Exploratory anchor-based analyses to determine minimal clinically important differences for the Brief Assessment of Cognition in Schizophrenia and the Schizophrenia Cognition Rating Scale.

In 30 seconds

This blinded, retrospective anchor-based analysis used data from INTERACT, a phase 2 randomized study of luvadaxistat, to estimate minimal clinically important differences (MCIDs) for cognition measures in participants with schizophrenia and cognitive impairment. Among 140 participants, interviewer-rated improvement was associated with greater change in both measures than no observable change. Distribution analyses suggested subgroup differences of 3.98 points for the Brief Assessment of Cognition in Schizophrenia (BACS) and 5.21 points for the Schizophrenia Cognition Rating Scale (SCoRS); a 4-point threshold was most sensitive to study treatment. The authors proposed potentially meaningful differences of 2–4 BACS points and 3–5 SCoRS points.

Key findings

• The improvement and no-change subgroups differed by 2.08 points in BACS change (P = 0.060) and 3.16 points in SCoRS change (P < 0.001), directionally favouring improvement.
• Distribution analyses indicated potentially significant subgroup differences of 3.98 points for BACS and 5.21 points for SCoRS.
• The authors proposed clinically meaningful differences of 2–4 points for BACS and 3–5 points for SCoRS; a 4-point threshold was most sensitive to study treatment.

Why it matters

Cognitive change in schizophrenia can be difficult to interpret clinically. These exploratory estimates offer potential reference points for judging whether changes on BACS and SCoRS may be meaningful, rather than relying only on statistical differences. They may also help inform interpretation of cognitive outcomes in future studies, while requiring confirmation beyond this analysis.

What to keep in mind

This was a blinded retrospective analysis of a cognitively impaired subsample from a phase 2 study, and the MCID estimates were exploratory.