MedAvante is dedicated to bringing certainty to the mental-health evaluations and assessments performed during CNS clinical trials. Our methodology provides advanced signal-detection solutions for diagnostic, screening, and symptom-severity assessments. By applying this methodology, MedAvante overcomes the major threat to achieving signal detection: Issues with the trial assessments’ validity and reliability caused by biases and interviewer variability. Our solution provides:
- Appropriate patient enrollment according to protocol
- No baseline score inflation for reduced placebo response
- Highest inter-rater reliability due to continuous calibration
Traditional CNS trials suffer high rates of failure, increasing rates of placebo response and inappropriate patient populations. Until recently, these problems were addressed through rater training at the outset of a trial. Periodic retraining during the study to minimize rater drift could further reduce the problem. However, up until now, this has not been operationally feasible as it does not address ‘noise’ introduced by bias and even the best trainers can each vary in their interpretation of the scales’ conventions and scoring criteria.
An entire cohort of study raters comprised of high quality interviewers can still result in poor reliability for the study as a whole, with an attendant negative effect on study power. That is because each site is managed independently and the sites do not continuously calibrate with each other on what are fundamentally subjective measurements. In addition, over the past ten years the number of investigative sites, with their attendant raters, has significantly increased to meet the patient needs of many studies contributing to study variability and signal noise.
Most importantly, the highest risk factors for a successful trial are posed by biases, specifically enrollment bias and the related baseline score inflation at study outset that can nullify the beneficial impact of even the highest quality interviewing.