Skip to main content
. 2017 Nov 13;17(15):1–167.

Table 2a:

Results of Economic Literature Review—Summary: Cost-Effectiveness of CBT for the Treatment of Major Depressive Disorder and/or Generalized Anxiety Disorder

Name, Year, Location Economic Analysis, Study Design, and Perspective Population and Comparator Interventions Results
Health Outcomes Costs Cost-Effectiveness
Wiles et al, 2016, United Kingdom68

  • Economic analysis: individual-level cost-utility analysis

  • Study design: long-term follow-up, multicentre, 1-year

  • Perspective: UK NHS and PPS

  • Time horizon: 3.8 years

  • Discount rate: 3.5%

  • Adults aged 18–75 years with treatment-resistant depression (BDI-II score ≥ 14), adherent to antidepressants for at least 6 weeks but with remaining severe symptoms

  • Total N: 214

  • Female (%): NR

  • Antidepressant use since 12-month follow-up: 72%

  • Intervention: CBT by a trained psychotherapist plus usual care (n = 116) in the original CoBalT RCT

  • Control: usual care (n = 98)

  • CBT: 12–18 sessions in the initial trial

  • Total QALYs (mean/person), CBT vs. usual care: 0.596 (SD: 0.17) vs. 0.544 (SD: 0.20)

  • Mean difference: 0.052 (0.003–0.102)

  • Currency, cost year: £, 2013

  • Total costs, CBT vs. usual care: £885 (SD: 937.92) vs. £604 (SD: 904.15)

  • Mean difference: £281 (32–531)

  • Base case analysis: ICER: £5,374/QALY

  • Probability of CBT being cost-effective long-term at £20,000 WTP/QALY: 92–94%
Koeser et al, 2015, United Kingdom69

  • Economic analysis: cost-utility analysis

  • Study design: decision analytic (modelling) analysis

  • Perspective: UK NHS

  • Time horizon: 24 months

  • Discount rate: 3.5%

  • Adults with moderate or severe MDD (HDRS-17 scores ≥ 14 and BDI scores ≥ 17)

  • Total N: NA

  • Model features: first-line treatment in 3-month acute phase, allowed for dropout but not for treatment augmentation or switching

  • Outcomes: remission, partial response, and no response; QALY at 12 and 24 months

  • Intervention 1: CBT only

  • Intervention 2: Combined treatment (CBT + pharmacotherapy)

  • Control: pharmacotherapy only

  • CBT: 16 sessions during the acute phase (3 months) + 2 booster sessions

  • Pharmacotherapy: citalopram 20 mg/day for 15 months

  • Total QALYs (mean per person), CBT vs. combined treatment vs. pharmacotherapy only: 0.1274 vs. 0.1274 vs. 1.236

  • Mean difference, CBT and combined treatment vs. pharmacotherapy only: 0.038 (95% CrI: −0.03–0.13) and 0.038 (95% CrI: −0.05 to 0.15): 0.052 (0.003–0.102)

  • Currency, cost year: £, 2012

  • Total costs, CBT vs. combined treatment vs. pharmacotherapy only: £4,418 vs. £5,060 vs. £3,645

  • Mean difference, CBT and combined treatment vs. pharmacotherapy only: £773 (95% CI: 470–1,036) and £1,415 (95% CI: 943–1,802) vs. £281 (95% CI: 32–531)

  • Base case analysis, CBT and combined treatment vs. pharmacotherapy only, ICER: £20,039/QALY, dominated

  • Probability of being cost-effective at £25,000 WTP/QALY: CBT vs. pharmacotherapy only, 20%; combined treatment vs. pharmacotherapy only, 43%
Hollinghurst et al, 2014, United Kingdom67; Wiles, 2014, United Kingdom (HTA report, duplicate publication)70

  • Economic analysis: individual-level cost-utility analysis

  • Study design: RCT, multicentre, 1-year (CoBalT)

  • Perspective: UK NHS and PPS

  • Time horizon: 1 year

  • Discount rate: 0%

  • Adults aged 18–75 years (mean 49.6 years) with treatment-resistant depression (BDI-II score ≥ 14, baseline score = 32), adherent to antidepressants for at least 6 weeks but still with severe symptoms

  • Total N: 469

  • Females (%): 72%

  • Sustained antidepressant use since the 12-month follow-up: 70%

  • Intervention: CBT by a trained psychotherapist + pharmacotherapy (n = 234)

  • Control: Usual care (pharmacotherapy) (n = 235)

  • CBT: 12–18 sessions, by psychotherapist; first session 90 minutes, the rest 1 hour

  • Salaried psychologists (£73/hour), overhead and noncontact time: 50%

  • Supervision carried out in groups after 2–3 sessions

  • Total QALYs (mean/person), CBT vs. usual care: 0.61 (SD: 0.22) vs. 0.55 (SD: 0.24)

  • Mean difference: 0.057 (0.015–0.01)

  • Currency, cost year: £, 2010

  • Total costs, CBT vs. usual care: £1,614 (SD: 1,100) vs. £763 (SD: 697)

  • Mean difference: £850 (683–1,017)

  • Base case analysis: ICER, £14,911/QALY

  • Probability of CBT being cost-effective long-term at £20,000 WTP/QALY and at £30,000 WTP/QALY: 74% and 91%

  • Cost-consequence analysis, CBT vs. usual care: £766 (SD: 967) vs. £786 (SD: 718)
Kafali et al, 2014, Puerto Rico/United States66

  • Economic analysis: individual-level cost-utility analysis

  • Study design: open-label RCT

  • Perspective: US, health care payer

  • Time horizon: 4 months

  • Discount rate: 0%

  • Latino adult residents aged 18+ years (96.1% < 65 years) with MDD, eligible for psychotherapy (PHQ-9 score > 10)

  • Total N: 257

  • Females (%): 82%

  • Sustained antidepressant use since the 12-month follow-up: 70%

  • Intervention: face-to-face CBT (n = 84) or phone CBT (n = 87) + usual care

  • Control: usual care with GP: pharmacotherapy or brief counselling (n = 86)

  • CBT: 5-session course/weekly, plus additional 2 sessions/biweekly (max. 8 sessions)

  • Face-to-face sessions, 1.5 hours; phone sessions, 1 hour

  • Delivered by a psychologist

  • Total QALYs (mean/person): NR

  • Mean difference: (1)

    Face-to-face CBT vs. usual care:

    PHQ-9: −2.30 (2) Phone CBT vs. usual care:

    PHQ-9: −2.98 (3) Phone vs. face-to-face CBT: PHQ-9: −0.79

  • Currency, cost year: USD, 2010

  • Total costs, CBT vs. usual care: NR

  • Mean difference:
    • (1)
      Face-to-face CBT vs. usual care: $731.86
    • (2)
      Phone CBT vs. usual care: $236.76
    • (3)
      Phone CBT vs. face-to-face CBT: −$501.18

  • Intervention costs:

    • (1)

      Face-to-face CBT: $136.63

    • (2)

      Phone CBT: $68.32

  • Nonintervention costs included GP, psychiatrist, psychologist, ED, medications

  • Base case analysis: ICER ($/additional unit change in PHQ-9):

    • (1)

      Face-to-face CBT vs. usual care: $318/1 unit change

    • (2)

      Phone CBT vs. usual care: $79/1 unit change

    • (3)

      Phone CBT vs. face-to-face CBT: −$634/1 unit change (cost-saving)

  • Phone CBT is as effective as face-to-face CBT in improving depression scale scores at significantly lower costs (P = .009)
Prukkanone et al, 2012, Thailand65

  • Economic analysis: cost-utility analysis

  • Study design: decision analytic (modelling) analysis

  • Perspective: Thailand, societal

  • Time horizon: 5 years

  • Discount rate: 3%

  • Adults with MDD

  • Total N: NA

  • Model features: discrete-event simulation model, 5 stepped-care scenarios that follow guidelines: 7-week acute-phase and 21-week maintenance-phase treatment

  • Modelled 3 treatment phases: acute, 8–12 weeks; continuation, 6 months; maintenance, up to 5 years

  • Outcomes: remission, relapse/recurrence, DALY

  • Intervention, 5 scenarios:

    • (1)

      CBT in acute phase

    • (2)

      CBT in maintenance phase

    • (3)

      Pharmacotherapy (fluoxetine) in acute phase

    • (4)

      Pharmacotherapy in continuation phase

    • (5)

      Pharmacotherapy in maintenance phase

  • Control: Do nothing

  • CBT: 10 sessions (8–12) during the acute phase and 5 booster sessions (4–6) during maintenance phase

  • Total DALYs averted/100,000 people with MDD, 5 scenarios:

    • (1)

      13,000 (95% CI: 9,000–20,000)

    • (2)

      20,000 (95% CI: 16,000–26,000)

    • (3)

      9,000 (95% CI: 8,000–11,000)

    • (4)

      14,000 (95% CI: 11,000–19,000)

    • (5)

      18,000 (95% CI: 15,000–24,000)

  • Currency, cost year: Thailand bahts, 2005

  • Total costs/100,000 people with MDD, 5 scenarios:

    • (1)

      ฿290 (95% CI: 180–410)

    • (2)

      ฿210 (95% CI: 1,600–300)a

    • (3)

      ฿70 (95% CI: 280–520)

    • (4)

      ฿460 (95% CI: 350–600)

    • (5)

      ฿680 (95% CI: 500–900)

  • Base case analysis, 5 scenarios vs. nothing, average CER (baht/DALY averted):

    • (1)

      ฿23,000 (95% CI: 10,000–36,000)

    • (2)

      ฿11,000 (95% CI: 8,000–14,000)

    • (3)

      ฿42,000 (95% CI: 32,000–57,000)

    • (4)

      ฿33,000 (95% CI: 26,000–44,000)

    • (5)

      ฿42,000 (95% CI: 32,000–57,000)

  • All interventions highly cost-effective, below 1 × GDP of ฿110,000 per capita
Holman et al, 2011, United Kingdom64

  • Economic analysis: individual-level cost-effectiveness analysis

  • Study design: open-label RCT

  • Perspective: UK NHS

  • Time horizon: 10 months

  • Discount rate: 0%

  • Older people (mean 74.1 years [SD: 7.0]) with a primary diagnosis of depression

  • Total N: 198

  • Females (%): 79%

  • 73.5% not taking antidepressants at baseline

  • Intervention 1: CBT

  • (n = 67)

  • Intervention 2: Talking therapy (n = 65)

  • Control: usual care (n = 66)

  • 7 sessions of CBT or talking therapy

  • Total QALYs (mean/person): NR

  • Average point reduction in BDI-II score (mean), CBT vs. usual care: 3.6 (0.7–6.5); CBT vs. talking therapy: 3.5 (0.3–6.5)

  • Currency, cost year: £, 2010

  • Total costs:

    • (1)

      CBT: £1,464 (1,198)

    • (2)

      Usual care: £1,037 (1,005)

    • (3)

      Talking therapy: £884 (537)

  • Mean difference, CBT vs. usual care: £427 (95% CI: 56–787); CBT vs. talking therapy: £580 (95% CI: 280–930)

  • Base case analysis: ICER, CBT vs. usual care: £120 per additional unit reduction in BDI-II score

  • Base case analysis: ICER, talking therapy vs. usual care): £167 per additional unit reduction in BDI-II score

  • Probability of CBT being cost-effective if £270 WTP per point reduction in BDI-II score: 90%
Sava et al, 2009, Romania62

  • Economic analysis: individual-level cost-effectiveness analysis

  • Study design: open-label RCT Perspective: Romania, societal Time horizon: 10 months Discount rate: 0%

  • Adults aged 35–39 years with MDD

  • Total N: 170

  • Females (%): 72%

  • Previous MDEs: 3–4

  • Intervention 1: CBT by a trained psychotherapist (n = 49)

  • Intervention 2: REBT (n = 48)

  • Control: fluoxetine, 20–40 mg/d (n = 44)

  • CBT: max. 20 individual 50-minute sessions over 14 weeks, plus 3 booster sessions in subsequent 6 months

  • REBT: max. 20 individual 50-minute sessions, plus 3 booster sessions in subsequent 6 months

  • Total QALYs: not measured

  • Mean change in depression-free days from baseline per month:

    • (1)

      CBT: 18.6 (9.6)

    • (2)

      REBT: 19.8 (9.0)

    • (3)

      Fluoxetine: 18.1 (8.7)

  • Currency, cost year: USD, 2007

  • Total costs:

    • (1)

      CBT: $505

    • (2)

      REBT: $518

    • (3)

      Fluoxetine: $667

  • Base case analysis, average ICERs:

    • (1)

      CBT: $26.44 per depression-free day, $1,638/QALY

    • (2)

      REBT: $23.77 per depression-free day, $1,734/QALY;

    • (3)

      Fluoxetine: $34.94 per depression-free day, $2,287/QALY
Sado et al, 2009, Japan63

  • Economic analysis: cost-utility analysis

  • Study design: decision analytic (modelling) analysis

  • Perspective: Japan, health care payer and societal

  • Time horizon: 1 year

  • Discount rate: 0%

  • Adults aged 18+ years with MDD Total N: NA

  • Age: NR

  • Model features: severe and moderate depression modelled separately

  • Outcomes: remission, response, and relapse at 3 and 6 months

  • Intervention: CBT + pharmacotherapy

  • Control: pharmacotherapy only

  • CBT: 3 months, 10 sessions (8–12) during acute phase; 5 booster sessions (4–6) during maintenance phase

  • Pharmacotherapy: 3 months paroxetine, 40 mg/day during acute phase; 6 months half-dose during maintenance phase

  • Both intervention and control: consultation with psychiatrist every 2 weeks

  • Total QALYs (mean/person), CBT + pharmacotherapy vs. pharmacotherapy only:

    • (1)

      Severe depression, 0.59 vs. 0.51

    • (2)

      Moderate depression, 0.74 vs. 0.71

  • Mean difference:

    • (1)

      Severe depression: 0.08

    • (2)

      Moderate depression: 0.03

  • Currency, cost year: JPY, 2005

  • Total costs, CBT + pharmacotherapy vs. pharmacotherapy only: ¥449,655 vs. ¥422,244

  • Mean difference: ¥27,411

  • Base-case analysis, health care perspective ICER, severe depression: ¥342,638/QALY; moderate depression: ¥913,700/QALY;

  • CBT + pharmacotherapy cost-effective at ¥6.75 million/QALY WTP

  • Probability of CBT+ pharmacotherapy being cost-effective at WTP: 98% for severe depression, 76% for moderate depression
Simon et al, 2006, United Kingdom61

  • Economic analysis: cost-utility analysis

  • Study design: decision analytic (modelling) analysis

  • Perspective: UK NHS

  • Time horizon: 15 months

  • Discount rate: 0%

  • Adults aged 18+ years with MDD

  • Total N: NA

  • Model features: severe depression and moderate depression modelled separately: 3 months initial treatment and 12 months of follow-up; included inpatient services, hospitalizations

  • Outcomes: remission, recurrence at12 months, QALYs

  • Intervention: CBT + pharmacotherapy

  • Control: Pharmacotherapy only

  • CBT: 16 sessions, 50 mins for over 3 months

  • Pharmacotherapy: fluoxetine, 40 mg/day

  • Total QALYs (mean/person), CBT + pharmacotherapy vs. pharmacotherapy only: severe depression, 0.63 vs 0.52; moderate depression, 0.89 vs. 0.84

  • Mean difference, severe depression: 0.11; moderate depression: 0.04

  • Currency and cost year: £, 2002–2003

  • Total costs, CBT + pharmacotherapy vs. pharmacotherapy only: £1,297 vs. £660

  • Mean difference: £637

  • Base-case analysis, health care perspective ICER, severe depression: £5,777/QALY; moderate depression: £14,540/QALY; cost per additionally treated patient: £4,056/QALY

  • Probability of being cost-effective at £30,000/QALY WTP, CBT + pharmacotherapy vs. pharmacotherapy only: 97%, severe depression; 88%, moderate depression
Revicki et al, 2005, United States60

  • Economic analysis: Individual-level cost-effectiveness analysis

  • Study design: open-label RCT

  • Perspective: US, health care payer perspective (Medicaid)

  • Time horizon: 12 months

  • Discount rate: 0%

  • Low-income women aged 18+ years with MDD, primary care setting

  • Mean age: CBT vs. pharmacotherapy vs. community referral: 29.8 years (SD: 7.9); 28.7 years (SD: 6.6); 29.5 years (SD: 9.1)

  • Total N: 267

  • Females (%): 100%

  • Uninsured: 63–67%

  • Intervention 1: CBT (n = 90)

  • Intervention 2: pharmacotherapy (n = 88)

  • Control (n = 89): community referral, education session, and referral to community services

  • CBT: 8 weekly sessions, group or individual by a licensed psychologist

  • 17% received additional course of CBT (8 sessions)

  • Pharmacotherapy: paroxetine hydrochloride (10–50 mg/day) or bupropion hydrochloride, 6 months

  • Days with depression (mean/person), CBT, pharmacotherapy vs. community referral: 251 (230–273), 258 (236–280) vs. 225 (206–244)

  • Mean difference: CBT, pharmacotherapy vs. community referral: 25.80 (P = .05), 39.7 (P = .005)

  • Currency, cost year: USD, 2002

  • Total outpatient costs, CBT and pharmacotherapy vs. community referral: $976 (SD: 90) and $1,020 (SD: 70) vs. $314 (SD: 48)

  • Mean difference, (1) CBT or (2) pharmacotherapy vs. community referral: (1) $636 (95% CI: 446–826); (2) $677 (95% CI: 484–870)

  • Base case analysis, ICER, CBT vs. control: $27.04 per depression-free day, $17,624/QALY; pharmacotherapy vs. control: $24.65 per depression-free day, $16,068/QALY
Scott et al, 2003, United Kingdom59

  • Economic analysis: individual-level cost-effectiveness analysis

  • Study design: single-blinded RCT

  • Perspective: UK NHS

  • Time horizon: 17 months

  • Discount rate: 6%

  • Adults aged 21–65 years with chronic depression (residual symptoms in past 8 weeks)

  • Mean age, intervention vs. control: 43.2 years (SD: 11.2) vs. 43.5 years (SD: 9.8)

  • Total N: 144

  • Females (%): 53%

  • Severe index MDE: 50%

  • Intervention: pharmacotherapy (min. 125 mg amitriptyline) + CBT by a trained psychotherapist + clinical management (n = 77)

  • Control: pharmacotherapy (min. 125 mg amitriptyline) + clinical management (n = 77)

  • CBT: 16 sessions by psychotherapist over 20 weeks with 2 booster sessions

  • Comparator: clinical management (30-minite appointments with a psychiatrist every 4 weeks during first 20 weeks and every 8 weeks during the next 48 weeks)

  • Relapse rate: CBT vs. control: 29% vs. 47% at 62 weeks (adjusted HR: 0.51, 95% CI: 0.32–0.92)

  • Currency, cost year: £, 1999

  • Total costs including CBT therapy, CBT vs. control: £1,898 (SD: 564) vs. £1,119 (SD: 700)

  • Mean difference: £779 (387–1,170)

  • Total costs of CBT only: £1,164 (1,084–1,244)

  • Base case analysis: ICER: £4,328 per relapse prevented, additional £12.50 per additional relapse-free day

  • Probability of CBT being cost-effective at £6,000 and £8,500 WTP per relapse avoided: 60% and 80%

Abbreviations: BDI-II, Beck Depression Inventory, second edition; CBT, cognitive behavioural therapy; CER, cost-effectiveness ratio; CoBalT, Cognitive Behavioural Therapy as an adjunct to pharmacotherapy for treatment-resistant depression in Primary Care; CrI, credible interval; DALY, disability-adjusted life-year; ED, emergency department; GP, general practitioner; HDRS-17, Hamilton Depression Rating Scale, 17 items; HR, hazard ratio; HTA, health technology assessment; ICER, incremental cost-effectiveness ratio; MDD, major depressive disorder; MDE, major depressive episode; NA, not applicable; NHS, National Health Service; NR, not reported; PHQ-9, Patient Health Questionnaire-9; PPS, Personal Public Service; REBT, rational emotive behaviour therapy; QALY, quality-adjusted life year; RCT, randomized controlled trial; SD, standard deviation; WTP, willingness-to-pay threshold.

a

This is the range reported; however, we believe there may have been a typographical error and that the correct range may be 160–300.