Cognitive Behavioral Therapy and Depression - TDMaddox

CBTandinsomniaArticle.pdf

Home >Psychology homework help >Cognitive Behavioral Therapy and Depression - TDMaddox

Regular Article

Cost-effectiveness of cognitive behavioral therapy for insomnia comorbid with depression: Analysis of a randomized controlled trial Norio Watanabe, MD, PhD,1,3 Toshiaki A. Furukawa, MD, PhD,5 Shinji Shimodera, MD, PhD,6

Fujika Katsuki, RN, PhD,4 Hirokazu Fujita, MD, PhD,6 Megumi Sasaki, PhD,7

Mitsuhiro Sado, MD, MSc2 and Michael L. Perlis, PhD8 1Department of Clinical Epidemiology, Translational Medical Center, National Center of Neurology and Psychiatry, 2Department of Neuropsychiatry, Keio University School of Medicine, Tokyo, 3Department of Psychiatry and Cognitive-Behavioral Medicine, Nagoya City University Graduate School of Medical Sciences, 4Department of Psychiatric and Mental Health Nursing, Nagoya City University School of Nursing, Aichi, 5Departments of Health Promotion and Human Behavior and of Clinical Epidemiology, Kyoto University Graduate School of Medicine/School of Public Health, Kyoto, 6Department of Neuropsychiatry, Kochi Medical School, Kochi University, Kochi, 7Health Care Center, Japan Advanced Institute of Science and Technology, Ishikawa, Japan, and 8Behavioral Sleep Medicine Program, Department of Psychiatry, University of Pennsylvania, Philadelphia, USA

Aim: Although the efficacy of cognitive behavioral therapy for insomnia has been confirmed, dissemi- nation depends on the balance of benefits and costs. This study aimed to examine the cost-effectiveness of cognitive behavioral therapy for insomnia consisting of four weekly individual sessions.

Methods: We conducted a 4-week randomized con- trolled trial with a 4-week follow up in outpatient clinics in Japan. Thirty-seven patients diagnosed as having major depressive disorder according to DSM-IV and suffering from chronic insomnia were randomized to receive either treatment as usual (TAU) alone or TAU plus cognitive behavioral therapy for insomnia. Effectiveness was evaluated as quality-adjusted life years (QALY) over 8 weeks’ time, estimated by bootstrapping of the observed total scores of the Hamilton Depression Rating Scale. Direct medical costs for cognitive behavioral therapy for insomnia and TAU were also evaluated. We cal- culated the incremental cost-effectiveness ratio.

Results: Over the 8 weeks of the study, the group receiving cognitive behavioral therapy for insomnia

plus TAU had significantly higher QALY (P = 0.002) than the TAU-alone group with an incremental value of 0.019 (SD 0.006), and had non-significantly higher costs with an incremental value of 254 (SD 203) USD in direct costs. The incremental cost- effectiveness ratio was 13 678 USD (95% confidence interval: −5691 to 71 316). Adding cognitive behav- ioral therapy for insomnia demonstrated an approxi- mately 95% chance of gaining one more QALY if a decision-maker was willing to pay 60 000 USD, and approximately 90% for 40 000 USD.

Conclusion: Adding cognitive behavioral therapy for insomnia is highly likely to be cost-effective for patients with residual insomnia and concomitant depression.

Key words: behavior therapy, cost–benefit analysis, depressive disorder, resource allocation, sleep initia- tion and maintenance disorders.

*Correspondence: Norio Watanabe, MD, PhD, Department of Clinical Epidemiology, Translational Medical Center, National Center of Neurology and Psychiatry, 4-1-1 Ogawa-higashi, Kodaira, Tokyo 187-8551, Japan. Email: [email protected] Trial registration: ClinicalTrials.gov Identifier: NCT00610259 http://www.clinicaltrials.gov Received 21 April 2014; revised 15 July 2014; accepted 3 September 2014.

Psychiatry and Clinical Neurosciences 2015; 69: 335–343 doi:10.1111/pcn.12237

mailto:[email protected]

http://www.clinicaltrials.gov/

THE PREVALENCE OF insomnia is very high witha prevalence rate of approximately 5–20% in the general population,1 which leads to increased use of health-care services and products as well as insomnia- related work absences and reduced productivity. The average annual per-person costs, including direct and indirect costs, have been estimated to be about 5000 USD for individuals with insomnia, and about 1400 USD for those presenting with insomnia symptoms.2

In addition, insomnia occurs comorbidly with many axis I disorders, especially for depression. The preva- lence rates of insomnia are estimated to be as high as 80–90% in untreated depressive patients.3,4

Pharmacological and psychological therapies for the treatment of insomnia have been developed and assessed for their efficacy. For psychotherapy, the effi- cacy of cognitive behavioral therapy for insomnia (CBT-I) has been well investigated. CBT-I is based on a multi-component approach that includes several modules,5 such as sleep hygiene education, sleep restriction and stimulus control as first-line interven- tions, and cognitive therapy, relaxation training and sleep compression as adjunctive ones.6 From results from several randomized controlled trials (RCT), the efficacy of CBT-I has been confirmed for primary insomnia7,8 as well as insomnia concomitant with depression.9,10

In practice, for the purpose of considering a strat- egy to disseminate an effective treatment to reduce the burden of insomnia, decisions about the value of insomnia treatment should be considered with the cost of the treatment. With regard to CBT for depres- sion, a large number of trials have examined the cost-effectiveness of the psychotherapy in various situations.11–13 However, to the best of our knowl- edge, the cost-effectiveness of CBT-I has been inves- tigated in only one cluster trial on hypnotic- dependent people with insomnia.14

In the present study, therefore, we aimed to examine the cost-effectiveness of CBT-I consisting of four weekly individual sessions in treatment for patients with residual insomnia and concomitant depression, by using the data from an efficacy trial of CBT-I we previously published.10

METHODS

Participants

Patients were recruited from 18 February 2008 to 9 April 2009 at three psychiatric outpatient depart-

ments in Japan. The patients were included in the trial if they were currently partially remitted and suf- fering from mild or moderate major depressive dis- order (diagnosed with the DSM-IV), despite having already been on maximum doses of two types of antidepressants for at least 4 weeks for the index episode. In addition to refractory depression, patients were only eligible for the present study if they pre- sented with chronic comorbid insomnia. Inclusion criteria included a score between 8 and 23 on the 17-item GRID-Hamilton Depression Rating Scale (HAMD)15 and a score of 8 or more on the Insomnia Severity Index (ISI).5,16,17 Psychotropic medications other than methylphenidate or modafinil, including antidepressants and hypnotics, and prescriptions for medical conditions were allowed and continued.

Study design and interventions

Participants were individually randomized to receive CBT-I plus treatment as usual (TAU) or TAU alone.

An independent statistician generated the random allocation sequences by the computer, using vari- able blocks and stratified by the severity of depres- sion and by study sites. Allocation sequences were kept centrally, and the allocation was provided by facsimile to each site upon notification of a patient’s enrolment.

The psychotherapy consisted of four weekly indi- vidual sessions, each lasting approximately 50 min, developed based on a published treatment manual for CBT-I.6 The treatment regimen was highly struc- tured, including modules of sleep hygiene education, introduction of the behavioral model of insomnia, sleep restriction, stimulus control, sleep titration, and relapse prevention.10 The regimen was provided to therapists in Japanese as a written manual. Therapists for the psychotherapy were five psychiatrists and a psychiatric nurse. They participated in a 2-day inten- sive training course on the psychotherapy before the study commencement, and received ongoing super- vision monthly thereafter. Patients allocated to the combination group were asked to self-administer these skills after the termination of the intervention sessions at 4 weeks until the final assessment at 8 weeks.

In TAU sessions, a patient met a physician (psy- chiatrist, none of the psychiatrists who conducted CBT-I) biweekly during which time they discussed their depression symptoms and insomnia and obtained medication. Each session typically lasted

336 N. Watanabe et al. Psychiatry and Clinical Neurosciences 2015; 69: 335–343

10 min. Changing types and doses of medication were not allowed in the first 4 weeks of the study unless rapid exacerbation of depression occurred. Physicians were allowed to discuss sleep hygiene but not sleep restriction or stimulus control for insomnia.

Assessment measures

Utility measures

Patients were assessed at baseline, at 4 (post- treatment) and at 8 weeks (1-month follow up). Patients who dropped out of the intervention were still asked to complete the assessments. Depression was assessed using the GRID-HAMD15 through a face- to-face semi-structured interview by blinded raters at each assessment.18 Patients were deemed as remitters for depression if their 17-item HAMD score was 7 or less, and as severely depressed if their HAMD score was 27 or more. This scale has been validated in Japan.19

In the primary analysis, the data from the HAMD were used to construct an outcome measure called depression-free days (DFD), as used in previous studies.20,21 In brief, if patients had an HAMD score of 27 or higher, they were assumed to be lacking a DFD; when scoring 7 or lower, they were assumed to have a full DFD; if they scored between 7 and 27, the day was weighted proportionally. To determine the number of DFD over the study period, the scores for the baseline and 4 weeks, and those for 4 weeks and 8 weeks were added and divided by 2, and then mul- tiplied by the number of days between assessments. The sums were totaled.

To compare the cost-effectiveness of the two arms, DFD were transformed into quality-adjusted days by using utility weights assigned to depression based on previous studies.20,21 The transformation assumes that a non-depressed person has a utility score of 1 (healthy) while a person meeting criteria for major depression has a utility score of 0.59. The utility scores were then transformed into quality-adjusted life years (QALY),22 which are a measure of disease burden, including both the quality and the quantity of life lived. The QALY is based on the number of years of life, and each year in perfect health is assigned the value of 1 down to a value of 0 for being deceased. As the duration of the present study was 8 weeks, a QALY score for a patient in the present study fell between 0.09 and 0.15.

To confirm the robustness of the results, we planned sensitivity analyses by means of changing

the calculation method of QALY to another one focusing on the categorized severity of depression.23

A QALY score was allocated to a patient according to the severity of depression validated and catego- rized in Japan based on the total score of the HAMD19 at each time-point of the assessments (i.e. 0.86 to remission, 0.74 to mild depression, 0.44 to moderate depression, and 0.30 to severe depression). The scores for the baseline and 4 weeks, and those for 4 weeks and 8 weeks were added and divided by 2, and then multiplied by the number of days between assessments. The sums were totaled.

Cost measures

Direct costs, which consisted of the costs for CBT-I, TAU for depression and insomnia, prescribed medi- cations, and hospital inpatient expense if any, were estimated in the present study. From a policy- makers’ points of view, indirect costs and educa- tional costs for therapists were not included in the cost measures. In the primary analysis (Approach 1), the cost for each session of the psychotherapy was presumed as the same as that of CBT for depres- sion (Table 1), because CBT-I has not been covered by the insurance system in Japan. Actual costs for medication and hospital inpatient expense, if any, were regrettably not recorded in the study. However, information about a dose transformed to a defined daily dose (DDD) for each class of antidepressants and hypnotics24 and duration of admission during the study was collected. The costs for these were imputed using costs for a representative drug (e.g. amitriptyline as a representative for tricyclic antide- pressants) in each class for medication in Japan, and an estimate from a survey using the national data for inpatient expense,25 respectively. As a result, tri- cyclic antidepressant per DDD was estimated at 0.286 USD, selective serotonin reuptake inhibitor at 1.900, serotonin and norepinephrine reuptake inhibitor at 1.245, other antidepressants at 0.947 and hypnotics at 0.810. Consultation fee at the out- patient clinic in the TAU group was estimated at 48.045 USD per session, defined by the national insurance system.

For sensitivity analyses, costs for each session of the psychotherapy were presumed according to different medical systems, and utilized to examine the robust- ness of the results from the primary analysis (Table 1).

Psychiatry and Clinical Neurosciences 2015; 69: 335–343 Cost-effectiveness of psychotherapy 337

Data management and analysis

The number of patients who should be included in the study was calculated prior to study commence- ment, based on a power analysis conducted for the ISI scores.10

For the primary analysis named as Approach 1 (Table 1), we included QALY calculated using DFD21

as the effectiveness, and costs for the psychotherapy presumed as the same as those for CBT for depres- sion, TAU, medication and inpatient expense. For the sensitivity analyses, QALY calculated using the sever- ity of depression23 were employed in Approaches 2 and 5. Costs of the psychotherapy presumed based on different medical systems were entered in Approaches 3, 4, 7 and 8. Hospital inpatient expenses, if any occurred, were considered as an important outcome but could be outliers and highly likely to bias the results to a large extent due to the small sample size in the present study. For sensitivity analyses, we planned to exclude participants who admitted to hospital during the study in Approaches 5, 6, 7 and 8.

Cost-effectiveness was evaluated by relating the dif- ferential cost per patient receiving either the psycho- therapy plus TAU or TAU alone to the differential effectiveness of each treatment in terms of QALY. The

incremental cost-effectiveness ratio (ICER) was calcu- lated as the difference in the cost divided by the difference in QALY.

Sampling uncertainty of the incremental cost per QALY was investigated by estimating a cost- effectiveness acceptability curve (CEAC). A CEAC can show the probability that an intervention is cost- effective compared with the alternative for a range of monetary values that a decision-maker may be willing to pay for a particular unit change in outcome.26 Both the ICER and the CEAC are derived from the joint distribution of incremental costs and incremental effectiveness, as estimated by 1000 non- parametric bootstrapping replicates of the observed data. The bootstrap re-sampling and inferential t-tests at a significant level of 0.05 for estimated costs and effectiveness between the groups were performed with Microsoft EXCEL 2010.

The protocol was approved by the Nagoya City University Institutional Review Board and the Ethical Review Board of Kochi Medical School. Atago Hospi- tal was approved to participate in the study by the latter review board, and the president of the hospital approved the protocol and study procedures. Thus, the study was approved by all the recruiting centers. With regard to participants, before enrolment in the study, interested individuals were contacted by study

Table 1 Presumed unit cost of CBT for insomnia and of inpatient care for estimating cost-effectiveness ratios

Approach

Number of patients included

Presumed location and situation to calculate unit cost of CBT for insomnia

Cost per session (USD)

Presumed hospital inpatient expense in Japan per day (USD)

1 (primary) 37 Japan, CBT for depression by a physician 43.210 135.103 2 37 Japan, CBT for depression by a physician 43.210 135.103 3 37 USA, typical face-to-face CBT for depression 100.000 135.103 4 37 Japan, 5 general short appointments at psychiatric

clinics (assuming that a physician meets 6 patients per hour)

169.753 135.103

5 35 Japan, CBT for depression by a physician 43.210 NA 6 35 Japan, CBT for depression by a physician 43.210 NA 7 35 USA, typical face-to-face CBT for depression 100.000 NA 8 35 Japan, 5 general short appointments at psychiatric

clinics (assuming that a physician meets 6 patients per hour)

169.753 NA

In Approaches 5, 6, 7 and 8, participants who admitted to hospital during the study were excluded for sensitivity analyses. Costs were calculated using an exchange rate on 25 June 2013: 1 USD equaled 97.2000 JPY, 0.6476 GBP, 0.7624 EUR, 1.0477 CAD, 1.0803 AUD and 0.9353 CHF. CBT, cognitive behavioral therapy; NA, not applicable.

338 N. Watanabe et al. Psychiatry and Clinical Neurosciences 2015; 69: 335–343

staff in a single room at hospital to maintain confi- dentiality. Study staff explained the purposes, proce- dures and potential risks or discomforts of the study to those who satisfied the eligibility criteria using the standard informed consent form, which emphasized that all potential participants who declined to partici- pate or otherwise did not participate were eligible for usual treatment and were not disadvantaged in any other way by not participating in the study. Following these instructions, written informed consent was obtained from all the participants included in the study.

RESULTS Thirty-seven patients satisfied the eligibility criteria, with 20 participants randomly assigned to receive combined treatment with CBT-I and TAU and 17 to receive TAU alone. Table 2 summarizes the clinical characteristics at baseline. During the study, antide- pressant dosage was changed for two participants each in the two groups. Hypnotic dosage was changed for two participants in the combination group and for none of the participants in the TAU- alone group. In the CBT-I-plus-TAU group, one patient discontinued the psychotherapy after report- ing that it was too difficult to comply with the pre- scribed sleep schedule. Beyond this reason, one subject in the intervention (received inpatient care for 25 days) and one in the control (for 18 days) groups were admitted to hospital due to exacerbation of depression. All the patients nevertheless completed all the study assessments at 8 weeks, thus no missing data were observed.

Figure 1 shows the incremental cost-effectiveness planes, which are scatter plots of the bootstrapped incremental costs and effectiveness pairs for all Approaches between the CBT-I-plus-TAU and the TAU-alone groups. A majority of plots were located in the upper right quadrant, which represented the posi- tion where adding CBT-I was more effective and more costly than TAU alone.

Effectiveness outcomes

QALY was statistically significantly higher (P = 0.002) in the CBT-I-plus-TAU group than in the TAU-alone group, at the incremental value of 0.019 (SD 0.006) for the primary analysis, Approach 1 (Table 3). In all the other analyses, QALY were also significantly higher in the intervention than those in the compara- tor. Incremental QALY appeared to be higher in Approaches 2 and 6, where those were estimated using the categorized severity of depression.

Cost outcomes

In the primary analysis (Approach 1), the total direct medical costs were not statistically significant (P = 0.404) but appeared to be higher in the CBT-I- plus-TAU group than in the TAU-alone group, at the incremental cost of 254 USD (SD 203) (Table 3). An incremental cost was the highest at 783 USD in Approach 4, where one session of CBT-I was pre- sumed for around 170 USD and inpatient expenses were included. Costs for the intervention group were significantly higher than the alternative, after

Table 2 Clinical characteristics of participants at baseline

Characteristic CBT-I + TAU (n = 20)

TAU alone (n = 17)

All patients (n = 37)

Age, mean (SD), years 52.9 (11.6) 47.8 (10.1) 50.5 (11.1) Sex, n (%)

Female 15 (75.0) 8 (47.1) 23 (62.2) Male 5 (25.0) 9 (52.9) 14 (37.8)

Duration of treatment for index episode, mean (SD), months 18.1 (11.1) 27.8 (46.5) 22.5 (32.4) Hamilton Depression Rating Scale, mean (SD) 15.0 (3.6) 16.8 (4.2) 15.8 (3.9) Insomnia Severity Index, mean (SD) 15.3 (4.7) 17.4 (3.3) 16.3 (4.2) Total antidepressant usage, mean (SD), DDD 1.7 (0.9) 1.5 (0.9) 1.6 (0.9) Hypnotic usage, mean (SD), DDD 0.7 (0.9) 1.1 (0.7) 0.9 (0.8)

CBT-I, cognitive behavioral therapy for insomnia; DDD, defined daily dose; TAU, treatment as usual.

Psychiatry and Clinical Neurosciences 2015; 69: 335–343 Cost-effectiveness of psychotherapy 339

1600

1400

1200

1000

800

600

400

200

−200−0.01 0.00

Approach 1 D

iff e re

n ce

in c

o st

( U

S D

)

0.01 0.02 0.03 0.04 0.05

−400

1600

1400

1200

1000

800

600

400

200

−200−0.01 0.00

Approach 3

D iff

e re

n ce

in c

o st

( U

S D

)

0.01 0.02 0.03 0.04 0.05

−400

1600

1400

1200

1000

800

600

400

200

−200−0.01 0.00

Approach 4

Difference in QALY

D iff

e re

n ce

in c

o st

( U

S D

)

0.01 0.02 0.03 0.04 0.05

−400

1600

1400

1200

1000

800

600

400

200

−200−0.01 0.00

Approach 2

D iff

e re

n ce

in c

o st

( U

S D

)

0.01 0.02 0.03 0.04 0.05

−400

1600

1400

1200

1000

800

600

400

200

−200−0.01 0.00

Approach 5

0.01 0.02 0.03 0.04 0.05

−400

1600

1400

1200

1000

800

600

400

200

−200−0.01 0.00

Approach 7

0.01 0.02 0.03 0.04 0.05

−400

1600

1400

1200

1000

800

600

400

200

−200−0.01 0.00

Approach 8

Difference in QALY

0.01 0.02 0.03 0.04 0.05

−400

1600

1400

1200

1000

800

600

400

200

−200−0.01 0.00

Approach 6

0.01 0.02 0.03 0.04 0.05

−400

Figure 1 Estimated incremental differences in costs and in effectiveness. Values were estimated from the trial data using 1000 bootstrap replicates. QALY, quality-adjusted life year.

340 N. Watanabe et al. Psychiatry and Clinical Neurosciences 2015; 69: 335–343

excluding participants who admitted to hospital during the study period (Approaches 5–8).

Cost-effectiveness outcomes

The ICER for the primary analysis (Approach 1) was 13 678 (95% confidence interval: −5691 to 71 316) USD/QALY (Table 3). Figures 2 and 3 show the CEAC of the primary (Approach 1) and the other

sensitivity analyses (Approaches 2–8), which repre- sent the probability for the addition of the psycho- therapy to be cost-effective for a range of potential maximum amounts that a decision-maker is willing to pay for one more QALY. For the primary analysis, adding CBT-I demonstrated an approximately 95% chance of being cost-effective when a decision- maker was willing to pay 60 000 USD for one more QALY, around 90% for 40 000 and around 70% for

Table 3 Costs, effectiveness and ICER, estimated using bootstrapping of the observed data

Approach

Effectiveness (QALY), mean (SD) Cost (USD), mean (SD) ICER (USD/QALY), mean (95%CI)CBT-I + TAU TAU Incremental CBT-I + TAU TAU Incremental

1 0.139 (0.004) 0.120 (0.004) 0.019 (0.006)* 702 (175) 448 (115) 254 (203) 13 678 (−5691 to 71 316) 2 0.108 (0.005) 0.084 (0.006) 0.024 (0.008)* 711 (170) 444 (108) 266 (201) 11 152 (−4108 to 65 443) 3 0.138 (0.004) 0.120 (0.005) 0.018 (0.006)* 937 (165) 435 (101) 501 (199) 27 252 (5220 to 128 572) 4 0.139 (0.004) 0.120 (0.005) 0.018 (0.006)* 1223 (181) 440 (109) 783 (213)* 42 929 (16 994 to 163 146) 5 0.141 (0.003) 0.120 (0.005) 0.020 (0.005)* 531 (17) 371 (32) 160 (36)* 7 828 (3527 to 18 613) 6 0.111 (0.004) 0.084 (0.006) 0.027 (0.007)* 531 (18) 372 (33) 159 (38)* 5 900 (2485 to 14 958) 7 0.141 (0.003) 0.121 (0.005) 0.020 (0.006)* 758 (18) 372 (33 387 (37)* 18 927 (11 586 to 44 439) 8 0.141 (0.003) 0.121 (0.005) 0.020 (0.005)* 1037 (18) 372 (34) 665 (38)* 32 780 (20 617 to 70 249)

*P < 0.05. The results are different from those of original data because the values here are based on the bootstrap calculation. CBT-I, cognitive behavioral therapy for insomnia; CI, confidence interval; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life years; TAU, treatment as usual.

0 20 000 40 000

ICER threshold (USD/QALY gained)

60 000 80 000 100 000

P ro

b a b ili

ty o

f n e w

in te

rv e n tio

n b

e in

g c

o st

-e ff e ct

iv e

0.0

0.2

0.4

0.6

0.1

0.3

0.5

0.7

0.9

0.8

1.0

Figure 2 Cost-effectiveness acceptability curves for Approaches 1–4. ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year. , Approach 1; , Approach 2; , Approach 3; , Approach 4.

0 20 000 40 000 60 000

ICER threshold (USD/QALY gained)

P ro

b a b ili

ty o

f n e w

in te

rv e n tio

n b

e in

g c

o st

-e ff e ct

iv e

80 000 100 000 0.0

0.2

0.4

0.6

0.1

0.3

0.5

0.7

0.9

0.8

1.0

Figure 3 Cost-effectiveness acceptability curves for Approaches 5–8. ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year. , Approach 5; , Approach 6; , Approach 7; , Approach 8.

Psychiatry and Clinical Neurosciences 2015; 69: 335–343 Cost-effectiveness of psychotherapy 341

20 000. The probability values were higher in sen- sitivity analyses, other than Approaches 3, 4, 7 and 8, where more costs for CBT-I were presumed than that in the primary analysis.

DISCUSSION To our knowledge, this paper represents the first study investigating the cost-effectiveness of CBT-I from results of an RCT of psychotherapy for insom- nia with concurrent depression. Adding CBT-I is highly likely to be cost-effective with the point esti- mate of the ICER of around 14 000 USD even for 8 weeks. Adding CBT-I demonstrated an approximately 95% chance of gaining one more QALY when a decision-maker is willing to pay for 60 000 USD, and around 90% for 40 000. The cost-effectiveness was robust when considering the results from the sensi- tivity analyses using a worst-case scenario. Given one QALY is often valued at 50 000–70 000 USD,27 the results reported herein demonstrate that patients with insomnia concurrent with depression benefit even for the 8 weeks up to which we followed the patients. The estimated QALY difference is limited to the trial period and any sustained effect of treatment is not included in the study. Considering that the effectiveness of the psychotherapy is highly unlikely to disappear immediately after the follow-up, mon- etary value of adding CBT-I appears very promising in terms of treatment for insomnia concurrent with depression.

Comparing our study with the previous cluster trial on CBT-I for hypnotic-dependent people with insom- nia,14 the present trial has several drawbacks, includ- ing smaller sample size, shorter duration of follow up, and study sites not being primary care clinics but outpatient clinics at university hospitals and a psy- chiatric hospital. However, although the previous trial reported a drop-out rate of 34% in patients during six sessions of CBT-I and of 29% at the 3-month follow-up assessment in patients in both the intervention and the control groups,28 the present study had a drop-out rate of 5% in patients receiving the psychotherapy and no drop-out at all assessment sessions, thus we have no missing data. This was achieved probably because of our enthusiastic follow ups of the patients.

Although the findings seem very promising, one may need to note some methodological limitations of the present study.

First, the sample sizes were small and concerns about the generalizability of the results may be raised. As a matter of fact, the costs were heavily affected by the periods of only two admissions. More- over, the study evaluated the patients up to 8 weeks only, and the long-term consequences were unclear. A further replication study with a larger sample and a longer period of follow up is needed to evaluate the outcomes with more confidence. Second, we could not determine whether the psychotherapy itself or careful watching of patients resulted in greater cost- effectiveness. We aimed to conduct the study to examine the added value of the psychotherapy to usual clinical care, but not to examine the effective- ness of the psychotherapy itself.

On the other hand, the strengths of our study include our method for calculation of the effective- ness. The cost-effectiveness of CBT-I has not often been investigated to date, probably due to difficulties in measuring the utility in patients with insomnia. In the present study, an objective measure for depres- sion severity, the HAMD, was used to estimate the effectiveness. The HAMD raters in the present study were successfully blinded to the group allocation,18

although raters in the previous trial were not blinded. In conclusion, adding CBT-I is highly likely to be

cost-effective for patients with residual insomnia and concomitant depression. Once the program and its cost-effectiveness are replicated in more trials with a larger sample and a longer period of follow up, CBT-I should be more widely disseminated to cover more patients in need.

ACKNOWLEDGMENTS This study was funded by a Grant-in-Aid for Scientific Research (No. 19230201) from the Ministry of Health, Labor and Welfare, Japan, and an Intramural Research Grant (25-8) for Neurological and Psychiat- ric Disorders of National Center of Neurology and Psychiatry, Japan. The authors have no conflicts of interests to declare that may be affected by the pub- lication of this article.

REFERENCES 1. Ohayon MM. Epidemiology of insomnia: What we know

and what we still need to learn. Sleep Med. Rev. 2002; 6: 97–111.

2. Daley M, Morin CM, LeBlanc M, Gregoire JP, Savard J. The economic burden of insomnia: Direct and indirect costs

342 N. Watanabe et al. Psychiatry and Clinical Neurosciences 2015; 69: 335–343

for individuals with insomnia syndrome, insomnia symp- toms, and good sleepers. Sleep 2009; 32: 55–64.

3. Carney CE, Segal ZV, Edinger JD, Krystal AD. A compari- son of rates of residual insomnia symptoms following pharmacotherapy or cognitive-behavioral therapy for major depressive disorder. J. Clin. Psychiatry 2007; 68: 254–260.

4. Sunderajan P, Gaynes BN, Wisniewski SR et al. Insomnia in patients with depression: A STAR*D report. CNS Spectr. 2010; 15: 394–404.

5. Morin CM, Espie CA. Insomnia: A Clinical Guide to Assess- ment and Treatment. Guilford Press, New York, 2004.

6. Perlis ML, Jungquist C, Smith MS, Posner D. Cognitive Behavioral Treatment of Insomnia: A Session-by-Session Guide. Springer Science, New York, 2005.

7. Montgomery P, Dennis J. Cognitive behavioural interven- tions for sleep problems in adults aged 60+. Cochrane Database Syst. Rev. 2003; (1): CD003161.

8. Morin CM, Vallieres A, Guay B et al. Cognitive behavioral therapy, singly and combined with medication, for persis- tent insomnia: A randomized controlled trial. JAMA 2009; 301: 2005–2015.

9. Manber R, Edinger JD, Gress JL, San Pedro-Salcedo MG, Kuo TF, Kalista T. Cognitive behavioral therapy for insom- nia enhances depression outcome in patients with comorbid major depressive disorder and insomnia. Sleep 2008; 31: 489–495.

10. Watanabe N, Furukawa TA, Shimodera S et al. Brief behav- ioral therapy for refractory insomnia in residual depres- sion: An assessor-blind, randomized controlled trial. J. Clin. Psychiatry 2011; 72: 1651–1658.

11. Simon GE, Ludman EJ, Rutter CM. Incremental benefit and cost of telephone care management and telephone psychotherapy for depression in primary care. Arch. Gen. Psychiatry 2009; 66: 1081–1089.

12. Kaltenthaler E, Shackley P, Stevens K, Beverley C, Parry G, Chilcott J. A systematic review and economic evaluation of computerised cognitive behaviour therapy for depression and anxiety. Health Technol. Assess. 2002; 6: 1–89.

13. Araya R, Flynn T, Rojas G, Fritsch R, Simon G. Cost- effectiveness of a primary care treatment program for depression in low-income women in Santiago, Chile. Am. J. Psychiatry 2006; 163: 1379–1387.

14. Morgan K, Dixon S, Mathers N, Thompson J, Tomeny M. Psychological treatment for insomnia in the regulation of long-term hypnotic drug use. Health Technol. Assess. 2004; 8: iii–iiv, 1–68.

15. Furukawa T, Akechi T, Ozaki N et al. GRID-HAMD-17. The Japanese Society of Clinical Neuropsychopharmacology, Tokyo, 2003.

16. Bastien CH, Vallieres A, Morin CM. Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med. 2001; 2: 297–307.

17. Munezawa T, Morin CM, Inoue Y, Nedate K. Develop- ment of the Japanese version of the Insomnia Severity Index (ISI-J). Jpn. J. Psychiatr. Treat. 2009; 24: 219– 225.

18. Watanabe N, Furukawa TA, Shimodera S et al. Can asses- sors in a psychotherapy trial be successfully blinded? Analysis of a randomized controlled trial on psycho- therapy for refractory insomnia in residual depression. Psychother. Psychosom. 2013; 82: 401–403.

19. Furukawa TA, Akechi T, Azuma H, Okuyama T, Higuchi T. Evidence-based guidelines for interpretation of the Ham- ilton Rating Scale for Depression (HAM-D). J. Clin. Psychopharmacol. 2007; 27: 531–534.

20. Pyne JM, Tripathi S, Williams DK, Fortney J. Depression- free day to utility-weighted score: Is it valid? Med. Care 2007; 45: 357–362.

21. Lave JR, Frank RG, Schulberg HC, Kamlet MS. Cost- effectiveness of treatments for major depression in primary care practice. Arch. Gen. Psychiatry 1998; 55: 645– 651.

22. Bennett KJ, Torrance GW. Measuring health state prefer- ences and utilities: Rating scale, time trade-off and stan- dard gamble techniques. In: Spilker B (ed.). Quality of Life and Pharmacoeconomics in Clinical Trials. Lipincott-Raven, Philadelphia, 1996; 253–266.

23. Sapin C, Fantino B, Nowicki ML, Kind P. Usefulness of EQ-5D in assessing health status in primary care patients with major depressive disorder. Health Qual. Life Outcomes 2004; 2: 20.

24. The WHO Collaborating Centre for Drug Statistics Meth- odology. ATC/DDD (Defined Daily Dose) Index 2009. 2009.

25. Sado M, Inagaki A, Yoshimura K, Korek A, Fujisawa D. Seishin shikkan no shakaiteki kosuto no suikei [Estimate of the Societal Cost Caused by Mental Illness]. Ministry of Health, Labour and Welfare of Japan, Tokyo, 2011.

26. Fenwick E, Byford S. A guide to cost-effectiveness acceptability curves. Br. J. Psychiatry 2005; 187: 106– 108.

27. McCrone P, Knapp M, Proudfoot J et al. Cost-effectiveness of computerised cognitive-behavioural therapy for anxiety and depression in primary care: Randomised controlled trial. Br. J. Psychiatry 2004; 185: 55–62.

28. Morgan K, Thompson J, Dixon S, Tomeny M, Mathers N. Predicting longer-term outcomes following psychological treatment for hypnotic-dependent chronic insomnia. J. Psychosom. Res. 2003; 54: 21–29.

Psychiatry and Clinical Neurosciences 2015; 69: 335–343 Cost-effectiveness of psychotherapy 343

Copyright of Psychiatry & Clinical Neurosciences is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.