Blake F. Dear Macquarie University Shehzad Ali University of York Judy B. Zou Carolyn N. Lorian Luke Johnston Matthew D. Terides Rony Kayrouz Macquarie University Britt Klein Federation University Australia, The Australian National University, and Swinburne University of Technology Milena Gandy Vincent J. Fogliati Macquarie University
The authors gratefully acknowledge the participants for their involvement and helpful comments. This research was enabled by a National Priority Driven Research Program Grant from beyondblue. The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication. BFD is supported by a National Health and Medical Research Council (NHMRC) Australian Public Health Fellowship. BFD, MDT, and SA had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Address correspondence to Blake F. Dear, Ph.D., eCentreClinic, Department of Psychology, Macquarie University, New South Wales, Australia; e-mail: [email protected]. 0005-7894/© 2014 Association for Behavioral and Cognitive Therapies. Published by Elsevier Ltd. All rights reserved.
Depression is a common and significant health problem among older adults. Unfortunately, while effective psychological treatments exist, few older adults access treatment. The aim of the present randomized controlled trial (RCT) was to examine the efficacy, long-term outcomes, and cost-effectiveness of a therapist-guided internet-delivered cognitive behavior therapy (iCBT) intervention for Australian adults over 60 years of age with symptoms of depression. Participants were randomly allocated to either a treatment group (n = 29) or a delayed-treatment waitlist control group (n = 25). Twenty-seven treatment group participants started the iCBT treatment and 70% completed the treatment within the 8-week course, with 85% of participants providing data at posttreatment. Treatment comprised an online 5-lesson iCBT
194
titov et al.
course with brief weekly contact with a clinical psychologist, delivered over 8 weeks. The primary outcome measure was the Patient Health Questionnaire-9 Item (PHQ-9), a measure of symptoms and severity of depression. Significantly lower scores on the PHQ-9 (Cohen’s d = 2.08; 95% CI: 1.38 – 2.72) and on a measure of anxiety (Generalized Anxiety Disorder-7 Item) (Cohen’s d = 1.22; 95% CI: 0.61 – 1.79) were observed in the treatment group compared to the control group at posttreatment. The treatment group maintained these lower scores at the 3-month and 12-month follow-up time points and the iCBT treatment was rated as acceptable by participants. The treatment group had slightly higher Quality-Adjusted Life-Years (QALYs) than the control group at posttreatment (estimate: 0.012; 95% CI: 0.004 to 0.020) and, while being a higher cost (estimate $52.9 l 95% CI: −23.8 to 128.2), the intervention was cost-effective according to commonly used willingness-to-pay thresholds in Australia. The results support the potential efficacy and cost-effectiveness of therapist-guided iCBT as a treatment for older adults with symptoms of depression. Australian and New Zealand Clinical Trials Registry: ACTRN12611000927921; https://www.anzctr.org.au/ Trial/Registration/TrialReview.aspx?id=343384. Keywords: depression; older adults; internet; cognitive behavior therapy (CBT); cost-effectiveness
DEPRESSION IS A MAJOR HEALTH ISSUE among older adults. Research indicates that 8% to 10% of people over 60 years of age experience clinically significant symptoms (Pirkis et al., 2009) but that depression is often underdiagnosed in older adults (Byers, Yaffe, Covinsky, Friedman, & Bruce, 2010; Gum, KingKallimanis, & Kohn, 2009). This is significant because depression in older adults is associated with poorer physical health (Brenes et al., 2008; Cockayne et al., 2011) and an increased risk of suicide (Grek, 2007). Unfortunately, despite the availability of effective treatments (Ayers, Sorrell, Thorp, & Wetherell, 2007; Gould, Coulson, & Howard, 2012; Nordhus & Pallesen, 2003; Scogin, Welsh, Hanson, Stump, & Coates, 2005), research indicates that the number of older adults seeking and receiving evidence-based treatment is low (Beekman et al., 2002; Wetherell, Lenze & Stanley, 2005). There are numerous barriers to traditional face-to-face psychological treatments and the most significant to older adults include stigma, mobility limitations, costs of treatment, and the limited number of trained therapists. Thus, there is considerable need for treatment approaches that overcome barriers and increase access to evidencebased treatment for older adults. Internet-delivered cognitive behavior therapy (iCBT) is one approach that may reduce barriers to psychological treatments (Andersson & Cuijpers,
2009; Andrews, Cuijpers, Craske, McEvoy, & Titov, 2010; Cuijpers et al., 2009). iCBT treatments provide the same information and teach similar skills as traditional face-to-face CBT treatments, but do so via the internet using structured materials. iCBT may be administered as a self-guided intervention, or may include therapist support provided via email and telephone (Titov, 2011). There is now considerable meta-analytic data supporting the efficacy of iCBT in treating adults with anxiety and depression (Andersson & Cuijpers, 2009; Andrews et al., 2010; Cuijpers et al., 2009). Evidence is also emerging indicating that the results of iCBT are comparable to those obtained in traditional face-to-face CBT (Cuijpers Donker, van Straten, Li, & Andersson, 2010; Kiropoulos et al., 2008) and that iCBT is cost-effective (Hedman, Ljotsson, & Lindfors, 2012). Clinical trials of computer-delivered CBT and iCBT have, however, focussed on adults aged 20 to 60 years to date (Crabb et al., 2012). However, the promising results of this research and the increasing use of the internet by older adults in many countries, including Australia (Ewing & Thomas, 2012), has triggered targeted research focusing on the efficacy of iCBT with older adults (e.g., adults 60 years and older) (Dear et al., 2013; Spek et al., 2007; Zou et al., 2012). For example, a randomized controlled trial (RCT; n = 301) targeting adults over the age of 50 (M = 55; SD = 4.6) with subclinical symptoms of depression revealed an iCBT intervention was at least as effective as group CBT in reducing symptoms of depression, and that both treatments were superior to a waitlist control group (Spek et al., 2007). More recently, a feasibility open trial evaluated an iCBT intervention for adults over the age of 60 (M = 63.4; SD = 5.08) with depression and found a large within-group effect size (Cohen’s d N 1.0) (Dear et al., 2013). While the results of this open trial are encouraging, a controlled research trial is needed to explore the efficacy and cost-effectiveness of iCBT for older adults with symptoms of depression. The present study was designed to examine the efficacy, longer-term outcomes, and cost-effectiveness of a new iCBT intervention, the Managing Your Mood Course, for older adults with symptoms of depression, using an RCT design comparing a treatment and a waitlist control group. The Managing Your Mood Course is a five-lesson intervention delivered over 8 weeks with telephone and email support from a therapist. It was hypothesized that: (a) the treatment group would report significantly reduced symptoms of depression at posttreatment compared with the waitlist-control group, (b) symptom reductions of the treatment group would be sustained at 3-month and 12-month follow-up, and (c) the iCBT treatment would be cost-effective based
internet cbt for older adults with depression
FIGURE 1
195
Participant flow chart.
on Quality-Adjusted Life-Years (QALYs) and commonly employed willingness-to-pay thresholds.
Method participants A total of 111 individuals applied to participate in the study via the eCentreClinic website (www. ecentreclinic.org) and 54 met the criteria for inclusion into the study. The inclusion criteria were as follows:
(a) resident of Australia; (b) 60 years of age and over; (c) reported that they have been assessed by a general practitioner or medical specialist to rule out a reversible physical cause for their depression. An initial additional criterion that applicants had a total score ≥ 10 on the Patient Health Questionnaire-9 Item (PHQ-9; Kroenke, Spitzer, & Williams, 2001) was removed during the trial due to the significant number of applicants who did not meet this criterion,
196
titov et al.
Table 1
Demographic Characteristics of the Treatment and Control Groups Variable
Gender Male Female Age Mean Range Marital Status Single/Never Married Married/ De Facto Separated/Divorced/Widowed Education None High School Trade / Technical Certificate Diploma / Degree Employment Full-time Part-time / Student Unemployed, retired or disabled Previous Mental Health Treatment Taking Medication Hours Per Week Using Internet Mean Range Confidence using the Internet Mean (rating 1 to 5) Range
Treatment Group
Control Group
n
%
n
%
5 22
18.5% 81.5%
9 16
36% 64%
64.52 (SD: 2.58) 61 to 69
66.16 (SD: 3.80) 62 to 76
Significance Statistics
χ 2 (1) = 2.02, p = .16
F = .011, p = .93
1 14 12
3.7% 51.9% 44.4%
4 9 12
16.0% 36.0% 48.0%
χ 2 (2) = 2.81, p = .25
1 6 3 17
3.7% 22.2% 11.1% 63.0%
2 9 6 8
8.0% 36.0% 24.0% 32.0%
χ 2(3) = 5.10, p = .16
2 7 18 18 13
7.4% 25.9% 66.7% 66.7% 48.1%
0 4 21 13 12
0.0% 16.0% 84.0% 52.0% 48.0%
χ 2 (2) = 2.98, p = .23
χ 2(1) = 1.16, p = .28 χ 2 (1) = 0.00, p = .99
15.26 (14.98) 2 to 60
16.28 (11.96) 3 to 50
F = .076, p = .78
2.89 (1.05) 1 to 5
2.88 (1.42 1 to 5
F b .001, p = .98
but who still reported that they felt depressed. Thereafter, all applicants were required to verbally confirm during the telephone interview that they were experiencing difficulties with depression and that they wanted treatment for these symptoms. The exclusion criteria were as follows: (a) current participation in CBT; (b) use of illicit drugs or consumption of more than three standard drinks/day; (c) current diagnosis of schizophrenia or bipolar disorder; (d) severe symptoms of depression (defined as a total score N 19 or responding N 2 to Question 9 (suicidal ideation) on the PHQ-9; (e) if taking medication for anxiety or depression, not having been on a stable dose for at least a month. Details of participant flow are shown in Figure 1. Successful applicants were contacted by the researchers who confirmed the applicant met the inclusion criteria and administered a structured diagnostic interview, the Mini International Neuropsychiatric Interview Version 5 (MINI), via telephone. Group allocation preceded the telephone interview. A total of 29 participants were randomized to the treatment group and 25 were randomized to the waitlist control group. A permuted block randomiza-
tion sequence was generated by an independent researcher at another institution using a random number generator (www.randomizer.org) and each allocation was kept in a sealed envelope until allocation, which was opened after an applicant met the inclusion criteria. The demographic characteristics of the sample are shown in Table 1. The study was approved by the Human Research Ethics Committee (HREC) of Macquarie University, Sydney, Australia, and the trial was registered on the Australian and New Zealand Clinical Trials Registry (ANZCTR) as ACTRN12611000927921.
design and measures The design comprised a CONSORT-revised compliant RCT comparing a treatment group with a delayed-treatment waitlist control group. Questionnaires were administered to the treatment group immediately prior to the treatment, immediately after the treatment, and again 3 months and 12 months posttreatment. All questionnaires were administered online. The MINI diagnostic assessments were administered to the treatment group at application and again at 3 months posttreatment. The waitlist
internet cbt for older adults with depression control group commenced treatment immediately after the treatment group completed the treatment and were administered a self-guided version of the course, that is, without therapist contact, as a part of another trial (Dear et al., submitted). The waitlist control group were administered questionnaires and diagnostic assessments at the same time points as the treatment group.
measures Patient Health Questionnaire-9 Item (PHQ-9; Kroenke et al., 2001) The PHQ-9 comprises 9 items measuring symptoms of major depressive disorder based on the DSM-IV criteria for depression. A total score ≥ 10 has been identified as an important threshold for identifying DSM-IV congruent depression with increasing scores indicating greater symptom severity (Kroenke et al., 2001). The PHQ-9 has been found to have good psychometric properties generally (Kroenke, Spitzer, Williams, & Löwe, 2010; Titov, Dear, McMillan, et al., 2011) and in older adults (Phelan et al., 2010). Cronbach’s alpha for the PHQ-9 in the current study was .84. Generalized Anxiety Disorder 7-Item Scale (GAD-7; Spitzer, Kroenke, Williams, & Löwe, 2006) The GAD-7 comprises 7 items measuring symptoms of anxiety (Löwe et al., 2008). A total score ≥ 8 has been shown to be the optimum sensitive and specific threshold for the possible presence of an anxiety disorder (Kroenke, Spitzer, Williams, Monahan, & Löwe, 2007). The GAD-7 has been found to exhibit good psychometric properties (Dear, Titov, Sunderland, et al., 2011; Spitzer et al., 2006). However, one recent study has found a lower cutoff score of 5 may be required when used with older adults (Wild et al., in press). Cronbach’s alpha for the GAD-7 in the current study was .86. EuroQol - 5 Dimensions - 5 Levels (EQ-5D-5L; Herdman et al., 2011) The EQ-5D-5L is a standard and widely used measure of health-related quality of life (Brazier, 2007; Herdman et al., 2011). It comprises 5 items that inquire about five dimensions of health: mobility, self-care, usual activities, pain and discomfort, and anxiety and depression. Responses to the five dimensions are weighted and summed to create a total score between 0 (i.e., indicating deceased) and 1 (i.e., indicating full-health). The EQ-5D-5L has been shown to have good convergent validity with the World Health Organization 5-item Well Being Questionnaire (Janssen et al., 2013). EQ-5D scores were converted into utility values using an algorithm for the Australian population (Norman et al., 2013).
197
Then, QALYs were calculated using the timeweighted average of the utility scores (Glick, 2007). Mini International Neuropsychiatric Interview Version 5.0.0 (MINI; Sheehan et al., 1998) The MINI is a brief diagnostic interview developed to determine the presence of current Axis-I disorders using DSM-IV diagnostic criteria. It has excellent interrater reliability (k = 0.88 to 1.00) and adequate concurrent validity with the Composite International Diagnostic Interview (World Health Organization, 1990).
intervention The Managing Your Mood Course is a new 8-week treatment intervention based on the general structure and core psychological principles included in other evidence-based CBT interventions developed by the authors (Dear, Titov, Schwencke, et al., 2011; Dear et al., 2013; Titov, Dear, Schwencke, et al., 2011; Titov et al., 2013; Titov et al., 2014; Zou et al., 2012). The course consists of five online lessons, five lesson summaries and homework assignments, regular automated reminder and notification emails, secure internal email-type messaging facility with a therapist, and case stories detailing the experiences of older adults recovering from depression (see Table S1). Each lesson is presented in a slide show format that includes a combination of didactic material (i.e., text-based instructions and information) and case-enhanced learning material (i.e., educational stories which demonstrate the application of skills and problem resolution). To facilitate gradual learning and mastery of skills, an additional week is provided for the more complex lessons (i.e., thought challenging and graded exposure), and participants are prevented from skipping lessons or progressing too quickly through the materials. Participants are instructed to read the lessons over 8 weeks according to a timetable. Homework assignments are provided for each lesson. Additional resources are provided during the 8 weeks that include materials about assertiveness communication skills, problem solving, managing worry, challenging beliefs, and sleep hygiene. therapists Two registered and experienced clinical psychologists (CL and JBZ) with doctoral qualifications provided all clinical contact with participants via weekly telephone calls or a secure email-type private messaging system. Both therapists were trained in the administration of MINI diagnostic assessments and in the treatment of depression. Two senior therapists, authors BFD and NT, provided scheduled weekly supervision sessions during which all cases were reviewed. Supervision was provided at other times as required. The therapist
198
titov et al.
attempted to contact participants each week during the course and tried to limit contact to 10 minutes, although more time could be used if clinically indicated. A research assistant provided administrative support to collate data and provide technical support. All contact with participants was recorded as was the total therapist time spent per participant. Therapists were not blinded to group status when administering the MINI diagnostic assessments either at assessment or at the 3-month follow-up. No audio recordings of therapist contact with participants were made.
statistical analyses All analyses were conducted using SPSS Version 22. Participants who did not start treatment were not included in analyses. Group differences in demographic data and pretreatment measures were analysed with one-way analyses of variance (ANOVAs) or chi-square tests (shown in Table 1). Mixed-models analyses employing an autoregressive covariance structure, maximum likelihood estimation and pretreatment scores as a covariate, were used to analyze changes on the symptom measures. These mixed models analyses were run separately for the overall sample and for the clinical subsample, with the latter comprising those participants scoring ≥ 10 on the PHQ-9 at pretreatment. The mixed-models approach is consistent with intention-to-treat analytic approaches under the assumption that data are missing at random. Effect sizes (Cohen’s d) and 95% confidence intervals were calculated for both within-group and between-group effects based on the observed means. The following criteria of clinical significance were used. First, reliable improvement was calculated using standard procedures for determining reliable change (Jacobson & Truax, 1991), which take into account the reliability and measurement error associated with the questionnaire being employed. Specifically, a participant was deemed to have made a reliable improvement if they scored above the total cutoff at pretreatment (i.e., ≥ 8 on the GAD-7 or ≥ 10 on the PHQ-9) and their symptoms improved by a reliable amount; that is, more than 5.20 or 3.53 on the PHQ-9 and the GAD-7, respectively (Gyani, Shafran, Layard, & Clark, 2013). Second, reliable recovery was determined to have occurred if a participant scored above the clinical cutoff at pretreatment, made a reliable improvement, and scored below the clinical cutoff at the posttreatment or follow-up time point of interest (Gyani et al., 2013). Importantly, these analyses were calculated separately for the PHQ-9 the GAD-7 and baseline observations were carried forward in these clinical significance analyses as a
conservative approach for controlling for missing data.
economic analysis The economic analysis was conducted from the perspective of the national health provider in Australia. The key outcome of the economic analysis is the incremental cost-effectiveness ratio (ICER), which is a ratio of the difference in costs (incremental costs) and difference in QALYs (incremental QALYs) between the treatment and control groups. Data on health-related resource use and health-related quality of life (HRQoL) were collected at baseline and Week 8 (i.e., end of treatment) and a resource use questionnaire recorded health-care visits over the last 8 weeks. Individual patient-level costs were calculated as the product of health-care resource use and relevant unit costs for: (a) primary and secondary health care consultations and admissions, (b) use of antidepressant and antianxiolytic medications, and (c) resource use associated with the iCBT treatment provided. The resource use associated with the iCBT treatment included the cost of the therapist’s time required, supervisor’s time, internet access as well as computer and telephone use, but did not include the cost of creating the iCBT intervention or the clinic’s software platform or website. Resource use was valued using costs from the Medicare Benefits Schedule (Australian Government Department of Health, 2013a) and the Pharmaceutical Benefits Scheme (Australian Government Department of Health, 2013b). The measure of health benefit used was the QALY. The health state descriptor measure was the EQ-5D-5L (Brazier, 2007; Herdman et al., 2011). EQ-5D scores were converted into utility values using an algorithm for the Australian population (Norman et al., 2013). Then QALYs were calculated using the time-weighted average of the utility scores (Glick, 2007). The incremental difference in costs and QALYs between groups was estimated using seemingly unrelated regression model, which takes account of the correlation between costs and QALYs. The analysis controlled for baseline covariates, including baseline utility, cost, and previous history of mental health treatment. The regression coefficient on the treatment variable in the cost and QALY equations represents the incremental difference in costs and QALYs, respectively. The ratio of these coefficients provided the ICER. The confidence interval around the ICER was estimated using 10,000 nonparametric bootstrap replicates, each representing a potential outcome of the economic analysis (Brazier, 2007). In order to evaluate cost-effectiveness of the treatment over a range of WTP thresholds, while taking account of uncertainty in incremental cost and QALY
199
internet cbt for older adults with depression estimates, we used the conventional cost-effectiveness acceptability curve (CEAC). The CEAC uses the bootstrap replicates to represent the probability of the intervention being cost-effective over a range of willingness to pay thresholds per QALY (Fenwick et al., 2004).
p = .001). Pairwise comparisons revealed the treatment group had significantly lower PHQ-9 and GAD-7 scores than the control group (p b .001) at posttreatment. The treatment group’s PHQ-9 and GAD-7 scores were significantly reduced at posttreatment, 3-month follow-up, and 12-month follow-up, compared to the pretreatment scores (p b .001). There were no significant differences from posttreatment to 3-month follow-up or from 3-month to 12-month follow-up on the PHQ-9 or the GAD-7 (ps N .05).
Results preliminary tests No differences were found between the treatment and control groups in demographic characteristics or in the proportions meeting diagnostic criteria (ps N .05). No differences were found between participants who did and did not complete posttreatment questionnaires (ps N .05). The numbers of treatment group participants meeting diagnostic criteria at pretreatment and at 3-month follow-up are shown in Table 2.
symptom outcomes for clinical sample Means and standard deviations for the clinical sample (PHQ-9 total scores ≥ 10) on the PHQ-9 and the GAD-7 are shown in Table 3. The mixedmodels analyses revealed a significant main effect for Time (F3,64 = 22.53, p b .001) and a significant Time × Group interaction, (F1,88 = 40.03, p b .001) on the PHQ-9. Similarly, a significant main effect for Time (F3, 47 = 14.39, p b .001) and a significant Time × Group interaction (F1, 75 =10.61, p = .002) were found on the GAD-7. Pairwise comparisons revealed the treatment group had significantly lower PHQ-9 and GAD-7 scores than the waitlist control group (p b .001) at posttreatment. The PHQ-9 and the GAD-7 scores of the treatment group at posttreatment, 3-month, and 12-month follow-up were all significantly lower than their pretreatment scores
symptom outcomes for overall sample Means and standard deviations for the PHQ-9 and the GAD-7 for the treatment and waitlist control groups are shown in Table 3. The mixed models analyses examining PHQ-9 scores revealed a significant effect for Time (F3, 90 = 17.35, p b .001) and Time × Group interaction (F1, 127 = 35.51, p b .001). Similarly, the mixed-models analyses examining GAD-7 scores revealed significant effects for Time (F3, 80 = 14.33, p b .001) and the Time × Group interaction (F1, 127 = 12.18,
Table 2
Frequency Data of Participants’ Diagnoses At time of application Treatment Group
MINI-Diagnosis MDE PAN/AG SP OCD PTSD GAD Number of diagnoses 0 1 2 3 4 5
3-Month Follow-upa
Control Group
Total
Treatment Group
n
%
n
%
n
%
n
%
16 4 5 2 2 10
59 14 18 7 7 37
19 1 4 0 3 7
79 4 16 4 12 29
35 5 9 3 5 17
68 9 17 5 9 33
7 2 2 1 1 7
25 7 7 3 3 25
7 9 7 1 2 1
25 33 25 3 7 3
4 13 2 3 2 0
16 54 8 12 8 0
11 22 9 4 4 1
21 43 17 7 7 2
17 5 3 0 1 1
62 18 11 0 3 3
Note. Baseline observations were carried forward for missing data. Abbreviations: MDE, Major Depressive Episode; PAN/AG, panic disorder/Agoraphobia; SP, social phobia; OCD, Obsessive Compulsive Disorder; PTSD, Post Traumatic Stress Disorder; GAD, Generalised Anxiety Disorder.
200
Table 3
Means, Standard Deviations and Effect Sizes (Cohen’s d) for the Observed and Estimated Marginal Means n
Observed Means (Completer Data) Pre
Post
Estimated Means (Mixed Models Data) 3-Month Follow-up
12-Month Pre Follow-up
Post
Effect sizes (Based On Observed Means) 3-Month Follow-up
12-Month Within Group Follow-up pre to post
Between Group Within Group Within Group post treatment pre to 3-month pre to 12-month
OVERALL SAMPLE PHQ-9 Treatment
11.21 4.29 (3.16) 5.55 (3.36) 5.57 (2.92) (3.41) 11.43 11.50 (3.14) – – (2.93)
1.63 (0.99 – 2.22) − 0.12 (− 0.67– 0.44)
2.08 (1.38 – 2.72)
1.25 (0.65 – 1.82) –
1.25 (0.65 – 1.82) –
GAD-7 Treatment
27 7.44 (5.15)
2.96 (2.87)
Control
25 8.28 (4.90)
7.64 (4.68)
7.57 (2.37) 7.82 (2.37)
3.47 (2.75) 3.43 (2.82) – –
1.07 1.22 (0.49 – 1.63) (0.61 – 1.79) 0.13 (− 0.42 – 0.69)
1.00 (0.42 – 1.55) –
1.12 (0.54 – 1.68) –
3.05 (3.44) 2.84 (2.63) – –
3.18 (2.57) 6.87 (2.54)
CLINICAL SAMPLE PHQ-9 (PHQ9 ≥ 10) Treatment 16 14.56 (4.39) Control 18 14.89 (3.04) GAD-7 (PHQ9 ≥ 10) Treatment 16 9.63 (5.54) Control 18 10.44 (3.83)
4.79 (2.32) 14.76 (4.26)
5.85 (4.52) –
5.64 (4.47) –
14.62 (2.86) 14.34 (.2.87)
5.16 (3.06) 14.09 (2.96)
6.57 (3.16) –
6.60 (3.36) –
2.91 2.79 (1.76 – 3.68) (1.85 – 3.81) 0.04 (− 0.62 – 0.69)
1.96 (1.08 – 2.75) –
2.02 (1.12 – 2.81) –
3.43 (3.34) 8.88 (4.51)
3.92 (3.92) –
3.45 (2.73) –
9.54 (2.58) 9.84 (2.59)
3.82 (2.76) 8.34 (2.66)
4.57 (2.87) –
4.10 (3.11) –
1.35 1.36 (0.55 – 2.08) (0.59 – 2.07) 0.37 (− 0.29 – 1.02)
1.19 (0.41 – 1.91) –
1.42 (0.61 – 2.15) –
Note. GAD-7; Generalised Anxiety Disorder 7-item. PHQ-9; Patient Health Questionnaire 9-item.
titov et al.
4.90 (4.05) 4.68 (4.47) 25 12.04 (5.42) 12.68 (5.48) – –
Control
27 11.04 (5.62) 3.96 (2.48)
201
internet cbt for older adults with depression (ps b .001). There were no significant changes in the treatment group from posttreatment to 3-month follow-up or from the 3-month to the 12-month follow-up (ps N .05), indicating that initial symptom reductions from pretreatment to posttreatment were maintained at follow-up.
reported that they were somewhat dissatisfied. No participants reported that they were very dissatisfied with the course. All responding participants (n = 20/20; 100%) indicated they would recommend the course to a friend and (n = 19/19; 100%) that the course was worth their time.
clinical contact Forty minutes of therapist time in total was required per participant for administrative purposes and for the administration of diagnostic interviews. The mean total therapist time per participant in the treatment group during treatment was 45.07 minutes (SD = 32.51). This time was spent sending and reading secure email messages and telephoning participants.
economic analysis After adjustment for baseline covariates in the regression analysis, the mean cost per participant in the treatment group was $52.0 more over the 8-week period (95% bias corrected CI: $-23.80 to $128.20) compared with the waitlist control group (Table 5). Moreover, after adjustment, individuals in the treatment group had, on average, slightly higher QALYs than individuals in the control group (difference in QALYs of 0.012; 95% CI: 0.004 to 0.020), but incurred higher costs than those participants allocated to the control group. The incremental cost-effectiveness ratio was $4,392 (95% bias-corrected CI: b 0 to $63,962) which means that, on average, the intervention will produce one additional QALY for additional cost of $4,392. Uncertainty in the cost-effectiveness estimates is represented in Figure S1. The cost-effectiveness plane shows that treatment is very likely to be more costly but more effective (with small probability of being more effective and less costly). The CEAC shows the probability of the treatment being costeffective at different WTP thresholds. There is 50% probability of treatment being cost-effective if the decision-maker is only willing to pay $4,392 per QALY (which is also the mean value of ICER) and
clinical significance The proportions of participants reporting reliable improvement and reliable recovery are shown in Table 4. Chi-square analyses indicated greater rates of reliable improvement and reliable recovery on the PHQ-9 and the GAD-7 in the treatment group compared to the control group (ps b .007). More than 65% of treatment group participants reported clinically reliable improvement and recovery in symptoms of depression at posttreatment. treatment satisfaction Of the 20 participants who provided feedback about the Course, 80% (n = 16) reported that they were satisfied or very satisfied with the course, 15% (n = 3) gave a rating of neutral and 5% (n = 1)
Table 4
Proportions Reporting Reliable Improvement and Reliable Recovery CLINICAL SAMPLE
n
PHQ-9 (PHQ-9 ≥ 10) Treatment 16 Control GAD-7 (GAD-7 ≥ 8) Treatment Control
17
16 17
Posttreatment
3- Month Follow-Up
12- Month Follow-Up
Reliable Improvement
χ2
Reliable Recovery
χ2
11 / 16 (68.7%) 1 / 17 (5.8%)
χ 2 = 14.81, p b .001
11 / 16 (68.7%) 0 / 17 (0%)
χ 2 = 18.29, p b .001
9 / 16 (56.2%) –
8 / 16 (50.0%) –
8 / 16 (50.0%) –
8 / 16 (50.0%) –
8 / 16 (50.0%) 3 / 17 (17.6%)
χ 2 = 8.76, p = .005
7 / 16 (43.7%) 2 / 17 (11.7%)
χ 2 = 8.36, p = .007
6 / 16 (37.5%) –
4/ 16 (25.0%) –
5 / 16 (31.2%) –
4/ 16 (25.0%) –
Reliable Reliable Reliable Reliable Improvement Recovery Improvement Recovery
Note. Diagnostic data was not available for one control group participant. Baseline observations were carried forward for missing data. A person was deemed to have made a reliable improvement if they scored above the total cut-off at pretreatment and their symptoms improved by a reliable amount. A person was deemed to have reliably recovered if they scored above the clinical cut-off at pretreatment, made a Reliable Improvement, and scored below the clinical cut-off at the posttreatment or follow-up time point of interest. GAD-7; Generalised Anxiety Disorder 7-item. PHQ-9; Patient Health Questionnaire 9-item.
202
titov et al.
Table 5
Adjusted Costs and QALYs for the Treatment and Control Groups at Posttreatment
Treatment Control Difference
Mean costs ($)
95% Bias Corrected Confidence Interval ($)
Mean QALYs
95% Bias Corrected Confidence Interval
198.6 146.6 52.0
155.5 - 292.3 80.8 - 250.6 − 23.8 - 128.2
0.114 0.102 0.012
0.002 - 0.150 − 0.009 - 0.140 0.004 - 0.020
Note. Adjusted for baseline utility, baseline cost (in the cost equation) and history of mental health problems. QALYs; quality-adjusted life-years. 1AUD = 0.937 USD.
it increases as decision maker’s WTP increases. The CEAC shows that treatment is very likely to be cost-effective at the commonly used willingness-topay threshold of $50,000/QALY in Australia (probability of being cost-effective is N 95%) (Fenwick et al., 2004; McCabe, Claxton, & Culyer, 2008; Figure S1).
Discussion This RCT examined a new iCBT intervention, the Managing Your Mood Course, for older adults experiencing symptoms of depression. It was hypothesized that (a) the treatment group would report significantly reduced symptoms of depression at posttreatment compared with the waitlist control group, (b) symptom reductions of the treatment group would be sustained at 3-month and 12-month follow-up, and (c) the iCBT treatment would be cost-effective based on QALYs and commonly employed willingness-to-pay thresholds. These hypotheses were supported. The treatment group reported lower symptoms of depression at posttreatment and these improvements were maintained at 3-month and 12-month follow-up. Moreover, the treatment group reported slightly higher QALYs and costs at posttreatment and the cost of the incremental cost-effectiveness ratio showed that the intervention is highly likely to be cost-effective at the commonly used willingness-to-pay threshold of $50,000/QALY in Australia. The findings of the present RCT are consistent with and extend the findings of a feasibility open trial which examined iCBT for older adults with low mood and depression (Dear et al., 2013). The present findings also extend the results of an earlier RCT (n = 301), which targeted a younger cohort of older adults (N 50 years of age) with subclinical symptoms of depression (Spek et al., 2007). These results also support the finding that psychological treatments including CBT are effective in older adults (Ayers et al., 2007; Gould et al., 2012; Nordhus & Pallesen, 2003; Scogin et al., 2005) and extend the positive results of therapist-guided iCBT for depression (Andersson & Cuijpers, 2009; Andrews et al., 2010) to older adults with depression. Thus, the present results support the potential of iCBT as an approach
for the dissemination of evidence-based psychological treatments to older adults with depression. Several findings from the present study deserve brief comment. First, consistent with the broader iCBT literature, relatively little therapist time (M = 45.07 minutes; SD = 32.51 minutes) was required per person during treatment compared with traditional face-to-face treatments. This suggests that the treatment model reported here may represent a cost-effective model for dissemination of psychological interventions in health services. Second, while the present study initially sought to recruit only participants with clinical level symptoms of depression as measured by the primary outcome measure (i.e., total score ≥ 10 on the PHQ-9; Kroenke et al., 2001), this criterion was abandoned due to concerns about false negatives. In particular, it was noted that applicants who were not meeting the cutoff were still describing significant and distressing symptoms of depression, raising questions about the relevant cutoffs for the PHQ-9 in older adults. However, while the removal of this cutoff criterion means some caution is required when interpreting the present results, it is notable that large clinical improvements were also found in the subgroup analyses that included only participants who met the initial cutoff criterion. Finally, while the intervention targeted symptoms of depression, significant clinical improvements were also observed on the measure of anxiety (Cohen’s d N 1.0), indicating that the benefits generalized to other symptom domains. This raises the possibility of modifying the existing intervention in order to intentionally create a transdiagnostic intervention for older adults that targets common and underlying symptoms of depressive and anxiety disorders (Barlow, Allen, & Choate, 2004; Titov, Dear, Johnston, & Terides, 2012). Some limitations of the current RCT are important to acknowledge. First, the use of a waitlist control group limits the conclusions that can be made about the exact cause of the observed treatment effect and the long-term differences in clinical outcomes. Second, although the present study provides encouraging support regarding iCBT for older adults with symptoms of depression, the participants in the trial were a younger cohort of older adults (mean age =
internet cbt for older adults with depression 65.31 years; SD = 3.30 years; range 61 to 76 years). It is therefore unclear how well the results of the present study will generalize to older cohorts. Third, the structured diagnostic interviewers were not blinded, which may have unintentionally biased the diagnostic process. Fourth, while there is evidence for their comparability (Hedman et al., 2010; Rohde, Lewinsohn, & Seeley, 1997), it is important to note that all questionnaires were administered via the internet and all diagnostic interviews were conducted via telephone rather than face-to-face. Fifth, although supportive, the cost-effectiveness analyses conducted were preliminary in nature and require replication with larger samples and longer follow-up periods. Finally, further studies involving a broader range and larger number of older adults are also needed in order to determine the characteristics of older adults who may or may not benefit from the iCBT approach. This is essential if iCBT is to be offered as a part of mental health services for older adults. In conclusion, large clinical improvements in symptoms of both depression and anxiety were observed following treatment with the Managing Your Mood iCBT intervention and were sustained at 3-month and 12-month follow-up. The treatment was associated with slightly higher QALYs and the preliminary cost-effectiveness analyses were broadly supportive of the iCBT intervention. Thus, the present study supports the potential utility of iCBT as an innovative and cost-effective approach to increasing access to treatment for older adults with symptoms of depression. Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.beth.2014.09. 008. Conflict of Interest Statement N. Titov and B. Dear are authors and developers of the Managing Your Mood Course, but derive no personal or financial benefit from it. N. Titov, B. Dear, and B. Klein are funded by the Australian Government to develop and provide a free national internet and telephone-delivered treatment service, the MindSpot Clinic (www.mindspot.org.au), for people with anxiety and depression.
References Andersson, G., & Cuijpers, P. (2009). Internet-based and other computerized psychological treatments for adult depression: A meta-analysis. Cognitive Behaviour Therapy, 38, 196–205. Andrews, G., Cuijpers, P., Craske, M. G., McEvoy, P., & Titov, N. (2010). Computer therapy for the anxiety and depressive disorders is effective, acceptable and practical health care: A meta-analysis. PLoS One, 5, e13196. Australian Government Department of Health. (2003a). Medicare benefits schedule book. ISBN: 978-1-74241-825-4. Australian Government Department of Health. (2003b). Pharmaceutical benefits schedule. Updated December 1, 2013. Retrieved December 3, 2013, from http://www.pbs. gov.au/pbs/home
203
Ayers, C. R., Sorrell, J. T., Thorp, S. R., & Wetherell, J. L. (2007). Evidence-based psychological treatments for late-life anxiety. Psychology and Aging, 22, 8–17. Barlow, D., Allen, L., & Choate, M. (2004). Toward a unified treatment for emotional disorders. Behavior Therapy, 35, 205–230. Beekman, A. T., Geerlings, S. W., Deeg, D. J., Smit, J. H., Schoevers, R. S., de Beurs, E., . . . van Tilburg, W. (2002). The natural history of late-life depression: A 6-year prospective study in the community. Archives of General Psychiatry, 59, 605–611. Brazier, J. (2007). Measuring and valuing health benefits for economic evaluation. New York: Oxford University Press. Brenes, G. A., Penninx, B. W., Judd, P. H., Rockwell, E., Sewell, D. D., & Wetherell, J. L. (2008). Anxiety, depression and disability across the lifespan. Aging and Mental Health, 12, 158–163. Byers, A. L., Yaffe, K., Covinsky, K. E., Friedman, M. B., & Bruce, M. L. (2010). High occurrence of mood and anxiety disorders among older adults: The National Comorbidity Survey Replication. Archives of General Psychiatry, 67, 489–496. Crabb, R. M., Cavanagh, K., Proudfoot, J., Learmonth, D., Rafie, S., & Weingradt, K. R. (2012). Is computerized cognitive-behavioural therapy a treatment option for depression in late-life? A systematic review. British Journal of Clinical Psychology, 51, 459–464. Cockayne, N. L., Glozier, N., Naismith, S. L., Christensen, H., Neal, B., & Hickie, I. B. (2011). Internet-based treatment for older adults with depression and co-morbid cardiovascular disease: Protocol for a randomised, double-blind, placebo controlled trial. BMC Psychiatry, 11, 10. Cuijpers, P., Donker, T., van Straten, A., Li, J., & Andersson, G. (2010). Is guided self-help as effective as face-to-face psychotherapy for depression and anxiety disorders? A systematic review and meta-analysis of comparative outcome studies. Psychological Medicine, 40, 1943–1957. Cuijpers, P., Marks, I. M., van Straten, A., Cavanagh, K., Gega, L., & Andersson, G. (2009). Computer-aided psychotherapy for anxiety disorders: A meta-analytic review. Cognitive Behaviour Therapy, 38, 66–82. Dear, B. F., Titov, N., Schwencke, G., Andrews, G., Johnston, L., Craske, M. G., & McEvoy, P. (2011). An open trial of a brief transdiagnostic internet treatment for anxiety and depression. Behaviour Research and Therapy, 49, 830–837. Dear, B. F., Titov, N., Sunderland, M., McMillan, D., Anderson, T., Lorian, C., & Robinson, E. (2011). Psychometric comparison of the Generalized Anxiety Disorder Scale-7 and the Penn State Worry Questionnaire for measuring response during treatment of generalised anxiety disorder. Cognitive Behaviour Therapy, 40, 216–227. Dear, B. F., Zou, J., Titov, N., Lorian, C., Johnston, L., Spence, J., . . . Knight, R. G. (2013). Internet-delivered cognitive behavioural therapy for depression: a feasibility open trial for older adults. Australian and New Zealand Journal of Psychiatry, 47, 169–176. Dear, B. F., Zou, J., Ali, S., Lorian, C., Johnston, L., Terides, M. D., … Titov, N. (submitted). Examining self-guided internet-delivered cognitive behaviour therapy for older adults with anxiety and depression: Two feasibility open trials. Manuscript submitted for publication. Ewing, S., & Thomas, J. (2012). CCi Digital Futures 2012: The Internet in Australia. Kelvin Grove, QLD: ARC Centre of Excellence for Creative Industries and Innovation. Fenwick, E., O'Brien, B., & Briggs, A. (2004). Cost-effectiveness acceptability curves— Facts, fallacies and frequently asked questions. Health Economics, 13, 11.
204
titov et al.
Glick, H. (2007). Economic evaluation in clinical trials. Oxford: Oxford University Press. Gould, R. L., Coulson, M. C., & Howard, R. J. (2012). Efficacy of cognitive behavioral therapy for anxiety disorders in older people: A meta-analysis and meta-regression of randomized controlled trials. Journal of the American Geriatrics Society, 60, 218–229. Grek, A. (2007). Clinical management of suicidality in the elderly: An opportunity for involvement in the lives of older patients. Canadian Journal of Psychiatry, 52(Suppl. 1), 47S–57S. Gum, A. M., King-Kallimanis, B., & Kohn, R. (2009). Prevalence of mood, anxiety, and substance-abuse disorders for older Americans in the National Comorbidity Survey-Replication. American Journal of Geriatric Psychiatry, 17, 769–781. Gyani, A., Shafran, R., Layard, R., & Clark, D. M. (2013). Enhancing recovery rates: Lessons from one year of IAPT. Behavior Research and Therapy, 51, 597–606. Hedman, E., Ljotsson, B., & Lindfors, N. (2012). Cognitive behaviour therapy via the internet: A systematic review of applications, clinical efficacy and cost-effectiveness. Expert Review of Pharmacoeconics and Outcomes Research, 12, 745–764. Hedman, E., Ljotsson, B., Ruck, C., Furmark, T., Carlbring, P., Lindefors, N., & Andersson, G. (2010). Internet administration of self-report measures commonly used in research on social anxiety disorder: A psychometric evaluation. Computers in Human Behaviour, 26, 736–740. Herdman, M., Gudex, C., Lloyd, A., Janssen, M., Kind, P., Parkin, D., Bonsel, G., & Badia, X. (2011). Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Quality of Life Research, 20, 1727–1736. Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach to defining meaningful change in psychotherapy research. Psychology, 59, 12–19. Janssen, M. F., Pickard, A.,S., Golicki, D., Gudex, C., Niewada, M., Scalone, L., Swinburn, P., & Busschbach, J. (2013). Measurement properties of the EQ-5D-5L compared with the EQ-5D-3L across eight patient groups: A multi-country study. Quality of Life Research, 22, 1717–1727. Kiropoulos, L., Klein, B., Austin, D. W., Gilson, K., Pier, C., Mitchell, J., & Ciechomski, L. (2008). Is internet-based CBT for panic disorder and agoraphobia as effective as face-toface? Journal of Anxiety Disorders, 22, 1273–1284. Kroenke, K., Spitzer, R. L., & Williams, J. B. (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16, 606–613. Kroenke, K., Spitzer, R. L., Williams, J. B., & Löwe, B. (2010). The Patient Health Questionnaire Somatic, Anxiety, and Depressive Symptom Scales: A systematic review. General Hospital Psychiatry, 32, 345–359. Kroenke, K., Spitzer, R. L., Williams, J. B., Monahan, P. O., & Löwe, B. (2007). Anxiety disorders in primary care: Prevalence, impairment, comorbidity, and detection. Annals of Internal Medicine, 146(5), 317–325. Löwe, B., Decker, O., Müller, S., Brähler, E., Schellberg, D., Herzog, W., & Herzberg, P. Y. (2008). Validation and standardization of the Generalized Anxiety Disorder Screener (GAD-7) in the general population. Medical Care, 46, 266–274. McCabe, C., Claxton, K., & Culyer, A. J. (2008). The NICE Cost-Effectiveness Threshold: What it is and what that means. PharmacoEconomics, 26(9), 12. Nordhus, I. H., & Pallesen, S. (2003). Psychological treatment of late-life anxiety: An empirical review. Journal of Consulting and Clinical Psychology, 71, 643–651.
Norman, R., Cronin, P. A., & Viney, R. C. (2013). A pilot discrete choice experiment to explore preferences for EQ-5D-5L health states. Applied Health Economics and Health Policy, 11, 287–298. Phelan, E., Williams, B., Meeker, K., Bonn, K., Frederick, J., Logerfo, J., & Snowden, M. (2010). A study of the diagnostic accuracy of the PHQ-9 in primary care elderly. BMC Family Practice, 11, 63. Pirkis, J., Pfaff, J., Williamson, M., Tyson, O., Stocks, N., Goldney, R., . . . Almeida, O. P. (2009). The community prevalence of depression in older Australians. Journal of Affective Disorders, 115, 54–61. Rohde, P., Lewinsohn, P. M., & Seeley, J. R. (1997). Comparability of telephone and face-to-face interviews in assessing axis I and II disorders. American Journal of Psychiatry, 154, 1593–1598. Scogin, F., Welsh, D., Hanson, A., Stump, J., & Coates, A. (2005). Evidence-based psychotherapies for depression in older adults. Clinical Psychology: Science and Practice, 12, 222–237. Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., . . . Dunbar, G. C. (1998). The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59 (Suppl. 20), 22–33. Spek, V., Nyklicek, I., Smits, N., Cuijpers, P., Riper, H., & Keyzer, J. (2007). Internet-based cognitive behaviour therapy for subthreshold depression in people over 50 years old: A randomiszed controlled trial. Psychological Medicine, 37, 1797–1806. Spitzer, R., Kroenke, K., Williams, J. B., & Löwe, B. (2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Archives of Internal Medicine, 166, 1092–1097. Titov, N. (2011). Internet-delivered psychotherapy for depression in adults. Current Opinion in Psychiatry, 24, 18–23. Titov, N., Dear, B. F., Johnston, L., Lorian, C., Zou, J., Wootton, B., McEvoy, P. M., Rapee, R. M. (2013). Improving adherence and clinical outcomes in self-guided internet treatment for anxiety and depression: randomised controlled trial. PLoS One, 262873. Titov, N., Dear, B. F., Johnston, L., McEvoy, P. M., Wootton, B., Terides, M. D., . . . Rapee, R. M. (2014). Improving adherence and clinical outcomes in self-guided internet treatment for anxiety and depression: A 12 month follow-up of a randomised controlled trial. PLoS One, e89591. Titov, N., Dear, B. F., Johnston, L., & Terides, M. D. (2012). Transdiagnostic internet treatment for anxiety and depression. Spanish Journal of Clinical Psychology, 17, 237–260. Titov, N., Dear, B. F., McMillan, D., Anderson, T., Zou, J., & Sunderland, M. (2011). Psychometric comparison of the PHQ-9 and BDI-II for measuring response during treatment of depression. Cognitive Behaviour Therapy, 40, 126–136. Titov, N., Dear, B. F., Schwencke, G., Andews, G., Johnston, L., Craske, M. G., & McEvoy, P. (2011). Transdiagnostic treatment for anxiety and depression: A randomised Controlled trial. Behaviour Research and Therapy, 49, 331–452. Wetherell, J. L., Lenze, E. J., & Stanley, M. A. (2005). Evidence-based treatment of geriatric anxiety disorders. Psychiatric Clinics of North America, 28, 871–896. Wild, B., Eckl, A., Herzog, W., Niehoff, D., Lechner, S., Maatouk, I., … Löwe, B. (in press). Assessing generalized anxiety disorder in elderly people using the GAD-7 and
internet cbt for older adults with depression GAD-2 scales: Results of a validation study. American Journal of Geriatric Psychiatry. World Health Organization. (1990). Composite International Diagnostic Interview, Version 1.0. Geneva: World Health Organization. Zou, J. B., Dear, B. F., Titov, N., Lorian, C. N., Johnston, L., Spence, J., . . . Sachdev, P. (2012). Brief internet-delivered
205
cognitive behavioural therapy for anxiety in older adults: A feasibility trial. The Journal of Affective Disorders, 26, 650–655. R E C E I V E D : March 5, 2014 A C C E P T E D : September 16, 2014 Available online 22 October 2014