[Skip to Content]
[Skip to Content Landing]
Figure.
CONSORT Diagram
CONSORT Diagram

CONSORT indicates Consolidated Standards of Reporting Trials.

Table 1.  
Baseline Characteristics of Youth in the Usual Care Group and the Reaching Out to Adolescents in Distress (ROAD) Intervention Groupa
Baseline Characteristics of Youth in the Usual Care Group and the Reaching Out to Adolescents in Distress (ROAD) Intervention Groupa
Table 2.  
Costs Associated With the Usual Care Group and the Intervention Group
Costs Associated With the Usual Care Group and the Intervention Group
Table 3.  
Intervention Costs and Cost-effectiveness Ratiosa
Intervention Costs and Cost-effectiveness Ratiosa
1.
Richardson  LP, Russo  JE, Lozano  P, McCauley  E, Katon  W.  The effect of comorbid anxiety and depressive disorders on health care utilization and costs among adolescents with asthma.  Gen Hosp Psychiatry. 2008;30(5):398-406.PubMedGoogle ScholarCrossref
2.
Lewinsohn  PM, Rohde  P, Seeley  JR.  Treatment of adolescent depression: frequency of services and impact on functioning in young adulthood.  Depress Anxiety. 1998;7(1):47-52.PubMedGoogle ScholarCrossref
3.
Knapp  M, McCrone  P, Fombonne  E, Beecham  J, Wostear  G.  The Maudsley long-term follow-up of child and adolescent depression, 3: impact of comorbid conduct disorder on service use and costs in adulthood.  Br J Psychiatry. 2002;180:19-23.PubMedGoogle ScholarCrossref
4.
Merikangas  KR, He  JP, Burstein  M,  et al.  Service utilization for lifetime mental disorders in U.S. adolescents: results of the National Comorbidity Survey–Adolescent Supplement (NCS-A).  J Am Acad Child Adolesc Psychiatry. 2011;50(1):32-45.PubMedGoogle ScholarCrossref
5.
Essau  CA.  Frequency and patterns of mental health services utilization among adolescents with anxiety and depressive disorders.  Depress Anxiety. 2005;22(3):130-137.PubMedGoogle ScholarCrossref
6.
Wu  P, Hoven  CW, Bird  HR,  et al.  Depressive and disruptive disorders and mental health service utilization in children and adolescents.  J Am Acad Child Adolesc Psychiatry. 1999;38(9):1081-1090.PubMedGoogle ScholarCrossref
7.
Wright  DR, Katon  WJ, Ludman  E,  et al.  The association of adolescent depressive symptoms with healthcare utilization and payer-incurred expenditures.  Acad Pediatr. 2016;16(1):82-89.PubMedGoogle ScholarCrossref
8.
Unutzer  J, Katon  WJ, Fan  MY,  et al.  Long-term cost effects of collaborative care for late-life depression.  Am J Manag Care. 2008;14(2):95-100.PubMedGoogle Scholar
9.
Katon  W, Russo  J, Lin  EH,  et al.  Cost-effectiveness of a multicondition collaborative care intervention: a randomized controlled trial.  Arch Gen Psychiatry. 2012;69(5):506-514.PubMedGoogle ScholarCrossref
10.
Katon  WJ, Richardson  L, Russo  J, Lozano  P, McCauley  E.  Quality of mental health care for youth with asthma and comorbid anxiety and depression.  Med Care. 2006;44(12):1064-1072.PubMedGoogle ScholarCrossref
11.
Simon  GE, Katon  WJ, VonKorff  M,  et al.  Cost-effectiveness of a collaborative care program for primary care patients with persistent depression.  Am J Psychiatry. 2001;158(10):1638-1644.PubMedGoogle ScholarCrossref
12.
Archer  J, Bower  P, Gilbody  S,  et al.  Collaborative care for depression and anxiety problems.  Cochrane Database Syst Rev. 2012;10:CD006525.PubMedGoogle Scholar
13.
Richardson  LP, Ludman  E, McCauley  E,  et al.  Improving outcomes for adolescents with depression in primary care: a randomized controlled trial.  JAMA. 2014;312(8):809-816.PubMedGoogle ScholarCrossref
14.
Asarnow  JR, Jaycox  LH, Duan  N,  et al.  Effectiveness of a quality improvement intervention for adolescent depression in primary care clinics: a randomized controlled trial.  JAMA. 2005;293(3):311-319.PubMedGoogle ScholarCrossref
15.
Richardson  LP, McCauley  E, Grossman  DC,  et al.  Evaluation of the Patient Health Questionnaire–9 Item for detecting major depression among adolescents.  Pediatrics. 2010;126(6):1117-1123.PubMedGoogle ScholarCrossref
16.
Richardson  LP, Rockhill  C, Russo  JE,  et al.  Evaluation of the PHQ-2 as a brief screen for detecting major depression among adolescents.  Pediatrics. 2010;125(5):e1097-e1103. doi:10.1542/peds.2009-2712.PubMedGoogle ScholarCrossref
17.
Knight  JR, Sherritt  L, Shrier  LA, Harris  SK, Chang  G.  Validity of the CRAFFT substance abuse screening test among adolescent clinic patients.  Arch Pediatr Adolesc Med. 2002;156(6):607-614.PubMedGoogle ScholarCrossref
18.
Drummond  MF, Sculpher  MJ, Torrance  GW, O’Briend  BJ, Stodart  GL.  Methods for the Economic Evaluation of Health Care Programmes. 3rd ed. New York, NY: Oxford University Press; 2005.
19.
Domino  ME, Burns  BJ, Silva  SG,  et al.  Cost-effectiveness of treatments for adolescent depression: results from TADS.  Am J Psychiatry. 2008;165(5):588-596.PubMedGoogle ScholarCrossref
20.
Mayes  TL, Bernstein  IH, Haley  CL, Kennard  BD, Emslie  GJ.  Psychometric properties of the Children’s Depression Rating Scale–Revised in adolescents.  J Child Adolesc Psychopharmacol. 2010;20(6):513-516.PubMedGoogle ScholarCrossref
21.
Domino  ME, Foster  EM, Vitiello  B,  et al.  Relative cost-effectiveness of treatments for adolescent depression: 36-week results from the TADS randomized trial.  J Am Acad Child Adolesc Psychiatry. 2009;48(7):711-720.PubMedGoogle ScholarCrossref
22.
Pyne  JM, Tripathi  S, Williams  DK, Fortney  J.  Depression-free day to utility-weighted score: is it valid?  Med Care. 2007;45(4):357-362.PubMedGoogle ScholarCrossref
23.
Lynch  FL, Hornbrook  M, Clarke  GN,  et al.  Cost-effectiveness of an intervention to prevent depression in at-risk teens.  Arch Gen Psychiatry. 2005;62(11):1241-1248.PubMedGoogle ScholarCrossref
24.
Kroenke  K, Spitzer  RL, Williams  JB.  The PHQ-9: validity of a brief depression severity measure.  J Gen Intern Med. 2001;16(9):606-613.PubMedGoogle ScholarCrossref
25.
Bird  HR, Andrews  H, Schwab-Stone  M,  et al.  Global measures of impairment for epidemiologic and clinical use with children and adolescents.  Int J Methods Psychiatr Res. 1996;6(4):295-307.Google ScholarCrossref
26.
Rubin  DB.  Multiple Imputation for Nonresponse in Surveys. New York, NY: Wiley; 1987.
27.
Simon  GE, Katon  WJ, Lin  EH,  et al.  Cost-effectiveness of systematic depression treatment among people with diabetes mellitus.  Arch Gen Psychiatry. 2007;64(1):65-72.PubMedGoogle ScholarCrossref
28.
Bureau of Labor Statistics. Consumer price index: 2014. http://data.bls.gov/timeseries/CUUR0000SAM. Accessed December 17, 2015.
29.
Grosse  SD.  Assessing cost-effectiveness in healthcare: history of the $50,000 per QALY threshold.  Expert Rev Pharmacoecon Outcomes Res. 2008;8(2):165-178.PubMedGoogle ScholarCrossref
30.
Burton  A, Billingham  LJ, Bryan  S.  Cost-effectiveness in clinical trials: using multiple imputation to deal with incomplete cost data.  Clin Trials. 2007;4(2):154-161.PubMedGoogle ScholarCrossref
31.
Stata Statistical Software: Release 12 [computer program]. College Station, TX: StataCorp LP; 2011.
32.
Gardner  W, Murphy  M, Childs  G,  et al.  The PSC-17: a brief Pediatric Symptom Checklist including psychosocial problem subscales: a report from PROS and ASPN.  Ambulatory Child Health. 1999;5:225-236.03.Google Scholar
33.
Haby  MM, Tonge  B, Littlefield  L, Carter  R, Vos  T.  Cost-effectiveness of cognitive behavioural therapy and selective serotonin reuptake inhibitors for major depression in children and adolescents.  Aust N Z J Psychiatry. 2004;38(8):579-591.PubMedGoogle ScholarCrossref
34.
Byford  S, Barrett  B, Roberts  C,  et al.  Cost-effectiveness of selective serotonin reuptake inhibitors and routine specialist care with and without cognitive behavioural therapy in adolescents with major depression.  Br J Psychiatry. 2007;191:521-527.PubMedGoogle ScholarCrossref
35.
Stikkelbroek  Y, Bodden  DH, Deković  M, van Baar  AL.  Effectiveness and cost effectiveness of cognitive behavioral therapy (CBT) in clinically depressed adolescents: individual CBT versus treatment as usual (TAU).  BMC Psychiatry. 2013;13:314.PubMedGoogle ScholarCrossref
36.
Stallard  P, Phillips  R, Montgomery  AA,  et al.  A cluster randomised controlled trial to determine the clinical effectiveness and cost-effectiveness of classroom-based cognitive-behavioural therapy (CBT) in reducing symptoms of depression in high-risk adolescents.  Health Technol Assess. 2013;17(47):vii-xvii, 1-109.PubMedGoogle ScholarCrossref
37.
Hollinghurst  S, Carroll  FE, Abel  A,  et al.  Cost-effectiveness of cognitive-behavioural therapy as an adjunct to pharmacotherapy for treatment-resistant depression in primary care: economic evaluation of the CoBalT Trial [published correction appears in Br J Psychiatry. 2014;204(3):245].  Br J Psychiatry. 2014;204(1):69-76.PubMedGoogle ScholarCrossref
38.
Tyrer  P, Cooper  S, Salkovskis  P,  et al.  Clinical and cost-effectiveness of cognitive behaviour therapy for health anxiety in medical patients: a multicentre randomised controlled trial.  Lancet. 2014;383(9913):219-225.PubMedGoogle ScholarCrossref
39.
Lynch  FL, Dickerson  JF, Clarke  G,  et al.  Incremental cost-effectiveness of combined therapy vs medication only for youth with selective serotonin reuptake inhibitor–resistant depression: Treatment of SSRI-Resistant Depression in Adolescents trial findings.  Arch Gen Psychiatry. 2011;68(3):253-262.PubMedGoogle ScholarCrossref
40.
Grochtdreis  T, Brettschneider  C, Wegener  A,  et al.  Cost-effectiveness of collaborative care for the treatment of depressive disorders in primary care: a systematic review.  PLoS One. 2015;10(5):e0123078. doi:10.1371/journal.pone.0123078.PubMedGoogle ScholarCrossref
41.
Ladapo  JA, Neumann  PJ, Keren  R, Prosser  LA.  Valuing children’s health: a comparison of cost-utility analyses for adult and paediatric health interventions in the US.  Pharmacoeconomics. 2007;25(10):817-828.PubMedGoogle ScholarCrossref
42.
Zuvekas  SH, Olin  GL.  Validating household reports of health care use in the Medical Expenditure Panel Survey.  Health Serv Res. 2009;44(5, pt 1):1679-1700.PubMedGoogle ScholarCrossref
Original Investigation
November 2016

The Costs and Cost-effectiveness of Collaborative Care for Adolescents With Depression in Primary Care Settings: A Randomized Clinical Trial

Author Affiliations
  • 1Department of Pediatrics, University of Washington School of Medicine, Seattle
  • 2Seattle Children’s Research Institute Center for Child Health, Behavior, and Development, Seattle, Washington
  • 3Group Health Research Institute, Seattle, Washington
  • 4Department of Psychiatry and Behavioral Sciences, University of Washington School of Medicine, Seattle
JAMA Pediatr. 2016;170(11):1048-1054. doi:10.1001/jamapediatrics.2016.1721
Key Points

Question  The Reaching Out to Adolescents in Distress (ROAD) collaborative care model has been found to be effective in treating adolescent major depressive disorder, but is it cost-effective?

Findings  A randomized clinical trial conducted at 9 primary care clinics in Washington State suggests that collaborative care results in an increase of 0.04 quality-adjusted life-year over usual care at $883 above usual care, for a mean incremental cost-effectiveness ratio of $18 239 per quality-adjusted life-year gained.

Meaning  Even by the most conservative standards, the ROAD collaborative care model is a cost-effective approach for treating adolescent depression.

Abstract

Importance  Depression is one of the most common adolescent chronic health conditions and can lead to increased health care use. Collaborative care models have been shown to be effective in improving adolescent depressive symptoms, but there are few data on the effect of such a model on costs.

Objective  To evaluate the costs and cost-effectiveness of a collaborative care model for treatment of adolescent major depressive disorder in primary care settings.

Design, Setting, and Participants  This randomized clinical trial was conducted between April 1, 2010, and April 30, 2013, at 9 primary care clinics in the Group Health system in Washington State. Participants were adolescents (age range, 13-17 years) with depression who participated in the Reaching Out to Adolescents in Distress (ROAD) collaborative care intervention trial.

Interventions  A 12-month collaborative care intervention included an initial in-person engagement session, delivery of evidence-based treatments, and regular follow-up by master’s level clinicians. Youth in the usual care control condition received depression screening results and could access mental health services and obtain medications through Group Health.

Main Outcomes and Measures  Cost outcomes included intervention costs and per capita health plan costs, calculated from the payer perspective using administrative records. The primary effectiveness outcome was the difference in quality-adjusted life-years (QALYs) between groups from baseline to 12 months. The QALYs were calculated using Child Depression Rating Scale–Revised scores measured during the clinical trial. Cost and QALYs were used to calculate an incremental cost-effectiveness ratio.

Results  Of those screened, 105 youths met criteria for entry into the study, and 101 were randomized to the intervention (n = 50) and usual care (n = 51) groups. Overall health plan costs were not significantly different between the intervention ($5161; 95% CI, $3564-$7070) and usual care ($5752; 95% CI, $3814-$7952) groups. Intervention delivery cost an additional $1475 (95% CI, $1230-$1695) per person. The intervention group had a mean daily utility value of 0.78 (95% CI, 0.75-0.80) vs 0.73 (95% CI, 0.71-0.76) for the usual care group. The net mean difference in effectiveness was 0.04 (95% CI, 0.02-0.09) QALY at $883 above usual care. The mean incremental cost-effectiveness ratio was $18 239 (95% CI, dominant to $24 408) per QALY gained, with dominant indicating that the intervention resulted in both a net cost savings and a net increase in QALYs.

Conclusions and Relevance  Collaborative care for adolescent depression appears to be cost-effective, with 95% CIs far below the strictest willingness-to-pay thresholds. These findings support the use of collaborative care interventions to treat depression among adolescent youth.

Trial Registration  clinicaltrials.gov Identifier: NCT01140464

Introduction

Depression is one of the most common chronic health conditions during adolescence. Studies1-6 have shown that youths with depression are more likely to use health care services and that depression is associated with increased health care use and costs, particularly among those with other chronic illnesses. Although there have been limited studies, the increased costs associated with depression appear to be due to spending across all areas of health care, including general medical services, specialty care, pharmacy use, and mental health services.1 These cost increases are not solely attributable to the costs of depression treatment and persist after controlling for measures of potential medical comorbidities.1,7

Among adults, improvements in depressive symptoms have been shown to be associated with decreased use of general medical care,8 and collaborative care models that improve outcomes for depression have been shown to be cost-effective.9-11 The collaborative care intervention model for depression includes patient engagement strategies, patient choice of treatment, provision of evidence-based treatments in the primary care setting, active follow-up by a depression care manager (DCM) to assist with treatment adherence, and stepped-care approaches to adjust treatment based on patient treatment response. While there have been more than 70 trials of the collaborative care model among adult populations,12 there have been only 2 trials of multicomponent collaborative care models among adolescent populations.13,14 In these studies,13,14 adaptations of the adult collaborative care model were found to improve the delivery of evidence-based care and reduce depressive symptoms among adolescents. To our knowledge, no studies have examined the cost-effectiveness of this model in the adolescent population. In this article, we describe an incremental cost-effectiveness evaluation of collaborative care treatment for adolescent depression.

Methods
Reaching Out to Adolescents in Distress Trial and Intervention

The Reaching Out to Adolescents in Distress (ROAD) study was a randomized trial of a collaborative care depression treatment program for adolescents with depressive disorders. The study methods (Supplement) are described in detail elsewhere.13 Participants were identified through population-based screening for depression among adolescents (age range, 13-17 years) who were enrolled in primary care at 9 clinics in Group Health, an integrated health care system in the Pacific Northwest. Study procedures were reviewed and approved by the Group Health Institutional Review Board.

Computerized records were used to identify all adolescents in the study age range who were registered for primary care at participating clinics. The parents of these potential participants were mailed a letter describing the study and offering them the opportunity to opt out of further contact. Parents were subsequently contacted by telephone to obtain oral informed consent. After parental consent was obtained, study staff contacted youths to obtain verbal assent and conduct depression screening using the 2-item Patient Health Questionnaire (PHQ-2) depressive screen. Adolescents who had a score of 2 or higher on the 2-item screening tool were asked to complete the full 9-item PHQ (PHQ-9) tool during the same telephone call. The PHQ-2 and PHQ-9 are validated screening tools for major depression among adolescents.15,16 Individuals with a PHQ-9 score of 10 or higher were invited to complete a baseline in-person assessment involving a second PHQ-9 assessment and the Kiddie Schedule for Affective Disorders and Schizophrenia diagnostic assessment for depressive disorders. Youths with a second PHQ-9 score of 10 or higher or who met diagnostic criteria for major depression on the Kiddie Schedule for Affective Disorders and Schizophrenia and had a Child Depression Rating Scale–Revised (CDRS-R) score of 42 or higher were invited to participate in the intervention study. Exclusions included those who did not speak English, individuals seeing a psychiatrist, and patients with a suicidal plan or recent attempt, bipolar disorder, alcohol or other drug misuse (CRAFFT17 score ≥4), or developmental delay (Figure).

Eligible youth were randomly assigned to receive usual care or a collaborative care treatment program based in the primary care clinic. The intervention was delivered by a master’s level DCM who served several primary care clinics. The DCMs provided intervention adolescents with a 1-hour engagement session, after which the youth and parent selected treatment with an antidepressant, psychotherapy, or both. Psychotherapy was provided by the DCMs in the clinic. Antidepressant medications were prescribed by primary care professionals and monitored by the DCMs. The DCMs regularly tracked patient response for all intervention youths using the PHQ-9 and received weekly caseload supervision from a psychiatrist, psychologist (E.L. and E.M.), and medical care professional (J.L. and L.P.R.). Symptom response was used to guide treatment advancement using stepped-care algorithms developed for the study.13 For the usual care group, patients, parents, and primary medical care professionals were sent a letter summarizing baseline depression interview results and recommending further assessment and treatment.

Effectiveness Outcomes

The primary effectiveness outcome for the present analysis was the change in quality-adjusted life-years (QALYs) from baseline to 12 months. The QALY is a composite measure of health that combines information on the duration of and the utility value associated with having a health condition.18 A utility value is a preference-based measure of health-related quality of life, measured on a scale ranging from 0.0 (death) to 1.0 (full health).18

Similar to the approach by Domino et al,19 daily depression severity was calculated using baseline, 6-month, and 12-month CDRS-R scores measured during the clinical trial. Scores were linearly interpolated between the time points to obtain a daily CDRS-R score. Youths with a CDRS-R score of 23 or less were considered nondepressed, with a score between 24 and 42 were considered mildly depressed, and with a score exceeding 42 were considered moderately to severely depressed.13,20 Daily utility values for no depression, mild depression, and moderate to severe depression (1.0, 0.8, and 0.6, respectively)19,21-23 were assigned to each youth based on his or her CDRS-R score. We then calculated the net difference in QALYs between groups during the study period by subtracting the mean daily utility value for the usual care group from the mean daily utility value for the intervention group.

In the ROAD intervention trial,13 survey data on depressive symptoms (including the CDRS-R) were missing at random for 18% of the sample at 6-month follow-up and for 20% of the sample at 12-month follow-up. In these cases, a module for multiple imputation with chained equations (Stata MI, version 12; StataCorp LP) was used to impute missing CDRS-R scores using data on baseline depressive symptom scores,20,24 the Columbia Impairment Scale score,25 and child age, sex, and race and parent educational level.26

Intervention Costs

Intervention costs were calculated using a microcosting approach, multiplying resource use by unit costs.18 Visit time related to care management and cognitive behavioral therapy sessions provided as part of the intervention were tracked throughout the study. Costs for intervention delivery were calculated using methods developed for collaborative care studies.9,27 First, we calculated an estimated cost per visit for each category of visit (eg, in-person with therapy, in-person without therapy, or telephone check-in). Subsequently, we calculated an estimated cost of intervention delivery for each individual by multiplying the number of visits in each category by the estimated cost for each type of visit. We also added a fixed $60 per patient cost for caseload supervision and information support, consistent with prior collaborative care studies.9,27

Cost per visit estimates were based on salary and fringe benefit rates for the DCMs, plus a 30% overhead rate for factors like space and administrative support. Using time accounting logs and reports from the DCMs, the estimated cost for an in-person visit with therapy (60-minute session, plus 45 minutes of administrative time) was $96. The estimated cost for an in-person visit without therapy (30-minute session, plus 30 minutes of administrative time) was $55. The estimated cost for a telephone check-in (15 minutes, plus 20 minutes of administrative time) was $32. Administrative time for each of these estimates included outreach efforts, preparation for the session, care coordination within the clinic, and record keeping. An example of the cost calculation for a participant is provided in the following equation:

[(8 In-Person With Therapy Visits × $96) + (2 In-Person Without Therapy Visits × $55)] + [(9 Telephone Check-in × $32) + $60] = $1226.

Visit-Related Costs

All intervention participants were patients at Group Health. We used reimbursement data obtained from Group Health to calculate other health expenditures. Group Health is an integrated care organization that coordinates both care and coverage for its members. Data on patient health care use and expenditures related to outpatient visits, inpatient visits, emergency department visits (including urgent care), prescription drugs, and diagnostic laboratory tests in the 12 months after study entry were collected from administrative data. Use and expenditures were categorized as care that was mental health related or not mental health related based on medical care professional specialty. Group Health expenditure data include estimates of direct costs related to administering patient health care, overhead costs in Group Health clinics, and costs related to approved out-of-plan services paid for by Group Health. Overhead costs consisted of costs related to facilities, payroll, and other administrative departments. Therefore, the cost analysis was conducted from the perspective of the payer. Costs were incurred between April 1, 2010, and April 30, 2013. All costs were inflated to 2014 US dollars using the medical care services component of the US consumer price index.28

Cost-effectiveness Analysis

We conducted the base case cost-effectiveness analysis from the payer perspective. Net costs (the net increase in costs from the intervention compared with usual care) and net effectiveness (the net difference in QALYs between the intervention and usual care groups) were used to calculate an incremental cost-effectiveness ratio (ICER) (net costs divided by net effectiveness).18 An ICER of $50 000 per QALY gained or below is likely to be considered cost-effective.29

Uncertainty Analyses

We used bootstrapping techniques to conduct uncertainty analyses to assess variability in our results from potential sampling bias.18 Because some of the CDRS-R–based effectiveness outcomes in the parent intervention study13 were derived from the trial data using multiple imputation, we first bootstrapped a nonimputed sample of effectiveness outcomes and then conducted multiple imputation on the bootstrapped sample, as recommended by Burton et al.30 Coefficients and standard errors for the 5 imputations were pooled according to rules by Rubin.26 Cost results were not imputed and were bootstrapped directly from the trial data. This process was repeated 1000 times, and the results were used to construct bias-corrected 95% CIs around cost-effectiveness estimates.18 All statistical analyses were conducted using a software program (Stata, version 12; StataCorp LP).31

Results

Of those screened, 105 youths met criteria for entry into the study, and 101 were randomized to the intervention (n = 50) and usual care (n = 51) groups (Figure). There were no differences in population demographic and health characteristics at baseline between the intervention and usual care groups (Table 1).

Costs

Intervention visit costs were $1475 (95% CI, $1230-$1695) per person (Table 2). In-person visits with therapy accounted for most intervention visit costs, with a mean cost of $939 (95% CI, $757-$1107) and a mean of 9.3 visits per child. Intervention youths had a mean of 4.5 in-person visits without therapy, at a mean cost of $261 (95% CI, $205-$329), and a mean of 6.3 telephone check-ins, at a mean cost of $212 (95% CI, $170-$258). There were no study-related expenses for the usual care group.

There were some differences between groups for certain categories of health plan costs. Inpatient and specialist health plan costs were substantially lower in the intervention group compared with the usual care group. Medical care professional costs were higher for the intervention group compared with the usual care group. However, overall health plan costs were not significantly different between the intervention ($5161; 95% CI, $3564-$7070) and usual care ($5752; 95% CI, $3814-$7952) groups. When both health plan costs and intervention costs were considered, the mean total intervention group costs were higher than the mean total usual care group costs. The net mean difference in total costs between the 2 groups was $883 (95% CI, −$920 to $3759). Intervention visit costs accounted for 22.2% of total costs for the intervention group (Table 2).

Effectiveness

Based on CDRS-R scores, the intervention group had a mean daily utility value of 0.78 (95% CI, 0.75-0.80), and the usual care group had a mean daily utility value of 0.73 (95% CI, 0.71-0.76). There were no significant differences in CDRS-R scores between groups at baseline. The net mean difference in effectiveness was 0.04 (95% CI, 0.02-0.09) QALY (Table 3).

Cost-effectiveness

The mean ICER was $18 239 (95% CI, dominant to $24 408) per QALY gained (dominant is defined below). In 25.9% (259 of 1000) of cases in bootstrapped uncertainty analyses used to generate 95% CIs, the intervention was both less expensive and more effective than usual care, meaning that the ICER (net costs divided by net effectiveness) was negative (Table 3). We did not report negative ICERs in 95% CIs because these ratios are difficult to interpret. That is, it is hard to know whether an ICER is negative because the numerator is negative (ie, the intervention is less expensive than usual care) or because the denominator is negative (ie, the intervention is less effective than usual care). Instead, in these cases, we say that the intervention is dominant, meaning that the intervention results in both a net cost savings and a net increase in QALYs.18

Discussion

Using health plan cost and use data and clinical depression scale scores, we determined that the collaborative care model of the ROAD intervention was more effective in reducing depressive symptoms and only marginally more expensive than usual care when implemented in an integrated health care environment. The incremental cost-effectiveness of the ROAD intervention is estimated to be $18 239 (95% CI, dominant to $24 408) per QALY gained, which would be considered cost-effective by even the most conservative cost-effectiveness threshold of $50 000 per QALY gained.29

Articles in the peer-reviewed literature have evaluated the cost-effectiveness of pharmacotherapy, cognitive behavioral therapy, or combination therapies to treat adolescent depression,19,21,33-39 although few analyses and treatments are directly comparable to our results because of differing timelines, comparison groups, costing perspectives, and outcomes.19,21,33-39 Owing to significant differences in health care systems, pharmaceutical costs, and clinical outcomes, we elected not to compare the results of our US-based economic evaluation with non-US studies.33-38 In addition, the Treatment of SSRI-Resistant Depression in Adolescents (TORDIA) trial39 focuses on a more severely affected population that was resistant to initial treatment and thus may not be comparable to the ROAD sample, who were identified in the primary care setting. Therefore, we excluded the TORDIA trial from our comparisons.

Based on our review of the literature, the trial with the most similar study population to ours was the Treatment for Adolescents With Depression Study (TADS),19,21 a 36-week clinical trial evaluating the use of fluoxetine and cognitive behavioral therapy to treat adolescent depression. Although the treatment duration in the TADS trial (9 months) was shorter than that in the ROAD trial13 (12 months), overall costs of treatment were similar. In the TADS trial, when considering only the health plan–related and not societal costs, treatment with fluoxetine alone, cognitive behavioral therapy, and combination therapy cost a mean of $7058, $6994, and $6046, respectively (2014 US dollars), per child (compared with $6636 for the ROAD trial). Because of differing study comparison groups and intervention duration, we elected not to compare the ICERs between the TADS trial and the ROAD trial. However, the ROAD intervention is cost-effective relative to other health plan–perspective adult collaborative care depression interventions, most of which have ICERs ranging from dominant to $24 569.40 Based on a 2007 review of similar pediatric health interventions that found a median ICER of $10 156 per QALY gained (2014 US dollars),41 the ROAD intervention would also be considered cost-effective in the context of other pediatric health interventions.

A strength of this study is that we used measures of costs and use in this analysis that were derived directly from health plan administrative records. Other studies19,21 have used a microcosting approach to estimate costs, multiplying parent-reported use by the Medicaid reimbursement rates for each visit costs, which may underestimate costs because Medicaid rates are typically lower than private insurance rates. In addition, parent-reported use is subject to recall bias and not always an accurate measure of actual use. While individuals are usually able to report emergency and inpatient use with some accuracy, outpatient use is often underreported in self-report surveys.42

Some limitations of this study should be noted. First, some CDRS-R scores were imputed because of missing data, but missingness was missing at random and was therefore unlikely to bias the results. Other measures of disease severity, including baseline CDRS-R, PHQ-9, and Columbia Impairment Scale scores, in addition to demographic factors, were used to impute missing values. Second, our sample was predominantly of white race and from a single health care organization in the Pacific Northwest of the United States, which may limit the generalizability of our findings to more diverse settings.

Third, unlike other studies, we did not collect data related to patient family out-of-pocket and time costs, which could be used to conduct an analysis from the societal perspective. In addition, societal costs related to school absenteeism and use of school-based counseling services were not measured. Other analyses suggest that family-incurred costs could represent up to 30% of total costs.21,39 Therefore, this study represents only a subset of costs, and the economic burden of depression may be even higher when societal costs are taken into account. Future studies should prospectively collect these data to better understand how collaborative care affects societal costs.

Fourth, utility values used in the analysis were derived from adult measures and may not precisely mimic adolescent health-related quality-of-life preferences. However, these measures have also been used in other economic analyses of interventions for adolescent depression,19,21 which allows for better comparability of the value of interventions. In addition, our assumed utility value of 0.6 is higher than the 0.4 used in some prior base-case evaluations,39 which may make the ROAD intervention appear less cost-effective than if we had used a lower utility value.

Conclusions

Collaborative care interventions have been shown to enhance the delivery of evidence-based treatments among depressed youth and to improve outcomes. The ROAD collaborative care intervention is likely to be cost-effective, with 95% CIs far below even the strictest willingness-to-pay thresholds and with ICERs within the range of similar adult collaborative care and pediatric health interventions. These findings support the use of collaborative care interventions to treat depression among adolescents.

Back to top
Article Information

Corresponding Author: Laura P. Richardson, MD, MPH, Department of Pediatrics, University of Washington School of Medicine, PO Box 5371, M/S CW8-6, Seattle, WA 98145-5005 (laura.richardson@seattlechildrens.org).

Accepted for Publication: May 23, 2016.

Published Online: September 19, 2016. doi:10.1001/jamapediatrics.2016.1721

Author Contributions: Drs Wright and Richardson had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: All authors.

Acquisition, analysis, or interpretation of data: Wright, Haaland.

Drafting of the manuscript: Wright.

Critical revision of the manuscript for important intellectual content: All authors.

Obtained funding: Richardson.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was funded by grant R01 MH085645-01A1 from the National Institute of Mental Health, National Institutes of Health (Dr Richardson, principal investigator).

Role of the Funder/Sponsor: The sponsor of this study had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: Wayne Katon, MD, University of Washington, Seattle, assisted with the development of the study design and analytic plan for this analysis.

References
1.
Richardson  LP, Russo  JE, Lozano  P, McCauley  E, Katon  W.  The effect of comorbid anxiety and depressive disorders on health care utilization and costs among adolescents with asthma.  Gen Hosp Psychiatry. 2008;30(5):398-406.PubMedGoogle ScholarCrossref
2.
Lewinsohn  PM, Rohde  P, Seeley  JR.  Treatment of adolescent depression: frequency of services and impact on functioning in young adulthood.  Depress Anxiety. 1998;7(1):47-52.PubMedGoogle ScholarCrossref
3.
Knapp  M, McCrone  P, Fombonne  E, Beecham  J, Wostear  G.  The Maudsley long-term follow-up of child and adolescent depression, 3: impact of comorbid conduct disorder on service use and costs in adulthood.  Br J Psychiatry. 2002;180:19-23.PubMedGoogle ScholarCrossref
4.
Merikangas  KR, He  JP, Burstein  M,  et al.  Service utilization for lifetime mental disorders in U.S. adolescents: results of the National Comorbidity Survey–Adolescent Supplement (NCS-A).  J Am Acad Child Adolesc Psychiatry. 2011;50(1):32-45.PubMedGoogle ScholarCrossref
5.
Essau  CA.  Frequency and patterns of mental health services utilization among adolescents with anxiety and depressive disorders.  Depress Anxiety. 2005;22(3):130-137.PubMedGoogle ScholarCrossref
6.
Wu  P, Hoven  CW, Bird  HR,  et al.  Depressive and disruptive disorders and mental health service utilization in children and adolescents.  J Am Acad Child Adolesc Psychiatry. 1999;38(9):1081-1090.PubMedGoogle ScholarCrossref
7.
Wright  DR, Katon  WJ, Ludman  E,  et al.  The association of adolescent depressive symptoms with healthcare utilization and payer-incurred expenditures.  Acad Pediatr. 2016;16(1):82-89.PubMedGoogle ScholarCrossref
8.
Unutzer  J, Katon  WJ, Fan  MY,  et al.  Long-term cost effects of collaborative care for late-life depression.  Am J Manag Care. 2008;14(2):95-100.PubMedGoogle Scholar
9.
Katon  W, Russo  J, Lin  EH,  et al.  Cost-effectiveness of a multicondition collaborative care intervention: a randomized controlled trial.  Arch Gen Psychiatry. 2012;69(5):506-514.PubMedGoogle ScholarCrossref
10.
Katon  WJ, Richardson  L, Russo  J, Lozano  P, McCauley  E.  Quality of mental health care for youth with asthma and comorbid anxiety and depression.  Med Care. 2006;44(12):1064-1072.PubMedGoogle ScholarCrossref
11.
Simon  GE, Katon  WJ, VonKorff  M,  et al.  Cost-effectiveness of a collaborative care program for primary care patients with persistent depression.  Am J Psychiatry. 2001;158(10):1638-1644.PubMedGoogle ScholarCrossref
12.
Archer  J, Bower  P, Gilbody  S,  et al.  Collaborative care for depression and anxiety problems.  Cochrane Database Syst Rev. 2012;10:CD006525.PubMedGoogle Scholar
13.
Richardson  LP, Ludman  E, McCauley  E,  et al.  Improving outcomes for adolescents with depression in primary care: a randomized controlled trial.  JAMA. 2014;312(8):809-816.PubMedGoogle ScholarCrossref
14.
Asarnow  JR, Jaycox  LH, Duan  N,  et al.  Effectiveness of a quality improvement intervention for adolescent depression in primary care clinics: a randomized controlled trial.  JAMA. 2005;293(3):311-319.PubMedGoogle ScholarCrossref
15.
Richardson  LP, McCauley  E, Grossman  DC,  et al.  Evaluation of the Patient Health Questionnaire–9 Item for detecting major depression among adolescents.  Pediatrics. 2010;126(6):1117-1123.PubMedGoogle ScholarCrossref
16.
Richardson  LP, Rockhill  C, Russo  JE,  et al.  Evaluation of the PHQ-2 as a brief screen for detecting major depression among adolescents.  Pediatrics. 2010;125(5):e1097-e1103. doi:10.1542/peds.2009-2712.PubMedGoogle ScholarCrossref
17.
Knight  JR, Sherritt  L, Shrier  LA, Harris  SK, Chang  G.  Validity of the CRAFFT substance abuse screening test among adolescent clinic patients.  Arch Pediatr Adolesc Med. 2002;156(6):607-614.PubMedGoogle ScholarCrossref
18.
Drummond  MF, Sculpher  MJ, Torrance  GW, O’Briend  BJ, Stodart  GL.  Methods for the Economic Evaluation of Health Care Programmes. 3rd ed. New York, NY: Oxford University Press; 2005.
19.
Domino  ME, Burns  BJ, Silva  SG,  et al.  Cost-effectiveness of treatments for adolescent depression: results from TADS.  Am J Psychiatry. 2008;165(5):588-596.PubMedGoogle ScholarCrossref
20.
Mayes  TL, Bernstein  IH, Haley  CL, Kennard  BD, Emslie  GJ.  Psychometric properties of the Children’s Depression Rating Scale–Revised in adolescents.  J Child Adolesc Psychopharmacol. 2010;20(6):513-516.PubMedGoogle ScholarCrossref
21.
Domino  ME, Foster  EM, Vitiello  B,  et al.  Relative cost-effectiveness of treatments for adolescent depression: 36-week results from the TADS randomized trial.  J Am Acad Child Adolesc Psychiatry. 2009;48(7):711-720.PubMedGoogle ScholarCrossref
22.
Pyne  JM, Tripathi  S, Williams  DK, Fortney  J.  Depression-free day to utility-weighted score: is it valid?  Med Care. 2007;45(4):357-362.PubMedGoogle ScholarCrossref
23.
Lynch  FL, Hornbrook  M, Clarke  GN,  et al.  Cost-effectiveness of an intervention to prevent depression in at-risk teens.  Arch Gen Psychiatry. 2005;62(11):1241-1248.PubMedGoogle ScholarCrossref
24.
Kroenke  K, Spitzer  RL, Williams  JB.  The PHQ-9: validity of a brief depression severity measure.  J Gen Intern Med. 2001;16(9):606-613.PubMedGoogle ScholarCrossref
25.
Bird  HR, Andrews  H, Schwab-Stone  M,  et al.  Global measures of impairment for epidemiologic and clinical use with children and adolescents.  Int J Methods Psychiatr Res. 1996;6(4):295-307.Google ScholarCrossref
26.
Rubin  DB.  Multiple Imputation for Nonresponse in Surveys. New York, NY: Wiley; 1987.
27.
Simon  GE, Katon  WJ, Lin  EH,  et al.  Cost-effectiveness of systematic depression treatment among people with diabetes mellitus.  Arch Gen Psychiatry. 2007;64(1):65-72.PubMedGoogle ScholarCrossref
28.
Bureau of Labor Statistics. Consumer price index: 2014. http://data.bls.gov/timeseries/CUUR0000SAM. Accessed December 17, 2015.
29.
Grosse  SD.  Assessing cost-effectiveness in healthcare: history of the $50,000 per QALY threshold.  Expert Rev Pharmacoecon Outcomes Res. 2008;8(2):165-178.PubMedGoogle ScholarCrossref
30.
Burton  A, Billingham  LJ, Bryan  S.  Cost-effectiveness in clinical trials: using multiple imputation to deal with incomplete cost data.  Clin Trials. 2007;4(2):154-161.PubMedGoogle ScholarCrossref
31.
Stata Statistical Software: Release 12 [computer program]. College Station, TX: StataCorp LP; 2011.
32.
Gardner  W, Murphy  M, Childs  G,  et al.  The PSC-17: a brief Pediatric Symptom Checklist including psychosocial problem subscales: a report from PROS and ASPN.  Ambulatory Child Health. 1999;5:225-236.03.Google Scholar
33.
Haby  MM, Tonge  B, Littlefield  L, Carter  R, Vos  T.  Cost-effectiveness of cognitive behavioural therapy and selective serotonin reuptake inhibitors for major depression in children and adolescents.  Aust N Z J Psychiatry. 2004;38(8):579-591.PubMedGoogle ScholarCrossref
34.
Byford  S, Barrett  B, Roberts  C,  et al.  Cost-effectiveness of selective serotonin reuptake inhibitors and routine specialist care with and without cognitive behavioural therapy in adolescents with major depression.  Br J Psychiatry. 2007;191:521-527.PubMedGoogle ScholarCrossref
35.
Stikkelbroek  Y, Bodden  DH, Deković  M, van Baar  AL.  Effectiveness and cost effectiveness of cognitive behavioral therapy (CBT) in clinically depressed adolescents: individual CBT versus treatment as usual (TAU).  BMC Psychiatry. 2013;13:314.PubMedGoogle ScholarCrossref
36.
Stallard  P, Phillips  R, Montgomery  AA,  et al.  A cluster randomised controlled trial to determine the clinical effectiveness and cost-effectiveness of classroom-based cognitive-behavioural therapy (CBT) in reducing symptoms of depression in high-risk adolescents.  Health Technol Assess. 2013;17(47):vii-xvii, 1-109.PubMedGoogle ScholarCrossref
37.
Hollinghurst  S, Carroll  FE, Abel  A,  et al.  Cost-effectiveness of cognitive-behavioural therapy as an adjunct to pharmacotherapy for treatment-resistant depression in primary care: economic evaluation of the CoBalT Trial [published correction appears in Br J Psychiatry. 2014;204(3):245].  Br J Psychiatry. 2014;204(1):69-76.PubMedGoogle ScholarCrossref
38.
Tyrer  P, Cooper  S, Salkovskis  P,  et al.  Clinical and cost-effectiveness of cognitive behaviour therapy for health anxiety in medical patients: a multicentre randomised controlled trial.  Lancet. 2014;383(9913):219-225.PubMedGoogle ScholarCrossref
39.
Lynch  FL, Dickerson  JF, Clarke  G,  et al.  Incremental cost-effectiveness of combined therapy vs medication only for youth with selective serotonin reuptake inhibitor–resistant depression: Treatment of SSRI-Resistant Depression in Adolescents trial findings.  Arch Gen Psychiatry. 2011;68(3):253-262.PubMedGoogle ScholarCrossref
40.
Grochtdreis  T, Brettschneider  C, Wegener  A,  et al.  Cost-effectiveness of collaborative care for the treatment of depressive disorders in primary care: a systematic review.  PLoS One. 2015;10(5):e0123078. doi:10.1371/journal.pone.0123078.PubMedGoogle ScholarCrossref
41.
Ladapo  JA, Neumann  PJ, Keren  R, Prosser  LA.  Valuing children’s health: a comparison of cost-utility analyses for adult and paediatric health interventions in the US.  Pharmacoeconomics. 2007;25(10):817-828.PubMedGoogle ScholarCrossref
42.
Zuvekas  SH, Olin  GL.  Validating household reports of health care use in the Medical Expenditure Panel Survey.  Health Serv Res. 2009;44(5, pt 1):1679-1700.PubMedGoogle ScholarCrossref
×