Establishing Minimal Clinically Important Difference in Depression

We used data (n = 1039) from three large multi-centre randomized controlled trials (RCTs), GenPoD, TREAD, and CoBalT (Table 1) (Lewis et al.2011 (link); Chalder et al.2012 ; Wiles et al.2013 (link)). Each RCT investigated treatment options for depression, used the BDI-II as their main outcome measure, included a ‘Global rating of change’ measure and followed participants over several months providing at least two time periods for analysis. The global ratings of change assessed change relative to the last assessment with a trial researcher, either baseline assessment or previous follow-up visit. Recruited participants in GenPoD and TREAD scored >14 or ⩾14 on the BDI-II, respectively, and fulfilled ICD-10 criteria for a primary diagnosis of depression. The CoBalT study recruited participants with treatment-resistant depression, so in addition to scoring ⩾14 on the BDI-II and fulfilling ICD-10 diagnosis of depression, participants had also adhered to an adequate dose of antidepressant medication for at least 6 weeks prior to entering the trial. The inclusion of CoBalT allowed us to assess the potential differences in MCID for depression that has not responded to antidepressants compared to patients with a new episode of depression (as recruited by GenPoD, TREAD) in primary care.
Table 1.

Sample characteristics

Study	RCT description	Time period	Time description	n	Age, years, mean (s.d.)	Female, n (%)	Depression duration >2 years, n (%)	BDI-II at first time point, mean (s.d.)	Mean BDI-II change, mean (s.d.)	Better, n (%)
GenPod	Reboxetine v. citalopram for depression	1	0–6 weeks	332	39.2 (12.7)	216 (65.1)	49 (14.8)	33.4 (9.1)	−14.8 (11.4)	258 (77.7)
		2	6–12 weeks	328	39.6 (12.8)	222 (67.7)	52 (15.9)	19.1 (10.6)	−4.2 (9.1)	240 (73.2)
TREAD	Exercise intervention v. care as usual for depression	1	0–4 months	288	40.9 (12.5)	191 (66.3)	33 (11.5)	31.8 (9.3)	−15.3 (11.8)	216 (75.0)
		2	4–8 months	213	42.5 (12.6)	150 (70.4)	24 (11.3)	16.2 (11.7)	−1.2 (9.8)	137 (64.3)
		3	8–12 months	203	42.9 (12.5)	142 (70.0)	25 (12.3)	14.3 (11.4)	−2.0 (8.3)	135 (66.5)
CoBalT	CBT as an adjunct to antidepressants v. care as usual for treatment resistant depression	1	0–6 months	418	49.8 (11.8)	307 (73.5)	247 (59.0)	31.6 (10.5)	−9.8 (12.8)	204 (48.9)
		2	6–12 months	385	50.1 (11.7)	285 (74.0)	224 (58.2)	21.4 (13.9)	−2.2 (10.6)	186 (48.3)

RCT, Randomized controlled trial; BDI-II, Beck Depression Inventory, 2nd edition.

Mean BDI-II change corresponds to improvements in depressive symptoms over time in all cases.

Participants in GenPoD were followed for a 12-week period, with follow-up data collection at 6 and 12 weeks. Participants completed the BDI-II and a global rating of change at both follow-ups. Participants in TREAD were followed for a 12-month period, with follow-up data collection points at 4, 8 and 12 months. The BDI-II and a global rating of change measure were collected at all three follow-ups. The CoBalT study followed participants for a 12-month period with data collection points at 3, 6, 9 and 12 months. Global ratings of change were asked at each of these follow-ups, but BDI-II data was collected at the 6- and 12-month follow-ups only. In CoBalT, patients completed the BDI-II and global rating scale as part of the follow-up questionnaire at 6 and 12 months completed primarily during a face-to-face appointment with a researcher. However, participants also completed the global rating scale at 3 and 9 months as part of a follow-up questionnaire administered over the telephone. When answering the global rating of change question at 6 and 12 months, it is likely, therefore, that the majority of patients in CoBalT rated their global change with reference to the telephone follow-up 3 months before. We present participant and study characteristics in Table 1.