How to Reduce Practice Effects in CIAS Cognitive Trials
How can clinicians and trialists structure cognitive assessment in schizophrenia studies to reduce practice and placebo effects?
Cognitive outcomes in schizophrenia studies can be distorted by repeated-test learning and placebo-related score inflation, making true treatment effects hard to detect. This guide applies to CIAS trials using serial cognitive testing and summarizes the specific assessment workflow used in this study to limit those effects before randomization.
-
Select stable schizophrenia participants before cognitive testing
Enroll adults aged 18 to 50 years with DSM-5 schizophrenia who are clinically stable enough for cognitive outcome assessment. In this study, patients needed PANSS scores of 5 or less on hallucinatory behavior and delusions, stable symptoms, and stable antipsychotic treatment for 3 months at screening, with a total antipsychotic dose of at least 6 mg risperidone equivalents.
-
Build a dedicated screening period before randomization
Use a 5-week screening phase that includes a screening visit plus baseline 1 and baseline 2 visits before treatment assignment. This separates eligibility confirmation from the final pretreatment cognitive anchor and creates room for repeated cognitive testing before the double-blind phase.
-
Administer the MCCB three times before randomization
Give the MCCB at screening, baseline 1, and baseline 2 before randomization rather than relying on a single pretreatment administration. The stated purpose in this study was to minimize training and practice effects during the treatment phase.
-
Use baseline 2 as the pretreatment reference point
After repeated prerandomization exposure to the cognitive battery, compare posttreatment scores with the baseline value obtained closest to randomization. In this trial, change from baseline was analyzed after the three pretreatment MCCB administrations, and additional covarying for change from screening to baseline 2 did not materially alter the primary result.
-
Interpret early postbaseline cognitive changes in light of practice-effect control
Check whether placebo-group improvement is attenuated at the first major follow-up. In this study, no practice effect was observed at the 12-week assessment, and only small treatment and placebo effects emerged by 24 weeks, supporting the usefulness of repeated prebaseline testing for containing early score inflation.
-
Reassess whether repeated prebaseline testing fits the cognitive target
Use caution if the treatment is specifically expected to improve episodic memory or retention. The authors note that repeated testing can itself enhance later performance on episodic memory tasks after long delays, so this approach may defeat the purpose when enhancement of memory retention is the treatment goal.
Clinical Considerations
- The article supports this workflow as a trial-design approach, not as a validated clinical care protocol for routine practice.
- Repeated prebaseline testing appeared to contain placebo response up to 12 weeks, but a small practice effect seemed to reemerge by week 24.
- The authors caution that repeated testing may not always be advisable, particularly for studies targeting episodic memory retention.
- The study had a relatively modest sample size and high dropout rate, which limit confidence in more subtle design effects.
Bottom Line
If you want cleaner cognitive outcome data in CIAS trials, administer the MCCB repeatedly before randomization, but avoid assuming that this strategy is ideal for every cognitive domain, especially episodic memory.
