Patient-rated troubling symptoms of depression instrument results correlate with traditional clinician- and patient-rated measures: A secondary analysis of a randomized, double-blind, placebo-controlled trial

Results In the PaRTS-D instrument, the four most frequently reported and troublesome symptoms were sadness (73.5%, severity 6.8), trouble concentrating (70.9%, 7.3), reduced involvement in pleasurable activities (61.9%, 7.3), and being tense or uptight (56.0%, 6.7). The improvement in PaRTS-D total score was significantly greater in risperidone-augmented compared with placebo-augmented patients at week 4 ( p = 0.034) and week 6 ( p = 0.007). Pearson correlations between the PaRTS-D scores and the measures of HRSD-17, CGI-S, PGIS, Q-LES-Q, and SDS were significant at both baseline and at week 6 LOCF ( p < 0.001 for each comparison). Limitations Results are from a post hoc analysis. Conclusions Significant correlations were observed between the PaRTS-D and other clinician- and patient-rated measures, with PaRTS-D being sensitive to the effects of treatment. These findings suggest that the PaRTS-D instrument is a reliable scale to assess antidepressant activity as experienced by the patients. Keywords Residual depression symptoms Major depressive disorder Risperidone Augmentation SSRI Selective serotonin reuptake inhibitors Atypical antipsychotic 1 Introduction Major depressive disorder (MDD) affects approximately 5% of the population at any point in time and accounts for more morbidity, mortality and heath care costs than any other psychiatric disorder.( Murray and Lopez, 1996; American Psychiatric Association, 2000; American Diabetes Association, 2004 ) Pharmacologic therapy is the most common approach to the treatment of depression, with the selective serotonin reuptake inhibitors (SSRIs) and serotonin norepinephrine reuptake inhibitors (SNRIs) generally considered the first line of depression treatment. The optimal assessment of antidepressant drug effect should include measure of the presence, severity, and troublesomeness of depressive symptoms. A wide range of instruments is available for this purpose. However, clinician-rated scales, such as the 17-item Hamilton Rating Scale for Depression (HRSD-17), Montgomery–Asberg Depression Rating Scale, and the Clinical Global Impressions of Severity (CGI-S) do not capture the presence and troublesomeness of symptoms from the patient's perspective. Further, patient-rated scales, such as the Patient Global Improvement Scale (PGIS), the Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q) ( Endicott et al., 1993 ), the Inventory of Depressive Symptomatology ( Rush et al., 2006a ), and the Sheehan Disability Scale (SDS) may not fully reflect the troublesomeness of symptoms or the patient's impression of their response to treatment. The Patient-Rated Troubling Symptoms for Depression (PaRTS-D) instrument was developed to provide a more individualized assessment, from the patient's perspective, of 8 commonly reported symptoms of depression. Symptoms rated on the PaRTS-D instrument are: sadness, feeling tense or uptight, reduced sleep, reduced appetite, trouble concentrating, reduced involvement in things that usually interest the subject, inability to feel emotions, and negative thoughts. In a double-blind, randomized, placebo-controlled study of patients with have already listed definition of MDD given risperidone augmentation to their antidepressant treatment, patients, via an interactive voice response system (IVRS), completed the PaRTS-D instrument as well as other standard clinician- and patient-rated instruments at baseline and at weekly intervals.( Mahmoud et al., 2007 ) This post hoc analysis of the trial results explored the relationship between the PaRTS-D instrument and the other established clinician- and patient-rating scales. 2 Methods 2.1 Analysis database The database of a published ( Mahmoud et al., 2007 ) large, randomized, multicenter study was used in this secondary analysis. In the primary study, adult outpatients, 18 to 65 years of age, with a primary clinical diagnosis of major depressive disorder according to the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV)( American Psychiatric Association, 2000 ) and an initial investigator-rated CGI-S score ≥ 4 ( Guy 1976 ) (a 0 to 7 scale measuring illness severity where 4 = moderately ill) who received antidepressant monotherapy for at least 4 weeks just prior to study enrollment were enrolled from 75 primary care and psychiatric centers (public and private) between October 19, 2004 and November 17, 2005. The primary study inclusion criteria were broad in order to maximize generalizability of the findings.( Fava et al., 2003; Rush et al., 2004, 2006b; Mahmoud et al., 2007 ), excluding only pregnant women, those with a serious suicidal risk or serious medical/neurologic illness, individuals with active substance or alcohol use disorders, and those currently receiving a tricyclic antidepressant, monoamine oxidase inhibitor, mood stabilizer, antiepileptic, or a centrally acting agent for the treatment of attention deficit disorder/attention deficit hyperactivity disorder or narcolepsy. During the first phase of the study, patients received open-label treatment with their current antidepressant monotherapy at dosages adjusted to levels considered optimal by product labeling. At the conclusion of the open-label phase, participants who were compliant with their antidepressant medication but nevertheless continuing to suffer persistent significant depressive symptoms [defined as an investigator-rated CGI-S score ≥ 4 and the patient-rated Carroll Depression Scale ( Nasr et al., 1984 ) total score ≥ 20] were eligible for entry into the 6-week double-blind phase. During this phase, patients continued their optimized original antidepressant regimen and were randomized to receive augmentation with once-daily risperidone or placebo in a 1:1 ratio. During the double-blind phase, the dose of risperidone or matching placebo was titrated using the schedule: 0.25 mg for the first 3 days, 0.5 mg on days 4 to 15, and 1.0 mg on days 16 to 28. On day 29 of the 6-week study, two treatment options were offered to patients continuing to have insufficient response: investigators could increase the dose of the double-blind treatment (to 2 mg/day of risperidone) or patients could discontinue the double-blind phase and receive open-label risperidone for an additional 4 weeks. The primary objective of the study was to demonstrate the efficacy of adjunctive risperidone treatment in patients with MDD suboptimally responsive to standard antidepressant therapy. In addition to the analysis of safety, a secondary objective of the study was to evaluate the efficacy of adjunctive risperidone in improving the symptoms of depression that are most troubling to patients as well as improve their quality of life and overall functioning. Changes from baseline in depression symptoms were assessed using established clinician- and patient-rating scales and the PaRTS-D instrument. 2.2 Clinical measurements The grid version of the HRSD-17 ( Hamilton, 1960; Kalali et al., 2002 ) and the CGI-S scales ( Guy, 1976 ), both clinician-rated instruments, were administered at each clinic visit (baseline [end of the open-label phase], and at weeks 1, 2, 4 and 6 of the double-blind phase) by study personnel trained in their use. In the HRSD-17, scores range from 0 to 52 with higher scores indicating more severe depression. Patients independently rated their response to augmentation treatment, using a touch-tone telephone IVRS at baseline and weekly thereafter. Patient-reported outcome was assessed with several validated instruments: the Q-LES-Q ( Endicott et al., 1993 ); Patient Global Improvement Scale (PGIS)—a 7-point scale ranging from 1 = very much improved to 7 = very much worse; and the SDS that measures 3 dimensions of social functioning (i.e., work/school, social life, and family life/home responsibilities) where higher scores denote greater functioning. The patients' most troubling symptoms of depression and their severity were assessed using the PaRTS-D at baseline and weekly thereafter using the IVRS system. At baseline, patients rated—from most to least troubling—8 specific core symptoms of depression from the Montgomery–Asberg Depression Rating Scale ( Montgomery and Asberg, 1979 ) (i.e., sadness, feeling tense or uptight, reduced sleep, reduced appetite, trouble concentrating, reduced involvement in things that usually interest you, inability to feel emotions, and negative thoughts) and then rated the severity of the 8 symptoms on a scale of 0 to 10 where 0 = absent, 1 to 3 = mild, 4 to 6 = moderate, 7 to 9 = marked, and 10 = extreme). A total severity score for all 8 symptoms was determined. The sum of the severity ratings of the 4 symptoms that were considered the most troubling was also determined. At weekly intervals, the severity of the 4 symptoms that ranked as most troubling for each patient were reassessed using the IVRS. The troublesomeness and severity of all 8 symptoms were rated again at weeks 4 and 6. Change from baseline in HRSD-17 total score at week 4 was the primary efficacy endpoint. After an additional 2 weeks of double-blind treatment, other efficacy parameters including treatment responders (> 50% reduction in HRSD-17 total score from baseline), remitters (HRSD-17 total score ≤ 7) and changes from baseline in the HRSD-17, CGI-S, and patient-rated measures of Q-LES-Q, PGIS, SDS and PaRTS-D were assessed. Adverse events were collected and reported throughout the study. 2.3 Statistical analysis The overall intent-to-treat (ITT) analysis set was defined as all randomly assigned subjects who took at least one dose of study medication. Change from baseline in HRSD-17 total score between groups was analyzed at each visit and last observation carried forward (LOCF) end point using analysis of covariance (ANCOVA) models with treatment, class of antidepressant therapy (strata), and the site as fixed effect design factors, and baseline HRSD-17 scores as a covariate. In addition to the ANCOVA model at Week 4 LOCF, the change from baseline in the HRSD-17 total score was analyzed using the repeated measures mixed effects ANCOVA model to assess the robustness of primary findings. Treatment group differences at baseline were tested using an analysis of variance (ANOVA) with effects for treatment, strata, and pooled site. Change and percent change from baseline and actual values for CGI-S and patient-rated measures were summarized, within-group differences were evaluated by paired t -test at each time point, and between-group comparisons analyzed using ANCOVA or ANOVA. Categorical outcomes were evaluated using the Cochran–Mantel–Haenszel (CMH) test, stratified by strata and/or site, or rank tests as appropriate. PaRTS-D scores were calculated as a percent of the maximum score possible for the four most troubling items as determined at baseline. Because there was no minimum PaRTS-D score required at baseline, individual scores were variable. In order to reduce the variability, an analysis was performed including only symptoms that were rated at least a “5” (i.e. moderately troubling) at baseline. For each subject, the percent change from maximum at baseline was then calculated, and change over time was measured. Pearson correlations examined the degree of association between the PaRTS-D total score and scores on the HRSD-17, CGI-S, PGIS, Q-LES-Q, and SDS. 3 Results 3.1 Patients Among the 463 screened patients, 362 (78.2%) completed the open-label antidepressant treatment phase and, of these, 274 met double-blind eligibility criteria and were randomized to risperidone ( n = 141) or placebo ( n = 133). In the ITT population, over 84% (226/268) completed the 6-week study with similar percentages of completers, premature discontinuers, and open-label risperidone continuers in both treatment groups ( Fig. 1 ). The risperidone- and placebo-augmentation groups did not differ significantly at baseline with respect to any demographic or depression-related characteristics ( Table 1 ). The persistence of depressive symptoms despite antidepressant therapy was supported by the HRSD-17 scores found in those who entered the open-label antidepressant treatment phase. Even after > 8 weeks of antidepressant therapy, mean HRSD-17 total scores of 24.6 ± 4.97 were observed, which, despite an additional 4 weeks of optimized antidepressant treatment, did not change significantly (HRSD-17 = 24.4 ± 4.95). 3.2 Treatment During the first 28 days, risperidone was administered at a mean modal dose of 0.89 mg ± 0.22 (range 0.25 to 1.0 mg), with nearly 80% (109/137, 79.6%) of patients receiving risperidone 1 mg/day at the week 4 endpoint. At day 29, dosage increases occurred in a similar percentage (∼ 20%) of those receiving double-blind risperidone or placebo augmentation. For weeks 1 to 6, the modal daily dose of risperidone was 1.12 ± 0.46 mg with that of placebo being 1.17 ± 0.47 mg dose equivalents. 3.3 Efficacy outcomes: patient-rated 3.3.1 Parts-D Of the 8 core depression symptoms, the four most frequently selected to be most troubling were sadness (73.5%, 197/268), trouble concentrating (70.9%, 190/268), reduced involvement in pleasurable activities (61.9%, 166/268), and tense or uptight (56.0%, 150/238), with the percentages of patients in each augmentation group reporting these with similar frequencies ( Table 1 ). The symptoms that were reported most frequently were also rated the most troubling with mean severity ratings of: trouble concentrating, 7.3; reduced involvement in pleasurable activities, 7.3; sadness, 6.8; and tense or uptight, 6.7. In the analyses of the maximum change from baseline in symptoms rated at least moderately troubling (score of 5 or greater) at baseline, both augmentation groups exhibited significant ( p < 0.001) reductions from baseline. Risperidone augmentation produced significantly greater improvement in these residual symptoms compared to placebo augmentation ( Fig. 2 ) at week 4 (− 24.9 ± 2.1 vs. − 19.5 ± 2.1, p = 0.034) and 6 LOCF (− 31.3 ± 2.0 vs. − 22.7 ± 2.0, p = 0.007). 3.3.2 Q-LES-Q At baseline, the mean patient-rated Q-LES-Q total scores, medication satisfaction item, and overall life satisfaction item scores were similar between the risperidone- and placebo-augmentation groups (46.3 and 45.5; 2.4 and 2.4, and 2.1 and 2.2; respectively). Both augmentation groups exhibited significant ( p < 0.001) improvements in Q-LES-Q scores from baseline with risperidone augmentation producing significantly greater levels of improvement in the total score as well as the subscales of medication satisfaction and overall life satisfaction compared to placebo augmentation at week 6 LOCF. ( Fig. 3 ) 3.3.3 PGIS Patients completed the PGIS at weekly intervals during risperidone and placebo augmentation. Both groups exhibited incremental improvements (decreases in score) in PGIS at each week, with risperidone producing a significantly greater improvement at the week 6 LOCF (2.8 ± 0.12 vs. 3.1 ± 0.12, p = 0.049) as well as selected weekly timepoints. ( Fig. 4 ) 3.3.4 SDS The SDS Total Score, Social Life Dimension, Family Life/Home Responsibility Dimension, and Work/School Dimension subscale scores improved from baseline in both the risperidone and placebo-augmentation groups. Significantly greater improvements in three of the four dimensions (exception of work/school dimension) were observed at the week 6 LOCF in risperidone-augmented patients compared to placebo-augmented patients. ( Fig. 5 ) 3.4 Symptom outcomes: investigator-rated 3.4.1 HRSD-17 Augmentation with risperidone or placebo produced improvements in HRSD-17 total score at each assessment; however, the decreases in HRSD-17 from baseline to weeks 4 and 6 LOCF were significantly greater with risperidone augmentation (− 8.5 ± 6.7 with risperidone vs. − 7.1 ± 6.3 with placebo, p = 0.027 at week 4 LOCF and − 10.5 ± 0.68 vs. − 8.06 ± 0.68, p = 0.004 at week 6 LOCF). At week 6 LOCF, remission of depression occurred in 19.7% (26/137) of those in the risperidone-augmentation group and 9.5% (12/131) of the placebo-augmentation group ( p = 0.016): treatment response was observed in 40.9% (54/137) of risperidone-augmentation and 28.6% (36/131) of placebo-augmentation patients ( p = 0.017). 3.4.2 CGI-S Both treatment groups improved from baseline on CGI-S score at week 6 LOCF, with a significantly greater improvement noted in those randomized to risperidone (− 1.3 vs. − 0.9, p = 0.002). 3.5 Correlations between parts-D and other outcomes Baseline and endpoint PaRTS-D total scores were significantly ( p < 0.001) correlated with investigator-rated measures of HRSD-17 (0.37 and 0.65, respectively) and CGI-S (0.24 and 0.6, respectively) as well as patient-rated measures of PGIS (no score at baseline, 0.67 at endpoint), Q-LES-Q (− 0.60, − 0.75, respectively), and SDS (0.66 and 0.80, respectively). 4 Discussion Overall, the results of this post hoc analysis of a large, randomized, placebo-controlled, double-blind trial of risperidone augmentation to antidepressant therapy in patients with residual symptoms despite at least 8 weeks of treatment are significant in that they document an improvement in outcome with augmentation therapy as well as provide evidence supporting the responsiveness and validity of the PaRTS-D tool. At baseline, the patients randomized to each treatment group were similar with respect to their demographic and depression-related characteristics as well as to their HRSD-17 scores and their PaRTS-D four most troubling symptoms prevalence and severity. In both the week 4 and week 6 LOCF analyses, the PaRTS-D change from baseline was greater in those randomized to risperidone augmentation than those augmented with placebo. The greater improvements from baseline with risperidone augmentation were also observed in the other clinician- and patient-rated outcome measures. Analyses between the improvements in the PaRTS-D and other patient-rated and clinician-rated outcomes revealed significant correlations between these instruments with more robust correlations observed between the PaRTS-D and other patient-rated outcomes as compared with other clinician-rated outcomes (i.e., the HRSD-17 and CGI-S). This suggests that the clinician's assessment of a patient's MDD status and residual symptoms differs from that of the patient's perception of their response to treatment. Limitations to the interpretation of these findings are that they are derived from acute data only, in a study of patients who required augmentation of their original therapy and may thus represent a more severely depressed population segment. Other limitations include that the findings are the result of a post hoc analysis; however, all of the residual symptom measurement tools assessed in this analysis were included in the original protocol and data analyses plans as secondary outcomes. In the past several years, depression remission and response has been the emphasis of pharmacologic treatment, as studies have clearly shown that the presence of residual symptoms is associated with a greater likelihood of full depression recurrence, greater decrements in disability, emotional and cognitive function as well as poorer physical functioning. ( Thase et al., 1992; Judd et al., 1998, 1999; Zimmerman et al., 2004 ) Unlike other common long-term disorders such as hypertension, diabetes, or heart disease there are no specific laboratory- or physical examination-based diagnostic tests to measure patients' response to MDD treatment. As a result, in treatment efficacy trials, response is measured and defined by the attainment of a threshold value on a clinician-rated outcome instrument, such as the HRSD-17 or the CGI-S. While clinician-rated instruments are frequently used in clinical trials, in clinical practice their use may be limited by the amount of time and effort required to complete them. ( Zimmerman et al., 2004 ) In recent years, a growing body of literature suggests that the use of patient-rated instruments provides a more valid as well as practical and sensitive assessment option that should be considered an essential part of the objective evaluation of pharmacologic response in those with MDD. ( Hickie et al., 2002; Zimmerman et al., 2004; Rush et al., 2005 ) In a recently published analyses of three self-reported measures of depression symptoms, Rush et al. (2005) found that two of the self-reported instruments (the 30-item Inventory of Depressive Symptomatology—Self-Report [IDS-SR30] and the 16-item Quick Inventory of Depressive Symptomatology—Self-Report [QIDS-SR16]) closely mirrored and confirmed the findings of the 24-item Hamilton Depression Rating Scale, and suggested that these self-reports may potentially substitute for clinician ratings. The investigators also stated that obtaining these self-rated assessments could be facilitated by the use of a telephone-based system. Like the findings of Rush et al. (2005) , high correlations were found between the PaRTS-D and other assessment tools, with this tool being accessed through an IVRS. In addition to increasing patient and healthcare provider convenience and the potential for routine monitoring, the PaRTS-D IVRS tool lends itself to electronic record keeping. In conclusion, the patient-rated PaRTS-D tool provides informative and treatment-sensitive assessments of residual symptoms and their severity that are most troubling to patients, and the impact of pharmacotherapy on these symptoms. Significant correlations were observed between the PaRTS-D and other clinician- and patient-rated measures, with PaRTS-D being sensitive to the effects of treatment. These findings indicate that the PaRTS-D instrument is a reliable scale, and contributes to the growing body of clinical evidence suggesting that patient-rated measures may be beneficial in the assessment of the effect of antidepressant activity. Role of funding source This study was supported by funding from Ortho-McNeil Janssen Scientific Affairs, LLC. The funding organization had a role in the study design. It was not involved in the conduct of the study or in the collection of the data. The funding source supported the analysis of the data and its interpretation. All authors had full access to the data files for the study. Conflict of interest Employment: G. J. Pandina and J. H. Wu (Johnson & Johnson Pharmaceutical Research & Development); I. Turkoz (Ortho-McNeil Janssen Scientific Affairs); M. J. Kujawa, R. A. Mahmoud and G. M. Gharabawi (work performed while employed at Ortho-McNeil Janssen Scientific Affairs); Consultancy: D.A. Revicki and L. Kleinman (Janssen, L.P.). Stock Ownership or options (other than mutual funds): G. J. Pandina, I. Turkoz, J. H. Wu, R.A. Mahmoud, M.J. Kujawa, G. M. Gharabawi (Johnson & Johnson). Acknowledgements Editorial support was provided by Susan Ruffalo, Pharm.D., MedWrite, Inc., Newport Coast, CA. References American Diabetes Association, 2004 American Diabetes Association Consensus development conference on antipsychotic drugs and obesity and diabetes Diabetes Care 27 2004 596 601 American Psychiatric Association, 2000 American Psychiatric Association Diagnosis and Statistical Manual of Mental Disorders 4th ed. 2000 American Psychiatric Press, Inc Washington, D.C. Text revision, DSM-IV Endicott et al., 1993 Endicott J. Nee J. Harrison W. Blumenthal R. Quality of life enjoyment and satisfaction questionnaire: A new measure Psychopharmacol. Bull. 29 2 1993 321 326 Fava et al., 2003 Fava M. Rush A.J. Trivedi M.H. Nierenberg A.A. Thase M.E. Sackeim H.A. Quitkin F.M. Wisniewski S. Lavori P.W. Rosenbaum J.F. Kupfer D.J. Background and rationale for the sequenced treatment alternatives to relieve depression (STAR⁎D) study Psychiatr. Clin. North. Am. 26 2003 457 494 Guy, 1976 Guy W. ECDEU Assessment manual for Psychopharmacology, Revised. E. US Department of Health, and Welfare Publication (ADM) 76–338 1976 National Institute of Mental Health Rockville, Maryland 218 222 Hamilton, 1960 Hamilton M. A rating scale for depression J. Neurol. Neurosurg. Psychiatry. 163 1960 28 40 Hickie et al., 2002 Hickie I.B. Andrews G. Davenport T.A. Measuring outcomes in patients with depression or anxiety: an essential part of clinical practice Med. J. Aust. 177 2002 205 207 Judd et al., 1998 Judd L.L. Akiskal H.S. Maser J.D. Zeller P.J. Endicott J. Coryell W. Paulus M.P. Kunovac J.L. Leon A.C. Mueller T.I. Rice J.A. Keller M.B. Major depressive disorder: A prospective study of residual subthreshold depressive symptoms as predictor of rapid relapse J. Affect. Disord. 50 2–3 1998 97 108 Judd et al., 1999 Judd L.L. Paulus M.P. Zeller P. The role of residual subthreshold depressive symptoms in early episode relapse in unipolar major depressive disorder Arch. Gen. Psychiatry. 56 1999 764 765 Kalali et al., 2002 Kalali A.H. Williams J.B. Kobak K.K. Engelhardt N. Evans K.R. Pearson J.D. Rothman M. Bech P. The new GRID HAM-D pilot testing and international field trials Int. J. Neuropsychopharma. 5 1 2002 S147 Mahmoud et al., 2007 Mahmoud R.A. Pandina G. Turkoz I. Kosik-Gonzalez C. Canuso C. Kujawa M. Gharabawi G.M. Risperidone for treatment—refractory major depressive disorder: A randomized trial Ann. Intern. Med. 147 9 2007 593 602 Montgomery and Asberg, 1979 Montgomery S.A. Asberg A. A new depression scale designed to be sensitive to change Br. J. Psychiatry. 134 1979 382 389 (April) Murray and Lopez, 1996 Murray C.J.L. Lopez A.D. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Disease, Injuries, and Risk Factors in 1990 and Projected to 2020 1996 Harvard University Press Cambridge, Mass Nasr et al., 1984 Nasr S.J. Altman E.G. Rodin M.B. Jobe T.H. Burg B. Correlation of the Hamilton and Carroll Depression Ratings Scales: a replication study among psychiatric outpatients J. Clin. Psychiatry. 45 4 1984 167 168 Rush et al., 2004 Rush A.J. Fava M. Wisniewski S.R. Lavori P.W. Trivedi M.H. Sackeim H.A. Thase M.E. Nierenberg A.A. Quitkin F.M. Kashner T.M. Kupfer D.J. Rosenbaum J.F. Alpert J. Stewart J.W. McGrath P.J. Biggs M.M. Shores-Wilson K. Lebowitz B.D. Ritz L. Niederehe G. Sequenced treatment alternatives to relieve depression (STAR⁎D): rationale and design Control. Clin. Trials. 25 2004 119 142 Rush et al., 2005 Rush A.J. Trivedi M.H. Carmody T.J. Ibrahim H.M. Markowitz J.C. Keitner G.I. Kornstein S. Arnow B. Klein D.N. Manber R. Dunner D.L. Gelenberg A.J. Kocsis J.H. Nemeroff C.B. Fawcett J. Thase M.E. Russell J.M. Jody D.N. Borian F.E. Keller M.B. Self-reported depressive symptom measures: sensitivity to detecting change in a randomized, controlled trial of chronically depressed, nonpsychotic outpatients Neuropsychopharmacology 30 2005 405 416 Rush et al., 2006a Rush A.J. Bernstein I.H. Trivedi M.H. Carmody T.J. Wisniewski S.R. Mundt J.C. Shores-Wilson K. Biggs M.M. Woo A. Nierenberg A.A. Fava M. An evaluation of the quick inventory of depressive symptomatology and the hamilton rating scale for depression: a sequenced treatment alternatives to relieve depression trial report Biol. Psychiatry. 59 6 2006 493 501 Rush et al., 2006b Rush A.J. Trivedi M.H. Wisniewski S.R. Stewart J.W. Nierenberg A.A. Thase M.E. Ritz L. Biggs M.M. Warden D. Luther J.F. Shores-Wilson K. Niederehe G. Fava M. Bupropion-SR, sertraline, or venlafaxine-XR after failure of SSRIs for depression N. Engl. J. Med. 354 2006 1231 1242 Thase et al., 1992 Thase M.E. Simons A.D. McGeary J. Cahalane J.F. Hughes C. Harden T. Friedman E. Relapse after cognitive behavior therapy of depression: potential implications for longer courses of treatment Am. J. Psychiatry. 149 8 1992 1046 1052 Zimmerman et al., 2004 Zimmerman M. Posternak M.A. Chelminski I. Using a self-report depression scale to identify remission in depressed outpatients Am. J. Psychiatry. 161 2004 1911 1913