The Potential for Digital Tech in Clinical Trials – How to Improve Engagement & Weed Out Deception

Contributed Commentary By Adam Hanina

September 23, 2019 | Clinical trials take years of dedicated work and cost billions of dollars. Despite the time commitment and spend, only 14% of drugs in clinical trials win FDA approval. For pharma companies, behavioral data during clinical trials could be the difference between success and failure, but these insights have been untapped thus far. Ensuring that the participants’ behavior aligns with the needs of the trial is essential to the successful development, approval, and commercialization of drugs. Biotech and pharmaceutical companies are increasingly recognizing the role digital technology can play in bridging the gap between the patient and the trial protocol. From predicting which participants are at risk of nonadherence at the start of the trial to notifying site managers when a participant missed a dose, digital technology like artificial intelligence (AI) can help avoid multi-$100 million mishaps by ensuring participants are complying to clinical trial protocols—ultimately reducing clinical trial deception, misleading data and associated costs, while also improving outcomes. 

Understanding the implications of a non-adherent participant 

A participant may not adhere to a trial’s protocol for myriad reasons—from simply forgetting to dose, to choosing not to take the medication due to unpleasant side effects, to intentionally not taking medication and only partaking in trials just for the micro-incentives—also known as the deceptive patient. Research shows that 20% of clinical trial participants never actually take the study drug. If participants take too much or too little of the drug, administer it at the wrong times or miss doses altogether, trial results will be skewed and put it at jeopardy to fail completely. On the contrary, if a trial is successful with the drug approved for market based on skewed results, it could threaten a patient’s safety in real-world situations. 

Beyond the reasons behind nonadherence, the manual assessments used today to determine one’s response and adherence to treatment, such as pill counts or self-reported assessments, are often subjective, easily-manipulated, and costly to manage. These approaches can provide false visibility into participant compliance or be abused by those who wish to willfully mislead sponsors, limiting the value of the trial’s results or lead to improper labeling once a drug is approved. 

It’s no secret that patient recruitment is a key factor in a trial’s success, but implementing digital technology can help. AI-enabled technology can standardize each aspect of a clinical trial including recruiting and retaining quality participants, ensuring adherence assessments are relevant to outcomes in the real world and activating timely interventions. 

Powering Trials With Digital Tech 

From the onset of a trial, data analytics can be leveraged in selecting quality participants by analyzing behavioral patterns and accurately predicting those who are unlikely to adhere to protocols. By leveraging these objective measurement tools to aggregate behavioral data during the screening period at the start of a trial, sponsors can better understand the type of patient enrolled into a trial, and exclude individuals who may raise red flags by exhibiting behavior associated with non-adherence. 

Once quality participants are enrolled, it’s critical to retain and support them to stay on track. While clinicians can conduct exams to capture non-obvious signs that a patient may be having an adverse effect to the drugs, they can only intervene during those in-person checkpoints. In reality, how a participant behaves in-between visits can tell an entirely different story. Advanced-AI algorithms can provide clinicians and site managers visibility into a patient’s engagement at home and identify reasons behind any patterns of nonadherence—allowing them to then intervene before patients drop out of the trial. 

These remote monitoring capabilities ensure patients follow medication regimens with reminders for when and what medications to take and notify clinicians when patients are late or miss a dose, so they are able to intervene in a timely manner. Because clinical trials can take years, ensuring long-term adherence helps drug companies gather and report more accurate data. Along with reminding patients of medication dosages, digital tech can analyze patient data from trials to identify trends, behaviors, and patterns of each individual, providing insights into the patient’s condition and allowing companies to create risk scores. 

In addition to nonadherence, AI and digital technologies are being used to visually identify patient deception. In research my team presented at the American Society of Clinical Psychopharmacology last year, we found that these subjects enroll in trials with no intention of participating and can account for anywhere between 2-30% of clinical trial patients (Shafner, L., Bardsley, R., Hall, G., & Hanina, A. (2018). Using Computer Vision and Machine Learning to Identify Patterns of Fraudulent Participant Activity in CNS Trials. American Society of Clinical Psychopharmacology Annual Scientific Meeting Proceedings). 

However, thanks to the power of AI, site managers are able to spot the behavior before it interferes with trial results. Ensuring patients have followed trial protocols and prescribed medication regimens helps site managers to identify deceptive behavior and provide pharma companies with more accurate data on the success of the drug. 

Digital Tech’s Broader Potential 

There’s infinite potential for digital technology in clinical trials as a risk reduction strategy. Digital technology can help address some of the reasons behind nonadherence—be it forgetfulness, side effects or deception—by remote monitoring capabilities and empowering patients to stay on track, while equipping administrators with actionable insights every step of the way. With these insights, pharma can reduce the costs of care and drug development, as well as improve trial and patient outcomes. 

Adam Hanina is co-founder and CEO of AiCure, a venture and NIH-funded behavioral data analytics company targeting the healthcare and life sciences industry. A healthcare entrepreneur, Mr. Hanina is a passionate advocate for the use of healthcare technology as a population health tool and has directed much of his previous work to this end. Prior to founding AiCure, Mr. Hanina was responsible for further developing Cerner Corporation’s European business development strategy and was a visiting fellow of eHealth at Imperial College in London, UK. He has spoken extensively on the topic of healthcare innovation at industry-leading conferences worldwide and held numerous advisory board positions, including for the Indian High Commission’s Healthcare, Pharmaceutical, & Biotech Committee and Deloitte Consulting’s Health Informatics Network for Europe. He is the primary inventor on over 30-awarded patents. He also published papers about leveraging artificial intelligence for patient engagement and acted as a subject-matter to the National Institutes of Health (NIH). He holds a MBA from the Wharton School of Business. He can be reached at adam.hanina@aicure.com