Ideal clinical trial results: What does success look like?

High quality clinical trial data for decision-making is the ultimate goal of running a study, but it can only be achieved if you define the right inputs. When you start with the ideal results based on the mechanisms of action, treatment profile and market positioning, you can then determine your ideal design to support those objectives and ultimately the ideal data to support the analysis:

The ideal results are our "executive requirements", which then drive the protocol design and ultimately the methods of data capture and data management. By the way, it is best to start with a focus on a high bar of success so that you are always stretching for your best results, best design and best data for every trial.

As an example:

• Ideal Results: We need to understand the impact of our treatment on heart rate as a common side effect of our class of drug; most available treatments have tachycardia as a side effect. Based on competitive analysis we have defined a specific target for safety as 20% or less participants with adverse events of mild tachycardia.

  
  

• Ideal Design: We capture daily heart rate as a cardiovascular endpoint, adding it to the protocol schedule of events for certain key visits (to ensure collection), then deploying a low burden wearable heart rate monitor to capture the data each day.

  
  

• Ideal Data: We ensure that heart rate results are present at baseline, >50% of interim treatment days and end of treatment for every participant; to mitigate risks of non-capture we put in place robust site and participant training, triggered notifications and risk-based monitoring based on key visits and week-on-week adherence thresholds.

The Quality by Design (QbD) approach considers the study design elements related to the endpoint: frequency of heart rate capture, choosing which technology is best for capturing the heart rate data, evaluating the burden of training on the device, supporting participants to keep the device charged, properly strapped and connected to Wi-Fi, and thinking through site monitoring and training to ensure setup/registration and any follow-up with participants is done as needed.

The Risk-Based Quality Management (RBQM) approach focuses on the QbD factors related to the endpoint data, and mitigates key risks that could impact it. This includes defining critical (targeted) monitoring and triggered (risk-based) monitoring. It also includes defining plans to process and clean the data, e.g. dashboards and reports showing missing critical data points, notifications for outlier values that trigger necessary safety reviews, and automated checks that fire when illogical values are entered.

...So the downstream activity of Quality by Design is based on the ideal study design, and the further downstream activity of Risk-Based Quality Management is based on capturing and processing ideal data. But to get to those downstream activities we need to understand our upstream ideal results, in other words "what does success look like for our treatment?".

Resources to help with considering ideal results:

• TPP template from HHS

• Clinical Leader article on protocol design with a good writeup of "objectives"

• A Drug Development Guide from Mt. Sinai

Back to the table of contents.