Major depressive disorder (MDD) often presents as a chronic and recurrent illness. It is the most common type of depression, affecting 6.6% of adults in the United States annually, with a lifetime prevalence of 16.2%.1 Patients with MDD are more likely to suffer from multiple comorbidities, including other mental illnesses or chronic conditions, compared with patients without MDD.2 MDD is also associated with increased risk for substance abuse, particularly alcohol abuse, leading to more severe depressive symptoms and impairment.3 Because of its clinical features, untreated or undertreated MDD can have significant negative impacts on an individual’s health, quality of life, and functional status, as well as work productivity and employment.2,4
Consequently, patients with MDD are among the highest users of healthcare and incur substantial indirect costs resulting from premature deaths, reduced productivity, and increased disability associated with the disease. 2,5 The economic burden of MDD was estimated to have amounted to $83.1 billion in 2000 in the United States.5 Of this total cost, $26.1 billion (31%) was for direct medical costs, $5.4 billion (7%) was for suiciderelated mortality costs, and the remaining $51.5 billion (62%) was for work-related costs.5
Effective treatment can substantially reduce the economic burden assumed by the healthcare system, employers, and society. Pharmacotherapy is a major type of treatment for MDD. Currently, the most widely prescribed antidepressants in the United States are selective serotonin reuptake inhibitors (SSRIs). These secondgeneration antidepressants have been shown to be effective and better tolerated than older antidepressants.6
Both escitalopram (Lexapro) and citalopram (Celexa) are SSRIs, and escitalopram is the pure S-enantiomer of the racemic citalopram. An enantiomer is composed of 2 compounds that are nonsuperimposable mirror images of each other. Citalopram is a racemate that comprises a 1:1 mixture of S(+)-enantiomer (escitalopram) and an R(−)-enantiomer (R-citalopram). It is the S-enantiomer that possesses the pharmacologic effect of the drug. The R-enantiomer in citalopram counteracts the activity of the S-enantiomer, which may be the underlying reason for the differences in the pharmacologic and clinical effects between escitalopram and citalopram.7-9
Citalopram was approved by the US Food and Drug Administration (FDA) for acute and maintenance treatment of adults with MDD in July 1998, and escitalopram was approved for acute and maintenance treatment of MDD in adults in August 2002. Citalopram became generic in October 2004.
Because citalopram and escitalopram are chemically related, payers may assume that they provide similar clinical benefits, and that the lower acquisition cost of generic citalopram will lead directly to reductions in overall treatment costs. However, citalopram and escitalopram are different medications and may not be interchangeable.
Many clinical trials and analyses have compared citalopram and escitalopram, and have consistently shown that escitalopram is more efficacious.10-14 In one study, escitalopram 10 mg/day was found to be at least as effective as citalopram 40 mg/day in an 8-week study of 491 outpatients with MDD.10 A placebo-controlled trial of 469 patients with moderate-to-severe depression revealed that by the end of 8 weeks of treatment, significantly more patients responded to escitalopram 10 to 20 mg daily than to citalopram 20 to 40 mg daily.11 Continuation of treatment with escitalopram has also demonstrated efficacy in preventing relapse of MDD.12,13 A double-blind trial of 280 patients with MDD randomized to escitalopram 20 mg/day or to citalopram 40 mg/day over 8 weeks demonstrated significant improvements in depression rating scales and in treatment response rates with escitalopram.14 Both groups demonstrated similar tolerability.14
Furthermore, an economic evaluation that was conducted parallel with a double-blind, randomized clinical trial in France showed that escitalopram is more costeffective than citalopram.12 Although conducted outside the United States, this economic evaluation has some value in understanding the relative cost-effectiveness of these 2 treatments, especially when similar evaluation is lacking in this country. Among ambulatory care patients with MDD followed for 8 weeks, escitalopram 20 mg/day with citalopram 40 mg/day.12 The cost differential was mainly a result of lower hospitalization costs.12
Driven by cost-containment considerations, healthcare systems have adopted a variety of policies designed to encourage physicians and patients toward lower-cost drugs, including use of generic drugs when they become available. However, without a comprehensive understanding of the total healthcare cost implications of these 2 antidepressants in the real world, promoting or mandating citalopram in drug formularies to replace escitalopram based solely on acquisition costs may not be cost-effective.
A comparison of the economic outcomes of patients treated with escitalopram and citalopram is valuable for decision makers to consider the total costs of MDD treatment. Although some published claims studies have compared escitalopram with other SSRIs as a group,8,9 to our knowledge very few real-world studies have compared escitalopram with citalopram—2 antidepressants that are often assumed to be equivalent as a result of their molecular structure. One analysis based on claims data recently compared clinical and economic profiles of escitalopram and citalopram in a geriatric population with MDD.15 The objective of this current study is to generalize this real-world comparison with adult patients with MDD.
In this current study, we analyzed claims data from the Ingenix Impact Database, which included complete medical and pharmacy claims for more than 25 million noncapitated managed care lives from more than 35 health plans, covering all census regions of the United States. Data elements used in the present analysis included enrollment records, patient demographics, inpatient and outpatient medical services, and pharmacy dispensing claims. The data collection period started January 1, 2003, because escitalopram was approved by the FDA for MDD treatment in August 2002.
Patients were included in the study if they had at least 1 inpatient claim or 2 medical claims of other types incurred on different dates that were associated with an MDD diagnosis (International Classification of Diseases, Ninth Revision, Clinical Modification 296.2x, 296.3x) between January 1, 2003, and June 30, 2005, and if they filled at least 1 prescription either for escitalopram or for citalopram during that same period. The first prescription filling date was defined as the index date. The antidepressant filled on the index date (either escitalopram or citalopram) was defined as the index drug.
Patients were classified into the citalopram group or the escitalopram group based on the index drug they received. Because this study was intended to compare citalopram and escitalopram, the brand-name and generic forms of citalopram were not distinguished. All patients who received either brand-name or generic citalopram and met the study criteria were included in the same citalopram group. Because the National Committee for Quality Assurance criteria for the appropriate treatment of MDD requires a minimum duration of 6 months, the study period was set to 6 months after therapy initiation.16 The baseline period was defined as the 6 months before the index date, and the study period was defined as the 6 months after the index date. Patients had to meet the following additional inclusion criteria: (1) they were aged ≥18 years as of the index date; (2) they were continuously enrolled in the plan for at least 12 months, including a minimum of 6 months before and 6 months after the index date; (3) they did not use any second-generation antidepressant drug (eg, SSRIs, serotonin-norepinephrine reuptake inhibitors, bupropion, nefazodone, trazodone, fluoxetine-olanza - pine, or mirtazapine) during the baseline period; and (4) they were not initiated on combination therapy (ie, did not use any second-generation antidepressants other than citalopram or escitalopram within 2 weeks after the index date).9,15
Treatment persistence. Treatment persistence was measured up to the time of discontinuation of therapy by the rate of discontinuation of the index drug during the study period. Patients were considered to discontinue the index drug if they did not have another refill within 45 days after the last day of the latest supply of the index drug recorded in the claims database.17 Patients who had discontinued and switched to another second-generation antidepressant were further identified, with switching defined as filling a prescription for another secondgeneration antidepressant within 45 days of the last day of the latest supply of the index drug.
Healthcare resource utilization. Healthcare re source utilization was observed in all-cause hospitalizations and emergency department visits during the study period. In particular, MDD-related hospitalizations and emergency department visits were also measured. A hospitalization or an emergency department visit was considered MDDrelated if it was associated with a primary or secondary diagnosis for MDD. The proportions of patients who used these types of services were estimated, as well as the number of hospitalization days and number of emergency department visits for each patient.
Healthcare costs. Healthcare costs were examined from the insurer’s perspective. Average costs per patient were estimated on the basis of payment data in the claims database for procedures, services, and prescription drugs, over the 6-month study period by the type of service (ie, professional services, including physicians and other medical professionals’ care; hospitalization cost; outpatient cost; emergency department cost; and prescription drug cost).
Total healthcare costs were calculated as the sum of total medical services costs and total prescription drug costs. In addition, MDD-related costs (total and by service type) were also calculated. A medical service cost was considered MDD-related if it was associated with a diagnosis of MDD. MDD-related pharmacy costs included the costs of antidepressants. Because MDD is often associated with multiple comorbidities,2 including other mental illnesses or chronic conditions, it is important to analyze not only MDD-related costs but also the total cost of disease.
Study outcomes were first compared between the citalopram and escitalopram groups using unadjusted analysis. Chi-square tests were used to compare the rates of discontinuation and discontinuation with switching to another second-generation antidepressant, hospitalization, and any emergency department visit during the study period. Wilcoxon tests were used to compare statistical differences in the number of hospitalization days, number of emergency department visits, and healthcare costs between the 2 groups.
The study outcomes were further compared using a multivariate analysis to control for differences in baseline characteristics between the 2 treatment groups. Patient baseline characteristics were measured during the baseline period and included age, sex, the Deyo adaptation of the Charlson Comorbidity Index (CCI),18 individual comorbidities selected from the literature, and baseline healthcare utilization and costs. (The CCI has become a common measure for comorbidity burden in observational studies, especially those using administrative claims databases. This index has been used in the literature related to depression and is intended as a measure of overall health outcomes of the patient, not as a measure of disease severity associated with depression.19)
To compare the rates of discontinuation and discontinuation with switching between the 2 groups, Cox proportional hazard models were used. Logistic regressions were applied to estimate the probability of hospitalization and emergency department visits during the study period. The number of hospitalization days and the number of emergency department visits were compared between the 2 groups using negative binomial regressions. Generalized linear model (GLM) regressions with log link and gamma distribution were used to analyze healthcare costs. For MDD-related total medical services costs, a 2-part model was used, because a large portion of patients had no hospital and medical costs.
To assess the robustness of the study results to variations in the price of antidepressant drugs, it was assumed that all second-generation antidepressants used by patients in the citalopram group had zero cost, whereas the prescription drug cost in the escitalopram group was kept unchanged. Under this assumption, the new total healthcare cost for the citalopram group was calculated and compared with the total healthcare cost of the escitalopram group using both Wilcoxon test and GLM regression as described above. The analysis represented a conservative estimation of the cost impact of escitalopram over citalopram by deducting all antidepressant drug costs from the total healthcare costs for the citalopram group.
A total of 14,677 patients met the inclusion criteria in this study, of which 10,465 were taking escitalopram, and 4212 were taking citalopram. Patient baseline characteristics are presented in Table 1.
Compared with patients using citalopram, patients using escitalopram had similar age (41.7 vs 42.1; P = .0131); fewer females (67.5% vs 69.5%; P = .019); a lower average CCI (0.77 vs 0.90; P = .005); a higher prevalence of generalized anxiety disorder, panic disorder, sleep disorder, and hyperlipidemia; and a lower prevalence of schizophrenia and cancer. In addition, patients taking escitalopram used more prescription drugs at baseline (4.54 vs 4.35; P = .003) and had more physician office visits (94.1% vs 91.0%; P <.001), but had fewer inpatient visits (13.5% vs 15.7%; P = .001), outpatient hospital visits (54.5% vs 58.6%; P <.001), and emergency department visits (22.0% vs 23.5%; P = .040) than patients taking citalopram. In addition, patients taking escitalopram had numerically lower baseline drug, medical, and total costs compared with patients taking citalopram, although the difference was significant only for the drug costs.
During the 6-month study period, 57.8% of patients taking escitalopram and 60.3% of those taking citalopram (P = .006) discontinued the index treatment (Table 2). The rate of discontinuation with switching to another second-generation antidepressant was also lower in the escitalopram group (24.7% vs 29.3%; P <.001). After controlling for differences in baseline characteristics, results of the unadjusted analysis were confirmed for escitalopram by hazard ratio of 0.94 for overall discontinuation (P = .006) and 0.83 for discontinuation and switching to another second-generation antidepressant (P <.001).
Healthcare Resource Utilization
Among patients in the escitalopram group, 11.6% experienced a hospitalization during the study period compared with 13.6% in the citalopram group (P = .001; Table 2). The rates of MDD-related hospitalizations were not significantly different between the 2 groups (4.0% vs 4.4%; P = .256). The escitalopram group also had shorter inpatient stays during the same period (1.0 vs 1.5 hospitalization days; P = .001). Results of the multivariate analysis were consistent with findings from the unadjusted analysis. Patients in the escitalopram group were less likely to be hospitalized (odds ratio [OR] 0.88; P = .036). In other words, the odds of patients in the escitalopram group being hospitalized was 12% lower than for patients in the citalopram group, and patients in the escitalopram group had 28% fewer hospitalization days than patients in the citalopram group (P = .001). The rates of MDDrelated hospitalizations were not significantly different between the 2 groups.
Similarly, there was a smaller proportion of patients with emergency department visits for any reason in the escitalopram group during the study period (19.8% vs 23.0%; P <.001; Table 2). Patients receiving escitalopram were also less likely to have an MDD-related emergency department visit (1.2% vs 1.9%; P = .001). These results remained significant after adjustment for baseline characteristics, both for emergency department visits for any reason (OR 0.83; P <.001) and for MDD-related emergency department visits (OR 0.60; P = .001).
On average, patients in the escitalopram group had 0.3 emergency department visits, whereas those in the citalopram group had 0.4 visits during the study period (P <.001). Results from the negative binomial model showed that patients in the escitalopram group had 19.0% fewer emergency department visits after controlling for differences in baseline characteristics (P <.001).
The average total healthcare cost per patient during the study period was $5551 in the escitalopram group, $1459 less than that in the citalopram group (P <.001; Table 3). Patients in the escitalopram group also had lower costs in every category. Approximately 92% of the difference in total healthcare costs was attributable to reduction in total medical services costs in the escitalopram group (−$1342; P <.001), of which $558 resulted from lower hospitalization costs (P <.001). Significantly lower costs in the escitalopram group were also manifested in outpatient visits (−$273; P <.001) and professional services (−$165; P = .001). Results from multivariate analyses indicate that the total healthcare costs were an average of $1174 less in the escitalopram group (P <.001), and that the average total medical services costs and total prescription drug costs were both significantly lower in the escitalopram group (−$972 and −$170, respectively; both P <.001).
MDD-related total healthcare costs during the study period were also lower in the escitalopram group. On average, patients in this group incurred $117 less per patient during the study period than those in the citalopram group (P <.001). The majority of the difference resulted from reduction in MDD-related total medical services costs (−$77; P = .014), and the rest of the difference was attributable to the lower antidepressant costs (−$40; P <.001). Among MDD-related total medical services costs, hospitalization, outpatient, and emergency department costs were all significantly lower in the escitalopram group. After controlling for differences in baseline characteristics, the adjusted MDD-related total healthcare costs were $109 lower in the escitalopram group (P = .003). The adjusted total antidepressant costs and MDD-related total medical services costs were $39 and $43 lower, respectively, in the escitalopram group (both P ≤.001).
Results from the sensitivity analysis showed that, even assuming zero costs for all second-generation antidepressants used by patients in the citalopram group, the average total healthcare costs were still lower in the escitalopram group ($5551 vs $6644; P = .010). After adjustment for baseline characteristics using a GLM model, patients in the escitalopram group still showed an average savings of $420 in total healthcare costs compared with patients in the citalopram group (P <.001; Figure).
Discussion In this present retrospective study, a large administrative claims database was used to compare treatment persistence, healthcare utilization, and costs for patients treated with escitalopram compared with patients treated with citalopram in a real-world setting. Although patients taking citalopram and escitalopram had different baseline comorbidity profiles, these differences were controlled for in this study with the multivariable regression analysis described above.
Results indicate that although a majority of patients in both treatment groups discontinued treatment within 6 months after the index date, patients treated with escitalopram had better treatment persistence as measured by overall discontinuation rate, as well as rate of discontinuation with switching to another second-generation antidepressant. These results are consistent with those in a recent clinical trial, which also showed a lower withdrawal rate in the escitalopram group compared with the citalopram group.14
More than one third of the difference in total healthcare costs was attributable to lower hospitalization costs in the escitalopram group. The results were consistent using both unadjusted analysis and multivariate analysis and were supported by the sensitivity analysis, in which it was assumed that all second-generation antidepressants had zero costs in the citalopram group.
These results are also consistent with findings from recent claims data analyses conducted in an elderly population with MDD.15 In that study of patients with MDD aged ≥65 years, those treated with escitalopram showed significantly better treatment persistence, fewer hospitalizations, and lower medical and total healthcare costs than patients treated with citalopram. 15 Similar to findings in the current study, most of the cost reduction was attributable to significantly lower hospitalizations and total medical costs.15
Treatment persistence not only contributes to the therapeutic success of MDD treatment but also to the cost-effectiveness of an antidepressant.20 Previous studies have provided some evidence that the lack of persistence is associated with higher total medical costs among patients with depression.21,22 Lack of persistence could prolong the disease episode and lead to a higher rate of recurrent events. Therefore, improved persis - tence will likely save the long-term cost of treating MDD. Furthermore, better persistence in patients with MDD and better outcomes may reduce the healthcare resource utilization and costs associated with comorbidities beyond depression itself. In this present analysis, a substantial portion of the difference in healthcare utilization and costs between the escitalopram and citalopram groups was not MDD-related. The cause of these differences cannot be determined from claims data. It may be related to the potential advantage of escitalopram in safety and effectiveness, but the extent of this factor is unknown.
The economic benefits of escitalopram over citalopram have been suggested by numerous cost-effectiveness analyses.23-27 Findings in this present study are consistent with such previous analyses. It should be noted that persistence, healthcare utilization, and healthcare costs are directly related to the clinical efficacy and the tolerability of a treatment. Thus, clinicians and thirdparty payers should carefully consider the relative benefits of different therapeutic options when making patient care and policy decisions rather than rely solely on drug acquisition costs.
This present study has several limitations. First, it is subject to the usual limitations associated with claims data as a result of the absence of detailed clinical information. Therefore, there may be selection bias if patients in one treatment group had more severe depression than patients in the second group. Such difference may have affected treatment persistence and healthcare utilization during the study period. However, to minimize selection bias, this study only included patients who were not treated with any other second-generation antidepressant during the baseline period, and every effort was made to control for differences in observed baseline characteristics.
Second, discontinuation was measured by prescription refills in the pharmacy claims database instead of the actual use of the medications. Although not the most accurate measurement of treatment persistence, pharmacy data have been extensively used in other studies of treatment persistence.28-30 To the extent that the measurement error was not systematically different between the 2 treatment groups, the results regarding treatment persis - tence in this study can be considered valid.
Drug sampling bias could be another limitation— because additional drug samples provided free of charge by the physicians cannot be observed in the claims data but they may influence the persistence measure. However, the direction of the bias is unclear. In addition, the claims database used did not report any information related to patient copayment for medications; such information was therefore not available for this analysis.
Finally, the study focuses on relatively short-term benefits of treatment with escitalopram versus treatment with citalopram, and thus does not provide empirical evidence on long-term benefits. Future research is warranted to consider long-term and indirect effects of these therapies on work productivity and disability.
Adding to previous literature, this study shows that the clinical and economic benefits of escitalopram are not only observed in clinical trials, but also manifested in the real-world setting. In this present study, compared with citalopram, escitalopram was associated with higher treatment persistence and lower MDD-related and overall healthcare utilization and costs among adult patients with MDD. Funding Source This study was funded by Forest Laboratories, Inc. Author Disclosure Statement Mr Greenberg, Ms Ben-Hamadi, Dr Eric Wu, Dr Yang, and Dr Andrew Yu are consultants to Forest Laboratories. Dr Erder was employed by Forest Laboratories at the time of the study.
- Kessler RC, Berglund P, Demler O, et al. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA. 2003;289:3095-3105.
- Bloom BS. Prevalence and economic effects of depression. Manag Care. 2004;13(6 suppl):9-16.
- Ostacher MJ. Comorbid alcohol and substance abuse dependence in depression: impact on the outcome of antidepressant treatment. Psychiatr Clin North Am. 2007;30:69-76.
- Croom KF, Plosker GL. Spotlight on the pharmacoeconomics of escitalopram in depression. CNS Drugs. 2004;18:469-473.
- Greenberg PE, Kessler RC, Birnbaum HG, et al. The economic burden of depression in the United States: how did it change between 1990 and 2000? J Clin Psychiatry. 2003;64:1465-1475.
- Weilburg JB. An overview of SSRI and SNRI therapies for depression. Manag Care. 2004;13:25-33.
- Fantino B, Moore N, Verdoux H, Auray JP. Cost-effectiveness of escitalopram vs. citalopram in major depressive disorder. Int Clin Psychopharmacol. 2007;22:107-115.
- Esposito D, Wahl P, Daniel G, et al. Results of a retrospective claims database analysis of differences in antidepressant treatment persistence associated with escitalopram and other selective serotonin reuptake inhibitors in the United States. Clin Ther. 2009;31:644-656.
- Wu EQ, Greenberg PE, Yang E, et al. Treatment persistence, healthcare utilisation and costs in adult patients with major depressive disorder: a comparison between escitalopram and other SSRI/SNRIs. J Med Econ. 2009;12:124-135.
- Burke WJ, Gergel I, Bose A. Fixed-dose trial of the single isomer SSRI escitalopram in depressed outpatients. J Clin Psychiatry. 2002;63:331-336.
- Lepola UM, Loft H, Reines EH. Escitalopram (10-20 mg/day) is effective and well tolerated in a placebo-controlled study in depression in primary care. Int Clin Psychopharmacol. 2003;18:211-217.
- Rapaport MH, Bose A, Zheng H. Escitalopram continuation treatment prevents relapse of depressive episodes. J Clin Psychiatry. 2004;65:44-49.
- Wade A, Despiegel N, Heldbo Reines E. Escitalopram in the long-term treatment of major depressive disorder. Ann Clin Psychiatry. 2006;18:83-89.
- Moore N, Verdoux H, Fantino B. Prospective, multicentre, randomized, doubleblind study of the efficacy of escitalopram versus citalopram in outpatient treatment of major depressive disorder. Int Clin Psychopharmacol. 2005;20:131-137.
- Wu E, Greenberg PE, Yang E, et al. Comparison of escitalopram versus citalopram for the treatment of major depressive disorder in a geriatric population. Curr Med Res Opin. 2008;24:2587-2595.
- Antidepressant medication management. The state of health care quality 2006. National Committee for Quality Assurance; Washington, DC; 2006. www.ncqa.org/ Portals/0/Publications/Resource%20Library/SOHC/SOHC_2006.pdf. Accessed February 28, 2011.
- McCombs JS, Luo M, Johnstone BM, Shi L. The use of conventional antipsychotic medications for patients with schizophrenia in a Medicaid population: therapeutic and cost outcomes over 2 years. Value Health. 2000;3:222-231.
- Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613-619.
- Birnbaum HG, Ben-Hamadi R, Greenberg PE, et al. Determinant of direct cost differences among US employees with major depressive disorders using antidepressants. Pharmacoeconomics. 2009;27:507-517.
- Patient compliance in depression. Based on a presentation by James Jefferson, MD. Am J Manag Care. 2000;6(2 suppl):S31-S38.
- Revicki DA, Simon GE, Chan K, et al. Depression, health-related quality of life, and medical cost outcomes of receiving recommended levels of antidepressant treatment. J Fam Pract. 1998;47:446-452.
- Thompson D, Buesching D, Gregor KJ, Oster G. Patterns of antidepressant use and their relation to costs of care. Am J Manag Care. 1996;2:1239-1246.
- Demyttenaere K, Hemels ME, Hudry J, Annemans L. A cost-effectiveness model of escitalopram, citalopram,and venlafaxine as first-line treatment for major depres