Placebell makes a step toward the characterization and the prediction of the placebo response and enables collection of more complete, pertinent and robust clinical data on new drugs in clinical development.
The magnitude of the placebo response has a negative influence in detecting true, statistically significant superiority, of active compounds compared to placebo. Placebell improves your assay sensitivity and facilitates your decision making process.
Placebell reduces the Placebo associated variance and thus significantly increases the study power to detect a true therapeutic effect.
By increasing assay sensitivity, you may reduce sample size and thus study duration and also avoid repeating clinical trials that show effects that do not quite reach statistical significance but strongly suggest a clinical effect.
Sample size reduction, and/or more accurate information, ultimately reduce your clinical development cost.
Positive impact of Placebell on clinical study power may support sample size reduction.
The current scheme
Formal comparisons between study arms, do not take into account individual patient’s placebo response, creating an increased variance of the placebo effect that blurs the results of clinical studies.
The Placebell effect
The Individual Placebo Response (IPR) differs from patient to patient. Placebell provides an Individual Placebo Response (IPR) prediction, then IPR may be used as a baseline covariate. This reduces the variance in both the placebo group and the treatment group.
Patient characteristics are strongly associated with the placebo response. An innovative, placebo-specific questionnaire has been developed that combines information from psychological traits, and personality, with disease information, demographic, and medical history.
proportions given as examples
Placebell solution is easy to implement. Basically patients are invited to fill out a questionnaire on paper or directly on a computer. The answers are computed and individual placebo responses calculated.
Based upon data from robust clinical trials, algorithms have been developed to predict individual placebo response (IPR) in new patients.
At database lock, a detailed report with all patients' individual placebo relative scores is sent to statistician group. This allows them to adjust for individual placebo response (as well as other baseline characteristics).
IPR : Individual Placebo Response
If stratification is needed, Placebell solution is linked to the Interactive Web Response System, which homogenously balances patients among treatment groups according to their individual scores.
In peripheral neuropathic pain (PNP), the placebo response has a pronounced effect and may account for as much as 60% of the drug response furthermore its pharmacological mechanism (via opioid and dopaminergic pathways) is well characterised.
The importance and the variability of this placebo response may jeopardize the ability to demonstrate the efficacy of the studied therapy.
Effectively managing of the placebo effect/response may allow an accurate detection of a positive response by a therapeutic agent which previously may have been missed.
We have studied placebo response in patients suffering from different PNP ethiology (from diabetes, trauma, cancer, …)
The Placebell model resulting from our studies, is a multivariate combination of baseline data: personality traits from a Tools4Patient validated questionnaire, medical history, demographics, and disease intensity measures. This model estimates the placebo score of PNP patients taking part in a clinical trial regardless of which group they are randomized to (active treatment or placebo).
This placebo score estimated at baseline could then be used as a placebo covariate in the statistical analyses to control for the placebo response and to reduce the impact of its variability.
The placebo covariate could also be used for the stratification to balance the placebo response in the study arms.
Use of Placebell may save a double-digit percentage of the variance and related sample size as well as reduce the Type 2 error.
Presented at the Promoting Statistical Insight 2017.
Poster #1.06 presented at the 1st Official Society for Interdisciplinary Placebo Studies conference.
Poster #171 presented at the 35th Americain Pain Society annual scientific meeting in 2016.
Reference DOI : 10.1016/j.jpain.2016.01.074
Poster #46 preented at the 2016 Develop Innovate Advance annual meeting.
Poster #159 presented at the 16th International Association for the Study of Pain world congress meeting in 2016.