|Evaluation of the cost effectiveness of escitalopram versus venlafaxine XR in major depressive disorder
|Fernandez J L, Montgomery S, Francois C
This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.
The use of escitalopram, a selective serotonin reuptake inhibitor, in the treatment of major depressive disorder (MMD). The comparator was venlafaxine RX, a serotonin noradrenaline (norepinephrine) reuptake inhibitor with established efficacy in MMD.
The study population comprised outpatients aged 18 to 85 years who fulfilled the Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV criteria for moderate to severe MDD, and were without suicidal tendencies. The patient's Montgomery-Asberg Depression Rating Scale (MADRS) total score was required to be higher than 18 at screening, 1 week before the start of the treatment, and at the start of treatment. Patients were excluded if they met any of the following criteria: history of mania or any bipolar disorder, schizophrenia or any psychotic disorder, or current evidence of obsessive-compulsive disorder, eating disorders, mental retardation, any pervasive development disorder or cognitive disorder. Patients taking medications thought likely to interfere with the study were also excluded.
The setting was primary care. The economic study was carried out in six European countries (Denmark, Finland, France, Germany, Spain and the UK).
Dates to which data relate
The effectiveness data were gathered for 8 weeks from the commencement of the study (2003). The use of health care services and days of sick leave were gathered for the 3 months before the start of the study (pre-study period) and for the 8 weeks after the start of the study. European 2003 prices were used to compute the costs.
Source of effectiveness data
The effectiveness data were derived from a single study.
Link between effectiveness and cost data
The costing was undertaken prospectively on the same patient sample as that used for the effectiveness study.
The authors did not report using power calculations to determine the sample size. A total of 293 patients were included in the trial. However, there was a lack of data for 42 patients (22 in the escitalopram arm and 20 in the venlafaxine arm) because either the physician or patient did not fill in the resource use questionnaire. Thus, the sample evaluated fell to 251. Of the 251 evaluable patients, 126 received escitalopram and 125 received venlafaxine. No significant differences, in terms of severity of depression, patient-related outcomes and co-morbidities, were observed between escitalopram- and venlafaxine-treated patients.
This was a randomised, double-blind, flexible-dose, multinational clinical trial. The follow-up period was 8 weeks. Full details of the study design were not presented in this paper, but elsewhere (Montgomery et al. 2004, see 'Other Publications of Related Interest' below for bibliographic details). At 8 weeks' follow-up, 245 patients reported valid cost information (4 escitalopram and 2 venlafaxine patients were lost relative to the pre-study period). Hence, the economic evaluation comprised 122 patients in the escitalopram group and 123 in the venlafaxine group.
Analysis of effectiveness
The analysis was conducted on the basis of treatment completers only. The primary health outcomes considered were the Quality of Life Depression Scale (QLDS) scores. No significant differences at baseline were observed between escitalopram- and venlafaxine-treated patients in terms of severity of depression, patient-related outcomes or co-morbidities.
The QLDS mean scores decreased from 18.6 to 12.4 for escitalopram-treated patients, (p<0.01), and from 18.8 to 12.1 for venlafaxine-treated patients, (p<0.01).
No statistically significant differences were observed between the groups.
The results of the study showed that patients in both groups experienced a significant improvement in quality of life, and that there were no significant differences between the two treatment groups when assessed by the QLDS or EQ-5D scores.
A multivariate modelling analysis was performed to identify the characteristics of patients who influenced the costs and to confirm findings from the perspective analysis of costs, controlling for patient characteristics. The variables included in the model before selection were costs at baseline, sociodemographic parameters (gender, age, marital status, living situation, employment status, educational level) and clinical characteristics at baseline (severity defined by MADRS, at least some problems with EQ-5D dimensions).
Measure of benefits used in the economic analysis
The measure of benefit used was the EQ-5D scores. A tariff of values for all EQ-5D states had been generated previously, using the time trade-off method, on a sub-set of these states from a sample of the English population.
The costs/quantity boundaries adopted were those of the health care system and society. Broad expenditure items included physician care, care by ancillary health care personnel, laboratory tests, clinical examinations and inpatient care. Eight co-prescriptions in the escitalopram group and nine in the venlafaxine group were for a hypnotic but, as no differences between groups in co-prescribed drugs were observed, co-medications were not included in the analysis. Health economics experts provided the prices used. These were based on national sources, except for the UK where health care costs were taken from Unit Costs of Health and Social Care published by the University of Kent. The quantities and the costs were reported separately. Discounting was not relevant because of the short follow-up period. The unit costs were adjusted to 2003 values using inflation rates (Consumer Price Index) for each country between 2001 and 2003.
Statistical analysis of costs
The results were presented in both a deterministic and a stochastic fashion. Differences in the costs were assessed using parametric (analysis of variance) and non-parametric (Kruskal-Wallis) tests, as well as bootstrapping.
The indirect costs were calculated using the human capital approach. As with the direct costs, discounting was not carried out. The patients provided information relating to patient-specific loss of production, 3 months prior to the study and during the follow-up period. Productivity losses were estimated according to the average wage in each country, based on the mean daily benefit paid by sickness insurance for the period off work.
The authors did not identify or explore any areas of uncertainty other than for the inclusion of sick leave costs (in order to assess the results from a societal perspective).
Estimated benefits used in the economic analysis
The EQ-5D mean scores improved from 0.52 to 0.78 for escitalopram-treated patients, (p<0.01), and from 0.54 to 0.77 for venlafaxine-treated patients, (p<0.01). No statistically significant differences were observed between the treatment groups.
The total health care costs were EUR 110 per patient in the escitalopram group and EUR 161 in the venlafaxine group.
The medication costs were EUR 62 in the escitalopram group and EUR 84 in the venlafaxine group.
The inpatient care (hospitalisation) costs were EUR 46 per patient in the venlafaxine group, whereas in the escitalopram group the costs were EUR 0.00.
When the data were adjusted for key cost-drivers, patients receiving escitalopram had statistically significantly lower health costs than those receiving venlafaxine (coefficient -0.34; p=0.007).
The direct costs for the average patient in the sample were 40% higher with venlafaxine than with escitalopram (95% confidence interval: 10 to 81).
Synthesis of costs and benefits
The incremental cost-effectiveness analysis was reported via the incremental cost-effectiveness ratio (ICER) confidence surface. Owing to the lack of significant differences in the efficacy of the two drugs, the analysis was not extended to the estimation of acceptability curves. An analysis of the ICER confidence surface demonstrated that health care costs were higher for the venlafaxine group than for the escitalopram group, and showed no between-group difference in the improvement of the EQ-5D score.
Escitalopram is as effective as venlafaxine in the treatment of major depressive disorder (MDD) and may be associated with lower costs from a societal and health care budget perspective.
CRD COMMENTARY - Selection of comparators
The rationale for the choice of the comparator was clear. Venlafaxine XR is a serotonin noradrenaline reuptake inhibitor with established efficacy in MDD, and is claimed to be more cost-effective than selective serotonin reuptake inhibitors. You must decide whether selective serotonin reuptake inhibitors are a widely used health technology in your own setting.
Validity of estimate of measure of effectiveness
This study was a double-blind, multinational, randomised controlled trial. It was unclear from the study whether the sample size was sufficiently large to obtain robust results. Only limited details of the study design were presented in this paper. With no details of the randomisation methods, patient selection, blinding or allocation concealment, it was difficult to fully assess the internal validity of the clinical trial. The QLDS scores would appear to represent a valid measure of effectiveness.
Validity of estimate of measure of benefit
The EQ-5D scores would appear to be valid measures of benefit. Moreover, the use of EQ-5D scores should also permit comparisons of the study results across different interventions.
Validity of estimate of costs
Given that both health care system and societal perspectives were adopted, direct and indirect costs were appropriately included. All relevant costs appear to have been included in the analysis. The price year was reported and the quantities and the prices were reported separately, thus enhancing the transferability of the results. Discounting was not relevant and was not conducted.
The authors acknowledged that larger sample sizes are required to increase the power of performed tests and to enable the detection of differences in costs between escitalopram and venlafaxine XR. The authors made extensive and detailed comparisons of their findings with those of other research, stating that this piece of research reflected the previous trend of similar effectiveness for both drugs but lower costs for patients in the venlafaxine group. The authors pointed out the fact that MDD epidemiology and patterns of medical practice are not substantially different between Western countries and, as the patients' characteristics were controlled, the multinational issues were more related to cost estimation procedures. Long-term cost-consequences of the treatments could not be fully captured because of the short duration of follow-up. The results were not reported selectively and the authors' conclusions reflected the scope of the analysis.
Implications of the study
The results of this research support the recommendation of new higher powered studies that might confirm the cost-savings achieved with escitalopram. The authors also pointed out that further investigation to establish whether escitalopram does have advantages over venlafaxine in terms of reducing hospitalisation costs would be of interest.
Fernandez J L, Montgomery S, Francois C. Evaluation of the cost effectiveness of escitalopram versus venlafaxine XR in major depressive disorder. PharmacoEconomics 2005; 23(2): 156-167
Other publications of related interest
Montgomery SA, Huusom AK, Bohmer J. A randomised study comparing escitalopram with venlafaxine XR in primary care patients with major depressive disorder. Neuropsychobiology 2004;50:57-64.
Hemels ME, Kasper S, Walter E, et al. Cost-effectiveness analysis of escitalopram: a new SSRI in the first-line treatment of major depressive disorder in Austria. Curr Med Res Opin 2004;20:869-78.
Hemels ME, Kasper S, Walter E, et al. Cost-effectiveness of escitalopram vs citalopram in the treatment of severe depression. Ann Pharmacother 2004:38:954-60.
Drummond M, Pand F. Transferability of economic evaluation results. In: McGuire A, editor. Economic evaluation in health care. New York: OHE; 2001. p.256-76.
Subject indexing assigned by NLM
Adolescent; Adult; Aged; Antidepressive Agents, Second-Generation /therapeutic use; Citalopram /therapeutic use; Comparative Study; Cost-Benefit Analysis; Cyclohexanols /therapeutic use; Depressive Disorder, Major /drug therapy /psychology; Double-Blind Method; Female; Health Care Costs; Humans; Male; Middle Aged; Prospective Studies; Quality of Life; Research Support, Non-U.S. Gov't
Date bibliographic record published
Date abstract record published