A retrospective analysis to estimate the healthcare resource utilization and cost associated with treatment-resistant depression in commercially insured US patients

Gang Li; Ling Zhang; Allitia DiBernardo; Grace Wang; John J Sheehan; Kwan Lee; Johan Reutfors; Qiaoyi Zhang

doi:10.1371/journal.pone.0238843

. 2020 Sep 11;15(9):e0238843. doi: 10.1371/journal.pone.0238843

A retrospective analysis to estimate the healthcare resource utilization and cost associated with treatment-resistant depression in commercially insured US patients

Gang Li ^1,^*, Ling Zhang ^1,^¤, Allitia DiBernardo ², Grace Wang ¹, John J Sheehan ³, Kwan Lee ¹, Johan Reutfors ⁴, Qiaoyi Zhang ¹

Editor: Fernando A Wilson⁵

¹Janssen Research & Development, LLC, Raritan, New Jersey, United States of America

²Janssen Research & Development, LLC, Titusville, New Jersey, United States of America

³Janssen Scientific Affairs, LLC, Titusville, New Jersey, United States of America

⁴Karolinska Institutet, Stockholm, Sweden

⁵University of Utah, UNITED STATES

Competing Interests: The authors have read the journal’s policy and have the following competing interests: GL, AD, GW, JS, KL, and QZ are employees of Janssen Research & Development, LLC and hold stock in the company. LZ was an employee of Janssen Research & Development, LLC, at the time this study took place and is currently an employee of Boehringer Ingelheim Pharmaceuticals, Inc. JR is in research collaboration with Janssen, AstraZeneca, AbbVie, and Pfizer, for which Karolinska Institutet has received grant support. JR has been a speaker for Eli Lilly and received unrestricted grant support from Schering-Plough. Wilson Joe, PhD, of MedErgy, provided editorial support for this manuscript. Editorial support was funded by Janssen Research & Development, LLC. This does not alter our adherence to PLOS ONE policies on sharing data and materials. There are no patents, products in development or marketed products associated with this research to declare.

^¤

Current address: Boehringer Ingelheim Pharmaceuticals, Inc, Ridgefield, Connecticut, United States of America

^✉

* E-mail: GLi@its.jnj.com

Roles

Gang Li: Conceptualization, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing

Ling Zhang: Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft

Allitia DiBernardo: Conceptualization, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing

Grace Wang: Conceptualization, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft

John J Sheehan: Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing

Kwan Lee: Formal analysis, Investigation, Methodology, Visualization, Writing – original draft

Johan Reutfors: Conceptualization, Investigation, Methodology, Visualization, Writing – review & editing

Qiaoyi Zhang: Conceptualization, Formal analysis, Investigation, Methodology, Visualization, Writing – review & editing

Fernando A Wilson: Editor

PMCID: PMC7485754 PMID: 32915863

Abstract

Objective

The economic burden of commercially insured patients in the United States with treatment-resistant depression and patients with non–treatment-resistant major depressive disorder was compared using data from the Optum Clinformatics™ claims database.

Methods

Patients 18–63 years on antidepressant treatment between 1/1/13 and 9/30/13, who had no treatment claims for depression 6 months before the index date (first antidepressant dispensing), and who had a major depressive disorder or depression diagnosis within 30 days of the index date, were included. Treatment-resistant depression was defined as receiving 3 antidepressant regimens during 1 major depressive disorder episode. Patients with treatment-resistant depression were matched with patients with non–treatment-resistant major depressive disorder at a 1:4 ratio using propensity score matching. The study consisted of 1-year baseline (pre-index) and 2-year follow-up (post index) periods. Cost outcomes were compared using a generalized linear model.

Results

2,370 treatment-resistant depression and 9,289 non–treatment-resistant major depressive disorder patients were included. In year 1 of the follow-up period, compared with non–treatment-resistant major depressive disorder, patients with treatment-resistant depression had: more emergency department visits (odds ratio = 1.39, 95% confidence interval = 1.24–1.56); more inpatient hospitalizations (odds ratio = 1.73, 95% confidence interval = 1.46–2.05); longer hospital stays (mean difference vs non–treatment-resistant major depressive disorder = 2.86, 95% confidence interval = 0.86–4.86 days); and more total healthcare costs (mean difference vs non–treatment-resistant major depressive disorder = US$3,846, 95% confidence interval = $2,855-$4,928). These patterns remained consistent in year 2 of the follow-up period.

Conclusion

Treatment-resistant depression was associated with higher healthcare resource utilization and costs versus non–treatment-resistant major depressive disorder in this commercially insured cohort of patients in the United States.

Introduction

Depression is a widespread, severely disabling disorder associated with impaired daily functioning, diminished quality of life, and increased mortality and healthcare utilization [1–4]. Healthcare costs such as outpatient medical services, pharmaceutical services, and inpatient services as well as indirect costs such as workplace presenteeism and absenteeism all contribute substantially towards the total burden of major depressive disorder [3]. In 2012, the US societal economic burden of major depressive disorder was estimated at $188 billion, which exceeded the US societal burden of cancer ($131 billion) and diabetes ($173 billion) [4].

The goal of major depressive disorder treatment is to achieve remission, a subclinical state where the patient is no more than mildly symptomatic, fully functional, and essentially indistinguishable from those without major depressive disorder [5–7]. However, as measured by the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study, which used a large, representative patient sample and presented a comprehensive view of nonresponse to depression treatment, approximately 30% of patients with major depressive disorder do not achieve remission even after adequate trials of 2 antidepressant treatments [5]. The term “treatment-resistant depression” refers to depression that does not respond to antidepressant therapy [6]. Although no consensus definition currently exists, the US Agency for Healthcare Research and Quality (AHRQ) and Food and Drug Administration (FDA) proposed a standard definition of treatment-resistant depression: failing to respond to a minimum of 2 antidepressants administered at an adequate dose, for an adequate duration [8,9]. There is currently no consensus regarding the definitions of ‘adequate’; a recent review on treatment-resistant depression found that most studies considered an adequate treatment duration to last for a minimum of 4 or 6 consecutive weeks, with the majority requiring ≥4 weeks [10,11]. In the current study, we defined adequate dose based upon the American Psychiatric Association Practice Guidelines for Treatment of Patients with Major Depressive Disorder [12] and approved recommended minimal dosage, and adequate duration was defined by an algorithm that required ≥29 days of prescription coverage (details outlined in Materials and Methods below). Results from STAR*D suggest that nonresponse to 2 adequate trials of established pharmacotherapy classes is an inflection point that predicts a poor prognosis with respect to low remission and high relapse rates, and is associated with higher rates of future medication intolerance [5,13].

Greater healthcare utilization and higher healthcare costs have been demonstrated in patients with treatment-resistant depression compared with non–treatment-resistant major depressive disorder [14–22]. According to three recent estimates, per-patient per-year direct healthcare costs in patients with treatment-resistant depression versus non–treatment-resistant major depressive disorder were US$6,709 higher among commercially-insured patients, $4,382 higher among Medicaid-insured patients, and $9,479 higher among US integrated delivery network–insured patients [14,20,22]. Among commercially-insured patients, indirect work loss–related costs were also US$1,811 greater in patients with treatment-resistant depression [14]. Although both treatment resistance and symptom severity are associated with increased direct and indirect costs in patients with major depressive disorder, treatment-resistant depression appears to be the primary contributor to the economic burden of depression [23]. This retrospective study was conducted to provide an updated estimate of the economic burden of patients with treatment-resistant depression compared with patients with non–treatment-resistant major depressive disorder using recent US data and a more comprehensive definition of treatment-resistant depression.

Materials and methods

Data source

This study was based on insurance claims data from the Optum Clinformatics™ Extended Data Mart (CEDM). CEDM stores medical and pharmacy benefit coverage records of commercial and Medicare Advantage health plan members. Data are routinely captured, verified, automated, and de-identified, providing a key information source for various research efforts.

Sample selection and study design

The study sample included patients, 18 to 63 years old, who had an antidepressant medication dispensed between January 1, 2013 and September 30, 2014 and had no claims for pharmacologic or nonpharmacologic depression treatments 6 months before the index date (the date of first antidepressant medication dispensing). The 12 months before the index date constituted the baseline period, and the 24 months after the index date constituted the follow-up period. Patients included in the analysis had a clinician’s diagnosis of major depressive disorder according the UnitedHealthcare Guidelines (International Classification of Diseases, 9th Edition [ICD-9] 296.2x and 296.3x [except for 296.25 and 296.30], 300.4, 309.0, 309.1, and 311) within ±30 days of the index date [24]. Eligible patients had to have ≥2 consecutive antidepressant medications dispensed (with a gap of ≤30 days after the index date for the 3 non–major depressive disorder diagnosis categories) to ensure some level of compliance. Except for patients who died during the study period, all patients were continuously enrolled in the health plan, for both pharmacy and medical benefits, during the baseline and follow-up periods.

Patients were excluded if they had an ICD-9 or International Classification of Diseases, 10th Edition (ICD-10) code for psychosis, schizophrenia, mania, or bipolar disorder, or an ICD-9 code for dementia at any time. Additionally, patients were excluded if they had a dispense of lithium, thyroid hormone (T3 or T4), an antipsychotic, or an anti-epileptic–type mood stabilizer, or if they received electroconvulsive treatment or transcranial magnetic stimulation during the 6 months before the index date.

Use of the database was reviewed by the New England Institutional Review Board (IRB) and was determined to be exempt from IRB approval, as this study did not involve human subjects research.

Treatment-resistant depression

Patients were identified as having treatment-resistant depression or non–treatment-resistant major depressive disorder using a claims-based algorithm. This study employed a definition of treatment-resistant depression based upon the AHRQ definition: depression that fails to respond to a minimum of 2 antidepressant treatments administered at an adequate dose and duration (referred to as drugs A and B below) [9]. A listing of treatments that were defined as antidepressants is shown in S1 Table. Accordingly, a patient with major depressive disorder was considered to have treatment-resistant depression if the patient received 3 antidepressant regimens, of adequate dose and duration for the first 2, in the current major depressive disorder episode. The first regimen was required to be an antidepressant, but the second and third regimens could be an antidepressant taken alone, with another antidepressant, or with an augmentation medication (anticonvulsant, antipsychotic, lithium, psychostimulant, or thyroid hormone) [12,25].

An adequate antidepressant dose was defined by the recommended minimal dosage in the American Psychiatric Association major depressive disorder practice guidelines [12] or in the US Food and Drug Administration-approved package inserts. Adequate duration was assessed using an algorithm to determine medication failure of drug A (first-line treatment) based on its treatment duration before the introduction of drug B. Drug A was considered a failure if drug B was introduced between 29 and 180 days or if drug A was augmented with drug B starting on Day 15 or later; the same algorithm was used to determine the failure of drug B based on the introduction of drug C.

Assessments

Patient demographic and baseline clinical characteristics were assessed and compared between treatment-resistant depression and non–treatment-resistant depression groups. Characteristics included: age group (18–24, 25–34, 35–44, 45–54, and 55–63 years); sex; index year (2013/2014); depression diagnosis (296.2, 296.3, 300, 309, or 311) within 30 days of the index date; diagnosis of anxiety, substance abuse, personality disorder, and post-traumatic stress disorder (PTSD) during the baseline period; and Elixhauser comorbidity score, calculated using diagnosis codes during the baseline period [26]. A propensity score was derived from the age on index date; gender; depression diagnosis code around index date; baseline diagnosis of anxiety, personality disorder, substance abuse, and/or PTSD; and Elixhauser comorbidity score. Patients with treatment-resistant depression were matched to those with non–treatment-resistant major depressive disorder using the propensity score at a 1:4 ratio with the greedy approach and calipers of width equal to 0.02.

Healthcare utilization and costs were estimated annually for 2 consecutive years during the follow-up period. Number of outpatient visits (which included office based and ambulatory hospital outpatient visits), proportion of patients with emergency department (ED) visits, proportion of patients with hospitalizations, and hospital length of stay (LOS; ie, the sum of hospital stay days from all hospitalizations during the one-year follow-up period) were assessed to measure resource utilization. Costs were estimated from the payer and patient perspectives. Medical costs to payers included claims for outpatient visits, ED visits, and hospitalizations; pharmacy costs to payers were the sum of pharmacy claims; and total costs to payers were the sum of medical costs and pharmacy costs to payers. Medical costs to patients were defined as the sum of deductibles, copayments, and coinsurance for all medical services; prescription costs to patients were defined as the sum of deductibles, copayments, and coinsurance for all prescription drugs; and total costs to patients were the sum of medical costs and prescription costs to patients. Total healthcare costs were defined as the sum of costs to payers and patients. All cost estimates were made using 2017 US$ rates.

Statistical analysis

Descriptive statistics were generated to summarize patient characteristics and outcome measures for treatment-resistant depression and non–treatment-resistant depression cohorts. Between-group comparisons of demographic and baseline clinical characteristics were made using chi-square tests for categorical variables and t-tests for continuous variables.

Annual resource utilization and costs associated with care for patients with treatment-resistant depression versus non–treatment-resistant major depressive disorder were compared using a generalized linear model and log-link function with negative binomial distribution for resource utilization and gamma distribution for cost (SAS GENMOD procedure), adjusted for the baseline value of the variable [12]. Estimates and 95% confidence intervals (CIs) were obtained using bootstrapping with 1000 iterations. In a sensitivity analysis, cost data were analyzed using a similar linear model with normal distribution. Additionally, a linear model with normal distribution was used to calculate differences in costs for patients with treatment-resistant depression versus non–treatment-resistant major depressive disorder, adjusting for the baseline variable, by depression diagnosis codes (major depressive disorder, ICD-9 296.X; dysthymic disorder, ICD-9 300.X; adjustment disorder, ICD-9 309.X; depressive disorder not otherwise specified, ICD-9 311.X). The odds ratio (OR) of hospitalization and ED visits for patients with treatment-resistant depression versus non–treatment-resistant major depressive disorder was estimated using a logistic regression model with repeated measurements (SAS GENMOD procedure) adjusted for the respective baseline value of the variable, year of follow-up, treatment-resistant depression status * year, and baseline Elixhauser score [27].

Two high dimensional covariate selection approaches [28] were implemented as part of sensitivity analyses to identify covariates that might potentially impact costs in addition to those pre-specified for the propensity score matching (see S1 Appendix for details).

Results

Patient disposition and baseline characteristics

Of 17,859 eligible patients diagnosed with major depressive disorder, 2,384 (13%) had treatment-resistant depression and 15,475 (87%) had non–treatment-resistant major depressive disorder (see patient disposition flow diagram in S1 Fig). Compared with patients with non–treatment-resistant major depressive disorder, patients with treatment-resistant depression were slightly younger, more likely to be female, and to have a history of anxiety or PTSD. After propensity score matching, 2,370 patients with treatment-resistant depression and 9,289 patients with non–treatment-resistant major depressive disorder were included in the analysis. Mean age after matching was 39.2 years, and 62% of patients were female. Baseline characteristics of patients in the 2 groups were comparable, except for the Elixhauser score (Table 1); therefore, Elixhauser score was adjusted in the healthcare utilization and cost analyses.

Table 1. Demographic and clinical characteristics of patients with treatment-resistant depression and non–treatment-resistant depression before and after the propensity score matching.

Characteristic	Unmatched					Matched
	Treatment-resistant depression (N = 2,384)		Non–treatment-resistant major depressive disorder (N = 15,475)			Treatment-resistant depression (n = 2,370)		Non–treatment-resistant major depressive disorder (n = 9,289)
	n	%	n	%	P	n	%	n	%	P
Age (years)	Mean 39.2	SD 13.0	Mean 40.1	SD 12.9	0.0021	Mean 39.2	SD 12.9	Mean 39.2	SD 12.8	0.9635
Age group (years)
18–24	472	20	2724	18		468	20	1746	19
25–34	440	18	2779	18		437	18	1770	19
35–44	585	25	3735	24		582	25	2335	25
45–54	547	23	3657	24		545	23	2132	23
55–63	340	14	2580	17	0.0082	338	14	1306	14	0.8111
Female sex	1481	62	9209	60	0.0154	1474	62	5758	62	0.8531
Comorbidities
Anxiety	659	28	3664	24	< .0001	646	27	2392	26	0.6138
Personality Disorder	17	1	53	<1	0.0813	8	<1	9	<1	0.7099
Substance Abuse	73	3	392	3	0.1056	71	3	227	2	0.1217
PTSD	37	2	179	1	< .0001	35	1	113	1	0.2195
Elixhauser score^a	Mean 1.8	SD 1.35	Mean 1.70	SD 1.26	0.0645	Mean 1.74	SD 1.33	Mean 1.68	SD 1.21	0.0293
Major depressive disorder diagnostic code
Major depressive disorder (ICD-9 296.X)	1003	42	5485	35		991	42	3831	41
Dysthymic disorder (ICD-9 300.X)	323	14	2307	15		322	14	1235	13
Adjustment disorder (ICD-9 309.X)	65	3	520	3		65	3	205	2
Depressive disorder NOS (ICD-9 311.X)	993	42	7163	46	< .0001	992	42	4018	43	0.3248

Open in a new tab

CI, confidence interval; SD, standard deviation.

^aThe Elixhauser score ranges from 0 to 30.

Healthcare utilization

Healthcare utilization was significantly and consistently higher in the treatment-resistant depression group than in the non–treatment-resistant major depressive disorder group (Table 2). In the first year of follow up, compared with non–treatment-resistant major depressive disorder, patients with treatment-resistant depression had: more emergency department visits (OR = 1.39, 95% CI = 1.24–1.56]) and more inpatient hospitalizations (OR = 1.73, 95% CI = 1.46–2.05]). In addition, the difference in adjusted predicted outcomes between patients with and without treatment-resistant depression was a 2.86-day longer length of hospital stay (difference vs non–treatment-resistant major depressive disorder, 95% CI = 2.86, 0.86–4.86 days) and 2.95 more outpatient visits (difference vs non–treatment-resistant major depressive disorder, 95% CI = 2.95, 2.48–3.43 visits). These patterns remained in the second year of follow up.

Table 2. Healthcare resource utilization per year during the study period.

Variable	Treatment-resistant depression		Non–treatment-resistant major depressive disorder		Treatment-resistant depression vs non–treatment-resistant major depressive disorder
Variable	n	%	n	%	Odds ratio	95% CI
Patients with ED visits in Year 1, %	628	26	2933	19	1.39	1.24	1.56
Patients with ED visits in Year 2, %	530	22	2714	18	1.27	1.13	1.43
Patients with inpatient hospitalization in Year 1, %	201	8	754	5	1.73	1.46	2.05
Patients with inpatient hospitalization in Year 2, %	163	7	737	5	1.43	1.19	1.73
Variable	Treatment-resistant depression		Non–treatment-resistant major depressive disorder		Treatment-resistant depression vs non–treatment-resistant major depressive disorder
Variable	n		n		Estimate of mean difference	95% CI
Hospital LOS in Year 1, number of days	8.75		5.90		2.86	0.86	4.86
Hospital LOS in Year 2, number of days	9.60		6.19		3.41	–0.43	7.25
Number of outpatient visits in Year 1	11.45		8.50		2.95	2.48	3.43
Number of outpatient visits in Year 2	7.39		5.59		1.79	1.35	2.24

Open in a new tab

CI, confidence interval; ED, emergency department; LOS, length of stay.

Healthcare costs

Consistent with the increased healthcare utilization observed in patients with treatment-resistant depression, costs were significantly higher in the treatment-resistant depression group (Table 3). The adjusted mean (95% CI) differences in total payer costs between the treatment-resistant depression and the non–treatment-resistant major depressive disorder groups were US$3,430 ($2,438-$4,478) for year 1 and US$2,191 ($1,031-$3,453) for year 2. Estimated between-group mean (95% CI) differences in patients’ total out-of-pocket costs were US$354 ($260-$457) for year 1 and US$184 ($91-$285) for year 2. For total healthcare costs, including both reimbursed costs and costs to patients, estimated between-group mean (95% CI) differences were US$3,846 ($2,855-$4,928) in year 1 and US$2,412 ($1,217-$3,713) in year 2. The results from the linear model with normal distribution were consistent: treatment-resistant depression patients had statistically significantly higher reimbursed costs as well as costs to patients (see S2 Table). The results from the two high dimensional covariate selection approaches that adjusted for additional covariates were also consistent (see S1 Appendix and tables and figure therein). Mean cost differences between patients with treatment-resistant depression versus non–treatment-resistant depression varied based on depression diagnosis code, but differences in sample sizes limit interpretation (see S3 Table).

Table 3. Comparison of costs per year between treatment-resistant depression and non–treatment-resistant major depressive disorder patients (US$)^a.

Variable	Treatment-resistant depression	Non–treatment-resistant major depressive disorder	Treatment-resistant depression vs non–treatment-resistant major depressive disorder
Variable			Adjusted mean difference	95% CI
Cost to payers
Medical cost in Year 1	9075	6125	2950	2051	3978
Medical cost in Year 2	8393	6621	1772	632	2958
Pharmacy cost in Year 1	2043	1507	535	300	789
Pharmacy cost in Year 2	2027	1664	362	58	720
Total cost to payers in Year 1	11014	7585	3430	2438	4478
Total cost to payers in Year 2	10175	7984	2191	1031	3453
Cost to patients
Medical cost in Year 1	1373	1019	444	347	556
Medical cost in Year 2	1207	1022	245	150	344
Prescription cost in Year 1	406	318	88	68	109
Prescription cost in Year 2	350	301	49	30	70
Total cost to patients in Year 1	1767	1323	354	260	457
Total cost to patients in Year 2	1499	1254	184	91	285
Total healthcare cost
Total healthcare cost in Year 1	12726	8881	3846	2855	4928
Total healthcare cost in Year 2	11591	9179	2412	1217	3713

Open in a new tab

CI, confidence interval.

^aMedical costs to payers included claims for outpatient visits, ED visits, and hospitalizations; pharmacy costs to payers were the sum of pharmacy claims; and total costs to payers were the sum of medical costs and pharmacy costs to payers. Medical costs to patients were defined as the sum of deductibles, copayments, and coinsurance for all medical services; prescription costs to patients were defined as the sum of deductibles, copayments, and coinsurance for all prescription drugs; and total costs to patients were the sum of medical costs and prescription costs to patients. Total healthcare costs were defined as the sum of costs to payers and patients.

Discussion

This study assessed healthcare utilization and costs of treatment-resistant depression, analyzing data from a US patient sample obtained from the CEDM database. During year 1 and year 2 following the index date, healthcare utilization was significantly higher in the treatment-resistant depression group than in the non–treatment-resistant major depressive disorder group. Consistent with this finding, patients with treatment-resistant depression had significantly higher reimbursed and out-of-pocket medical, pharmacy, and total healthcare costs.

Multiple prior studies of treatment-resistant depression have employed a range of criteria to define the treatment-resistant depression patient population. These criteria typically include some combination of the following: clinical diagnosis of depression, number of treatments used (>2 to ≥4), use of specific medications, time on medication(s), upward titration of medication(s), use of optimization strategies, and results from questionnaires [14–22,29]. Therefore, it is not surprising to observe the variance in reported percentages of patients with major depressive disorder who were treatment-resistant, which ranged between 11% and 30%; in most studies, patients with treatment-resistant depression were predominately female (64% to 74%) and between 35 and 55 years of age, although in 1 study [17] only 41% of patients with treatment-resistant depression were female. Given the lack of a consensus treatment-resistant depression definition, we adopted an evidence-based, comprehensive definition based on the AHRQ definition employed in the STAR*D trial: failure to respond to 2 oral antidepressant treatments of adequate duration and dose.

In this study, mean total healthcare costs to payers in year 1 and year 2, respectively, were 45% and 27% higher for patients with treatment-resistant depression compared with those with non–treatment-resistant major depressive disorder. This result is consistent with prior work across a range of patient populations, which found a 25%-134% higher burden among those with treatment-resistant depression versus those with non–treatment-resistant major depressive disorder [14–22]. Although not absolute, in general those analyses that assessed some component(s) of the indirect burden of major depressive disorder, such as productivity, identified larger percentage increases in the burden, suggesting that the incremental indirect burden of major depressive disorder among those with treatment-resistant depression is larger than the incremental direct burden.

One unexpected finding of the current is study is that estimated differences in healthcare utilization and costs between the treatment-resistant depression and non–treatment-resistant major depressive disorder groups were generally smaller in year 2 than in year 1. Specifically, we found that the number of outpatient visits and proportions of patients who had ED visits and inpatient hospitalization were all reduced during year 2 compared with year 1 for both treatment-resistant depression and non–treatment-resistant major depressive disorder groups, and the estimated mean differences were also slightly reduced. Although hospital LOS did not follow that trend, the estimated mean difference between treatment-resistant depression and non–treatment-resistant major depressive disorder groups was not significant in year 2. As episodes of treatment-resistant depression are generally longer than those of non–treatment-resistant major depressive disorder [30], it was anticipated that the incremental burden of treatment-resistant depression would remain constant in year 2 or increase compared with year 1. However, the episodic nature of major depressive disorder, or the possibility that over time and multiple medication changes switches, patients may eventually find an effective treatment may help to explain this relative decrease; however, it is unclear whether these year-to-year differences are clinically meaningful and further investigation is needed.

In contrast to most previous studies, which focused on healthcare costs reimbursed by payers, the current study also examined patients’ out-of-pocket costs. Patients with treatment-resistant depression had out-of-pocket medical and pharmacy costs of US$1,323 in year 1 and US$1,254 in year 2. Compared with non–treatment-resistant major depressive disorder, costs for treatment-resistant depression represented increases of US$354 in year 1 and US$184 in year 2. These costs are likely to represent a substantial burden for many patients with treatment-resistant depression. In the STAR*D study, participants reported high unemployment rates, ranging from 36% for patients who responded to step 1 treatment to 47% for patients who progressed to step 4 [5]. In another STAR*D analysis, patients with treatment-resistant depression demonstrated lower vocational productivity than patients with non–treatment-resistant major depressive disorder [31]. A claims-based study found that employees with treatment-resistant depression had an average of 35.8 work loss days per year, which was 1.7 times the rate of work loss days in employees with non–treatment-resistant major depressive disorder and 6.2 times that of those without major depressive disorder [14]. Thus, the higher out-of-pocket healthcare costs associated with treatment-resistant depression shown in the current analysis may represent a considerable financial hardship for this vulnerable population.

This study has several limitations. Data were from a claims database, which captures diagnoses recorded for reimbursement purposes rather than clinical diagnoses. Depression may be underreported in claims data for various reasons such as social stigma and financial incentives to bill for general medical disorder management. Diagnoses were based on individual physicians’ clinical judgment and did not receive additional validation. Medication changes suggest treatment failure, but it is not possible to disentangle switches due to lack of efficacy or tolerability, or patient choice. In the absence of full medical histories, patients’ major depressive disorder previous history, such as years of diagnosed major depressive disorder and number of major depressive disorder episodes, was not captured. Furthermore, results obtained using the Optum Clinformatics^TM database may not generalize beyond patients with employer-sponsored commercial insurance and Medicare Advantage insurance.

Conclusions

The results of this retrospective study suggest that patients with treatment-resistant depression have significantly greater healthcare utilization than matched patients with non–treatment-resistant major depressive disorder. This difference in healthcare utilization translates into significantly higher reimbursed and out-of-pocket medical, pharmacy, and overall costs for patients with treatment-resistant depression than those with non–treatment-resistant major depressive disorder.

Supporting information

S1 Appendix. High dimensional covariate selection approaches.

(DOCX)

Click here for additional data file.^{(99.5KB, docx)}

S1 Fig. Patient disposition.

(DOCX)

Click here for additional data file.^{(93.7KB, docx)}

S1 Table. List of antidepressant medications and minimum adequate dose.

(DOCX)

Click here for additional data file.^{(35.3KB, docx)}

S2 Table. Comparison of costs per year (US$) between treatment-resistant depression and non–treatment-resistant major depressive disorder patients by the model with gamma log link vs a linear model^a.

(DOCX)

Click here for additional data file.^{(34.6KB, docx)}

S3 Table. Difference in least square means of costs per year (US$) between treatment-resistant depression and non–treatment-resistant major depressive disorder patients by depression diagnosis codes (from linear models)^a.

(DOCX)

Click here for additional data file.^{(35.1KB, docx)}

S4 Table. Demographic and clinical characteristics of patients with treatment-resistant depression and non–treatment-resistant depression before and after the propensity score matching at 1:1 ratio.

(DOCX)

Click here for additional data file.^{(36.6KB, docx)}

S5 Table. Healthcare resource utilization per year during the study period (matched at 1:1 ratio).

(DOCX)

Click here for additional data file.^{(33.3KB, docx)}

S6 Table. Comparison of costs per year between treatment-resistant depression and non–treatment-resistant major depressive disorder patients (US$; matched at 1:1 ratio)^a.

(DOCX)

Click here for additional data file.^{(34.4KB, docx)}

Data Availability

The data for these analyses were made available to the authors by third-party license from Optum, a commercial data provider in the US, and Janssen Pharmaceuticals (who have a license for analysis of the Optum Clinformatics™ Extended Data Mart [CEDM]). As such, the authors cannot provide the raw data themselves. Other researchers could access the data by purchase through Optum; and the inclusion criteria specified in the Methods section would allow them to identify the same cohort of patients we used for these analyses. Interested individuals may visit https://www.optum.com/solutions/life-sciences.html for more information on accessing Optum CEDM data. We confirm that no authors had special privileges to access data from Optum via third-party license, and that other researchers would be able to access the data in the same manner as the authors.

Funding Statement

Janssen Research & Development, LLC provided funding for this study in the form of salaries for GL, LZ, AD, GW, JJS, KL, and QZ, as well as funding for editorial support. Schering-Plough also provided funding in the form of unrestricted grant support to JR. The specific roles of these authors are articulated in the ‘author contributions’ section. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Hasin DS, Sarvet AL, Meyers JL, Saha TD, Ruan WJ, Stohl M, et al. Epidemiology of adult DSM-5 major depressive disorder and its specifiers in the United States. JAMA Psychiatry. 2018;75: 336–346. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Eaton WW, Martins SS, Nestadt G, Bienvenu OJ, Clarke D, Alexandre P. The burden of mental disorders. Epidemiol Rev. 2008;30: 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Greenberg PE, Fournier AA, Sisitsky T, Pike CT, Kessler RC. The economic burden of adults with major depressive disorder in the United States (2005 and 2010). J Clin Psychiatry. 2015;76: 155–162. [DOI] [PubMed] [Google Scholar]
4.Mrazek DA, Hornberger JC, Altar CA, Degtiar I. A review of the clinical, economic, and societal burden of treatment-resistant depression: 1996–2013. Psychiatr Serv. 2014;65: 977–987. [DOI] [PubMed] [Google Scholar]
5.Rush AJ, Trivedi MH, Wisniewski SR, Nierenberg AA, Stewart JW, Warden D, et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am J Psychiatry. 2006;163: 1905–1917. [DOI] [PubMed] [Google Scholar]
6.Murphy JA, Sarris J, Byrne GJ. A review of the conceptualisation and risk factors associated with treatment-resistant depression. Depress Res Treat. 2017:4176825. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Ballenger JC. Clinical guidelines for establishing remission in patients with depression and anxiety. J Clin Psychiatry. 1999;60: 29–34. [PubMed] [Google Scholar]
8.US Food and Drug Administration. Major depressive disorder: developing drugs for treatment. Guidance for industry. Available from: https://www.fda.gov/media/113988/download
9.Gaynes BN, Asher G, Gartlehner G, Hoffman V, Green J, Boland E, et al. Definition of treatment-resistant depression in the Medicare population. Technology assessment program. Project ID: PSYT0816. (Prepared by RTI–UNC Evidence-Based Practice Center under Contract No. HSA290201500011I_HHSA29032006T). Rockville: Agency for Healthcare Research and Quality; 2018. [PubMed]
10.Thase ME, Blomgren SL, Birkett MA, Apter JT, Tepner RG. Fluoxetine treatment of patients with major depressive disorder who failed initial treatment with sertraline. J Clin Psychiatry. 1997;58: 16–21. [DOI] [PubMed] [Google Scholar]
11.Gaynes BN, Lux L, Gartlehner G, Asher G, Forman-Hoffman V, Green J, et al. Defining treatment-resistant depression. Depress Anxiety. 2020;37: 134–145. [DOI] [PubMed] [Google Scholar]
12.Gelenberg AJ, Freeman MP, Markowitz JC, Rosenbaum JF, Thase ME, Trivedi MH, et al. Practice guideline for the treatment of patients with major depressive disorder. 3rd ed American Psychiatric Association; 2010. [Google Scholar]
13.Conway CR, George MS, Sackeim HA. Toward an evidence-based, operational definition of treatment-resistant depression: when enough is enough. JAMA Psychiatry. 2017;74: 9–10. [DOI] [PubMed] [Google Scholar]
14.Amos TB, Tandon N, Lefebvre P, Pilon D, Kamstra RL, Pivneva I, et al. Direct and indirect cost burden and change of employment status in treatment-resistant depression: a matched-cohort study using a US commercial claims database. J Clin Psychiatry. 2018;79: 17m11725. [DOI] [PubMed] [Google Scholar]
15.Gibson TB, Jing Y, Smith Carls G, Kim E, Bagalman JE, Burton WN, et al. Cost burden of treatment resistance in patients with depression. Am J Manag Care. 2010;16: 370–377. [PubMed] [Google Scholar]
16.Greenberg P, Corey-Lisle PK, Birnbaum H, Marynchenko M, Claxton A. Economic implications of treatment-resistant depression among employees. Pharmacoeconomics. 2004;22: 363–373. [DOI] [PubMed] [Google Scholar]
17.Ivanova JI, Birnbaum HG, Kidolezi Y, Subramanian G, Khan SA, Stensland MD. Direct and indirect costs of employees with treatment-resistant and non-treatment-resistant major depressive disorder. Curr Med Res Opin. 2010;26: 2475–2484. [DOI] [PubMed] [Google Scholar]
18.Olchanski N, McInnis Myers M, Halseth M, Cyr PL, Bockstedt L, Goss TF, et al. The economic burden of treatment-resistant depression. Clin Ther. 2013;35: 512–522. [DOI] [PubMed] [Google Scholar]
19.Olfson M, Amos TB, Benson C, McRae J, Marcus SC. Prospective service use and health care costs of Medicaid beneficiaries with treatment-resistant depression. J Manag Care Spec Pharm. 2018;24: 226–236. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Pilon D, Sheehan JJ, Szukis H, Singer D, Jacques P, Lejeune D, et al. Medicaid spending burden among beneficiaries with treatment-resistant depression. J Comp Eff Res. 2019;8: 381–392. [DOI] [PubMed] [Google Scholar]
21.Pilon D, Joshi K, Sheehan JJ, Zichlin ML, Zuckerman P, Lefebvre P, et al. Burden of treatment-resistant depression in Medicare: a retrospective claims database analysis. PLoS One. 2019;14: e0223255. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Pilon D, Szukis H, Joshi K, Singer D, Sheehan JJ, Wu JW, et al. US integrated delivery networks perspective on economic burden of patients with treatment-resistant depression: a retrospective matched-cohort study. Pharmacoecon Open. 2020;4: 119–131. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Fostick L, Silberman A, Beckman M, Spivak B, Amital D. The economic impact of depression: resistance or severity? Eur Neuropsychopharmacol. 2010;20: 671–675. [DOI] [PubMed] [Google Scholar]
24.UnitedHealthcare. Guideline: ICD-9 code–drug match POS prior authorization. Available from: https://www.uhccommunityplan.com›icd-9_code-drug_match_fl_fhk
25.Al-Harbi KS. Treatment-resistant depression: therapeutic trends, challenges, and future directions. Patient Prefer Adherence. 2012;6: 369–388. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43: 1130–1139. [DOI] [PubMed] [Google Scholar]
27.SAS Institute Inc. SAS/STAT^® 9.1 user’s guide. Cary: SAS Institute Inc.; 2004. [Google Scholar]
28.Schneeweiss S, Rassen JA, Glynn RJ, Avorn J, Mogun H, Brookhart MA. High-dimensional propensity score adjustment in studies of treatment effects using health care claims data. Epidemiology. 2009;20: 512–522. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Reutfors J, Andersson TM, Brenner P, Brandt L, DiBernardo A, Li G, et al. Mortality in treatment-resistant unipolar depression: a register-based cohort study in Sweden. J Affect Disord. 2018;238: 674–679. [DOI] [PubMed] [Google Scholar]
30.Kubitz N, Mehra M, Potluri RC, Garg N, Cossrow N. Characterization of treatment resistant depression episodes in a cohort of patients from a US commercial claims database. PLoS One. 2013;8: e76882. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.DiBernardo A, Lin X, Zhang Q, Xiang J, Lu L, Jamieson C, et al. Humanistic outcomes in treatment resistant depression: a secondary analysis of the STAR*D study. BMC Psychiatry. 2018;18: 352. [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0238843.r001

Decision Letter 0

Fernando A Wilson

25 Mar 2020

PONE-D-19-34560

A retrospective analysis to estimate the healthcare resource utilization and cost associated with treatment-resistant depression

PLOS ONE

Dear Dr. Li,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Reviewer comments are included below.

We would appreciate receiving your revised manuscript by May 09 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Fernando A. Wilson, PhD

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

* In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. Thank you for stating the following in the Competing Interests section:

"Authors GL, LZ, AD, GW, JS, KL, and QZ are employees of Janssen Research & Development, LLC and hold stock in the company. JR is in research collaboration with Janssen, AstraZeneca, Abbvie, and Pfizer, for which Karolinska Institutet has received grant support. JR has been a speaker for Eli Lilly and received unrestricted grant support from Schering-Plough. Wilson Joe, PhD, of MedErgy, provided editorial support for this manuscript. Editorial support was funded by Janssen Research & Development, LLC."

i) Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

ii) Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

Additional Editor Comments (if provided):

- In your revised manuscript, verify that PLOS ONE formatting requirements are addressed. (https://journals.plos.org/plosone/s/submission-guidelines)

- In the text, reference numbers should be enclosed in square brackets (https://journals.plos.org/plosone/s/submission-guidelines#loc-references)

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: No

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This study used 2013 Optum Clinformatics claims data to examine the association of treatment-resistant depression with healthcare resource utilization and costs in the U.S. The authors found that patients with treatment-resistant depression were more likely to receive inpatient, outpatient, and emergency care, and to have higher healthcare costs during the follow-up period after the index date.

[1] Major

None.

[2] Minor

1. P.9, line157. The authors may want to perform a sensitivity analysis (i.e., 1:1 propensity score matching), given that they used the propensity score at a 1:4 ratio, and the literature suggests that increasing the number of untreated subjects matched to each treated subject tends to increase the bias in the estimated treatment effect (e.g., Austin PC, Am J Epidemiol. 2010).

2. P.9, line 161. Please clarify whether hospital length of stay (LOS) is total LOS during the follow-up period or LOS per admission.

3. Please clarify the terminology and/or definitions. For example,

(1) P.9., lines 162-167. “Payer” seems to mean insurer and “costs to payer” seems to mean insurance coverage/reimbursement in this context, given that total health care costs are sum of “payer” costs and patients’ out-of-pocket costs. The authors may want to clarify this throughout the text and in Table 3.

(2) P.9., lines 162-167. The authors defined (a) “payer” costs over service types (e.g., outpatient, inpatient, ED), (b) patient cost as the sum of the patient’s out-of-pocket costs (e.g., deductible, copayment), and (c) total healthcare costs as the sum of “payer” and patient costs. This is confusing because (i) both “payers” and patients are types of payers and (ii) both of them pay costs across various service types (i.e., “payer” costs are the amount covered by insurers, and patient costs are the amount not covered by insurers).

(3) Pages 14 and 16, and Tables 2 and 3. The interpretation of the effects of treatment with resistant depression on outpatient visits, hospital LOS, and costs was somewhat unclear (e.g., “estimated mean differences” or “estimate of mean difference”). To improve the interpretability of the estimates, the authors may want to define the effects as average marginal effect (i.e., difference in adjusted predicted outcomes between patients with and without treatment-resistant depression) and use this terminology consistently in the text and tables.

Reviewer #2: Thank you for your paper. As you will see below, I have concerns about this study's operational definition of treatment resistant depression (TRD), the inclusion of both privately insured persons and Medicare Advantage beneficiaries without distinguishing between them in the analyses, and the dissimilarity between the two matched groups in terms of the comorbidity index.

----------------------

1) Abstract and title: Please include information about the subjects (e.g., persons with private or Medicare Advantage insurance coverage). Payer and patient expenditures would be expected to vary greatly by insurance type, so this is important information.

2) Page 4 paragraph 1: The second sentence indicates that the estimated economic burden of MDD in the US in 2010 was US$210.5 billion, but the last sentence indicates that the societal economic burden of MDD in 2012 was $188 billion. This inconsistency is not explained and thus raises questions. Is MDD becoming less economically burdensome or is this due to methodological differences? Please rework this first paragraph so the reader is not distracted by the inconsistency in past research. (e.g., provide more information, or the last sentence could become less specific, simply stating that the economic burden of MDD has been estimated to exceed that of cancer and diabetes but providing no specific numbers).

3) Page 4-5 lines 67 through 70: You indicate that reference number 10 (by Al-Harbi) proposes that an adequate duration is treatment for >=4 weeks with >=3 weeks on an adequate dose. Reading this reference, it appears that this suggestion was not Al-Harbi's but was put forth in a paper cited by Al-Harbi in their introduction section (reference #9 in that paper - Thase ME, Rush JA. Treatment-resistant depression. In: Bloom FE, Kupfer DJ, editors. Psychopharmacology. New York, NY: Raven; 1995.). However, Al-Harbi goes on to review the literature and ultimately concludes in the Discussion section that "It seems that depression should only be considered drug resistant after at least 6 weeks of two trials of antidepressant therapy" (page 383). Al-Harbi also discusses the 6 week time period in "Optimization of Antidepressants" section on page 374. Thus, your statement that Al-Harbi proposes a >=4 week time period appears to be inaccurate -- >-4 weeks actually contradicts the conclusions of the review.

4) Page 5 lines 70 through 73: You say that your methodology for adequate dose is based on the Massachusetts General Hospital (MGH) Antidepressant Treatment Response Questionnaire (ATRQ). However, you do not use the Massachusetts General Hospital Antidepressant Treatment Response Questionnaire's definition of adequate duration: According to the article you cite (#11), "The MGH ATRQ defines 6 weeks on an adequate dose of antidepressant medication as an adequate duration of treatment." Please change your definition accordingly and rework your analyses. Doing so would be consistent both with the Al-Harbi article you cite, the MGH ATRQ that you reference, and numerous recent studies that used claims data to examine the costs of treatment resistant depression, including but not limited to one recently published in PLOS ONE (see #5 below). Others include Amos et al 2018 "Direct and Indirect Cost Burden and Change of Employment Status in Treatment Resistant Depression"; Benson et al 2020, "An evaluation of the clinical and economic burden among older adult Medicare-covered beneficiaries with treatment resistant depression"; Pilon et al 2019, "Medicaid spending burden among beneficiaries with treatment-resistant depression"; Pilon 2019, "US integrated delivery networks perspective on economic burden of treatment resistant depression: retrospective matched cohort study." This is not a complete list - I am only providing a few examples.

5) Page 6 line 101: Presumably you only include persons 18-63 (rather than 18-64, which is a more typical age range) because you require two years of continuous eligibility during the follow-up period and you wanted to limit the study to working-age adults. However, it is unclear why you limited the age range to working-age adults given that your data source includes both privately insured persons and Medicare Advantage members (according to lines 95-98 on page 6). Over 85% of Medicare Advantage members were >=65 during the period you describe, and previous research indicates that TRD is a burden in the Medicare population >=65 years of age (see Pilon et al 2019, "Burden of treatment-resistant depression in Medicare: A retrospective claims database analysis," PLOS ONE and Benson et al 2020, "An evaluation of the clinical and economic burden among older adult Medicare-covered beneficiaries with treatment resistant depression," Am J of Geriatric Psychiatry). Further, Medicare beneficiaries under age 65 are likely to be unlike the privately insured persons that are included in your data: They're only eligible because they are receiving Social Security Disability Insurance (SSDI) payments or were diagnosed with end-stage renal disease (ESRD) or amyotrophic lateral sclerosis (ALS). Given all of this, please do one of the following: expand your age range or exclude Medicare Advantage beneficiaries from your analysis.

6) Methods: If you expand your age range and retain Medicare Advantage beneficiaries in your analysis, please include the insurance type (MA or private) in the characteristics that were assessed and compared between the two groups. Please also include this in the propensity score matching -- the costs and patient characteristics would be expected to differ greatly for the two groups.

7) Page 6 line 107: Please provide a citation for the UnitedHealthcare Guidelines. It is unusual to include adjustment disorder diagnoses within the major depression disorder diagnostic group - explain/justify.

8) Page 8 line 135-136 - Please include a citation justifying your inclusion of non-antidepressant medications (it is justifiable, but there should be a citation)

9) Page 8 line 137-139 - Please create a supplemental file that defines the specific recommended minimal dosages for each medication. Doing so is consistent with past claims-based research on the same topic, and it enables other researchers to replicate and/or build on your study

10) Page 8 line 149 - your age ranges in this sentence include persons 55-64 (but you excluded 64 year olds) and >=65 year olds (but these persons were not included in the study). Please update this language as needed depending on how you approach the change requested in #5 above.

11) Page 9 line 159 through 163 - Please clarify what is included in "outpatient visits" -- is this ambulatory hospital outpatient visits, office-based visits, or both? Also, clarify "medical claims" given that outpatient, inpatient, ED and pharmacy are listed separately.

12) Page 9 line 164 through 165 - You say that patient costs were the sum of deductibles, copayments and coinsurance and you mention procedures. Are patient costs for prescription pharmaceuticals included? If so, I recommend that you reword: Patient costs were defined as the sum of deductibles, copayments and coinsurance for all medical and pharmacy services and supplies paid through patients' insurance benefits (or something similar). If not, please explain the decision to exclude out of pocket pharmacy costs.

13) Page 8 line 141 through 144: Please see feedback #4 above regarding the definition of "adequate duration." This operational definition is questionable; a change is needed.

14) Page 8 line 150 - patients often have multiple types/categories of depression diagnoses in claims, even in a short period of time. If a person had >1 depression diagnosis within 30 days of the index date, how were they categorized into a single group?

15) Page 7 line 152 - please provide a citation for the version of the Elixhauser comorbidity score that you used

16) Page 11 - the propensity score-matched data differed on Elixhauser score, suggesting that the propensity score matching was not wholly successful in rendering the two groups similar in terms of the important variables associated with costs. This is a significant issue. Please justify the decision to adjust for this variable rather than tightening the matching logic for the propensity score matching, or rework to tighten the matching logic. If not reworked, please include a discussion of this issue in the limitations section.

17) Page 21 line 294 - You describe the high unemployment rates of persons with TRD, but at the same time your sample primarily consists of persons with employer sponsored insurance (as that is what is most prevalent in the Optum data) and you require continuous enrollment in the health plan (and thus you're requiring continuous employment). Please add this to the limitations section -- your study may not be representative of many persons with TRD. Instead, it represents those able to be continuously employed, which may be those with less severe forms of TRD.

18) Page 21 last paragraph - TRD is defined solely on medications and does not take into account other treatment strategies for depression, including ECT, rTMS, VNS, or psychotherapy. Please add this to the limitations section -- not all TRD may be identified based on a medication-only algorithm.

19) General comment on discussion section - it is unclear what your paper adds to the existing literature given the large number of studies that already explore this topic. Please emphasize what is new/different/notable about your study and explain the importance of the new information provided by your study.

20) General comment on discussion/other sections in terms of references - you do not look to many of the most recent articles on the costs of TRD in your discussion and other sections of the manuscript. See a few listed above, and this is not a complete list. Please update your literature review and update your paper accordingly.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Sep 11;15(9):e0238843. doi: 10.1371/journal.pone.0238843.r002

Author response to Decision Letter 0

23 Jul 2020

Comments

Journal Requirements

Comment 1: Please ensure that your manuscript meets PLOS ONE’s style requirements, including those for file naming. The PLOS ONE style templates can be found at

Response: Formatting updates have been made so that the manuscript meets PLOS ONE style requirements.

Comment 2: PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ.

Response: The corresponding author’s ORCID iD has been verified in Editorial Manager.

Comment 3: We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

*In your revised cover letter, please address the following prompts:

We will update your Data Availability statement on your behalf to reflect the information you provide.

Response: There are legal restrictions on sharing the de-identified data set used in our analysis. The data were made available via a third-party license from Optum; Janssen Pharmaceuticals has a license for analysis of the Optum Clinformatics™ Extended Data Mart (CEDM). As such, we cannot provide the raw data themselves. However, other researchers can access the data by purchase through Optum. Those who are interested may visit https://www.optum.com/solutions/life-sciences.html for more information on accessing Optum CEDM data. We confirm that we had no special privileges to access data from Optum via third-party license, and that other researchers would be able to access the data in the same manner.

Comment 4: Thank you for stating the following in the Competing Interests section:

“Authors GL, LZ, AD, GW, JS, KL, and QZ are employees of Janssen Research & Development, LLC and hold stock in the company. JR is in research collaboration with Janssen, AstraZeneca, Abbvie, and Pfizer, for which Karolinska Institutet has received grant support. JR has been a speaker for Eli Lilly and received unrestricted grant support from Schering-Plough. Wilson Joe, PhD, of MedErgy, provided editorial support for this manuscript. Editorial support was funded by Janssen Research & Development, LLC.”

ii) Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

Response: We have amended the competing interests section of the manuscript to state that “These interests do not alter the authors’ adherence to PLOS ONE policies on sharing data and materials.”

Editor Comments

Comment 1: In your revised manuscript, verify that PLOS ONE formatting requirements are addressed. (https://journals.plos.org/plosone/s/submission-guidelines)- In the text, reference numbers should be enclosed in square brackets (https://journals.plos.org/plosone/s/submission-guidelines#loc-references).

Response: Formatting updates have been made so that the manuscript meets PLOS ONE style requirements. The revised manuscript uses square brackets, rather than parentheses, to enclose in-text citations.

Reviewer #1 Comments

Comment 1: P.9, line157. The authors may want to perform a sensitivity analysis (i.e., 1:1 propensity score matching), given that they used the propensity score at a 1:4 ratio, and the literature suggests that increasing the number of untreated subjects matched to each treated subject tends to increase the bias in the estimated treatment effect (e.g., Austin PC, Am J Epidemiol. 2010).

Response: Thank you very much for pointing out a potential concern with our methodology and for the suggestion to use a 1:1 ratio in propensity score matching (PSM). The results for a 1:1 ratio are consistent with those from a 1:4 ratio for PSM. The results of 1:1 ratio matching are presented in three new tables in the Supporting Information (S6, S7, and S8) as the following:

Comment 2: P.9, line 161. Please clarify whether hospital length of stay (LOS) is total LOS during the follow-up period or LOS per admission.

Response: The LOS is the sum of hospital stay days from all hospitalizations during one-year follow-up period. This is now clarified in the Methods section (p 8-9; new text in red font):

“Number of outpatient visits (which included office based and ambulatory hospital outpatient visits), proportion of patients with emergency department (ED) visits, proportion of patients with hospitalizations, and hospital length of stay (LOS; ie, the sum of hospital stay days from all hospitalizations during the one-year follow-up period) were assessed to measure resource utilization.”

Comment 3: Please clarify the terminology and/or definitions. For example:

Response: Thank you for the suggestion for clarifications. The sentences have been revised as follows (p 9):

“Medical costs to payers included claims for outpatient visits, ED visits, and hospitalizations; pharmacy costs to payers were the sum of pharmacy claims; and total costs to payers were the sum of medical costs and pharmacy costs to payers. Medical costs to patients were defined as the sum of deductibles, copayments, and coinsurance for all medical services; prescription costs to patients were defined as the sum of deductibles, copayments, and coinsurance for all prescription drugs; and total costs to patients were the sum of medical costs and prescription costs to patients. Total healthcare costs were defined as the sum of costs to payers and patients.”

This text has also been added as a footnote below Table 3 for additional clarity.

Response: Thank you for the suggestion; please see the above response.

Response: We agree with the Reviewer and have adjusted the language in the text accordingly (please see p 14-18).

Reviewer #2 Comments

General comment: Thank you for your paper. As you will see below, I have concerns about this study's operational definition of treatment resistant depression (TRD), the inclusion of both privately insured persons and Medicare Advantage beneficiaries without distinguishing between them in the analyses, and the dissimilarity between the two matched groups in terms of the comorbidity index.

Response: Thank you for pointing out the distinction between private versus Medicare Advantage insurance coverage. The focus of this study was on patients aged 18 to 63 years at index (see the abstract, ‘Sample selection and study design’ section, and Table 1). We have clarified this point in the text and limitations section of the Discussion.

The text in the ‘Assessments’ section (p 8) now reads:

“Characteristics included: age group (18–24, 25–34, 35–44, 45–54, and 55–63 years); sex; index year…”

The Discussion now includes the following limitation (p 22):

“Furthermore, results obtained using the Optum ClinformaticsTM database may not generalize beyond patients with employer-sponsored commercial insurance and Medicare Advantage insurance.”

Additionally, we checked the insurance type for the patients in this study and confirmed that only 5 out of 9289 non-TRD patients had both commercial and Medicare insurance; all TRD patients had commercial insurance.

Comment 1: Abstract and title: Please include information about the subjects (e.g., persons with private or Medicare Advantage insurance coverage). Payer and patient expenditures would be expected to vary greatly by insurance type, so this is important information.

Response: Thank you for the suggestion. As described in the previous response, we have confirmed that all the patients in the study cohort had commercial insurance coverage and only 5 non-TRD patients had additional Medicare Advantage coverage. We do not expect this will change our results substantially.

The manuscript title has been revised as follows:

“A retrospective analysis to estimate the healthcare resource utilization and cost associated with treatment-resistant depression in commercially insured US patients”

Additionally, the abstract objective now reads:

“The economic burden of commercially insured patients in the United States with treatment-resistant depression and patients with non–treatment-resistant major depressive disorder was compared using data from the Optum Clinformatics™ claims database.”

Comment 2: Page 4 paragraph 1: The second sentence indicates that the estimated economic burden of MDD in the US in 2010 was US$210.5 billion, but the last sentence indicates that the societal economic burden of MDD in 2012 was $188 billion. This inconsistency is not explained and thus raises questions. Is MDD becoming less economically burdensome or is this due to methodological differences? Please rework this first paragraph so the reader is not distracted by the inconsistency in past research (e.g., provide more information, or the last sentence could become less specific, simply stating that the economic burden of MDD has been estimated to exceed that of cancer and diabetes but providing no specific numbers).

Response: Thank you for bringing this distraction to our attention. The introductory paragraph has been reworked and now reads as follows (p 4):

“Depression is a widespread, severely disabling disorder associated with impaired daily functioning, diminished quality of life, and increased mortality and healthcare utilization [1-4]. Healthcare costs such as outpatient medical services, pharmaceutical services, and inpatient services as well as indirect costs such as workplace presenteeism and absenteeism all contribute substantially towards the total burden of major depressive disorder [3]. In 2012, the US societal economic burden of major depressive disorder was estimated at $188 billion, which exceeded the US societal burden of cancer ($131 billion) and diabetes ($173 billion) [4].”

Comment 3: Page 4-5 lines 67 through 70: You indicate that reference number 10 (by Al-Harbi) proposes that an adequate duration is treatment for >=4 weeks with >=3 weeks on an adequate dose. Reading this reference, it appears that this suggestion was not Al-Harbi's but was put forth in a paper cited by Al-Harbi in their introduction section (reference #9 in that paper - Thase ME, Rush JA. Treatment-resistant depression. In: Bloom FE, Kupfer DJ, editors. Psychopharmacology. New York, NY: Raven; 1995.). However, Al-Harbi goes on to review the literature and ultimately concludes in the Discussion section that "It seems that depression should only be considered drug resistant after at least 6 weeks of two trials of antidepressant therapy" (page 383). Al-Harbi also discusses the 6 week time period in "Optimization of Antidepressants" section on page 374. Thus, your statement that Al-Harbi proposes a >=4 week time period appears to be inaccurate -- >-4 weeks actually contradicts the conclusions of the review.

Response: There are indeed varying definitions and suggestions for what is an “adequate duration” for a treatment trial. After further critical reading of the Al-Harbi (2012) article, he seems not to have done a thorough review of what duration would be optimal. In fact, it seems somewhat illogical that he refers to both Thase & Rush (1995) and Thase et al. (Thase ME, Blomgren SL, Barkett MA, et al. Fluoxetine treatment of patients with major depressive disorder who failed to initial treatment with sertraline. J Clin Psychiatry. 1997;58:16-21) to support the claim cited above that at least six weeks should be required – because in fact those two references do not at all reflect the full spectrum of studies which have investigated this.

Another recent, and much more thorough, review is by Gaynes et al. 2020 (Gaynes BN, Lux L, Gartlehner G, et al. Defining treatment-resistant depression. Depress Anxiety. 2020;37(2):134-145). They reviewed a large number of studies and wrote “Experts do not agree on how to define an adequate dose and adequate duration. Typically, the minimum duration cited is 4 weeks.” Further, they write (in section 3.2.2): “…we then determined whether the investigators had confirmed the duration: that is, clarified that patients previously received what KQ 1 [KQ1: What definitions of TRD appear in these sources? Do definitions converge on the best one?] had indicated was an adequate dose. In KQ 1, approximately one-half of the eligible reviews and guidelines identified a minimum of 4 weeks of treatment; the other half identified it as 6 weeks. We defined an adequate dose here as 4 weeks because one primary tool to confirm the adequacy of dose and duration, the ATHF, required at least 4 weeks to be considered as adequate duration. Of 185 studies, 146 (79%) considered in their selection criteria whether the patient had been treated previously with an adequate dose; 112 (61%) systematically confirmed that the dose was adequate by specifying dosage levels. Of all 185 studies, 150 (81%) considered in their selection criteria whether prior treatments were of adequate duration; 128 (69%) systematically confirmed that the duration was adequate (≥4 weeks of treatment).”

In conclusion, we no longer consider Al-Harbi to be a preferred reference for “adequate duration” but instead Gaynes et al. 2020, and we have revised the text accordingly (p 4-5):

“Although no consensus definition currently exists, the US Agency for Healthcare Research and Quality (AHRQ) and Food and Drug Administration (FDA) proposed a standard definition of treatment-resistant depression: failing to respond to a minimum of 2 antidepressants administered at an adequate dose, for an adequate duration [8,9]. There is currently no consensus regarding the definitions of ‘adequate’; a recent review on treatment-resistant depression found that most studies considered an adequate treatment duration to last for a minimum of 4 or 6 consecutive weeks, with the majority requiring ≥4 weeks [10,11]. In the current study, we defined adequate dose based upon the American Psychiatric Association Practice Guidelines for Treatment of Patients with Major Depressive Disorder [12] and approved recommended minimal dosage, and adequate duration was defined by an algorithm that required ≥29 days of prescription coverage (details outlined in Materials and Methods below). Results from STAR*D suggest that nonresponse to 2 adequate trials of established pharmacotherapy classes is an inflection point that predicts a poor prognosis with respect to low remission and high relapse rates, and is associated with higher rates of future medication intolerance [5,13].”

Comment 4: Page 5 lines 70 through 73: You say that your methodology for adequate dose is based on the Massachusetts General Hospital (MGH) Antidepressant Treatment Response Questionnaire (ATRQ). However, you do not use the Massachusetts General Hospital Antidepressant Treatment Response Questionnaire's definition of adequate duration: According to the article you cite (#11), "The MGH ATRQ defines 6 weeks on an adequate dose of antidepressant medication as an adequate duration of treatment." Please change your definition accordingly and rework your analyses. Doing so would be consistent both with the Al-Harbi article you cite, the MGH ATRQ that you reference, and numerous recent studies that used claims data to examine the costs of treatment resistant depression, including but not limited to one recently published in PLOS ONE (see #5 below). Others include Amos et al 2018 "Direct and Indirect Cost Burden and Change of Employment Status in Treatment Resistant Depression"; Benson et al 2020, "An evaluation of the clinical and economic burden among older adult Medicare-covered beneficiaries with treatment resistant depression"; Pilon et al 2019, "Medicaid spending burden among beneficiaries with treatment-resistant depression"; Pilon 2019, "US integrated delivery networks perspective on economic burden of treatment resistant depression: retrospective matched cohort study." This is not a complete list - I am only providing a few examples.

Response: In regards to the first part of the Reviewer’s comment (ie, adequate duration), we appreciate the Reviewer pointing out the inaccurate citation of Chandler et al. 2010; in fact, we did not follow the Massachusetts General Hospital Antidepressant Treatment Response Questionnaire. We apologize for the confusion and have revised the text accordingly (p 4-5):

“In the current study, we defined adequate dose based upon the American Psychiatric Association Practice Guidelines for Treatment of Patients with Major Depressive Disorder [12] and approved…”

The duration concern is answered in our response to this Reviewer’s comment #3. Briefly, Al-Harbi was not consistent and did not provide a recommendation, mentioning both 4 and 6 weeks.

To provide some additional context: we started our TRD research in 2016 and explored variations in the definition of ‘adequate duration’ in a feasibility study based on the data available through the end 2015. We set the upper limit as 180 days, i.e., if a patient was on the same treatment for 180 days or fewer, this patient would not be considered to have TRD. This consideration was based on the STAR*D study in which <1% of patients were counted as a treatment failure with a duration of >180 days. We then examined the impact of the choice of the lower limit of the ‘adequate duration’ definition at 14, 28, and 42 days; the result was TRD rates of 11.79%, 10.91%, and 10.17% (please see the following figure). This suggested minimal differences among the choices of the lower limit.

In regards to the second part of the Reviewer’s comment (ie, patient population), this study was an analysis of commercially insured patients (as described in our response to this Reviewer’s comment #1).

Comment 5: Page 6 line 101: Presumably you only include persons 18-63 (rather than 18-64, which is a more typical age range) because you require two years of continuous eligibility during the follow-up period and you wanted to limit the study to working-age adults. However, it is unclear why you limited the age range to working-age adults given that your data source includes both privately insured persons and Medicare Advantage members (according to lines 95-98 on page 6). Over 85% of Medicare Advantage members were >=65 during the period you describe, and previous research indicates that TRD is a burden in the Medicare population >=65 years of age (see Pilon et al 2019, "Burden of treatment-resistant depression in Medicare: A retrospective claims database analysis," PLOS ONE and Benson et al 2020, "An evaluation of the clinical and economic burden among older adult Medicare-covered beneficiaries with treatment resistant depression," Am J of Geriatric Psychiatry). Further, Medicare beneficiaries under age 65 are likely to be unlike the privately insured persons that are included in your data: They're only eligible because they are receiving Social Security Disability Insurance (SSDI) payments or were diagnosed with end-stage renal disease (ESRD) or amyotrophic lateral sclerosis (ALS). Given all of this, please do one of the following: expand your age range or exclude Medicare Advantage beneficiaries from your analysis.

Response: The focus of this study was on privately insured patients. The age restriction used in the analysis would eliminate Medicaid patients. Additional details on the associated clarifications made in the revised manuscript are described in our responses to this Reviewer’s general comment and comment #1.

Comment 6: Methods: If you expand your age range and retain Medicare Advantage beneficiaries in your analysis, please include the insurance type (MA or private) in the characteristics that were assessed and compared between the two groups. Please also include this in the propensity score matching -- the costs and patient characteristics would be expected to differ greatly for the two groups.

Response: We have confirmed that all the patients included in this cohort study had commercial insurance coverage, as described in more detail in our responses to this Reviewer’s general comment and comment #1.

Comment 7: Page 6 line 107: Please provide a citation for the UnitedHealthcare Guidelines. It is unusual to include adjustment disorder diagnoses within the major depression disorder diagnostic group - explain/justify.

Response: Please see reference #24, which is cited at the end of the sentence in question. This reference is also embedded below and accessible via this link: https://www.uhccommunityplan.com/assets/healthcareprofessionals/pharmacyprogram/FL-Pharmacy/ICD-9_Code-Drug_Match_FL_FHK.pdf.

Comment 8: Page 8 line 135-136 - Please include a citation justifying your inclusion of non-antidepressant medications (it is justifiable, but there should be a citation).

Response: We have added two references, Al-Harbi 2012 and Gelenberg et al. 2010 (references #12 and 25), on page 7. Al-Harbi 2012 lists all these drugs as augmentation therapies in Table 3, and the drugs were also discussed in Gelenberg et al. 2010.

Comment 9: Page 8 line 137-139 - Please create a supplemental file that defines the specific recommended minimal dosages for each medication. Doing so is consistent with past claims-based research on the same topic, and it enables other researchers to replicate and/or build on your study.

Response: Such a table was included in the original submission as supporting information. Please see S1 Table (‘List of antidepressant medications and minimum adequate dose’), which includes the following footnotes:

“aStarting doses were based on the recommended starting dose indicated in the American Psychiatric Association (APA) Practice Guidelines for Treatment of Patients with Major Depressive Disorder, 3rd edition, 2010 (https://psychiatryonline.org/pb/assets/raw/sitewide/practice_guidelines/guidelines/mdd.pdf).

bStarting doses for other antidepressant medications not included in the APA Practice Guidelines for Treatment of Patients with Major Depressive Disorder were based on the starting doses indicated in the label (http://www.accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm).

cOther selected medications from the database include an antidepressant-antipsychotic combination treatment indicated for treatment resistant depression, a selected antianxiety agent, and other agents not approved for use in the United States (US Food and Drug Administration. Drugs@FDA: FDA-approved drugs. Available from: https://www.accessdata.fda.gov/scripts/cder/daf/).

dNot applicable; not approved for use in the United States.”

Comment 10: Page 8 line 149 - your age ranges in this sentence include persons 55-64 (but you excluded 64 year olds) and >=65 year olds (but these persons were not included in the study). Please update this language as needed depending on how you approach the change requested in #5 above.

Response: This was a typo and has been corrected. The text in the ‘Assessments’ section (p 8) now reads:

“Characteristics included: age group (18–24, 25–34, 35–44, 45–54, and 55–63 years); sex; index year…”

Comment 11: Page 9 line 159 through 163 - Please clarify what is included in "outpatient visits" -- is this ambulatory hospital outpatient visits, office-based visits, or both? Also, clarify "medical claims" given that outpatient, inpatient, ED and pharmacy are listed separately.

Response: We appreciate this suggestion and the following clarifications have been added (please see p 8-9):

• Outpatient visits included office based and ambulatory hospital outpatient visits based on place of service variable in data

• Medical claims included claims for outpatient visits, ED visits, hospitalization by place of service variable in data

• Pharmacy claims were the prescription claims from pharmacies, excluding in-hospital medication administration

Comment 12: Page 9 line 164 through 165 - You say that patient costs were the sum of deductibles, copayments and coinsurance and you mention procedures. Are patient costs for prescription pharmaceuticals included? If so, I recommend that you reword: Patient costs were defined as the sum of deductibles, copayments and coinsurance for all medical and pharmacy services and supplies paid through patients' insurance benefits (or something similar). If not, please explain the decision to exclude out of pocket pharmacy costs.

Response: Thank you; the text has been modified as you suggested. The text has been revised as follows (p 9):

This text has also been added as a footnote below Table 3 for additional clarity.

Comment 13: Page 8 line 141 through 144: Please see feedback #4 above regarding the definition of "adequate duration." This operational definition is questionable; a change is needed.

Response: Please see our response to this Reviewer’s comments #3 and #4.

Comment 14: Page 8 line 150 - patients often have multiple types/categories of depression diagnoses in claims, even in a short period of time. If a person had >1 depression diagnosis within 30 days of the index date, how were they categorized into a single group?

Response: We ordered the 4 categories by their seriousness, from high to low as: major depressive disorder (ICD-9 296.X), dysthymic disorder (ICD-9 300.X), adjustment disorder (ICD-9 309.X), and depressive disorder NOS (ICD-9 311.X). In case a patient had 2 categories, this patient was assigned the more serious category. Few patients had codes in multiple categories.

Comment 15: Page 7 line 152 - please provide a citation for the version of the Elixhauser comorbidity score that you used.

Response: The reference is: Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43: 1130-1139. This reference is now cited as reference #26.

Comment 16: Page 11 - the propensity score-matched data differed on Elixhauser score, suggesting that the propensity score matching was not wholly successful in rendering the two groups similar in terms of the important variables associated with costs. This is a significant issue. Please justify the decision to adjust for this variable rather than tightening the matching logic for the propensity score matching, or rework to tighten the matching logic. If not reworked, please include a discussion of this issue in the limitations section.

Response: Thank you for the great feedback. The small p-value was due to large sample sizes. The mean differences were 0.1 and 0.07 before and after matching, respectively, and were small. In addition, we checked the quality of matching using ‘standardized difference’, which has become a standard metric replacing the p-value approach. A value of the absolute standardized difference <0.1 is considered as comparable between groups by Austin (Austin, PC, Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples, Statist. Med. 2009; 28:3083–3107 (www.interscience.wiley.com) DOI: 10.1002/sim.3697). The following table suggests that TRD and non-TRD were comparable before matching. The matching did improve the balance among most variables as measured by ‘standardized difference’. Nevertheless, we reported results as planned in the study protocol.

Comment 17: Page 21 line 294 - You describe the high unemployment rates of persons with TRD, but at the same time your sample primarily consists of persons with employer sponsored insurance (as that is what is most prevalent in the Optum data) and you require continuous enrollment in the health plan (and thus you're requiring continuous employment). Please add this to the limitations section -- your study may not be representative of many persons with TRD. Instead, it represents those able to be continuously employed, which may be those with less severe forms of TRD.

Response: We appreciate this comment and have acknowledged this limitation with the following statement (p 21):

“Furthermore, results obtained using the Optum ClinformaticsTM database may not generalize beyond patients with employer-sponsored commercial insurance and Medicare Advantage insurance.”

Comment 18: Page 21 last paragraph - TRD is defined solely on medications and does not take into account other treatment strategies for depression, including ECT, rTMS, VNS, or psychotherapy. Please add this to the limitations section -- not all TRD may be identified based on a medication-only algorithm.

Response: We used a medication-based algorithm to mimic the TRD definition that has been applied by the US FDA and EU EMA in their current regulatory reviews of TRD treatments. For patients with TRD, psychotherapy might not be sufficient to control the depressive symptom without an antidepressant treatment, so we did not count psychotherapy alone as a line of treatment. We consider this is a conservative approach. The non-pharmacologic procedures like ECT, rTMS, and VNS, are usually used in the later lines after multiple pharmacologic treatment failures. We expect most of these patients have been captured in our study cohort who have failed at least two antidepressant treatments in the current depression episode.

Comment 19: General comment on discussion section - it is unclear what your paper adds to the existing literature given the large number of studies that already explore this topic. Please emphasize what is new/different/notable about your study and explain the importance of the new information provided by your study.

Response: This is a very good suggestion. We have emphasized an important aspect of our analysis, which is the data on patients’ out-of-pocket healthcare costs. The paragraph in the Discussion section on page 21 now reads:

“In contrast to most previous studies, which focused on healthcare costs reimbursed by payers, the current study also examined patients’ out-of-pocket costs. Patients with treatment-resistant depression had out-of-pocket medical and pharmacy costs of US$1,323 in year 1 and US$1,254 in year 2. Compared with non–treatment-resistant major depressive disorder, costs for treatment-resistant depression represented increases of US$354 in year 1 and US$184 in year 2. These costs are likely to represent a substantial burden for many patients with treatment-resistant depression. In the STAR*D study, participants reported high unemployment rates, ranging from 36% for patients who responded to step 1 treatment to 47% for patients who progressed to step 4 [5]. In another STAR*D analysis, patients with treatment-resistant depression demonstrated lower vocational productivity than patients with non–treatment-resistant major depressive disorder [31]. A claims-based study found that employees with treatment-resistant depression had an average of 35.8 work loss days per year, which was 1.7 times the rate of work loss days in employees with non–treatment-resistant major depressive disorder and 6.2 times that of those without major depressive disorder [14]. Thus, the higher out-of-pocket healthcare costs associated with treatment-resistant depression shown in the current analysis may represent a considerable financial hardship for this vulnerable population.”

Comment 20: General comment on discussion/other sections in terms of references - you do not look to many of the most recent articles on the costs of TRD in your discussion and other sections of the manuscript. See a few listed above, and this is not a complete list. Please update your literature review and update your paper accordingly.

Response: We appreciate this comment and have added recent articles describing costs related to TRD to the Introduction and Discussion sections of our manuscript. These articles are:

20. Pilon D, Sheehan JJ, Szukis H, Singer D, Jacques P, Lejeune D, et al. Medicaid spending burden among beneficiaries with treatment-resistant depression. J Comp Eff Res. 2019;8: 381-392.

21. Pilon D, Joshi K, Sheehan JJ, Zichlin ML, Zuckerman P, Lefebvre P, et al. Burden of treatment-resistant depression in Medicare: a retrospective claims database analysis. PLoS One. 2019;14: e0223255.

22. Pilon D, Szukis H, Joshi K, Singer D, Sheehan JJ, Wu JW, et al. US integrated delivery networks perspective on economic burden of patients with treatment-resistant depression: a retrospective matched-cohort study. Pharmacoecon Open. 2020;4: 119-131.

The Introduction has been updated as follows (p 5):

“According to three recent estimates, per-patient per-year direct healthcare costs in patients with treatment-resistant depression versus non–treatment-resistant major depressive disorder were US$6,709 higher among commercially-insured patients, $4,382 higher among Medicaid-insured patients, and $9,479 higher among US integrated delivery network–insured patients [14,20,22].”

The Discussion has been updated as follows (p 19-20):

“In this study, mean total healthcare costs to payers in year 1 and year 2, respectively, were 45% and 27% higher for patients with treatment-resistant depression compared with those with non–treatment-resistant major depressive disorder. This result is consistent with prior work across a range of patient populations, which found a 25%-134% higher burden among those with treatment-resistant depression versus those with non–treatment-resistant major depressive disorder [14-22].”

Attachment

Submitted filename: Response to Reviewers.docx

Click here for additional data file.^{(253.1KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0238843.r003

Decision Letter 1

Fernando A Wilson

26 Aug 2020

A retrospective analysis to estimate the healthcare resource utilization and cost associated with treatment-resistant depression in commercially insured US patients

PONE-D-19-34560R1

Dear Dr. Li,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Fernando A. Wilson, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

**********

6. Review Comments to the Author

Reviewer #1: [1] Major

None.

[2] Minor

None. All comments have been adequately addressed in the revised text and response letter.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

PLoS One. doi: 10.1371/journal.pone.0238843.r004

Acceptance letter

Fernando A Wilson

1 Sep 2020

PONE-D-19-34560R1

A retrospective analysis to estimate the healthcare resource utilization and cost associated with treatment-resistant depression in commercially insured US patients

Dear Dr. Li:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Fernando A. Wilson

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Appendix. High dimensional covariate selection approaches.

(DOCX)

Click here for additional data file.^{(99.5KB, docx)}

S1 Fig. Patient disposition.

(DOCX)

Click here for additional data file.^{(93.7KB, docx)}

S1 Table. List of antidepressant medications and minimum adequate dose.

(DOCX)

Click here for additional data file.^{(35.3KB, docx)}

(DOCX)

Click here for additional data file.^{(34.6KB, docx)}

(DOCX)

Click here for additional data file.^{(35.1KB, docx)}

(DOCX)

Click here for additional data file.^{(36.6KB, docx)}

S5 Table. Healthcare resource utilization per year during the study period (matched at 1:1 ratio).

(DOCX)

Click here for additional data file.^{(33.3KB, docx)}

S6 Table. Comparison of costs per year between treatment-resistant depression and non–treatment-resistant major depressive disorder patients (US$; matched at 1:1 ratio)^a.

(DOCX)

Click here for additional data file.^{(34.4KB, docx)}

Attachment

Submitted filename: Response to Reviewers.docx

Click here for additional data file.^{(253.1KB, docx)}

Data Availability Statement

[pone.0238843.ref001] 1.Hasin DS, Sarvet AL, Meyers JL, Saha TD, Ruan WJ, Stohl M, et al. Epidemiology of adult DSM-5 major depressive disorder and its specifiers in the United States. JAMA Psychiatry. 2018;75: 336–346. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238843.ref002] 2.Eaton WW, Martins SS, Nestadt G, Bienvenu OJ, Clarke D, Alexandre P. The burden of mental disorders. Epidemiol Rev. 2008;30: 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238843.ref003] 3.Greenberg PE, Fournier AA, Sisitsky T, Pike CT, Kessler RC. The economic burden of adults with major depressive disorder in the United States (2005 and 2010). J Clin Psychiatry. 2015;76: 155–162. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref004] 4.Mrazek DA, Hornberger JC, Altar CA, Degtiar I. A review of the clinical, economic, and societal burden of treatment-resistant depression: 1996–2013. Psychiatr Serv. 2014;65: 977–987. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref005] 5.Rush AJ, Trivedi MH, Wisniewski SR, Nierenberg AA, Stewart JW, Warden D, et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am J Psychiatry. 2006;163: 1905–1917. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref006] 6.Murphy JA, Sarris J, Byrne GJ. A review of the conceptualisation and risk factors associated with treatment-resistant depression. Depress Res Treat. 2017:4176825. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238843.ref007] 7.Ballenger JC. Clinical guidelines for establishing remission in patients with depression and anxiety. J Clin Psychiatry. 1999;60: 29–34. [PubMed] [Google Scholar]

[pone.0238843.ref008] 8.US Food and Drug Administration. Major depressive disorder: developing drugs for treatment. Guidance for industry. Available from: https://www.fda.gov/media/113988/download

[pone.0238843.ref009] 9.Gaynes BN, Asher G, Gartlehner G, Hoffman V, Green J, Boland E, et al. Definition of treatment-resistant depression in the Medicare population. Technology assessment program. Project ID: PSYT0816. (Prepared by RTI–UNC Evidence-Based Practice Center under Contract No. HSA290201500011I_HHSA29032006T). Rockville: Agency for Healthcare Research and Quality; 2018. [PubMed]

[pone.0238843.ref010] 10.Thase ME, Blomgren SL, Birkett MA, Apter JT, Tepner RG. Fluoxetine treatment of patients with major depressive disorder who failed initial treatment with sertraline. J Clin Psychiatry. 1997;58: 16–21. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref011] 11.Gaynes BN, Lux L, Gartlehner G, Asher G, Forman-Hoffman V, Green J, et al. Defining treatment-resistant depression. Depress Anxiety. 2020;37: 134–145. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref012] 12.Gelenberg AJ, Freeman MP, Markowitz JC, Rosenbaum JF, Thase ME, Trivedi MH, et al. Practice guideline for the treatment of patients with major depressive disorder. 3rd ed American Psychiatric Association; 2010. [Google Scholar]

[pone.0238843.ref013] 13.Conway CR, George MS, Sackeim HA. Toward an evidence-based, operational definition of treatment-resistant depression: when enough is enough. JAMA Psychiatry. 2017;74: 9–10. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref014] 14.Amos TB, Tandon N, Lefebvre P, Pilon D, Kamstra RL, Pivneva I, et al. Direct and indirect cost burden and change of employment status in treatment-resistant depression: a matched-cohort study using a US commercial claims database. J Clin Psychiatry. 2018;79: 17m11725. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref015] 15.Gibson TB, Jing Y, Smith Carls G, Kim E, Bagalman JE, Burton WN, et al. Cost burden of treatment resistance in patients with depression. Am J Manag Care. 2010;16: 370–377. [PubMed] [Google Scholar]

[pone.0238843.ref016] 16.Greenberg P, Corey-Lisle PK, Birnbaum H, Marynchenko M, Claxton A. Economic implications of treatment-resistant depression among employees. Pharmacoeconomics. 2004;22: 363–373. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref017] 17.Ivanova JI, Birnbaum HG, Kidolezi Y, Subramanian G, Khan SA, Stensland MD. Direct and indirect costs of employees with treatment-resistant and non-treatment-resistant major depressive disorder. Curr Med Res Opin. 2010;26: 2475–2484. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref018] 18.Olchanski N, McInnis Myers M, Halseth M, Cyr PL, Bockstedt L, Goss TF, et al. The economic burden of treatment-resistant depression. Clin Ther. 2013;35: 512–522. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref019] 19.Olfson M, Amos TB, Benson C, McRae J, Marcus SC. Prospective service use and health care costs of Medicaid beneficiaries with treatment-resistant depression. J Manag Care Spec Pharm. 2018;24: 226–236. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238843.ref020] 20.Pilon D, Sheehan JJ, Szukis H, Singer D, Jacques P, Lejeune D, et al. Medicaid spending burden among beneficiaries with treatment-resistant depression. J Comp Eff Res. 2019;8: 381–392. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref021] 21.Pilon D, Joshi K, Sheehan JJ, Zichlin ML, Zuckerman P, Lefebvre P, et al. Burden of treatment-resistant depression in Medicare: a retrospective claims database analysis. PLoS One. 2019;14: e0223255. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238843.ref022] 22.Pilon D, Szukis H, Joshi K, Singer D, Sheehan JJ, Wu JW, et al. US integrated delivery networks perspective on economic burden of patients with treatment-resistant depression: a retrospective matched-cohort study. Pharmacoecon Open. 2020;4: 119–131. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238843.ref023] 23.Fostick L, Silberman A, Beckman M, Spivak B, Amital D. The economic impact of depression: resistance or severity? Eur Neuropsychopharmacol. 2010;20: 671–675. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref024] 24.UnitedHealthcare. Guideline: ICD-9 code–drug match POS prior authorization. Available from: https://www.uhccommunityplan.com›icd-9_code-drug_match_fl_fhk

[pone.0238843.ref025] 25.Al-Harbi KS. Treatment-resistant depression: therapeutic trends, challenges, and future directions. Patient Prefer Adherence. 2012;6: 369–388. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238843.ref026] 26.Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43: 1130–1139. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref027] 27.SAS Institute Inc. SAS/STAT^® 9.1 user’s guide. Cary: SAS Institute Inc.; 2004. [Google Scholar]

[pone.0238843.ref028] 28.Schneeweiss S, Rassen JA, Glynn RJ, Avorn J, Mogun H, Brookhart MA. High-dimensional propensity score adjustment in studies of treatment effects using health care claims data. Epidemiology. 2009;20: 512–522. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238843.ref029] 29.Reutfors J, Andersson TM, Brenner P, Brandt L, DiBernardo A, Li G, et al. Mortality in treatment-resistant unipolar depression: a register-based cohort study in Sweden. J Affect Disord. 2018;238: 674–679. [DOI] [PubMed] [Google Scholar]

[pone.0238843.ref030] 30.Kubitz N, Mehra M, Potluri RC, Garg N, Cossrow N. Characterization of treatment resistant depression episodes in a cohort of patients from a US commercial claims database. PLoS One. 2013;8: e76882. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238843.ref031] 31.DiBernardo A, Lin X, Zhang Q, Xiang J, Lu L, Jamieson C, et al. Humanistic outcomes in treatment resistant depression: a secondary analysis of the STAR*D study. BMC Psychiatry. 2018;18: 352. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

A retrospective analysis to estimate the healthcare resource utilization and cost associated with treatment-resistant depression in commercially insured US patients

Gang Li

Ling Zhang

Allitia DiBernardo

Grace Wang

John J Sheehan

Kwan Lee

Johan Reutfors

Qiaoyi Zhang

Roles

Abstract

Objective

Methods

Results

Conclusion

Introduction

Materials and methods

Data source

Sample selection and study design

Treatment-resistant depression

Assessments

Statistical analysis

Results

Patient disposition and baseline characteristics

Table 1. Demographic and clinical characteristics of patients with treatment-resistant depression and non–treatment-resistant depression before and after the propensity score matching.

Healthcare utilization

Table 2. Healthcare resource utilization per year during the study period.

Healthcare costs

Table 3. Comparison of costs per year between treatment-resistant depression and non–treatment-resistant major depressive disorder patients (US$)a.

Discussion

Conclusions

Supporting information

Data Availability

Funding Statement

References

Decision Letter 0

Fernando A Wilson

Roles

Author response to Decision Letter 0

Decision Letter 1

Fernando A Wilson

Roles

Acceptance letter

Fernando A Wilson

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Table 3. Comparison of costs per year between treatment-resistant depression and non–treatment-resistant major depressive disorder patients (US$)^a.