What is the Difference Between 'Glass Box' and 'Black Box' Clinical Intelligence?

Clinical trials continue to transform due to recent advancements in technology. Jiang Li, CEO of Vivalink, spoke with Pharmaceutical Executive about these advancements, along with the new issues that trial designers must keep in mind.

Pharmaceutical Executive: What is the difference between 'glass box' and 'black box' clinical intelligence?
Jiang Li: The glass box aspect is the transparency, auditability, and traceability of odata collected in the clinical trial process. What we see in the industry right now is a lot of vendors provide a black box approach. Or they may have their proprietary score. It's very hard for a clinician or PI to verify the underlining physiology.

However, in the glass box approach, the raw and timestamped data (such as physiological signals like ECG, temperature, or blood pressure) are used to generate the final synthesized insight. That's what we mean by glass box with the raw data with timestamps all provided together with the insights.

PE: How does patient adherence impact both the science and finance of clinical trials?
Li: Obviously, adherence is a key success factor. What you can say is that the missing data of the clinical trial is really the data capital. A trial sponsor invests a certain amount of dollars, and if there's a lot of missing data, it's a waste of the budget.

The adherence is, ultimately, a denominator for their ROI. The missing data with the core adherence creates some data gaps in the analytics, and it could impact the ability for the sponsor or PI to understand the trends of the data as well as what's going on with the patient.

Continuous monitoring is a new frontier or weapon for the sponsors to have better insights and understanding what's going on with trial participants. The evidence generation is moving from snapshots to real time reality.

If trial participants only go to the clinic and they capture data over there for 20 minutes, versus capturing the data for continuously for one month, it's 20 to 30 times more coverage. That's really increased the insights and the span of understanding of the patient physiologic information.

PE: Why is the failure rate for DCTs so high?
Li: In a traditional sense, DCT has a long history, but it really started booming in the pandemic days.

One key failure mechanism is the logistic implementation for the DCT trials. A lot of times, when you go to a global trial, there's a lot of logistic management for DCT. Let's say you must ship the sensor case to 10 different countries to 100 sites. There's a lot of logistic work to be done.

Furthermore, patient adherence to those sensors must be really good.