Opportunities of Digitalization in Clinical Trials
Most industries have been transformed or at least revitalized by digitalization over the last decade. What about healthcare and most specifically here, clinical trials?
It seems that we are still in the early years of digital innovation in clinical trials and I thought it would be interesting to wonder about some untapped opportunities where technology could bring value in trials.
TL;DR — Key takeaways
- Finding the appropriate clinical trial for a patient could be made partially automatic thanks to Electronic Health Record systems, therefore increasing the number of patients involved in clinical research
- Connected medical devices and virtual reality can help the patient engage more proactively in his trial and therefore reduce drop-out.
- Recurrent users of online cognitive tests could be the premise of a pool of eligible patients for clinical trials related to neurodegenerative diseases.
1 — Automated suggestion of trials to patients during medical visits using Electronic Health Record
This opportunity is a promising match between two parts of the healthcare landscape maturing technology-wise.
What are the issues at stake?
Suggesting and presenting adequate clinical trials to a patient at an appropriate time is burdensome both for the practitioner and the patient.
The practitioner needs to stay aware of:
- upcoming clinical trials for a given disease
- eligibility criteria of each trial
- burden of each trial, e.g. will his patient needs to drive 300km each month to have his measures taken?
The patient needs to understand the rationale and implications of a clinical trial within a 30 minutes visit and is expected to give his consent almost right away, which is source of stress and confusion.
How can digitalization help?
As Electronic Health Record (EHRs, or Electronic Medical Records) are being progressively adopted by practitioners and healthcare facilities, some of the patient’s health data will be available for digital queries on demand.
We already have fairly comprehensive open databases for clinical trials, such as clinicaltrials.gov. However, eligibility criteria are loosely structured, in free form text, which makes it very hard to programmatically answer queries like “What are the clinical trials that this patient would be eligible to?”
We would therefore need to apply natural language processing techniques on the free form text to extract structured criteria, that can be queried against. Some initiatives addressing this issue have already been open-sourced.
An integration between the EHR system and database of clinical trials could then enable something along the lines:
When a patient is in a medical visit with a practitioner, the practitioner enters new exam results in EHR and instantly gets a notification if there are clinical trials appropriate for that patient
This would require no action from the practitioner, which is crucial given the time constraints of most medical visits.
2 — Real-time data collected through gamification to reduce patient dropout
What are the issues at stake?
Once a patient is included in a clinical trial, practitioners and sponsors (which are often the pharmaceutical companies) need to monitor health outcomes and potential adverse events or lack of motivation that might lead him to leave the clinical trial or deviate from his treatment indications.
The problem is substantial, with average dropout rate across all clinical trials being around 30%, and can drastically reduce the statistical power of a clinical trial — with even more statistical harm for Intention-to-treat analyses, and thereby indirectly impede access to new treatments. For patients that have drop-out, the trial was a loss of time and a source of burden and fatigue.
Quoting an analysis from PatientCentra:
One of the major reasons patients drop out of a study is because of the time commitment. Data collection can be a prominent source of wasted time for patients because redundancies are often present.
The problem of patient retention seems currently neglected, as visible in this study of the U.S. Department of Health and Human Services that shows that the patient retention costs only account for 0.22% of the overall budget of Phase III clinical trials.
How can digitalization help?
- Connected medical devices
The offer of connected devices, for sport, health or lifestyle, is getting more diverse every year. Some of these devices are directly addressing health monitoring at home, from cardiovascular check-up to weight and body composition.
For patients, using connected devices at home implies:
- Some routine visits during the trial could be avoided, given the connected device is measuring enough markers and is clinically validated.
- In return the patient can get personalised medical feedbacks in near-real time based on his measures and proactively seek refined feedbacks by adding more data.
For practitioners and pharmaceutical companies, having patients using connected devices at home during a clinical trial implies:
- Continuous stream of data coming from patients, enabling monitoring at a lower cost and reduced burden of data entry
- The stream of data could enable machine learning models to predict dropout probability for a given patient given its activity over the last period and therefore deliver tailored guidance and messages to convince him out of quitting.
The quantitative impact of wearables and connected devices has yet to be measured in large clinical trials. Some initiatives were not conclusive, and the broader state of the art is greatly summarised in this article.
It’s interesting to note that some contract research organizations are already proposing support for clinical trials including wearables.
- Virtual Reality for functional exercises
If we push the concept of connected devices further, we could design virtual reality (VR) games that would mimic actual motor or functional exercises or even neuropsychological games.
An example of disease that could benefit from such approach is Multiple Sclerosis, where affected patients are young- hence digital friendly, and affected physically so they routinely need motor and functional exercises.
In the context of a clinical trial it would mean:
For patients,
- More control and engagement over their treatment with the fun and reward system induced by games, with the health-related benefits of doing these exercises frequently.
- Games could be adaptive and evolve as the patient master each exercise, leading to routines tailored to the patient abilities
For practitioners and pharmaceutical companies,
- Higher engagement of patients and therefore reduced dropout rate
- If the VR game is pushed to the status of medical device, results from the exercises in VR could be seen as secondary or exploratory outcomes of interest in the clinical trial
Recent initiatives to let patients contribute to clinical research at home, like FloodLightOpen, give examples of exercises that could be integrated in a VR setup.
3 — Pool of pre-symptomatic patients to accelerate clinical trials
What are the issues at stake?
The most complex neurodegenerative — and most medical conditions actually — diseases are heterogeneous which makes it hard to predict the course for a given patient.
For Alzheimer’s, the vast majority of drugs have failed (>99% of trials showing no drug-placebo difference) to revert damages made to the brain or restore cognitive functions.
The mechanism of action of Alzheimer’s is not entirely clear and several (probably non-mutually exclusive) theories still co-exist at the moment. However one clinical strategy is gaining traction to push effective drugs to the market and consists in targeting patients in earlier stage of the disease, up until pre-symptomatic patients. By targeting younger and healthier patients, pharmaceutical companies and scientists hope to inflect the disease course before irreversible damages.
However recruiting patients in very early stages of a disease is a considerable challenge, partly because:
- These patients might not be managed yet by a physician and therefore are under the radar of medical practice and clinical research
- Early stage also implies more similarity between a patient that will later on develop Alzheimer’s and a patient that will remain healthy despite minor (and quasi-normal) cognitive dysfunctions, making the eligibility criteria a bit weaker.
How can digitalization help?
For neurodegenerative diseases, cognitive tests are used as eligibility criteria for clinical trials.
Many cognitive tests have been computerized and can now be used online at home.
We could push this further by:
- Having healthy (at least pre-symptomatic) people of advanced age regularly do these online tests at home
- Based on the temporal evolution of their scores, the solution could detect and flag “at risk” individuals. Possibly, messages could be delivered to the user to advise him to consult a general practitioner.
- Besides, the community could also be a pool of potentially eligible patients for new clinical trials in neurological diseases. This could significantly accelerate recruitment phases in such trials.
One of the challenge of such platform is that it should be engaging and user-friendly to make it intuitive for people of advanced age to use the platform.
Initiatives like BrainHealthRegistry are a great step towards this goal.
In following posts, we will look at the adoption of digital solutions for other actors of clinical research, like sites or contract research organizations.
