Treatment and Healthcare Utilization Among Commercially Insured Patients With Major Depressive Disorder

Among commercially insured patients with major depressive disorder (MDD), desvenlafaxine use was associated with lower MDD-specific and total costs of care compared with duloxetine and venlafaxine.
Published Online: June 04, 2013
Debra Eisenberg, PhD; Keith E. Isenberg, MD; Tao Gu, PhD; Jose Ma J. Alvir, DrPH; Cedric O’Gorman, MD; and Robert J. Sanchez, PhD
Background: The antidepressants desvenlafaxine, escitalopram, duloxetine, and venlafaxine XR have been shown to effectively treat major depressive disorder (MDD).

Objectives: To evaluate treatment patterns, healthcare resource utilization, and cost associated with use of these agents to treat depression.

Study Design: Retrospective cohort study utilizing administrative medical and pharmacy claims data.

Methods: Patients with 1 or more medical claims with the diagnosis of MDD and 1 or more pharmacy claims for antidepressant medication were identifi ed between January 1, 2009, and November 30, 2009. Patients were followed longitudinally for 12 months before and after the index date (date of fi rst antidepressant prescription claim).

Results: A total of 45,913 patients were included in the analysis (desvenlafaxine 6.5%, escitalopram 47.1%, duloxetine 25.4%, venlafaxine XR 21.0%). MDD-specifi c healthcare cost was lowest for desvenlafaxine ($1948), followed by escitalopram ($2065), duloxetine ($2867), and venlafaxine ($3078). Likewise, total cost of care was lowest for desvenlafaxine ($11,480), followed by escitalopram ($11,879), venlafaxine ($13,380), and duloxetine ($16,384). Generalized linear models that controlled for clinical and demographic variables (reference drug, desvenlafaxine) indicated a signifi cant difference in MDD-related cost (duloxetine cost ratio [CR] = 1.270, venlafaxine CR = 1.388; P <.0001 for both) and total cost (duloxetine CR = 1.139, venlafaxine CR = 1.051; P <.05 for both; escitalopram CR = 0.946, P = .0017).

Conclusion: In this population, there appears to be an association between use of desvenlafaxine and lower postindex MDDspecifi c and total cost of care, compared with use of duloxetine and venlafaxine.

Am J Pharm Benefits. 2013;5(3):e55-e65
This study evaluated treatment patterns, healthcare resource utilization, and cost associated with desvenlafaxine, escitalopram, duloxetine, and venlafaxine XR for treatment of major depressive disorder (MDD).

  • Venlafaxine XR and desvenlafaxine were associated with the highest and lowest costs, respectively, for 12 months of MDD-related healthcare expenses, with no apparent difference in the course of the depressive episode.

  • In all cohorts except escitalopram, cost of mental health prescription medication was greater than cost of MDD-related medical treatment.

  • MDD-related medical expenditures decreased preindex to postindexfor the desvenlafaxine and duloxetine cohorts, demonstrating a positive impact of treatment on costs.
Major depressive disorder (MDD) is a disabling condition that signifi cantly impacts overall health and healthcare resource utilization and cost. While part of the burden of care is due to the fact that only half of patients are accurately diagnosed, other factors such as increased comorbid conditions and type of antidepressant treatment play a role in the utilization and cost burden.1,2 Along with the direct costs of office visits, diagnostic tests, treatment, and medications, MDD is associated with indirect costs such as marked reductions in productivity and high absentee rates.3

In a large employer-based study, an evaluation of the direct (pharmacy utilization, professional and outpatient visits) and indirect (absenteeism and presenteeism) costs of disease found that the combination of depression, anxiety, and emotional disorders was more costly than any of the traditional cost drivers including coronary artery disease, diabetes, asthma, chronic obstructive pulmonary disease, and congestive heart failure.4 Other emotional and physical symptoms (eg, cognitive impairment, headache, abdominal pain, fatigue) often occur in conjunction with depression, further increasing healthcare costs.5,6 Previous research estimated that workers with depression cost $31 billion in lost productivity compared with workers without depression.7

Patients with depression have been shown to have higher incidences of comorbid conditions, including anxiety disorders, which add to their overall healthcare costs.8,9 There have also been differences in healthcare utilization and costs in patients with well-controlled depression versus those with uncontrolled depression, thus suggesting the importance of selecting the proper antidepressant for symptom control. Because of these issues, total healthcare costs may be as much as 50% higher in people with depression than in their nondepressed counterparts.10

The elevated healthcare costs associated with MDD underscore the importance of treatment choice, and a rapid and robust response to a therapy is likely to decrease the risk of treatment discontinuation or failure.11 In a study of escitalopram versus 3 generic selective serotonin uptake inhibitors (SSRIs) for the treatment of MDD, Esposito and colleagues11 hypothesized that the higher treatment costs associated with escitalopram would be offset by lower total healthcare costs. The results demonstrated that despite significantly higher pharmacy and inpatient costs of escitalopram compared with the other SSRIs, escitalopram was associated with lower total healthcare costs than the generic SSRIs.11

While depression can be successfully treated with a variety of antidepressant medication options, few studies have evaluated the impact of the various brand treatment options on healthcare resource utilization and overall health expenditures. Therefore, the purpose of this study was to assess treatment patterns, including patient adherence and persistence, and healthcare resource utilization and cost for patients using desvenlafaxine, escitalopram, duloxetine, and extended-release venlafaxine (venlafaxine XR) for the treatment of MDD.


This retrospective cohort study utilized administrative medical and pharmacy claims data from the HealthCore Integrated Research Database, which includes longitudinal claims data for 14 geographically dispersed health plans in the United States, covering approximately 35.5 million people. The study population consisted of patients aged 18 to 64 years with at least 1 medical claim for MDD (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] codes 296.2x-296.3x) any time within the study period and at least 1 prescription claim for desvenlafaxine, escitalopram, duloxetine, or venlafaxine XR from January 1, 2009, through November 30, 2009. All patients were required to have continuous eligibility for a period of 12 months before and 12 months after the index date, which was defined as the date of the first claim for antidepressant medication. Patients were excluded if they had bipolar disorder (ICD-9-CM codes 296.0x, 296.1x, 296.4x-296.9x) at any time during the index period or if they had multiple study medications prescribed on the index date.

Patients were characterized by demographics and physician specialty (psychiatrist vs nonpsychiatrist). Baseline comorbidities were assessed by using the Deyo-Charlson Comorbidity Index (DCI). Treatment patterns, including adherence and persistence with prescribed treatment regimens, were assessed during the postindex period. The proportions of patients with different treatment patterns were calculated and compared between the index drug and all other antidepressants in the postindex period. “New starts” were defined as patients who had no antidepressant medication in the preindex period; “continued” was defined as patients who were on index therapy for the entire 12-month postindex period with no augmentation; “discontinued” was defined as patients who stopped use of their index therapy and did not have any subsequent antidepressants during the 12-month postindex period; “switched” was defined as patients who stopped use of their index medication and changed to anonindex antidepressant medication (first switch only); and “augmented” was defined as patients who added an SSRI, selective serotonin-norepinephrine reuptake inhibitor (SNRI), or atypical antipsychotic medication without stopping the index drug use (first augment only; as adapted by Connolly and Thase12).

Adherence was defined as the extent to which patients took medications as prescribed by their healthcare provider with respect to medication timing, dosage, and frequency, and was measured using the proportion of days covered (PDC).13 The index drug PDC was defined as the number of days with drug on hand divided by 365 X 100. Proportion of days covered is an indicator of the level of patient adherence during a time period defined by the study. Patients were classified as adherent (PDC >80%) or nonadherent (PDC <80%).

To assess the impact of depression medications on total healthcare and MDD-related resource utilization and cost, healthcare costs were identified using the administrative claims for inpatient hospitalization, emergency department visits, outpatient services, physician office visits, psychiatrist office visits, other mental health provider visits, long-term care facility services, and pharmacy claims during the postindex period. The MDD-related resource utilization and cost were defined as any utilization or cost claim with an ICD-9-CM code for MDD (296.2x-296.3x). Total healthcare was defined as a summation of all healthcare resources utilized in the specified time frame (eg, preindex or postindex period). All costs were adjusted to 2010 US dollars.

Statistical Analysis

Descriptive statistics, means (standard deviation [SD]), median, and relative frequencies were reported for continuous and categorical data. Continuous characteristics were compared using appropriate analysis of variance methodology (both parametric and nonparametric). Categorical variables were compared using x2 tests. For all analyses, a 2-tailed level of significance was set at the .05 level.

Generalized linear models (GLMs) were used to examine the association between healthcare utilization (including office visits, other outpatient services, hospital inpatient admission, emergency department visits, and long-term care services) and the selected antidepressant medications while controlling for other patient demographic and clinical characteristics. Additionally, GLMs with the log link function and a gamma distribution were used to examine the association between healthcare costs and selected antidepressant medications while controlling for other patient demographic and clinical characteristics, and preindex medical spending.


Patient Characteristics

From an initial number of 1,473,481 patients with at least 1 pharmacy claim for any antidepressant drug, 45,913 patients met all study eligibility requirements (Figure 1). Of these, 2997 (6.5%) were indexed on desvenlafaxine, 21,615 (47.1%) on escitalopram, 11,674 (25.4%) on duloxetine, and 9627 (21.0%) on venlafaxine XR (Table 1).

The overall mean preindex comorbidity score was 0.48 (SD 1.14), with the highest mean DCI score (0.58 ± 1.20) seen in the duloxetine cohort (P <.0001 between all cohorts) (Table 1) Across all cohorts, 32.3% of patients had comorbid depression (excluding MDD) with the highest percentage in the desvenlafaxine cohort (39.3%) (P <.0001 between all cohorts). Also, within the total population, 25.7% of patients had anxiety disorder, 5.0% had substance abuse disorder, and 4.0% had insomnia. For non–mental health comorbidities within the total population, 7.2% had congestive heart failure, 10.3% had diabetes mellitus, 35.8% had hypertension, and 8.4% had fibromyalgia. The duloxetine cohort had a higher rate of congestive heart failure (9.2%), diabetes mellitus (13.6%), hypertension (42.5%), fibromyalgia (14.9%), and diabetic peripheral neuropathy (1.9%) compared with the other cohorts (P <.0001 between all cohorts).

PDF is available on the last page.
Recommended Articles
Measure definitions to estimate the effects of cost sharing on adherence to medications prescribed together (ie, concurrent adherence) are lacking.
The study found that the geriatric population in the United States receives prescriptions for mental health drugs at more than twice the rate that younger adults do, but they present a lower rate of seeking psychiatric care.
Healthcare spending growth between 2014 and 2024 is projected to be substantially lower than the 3 decades prior to 2008, according to a new report from CMS. In addition, the average premium for a basic Medicare Part D prescription plan will remain stable in 2016.
Patient characteristics and product mix play an important role in determining the overall use and costs of comparable insulins across different manufacturers.
Urologists at Cancer Research UK have identified 5 distinct genomic signatures in prostate cancer that can have important implications on treatment decisions.