|Duloxetine in acute major depression: review of comparisons to placebo and standard antidepressants using dissimilar methods
|Girardi P, Pompili M, Innamorati M, Mancini M, Serafini G, Mazzarini L, Del Casale A, Tatarelli R, Baldessarini RJ
The review found that duloxetine appeared more effective than placebo for acute major depression in adults, and was comparable to selective serotonin-reuptake inhibitors, but that more research is needed. In view of poor reporting of review methods and adverse event outcomes, and lack of statistical evidence that the trials were suitable for pooling, the authors’ conclusions may not be reliable.
To evaluate the efficacy and tolerability of duloxetine for acute major depression in adults.
MEDLINE was searched to August 2008. Search terms were reported. Reference lists of articles retrieved were searched and trials were sought from the drug manufacturer. The search was limited to unpublished studies available from the drug manufacturer and published peer-reviewed studies.
Randomised controlled trials (RCTs) of duloxetine treatment for adults with acute major depressive disorder, diagnosed by current diagnostic criteria, were eligible for inclusion. Outcomes of interest in the review were response (i.e. at least 50% reduction of depression severity on a standard depression rating scale), remission (as defined below), drop-out (i.e. treatment discontinuation), adverse events, such as sustained hypertension (as defined in the review), and dose-effect relationship.
The mean age of participants in the included trials ranged from 40 to 73 years (mean 44 years). Most participants were female (range across study groups 52% to 100%, where reported). All the studies used either the Hamilton Depression Rating Scale (mean baseline participant score was 20.4 points) or the Montgomery-Asberg Depression Rating Scale (mean baseline participant score was 40 points). Remission was defined as 7 or fewer points on these scales. A mean dose of 73.5mgs of duloxetine was used (range 40 to 120mgs). Control groups received either placebo or one of the following serotonin-reuptake inhibitors: paroxetine, S-citalopram, fluoxetine or venlafaxine. Trial duration ranged from six to 12 weeks (mean 8.2 weeks). Nearly all trials were designed and sponsored by the drug manufacturer.
Study selection was conducted independently by two reviewers, with disagreements resolved by a third.
Assessment of study quality
The validity of published trials was assessed using the Jadad Scale (which assessed the adequacy of reported randomisation, double-blinding, and withdrawals or drop-outs). Trials were included only if they scored at least 4 of a maximum 5 points.
Validity assessment was conducted by at least two reviewers.
Event rates in the two groups of each trial were extracted using data from mixed-model-repeated-measures (MMRM) and last-observation-carried forward-methods (LOCF), where possible. Rate ratios (RRs) and their 95% confidence intervals (CIs) were extracted or calculated for mixed-model-repeated-measures and last-observation-carried-forward analyses, where possible.
The authors did not state how many reviewers performed the data extraction.
Methods of synthesis
Event rates in individual trials were pooled by simple addition to calculate crude pooled rates across trials. Trials were also combined using the DerSimonian and Laird random-effects model to calculate pooled rate ratios and numbers-needed-to-treat (NNT) or numbers-needed-to-harm (NNH), with 95% confidence intervals. Linear correlations were used to test for relationships between response and remission, and between mean daily dose of duloxetine and outcomes. Heterogeneity was assessed using the Q statistic.
Results of the review
Seventeen RCTs (22 comparisons) were included in the review (n reported in text as 6,106 non-overlapping patients, range 33 to 337).
Response rates were significantly higher with duloxetine than placebo (RR 1.42, 95% CI 1.31 to 1.53; NNT 6.7; 17 comparisons; LOCF analysis); mixed-model-repeated-measures analysis (11 comparisons) yielded a similar rate ratio. Remission rates over seven to 12 weeks were also significantly higher with duloxetine than placebo (RR 1.42, 95% CI 1.28 to 1.56; NNT 9.2; 17 comparisons; LOCF analysis); mixed-model-repeated-measures analysis (11 comparisons) yielded an rate ratio of 1.61 (95% CI 1.41 to 1.84).
There was minimal difference in response rate between duloxetine and serotonin-reuptake inhibitors, whether analysed by last-observation-carried-forward (16 comparisons) or mixed-model-repeated-measures (four comparisons).
There was a statistically significant correlation between duloxetine dose and response rate (17 comparisons; p=0.001), but no significant correlation between dose and remission rate.
Adverse effects were significantly more common with duloxetine than placebo (RR 1.18, 95% CI 1.14 to 1.32; NNH 8.1; 15 comparisons). The drop-out rate due to adverse events was significantly higher with duloxetine than either placebo (RR 2.16, 95% CI 1.55 to 3.00; NNH 20.3; 15 comparisons) or serotonin-reuptake inhibitors (RR 1.57, 95% CI 1.27 to 1.93; NNH 31.7; 14 comparisons).
Other findings were reported in the review.
Duloxetine appeared to be more effective than placebo for acute major depression in adults, and of similar efficacy to serotonin-reuptake inhibitors. More research is needed.
The objectives of the review were clear and relevant sources were searched for studies, although no search start date was specified and it was unclear whether the search was limited by language. It was also unclear why unpublished studies were eligible for inclusion only if they were supplied by the drug manufacturer; this may have created a potential for bias. Publication bias was not formally assessed. Steps were taken to minimise the risk of reviewer error and bias by more than one reviewer independently undertaking study selection; it appeared that this precaution also applied to validity assessment. However, the method used for data extraction was not described.
The Jadad scale addressed some important quality criteria, but no details were reported on individual trial results or on other quality components, such as allocation concealment and follow-up rates. It was unclear whether quality criteria for study inclusion were applied to the unpublished trials. All the trials were short term, which limited the clinical applicability of their findings. The authors pooled trial results by simple addition of raw data, which was unreliable, as it did not allow for differences between trials. Trials were also pooled using appropriate statistical techniques to combine the data and test for heterogeneity. However, the results of the heterogeneity assessment were not reported and for most analyses the forest plots were not presented. These factors made it difficult to ascertain the homogeneity and reliability of the pooled results. The authors did not comment in their conclusions on the tolerability of duloxetine, or the significantly higher rate of drop-outs due to adverse effects associated with duloxetine compared to serotonin-reuptake inhibitors.
In view of poor reporting of review methods and adverse effect outcomes, and the lack of statistical evidence that the trials were suitable for pooling, the authors’ conclusions may not be reliable.
Six of the nine review authors disclosed past or current links with pharmaceutical companies (as employees, consultants or researchers).
Implications of the review for practice and research
Practice: The authors did not state any implications for practice.
Research: The authors stated that more studies are needed to determine the effectiveness and tolerability of duloxetine compared to other antidepressants, the optimum dosage of duloxetine and whether discontinuation should be abrupt or tapered. Future studies should also investigate the use of duloxetine for preventing relapse of depression, and its use for indications other than depression.
Part funded by a grant from the Bruce J Anderson Foundation and the McLean Hospital Private Donors’ Research Fund for Psychopharmacology Research.
Girardi P, Pompili M, Innamorati M, Mancini M, Serafini G, Mazzarini L, Del Casale A, Tatarelli R, Baldessarini RJ. Duloxetine in acute major depression: review of comparisons to placebo and standard antidepressants using dissimilar methods. Human Psychopharmacology 2009; 24(3): 177-190
Subject indexing assigned by NLM
Antidepressive Agents /chemistry /therapeutic use; Depressive Disorder, Major /drug therapy; Duloxetine Hydrochloride; Humans; Placebos /therapeutic use; Randomized Controlled Trials as Topic /methods /standards; Thiophenes /chemistry /therapeutic use
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.