Ds252 1.2

The British Journal of Psychiatry (2009)194, 252–259. doi: 10.1192/bjp.bp.108.057554 Current episode duration, weeks: mean (s.d.) Montgomery–A˚sberg Depression Rating Scale Hamilton Rating Scale for Depression–17 SSRI, selective serotonin reuptake inhibitor.
Antidepressant dosage in milligrams by study week 12 for MADRS (w2(1)=5.87, P=0.015) and BDI (w2(1)=7.14, Response and remission with last observation Analysis of clinical trials with missing data has traditionally been performed with the last observation carried forward (LOCF)procedure. It has been demonstrated that this approach introduces We caution against the interpretation of the significant differences a systematic bias and underestimates the level of uncertainty.1,2 in LOCF analysis for several reasons. First, unlike the mixed-effect Therefore, LOCF is not considered a valid analysis by many models, the LOCF analysis does not take into account variations experts in the field.3 However, some researchers still prefer LOCF between study centres. Second, LOCF introduces systematic bias analysis for the purposes of simplicity and comparability with in the context of drop out.1–3 Indeed, the results of LOCF analysis previous reports. Therefore, we report Montgomery–A˚sberg replicate a previous report by Joyce et al5 and illustrate the effect Depression Rating Scale (MADRS), Hamilton Rating Scale for of differential attrition on the LOCF analysis. The meaning of Depression–17 (HRSD) and Beck Depression Inventory (BDI) at differences in ‘remission’ in the absence of any significant week 8 and week 12 with the LOCF in cases of drop out, switching differences in ‘response’ or in the rate of change is questionable.
or missing data. We report the rates of response (50% symptom This discrepancy between analyses is explained by inequality in reduction) and remission (HRSD=7; MDRS=10; BDI=10) accord- baseline severity among non-randomly allocated participants and differential attrition among randomly allocated participants.
Lane P. Handling drop-out in longitudinal clinical trials: a comparison of theLOCF and MMRM approaches. Pharm Stat 2008; 7: 93–106.
Depression severity levels at weeks 8 and 12 with missing valuesinputted following the LOCF rule are presented in online Table Mallinckrodt CH, Clark WS, David SR. Accounting for dropout bias usingmixed-effects models. J Biopharm Stat 2001; 11: 9–21.
DS3. There were no significant between-drug differences in Leon AC, Mallinckrodt CH, Chuang-Stein C, Archibald DG, Archer GE, Chartier response (defined as 50% score reduction on MADRS, HRSD–17 K. Attrition in randomized controlled clinical trials: methodological issues in or BDI). More participants allocated to escitalopram reached cri- psychopharmacology. Biol Psychiatry 2006; 59: 1001–5.
teria for remission at week 8 on MADRS (w2(1)=4.02, P=0.045), Frank E, Prien RF, Jarrett RB, Keller MB, Kupfer DJ, Lavori PW, et al.
HDRS (w2(1)=5.69, P=0.017) and BDI (w2(1)=5.69, P=0.017).
Conceptualization and rationale for consensus definitions of terms in major More participants allocated to escitalopram reached criteria for depressive disorder. Remission, recovery, relapse, and recurrence. Arch GenPsychiatry 1991; 48: 851–5.
(w2(1)=11.05, P=0.001) and on BDI (w2(1)=3.84, P=0.050).
Joyce PR, Mulder RT, Luty SE, Sullivan PF, McKenzie JM, Abbott RM, et al.
Patterns and predictors of remission, response and recovery in major Analysis of the randomised sample confirmed differences in rates depression treated with fluoxetine or nortriptyline. Aust N Z J Psychiatry of remission at week 8 on BDI (w2(1)=5.57, P=0.018) and at week Descriptive statistics, response and remission with last observation carried forward (LOCF) procedure MADRS, Montgomery–A˚sberg Depression Rating Scale; HDRS–17, Hamilton Rating Scale for Depression (17 items); BDI, Beck Depression Inventory.

Source: http://tnu.dk/fileadmin/CPG/Psykiatrisk_genetik/Depression/GENDEP_Supplementary_tables_differential_efficacy.pdf