At any one time, there are around 50,000 clinical trials taking place worldwide. Clinical trials are vitally important for pharma companies, accounting for up to 40% of research budgets in many cases. However, large numbers of trials that are proposed fail to recruit patients, and some estimates suggest that up to half of those proposed fail to recruit any participants at all. Of trials that start, up to 85% fail to retain enough participants for the results to be meaningful.
This does not, however, mean that patients are not keen to find clinical trials. For many critically ill patients, especially those for whom current treatments are either unsuitable or simply not working, clinical trials represent a lifeline. Many are desperate to find a suitable trial. Recruiters have been using social media as a way to find trial participants for some time, but it is a bit ‘hit and miss’.
There is, therefore, huge appetite on both sides of clinical trial recruitment to find a better way to match patients with suitable trials. But how?
An answer to the question
Microsoft thinks it may have the answer, or at least part of the answer. Its research lab in Israel has recently developed a Clinical Trials Bot. Starting as a hackathon project, the bot now allows patients and doctors to search for studies about a particular disease. By asking a series of text questions, the bot can assess the patients’ match with the criteria for each trial, and suggest links to the trials that best match the patient’s need. Pharma companies can also use it to find potential trial participants.
Like other uses of bots in healthcare, the bot uses a form of artificial intelligence, machine reading. It assesses the criteria for each clinical trial, and then uses that to ‘decide’ which questions to ask patients, and how to match them to possible trials. For example, a patient might search for ‘trials for woman in California with breast cancer’. The bot would ask questions about the level of cancer—for example, is it metastatic?—and how far the patient is prepared to travel. In each case, the patient is offered five possible choices of answer. As they select answers, the bot chooses the next question. Each question and answer allows the bot to refine the list of possible clinical trials.
At this stage, the bot is still more or less experimental. It is certainly not going into production immediately. However, Microsoft is already clear that it does not see it having a future as a stand-alone Microsoft product. Instead, it is talking to pharma companies about how they could use the bot as a way to find trial participants, and to other partners about its use as a tool for patients. The future of the bot is therefore not very clear, but there is no question that it has potential.
Solving real-world problems
There are a number of other examples of technology having potential in healthcare, however. Not all of them have delivered. Telehealth and telemedicine have been touted as the ‘next best thing’, the option that will allow us to deliver ‘more for less’ in healthcare, and improve patients’ health without needing more doctors or nurses. However, many of these advances remain confined to small areas or local projects. They are certainly not in widespread use. Even apps like Babylon, promising ‘a doctor on your phone’, are not nearly as ubiquitous as the developers had hoped.
Some commentators have suggested that where technology fails to take off, it is because it is not really solving a problem. This is not strictly true, however. Telehealth and apps are definitely solving a problem, but not necessarily for patients.
Telehealth helps healthcare providers. It allows them to look after more patients, with fewer healthcare practitioners. The benefits for patients, however, are less clear. Babylon and similar apps tend to focus on younger people, who are not the main users of health services. Over 70% of healthcare resources are used by around 20% of the population, usually older people with chronic problems. This group are unlikely to be suitable for healthcare delivery via an app.
For technology to take off in healthcare, it must be acceptable and beneficial to patients as well as clinicians and providers. It is early days yet, but the Clinical Trials Bot might just tick that box. Let us hope that it delivers on its potential, and that Microsoft finds suitable partners for its ongoing development.