Automation Can Solve Diversity Issues in Clinical Trial Recruitment

Matt Wilson uMed clinical trial

How can we make clinical trial recruitment more reflective of the real world? Technology is coming to the rescue. 

The quest for diversity in clinical trials goes to the heart of patient care. Why are some demographic groups over-represented in studies when others are more difficult to recruit? 

Studies that fail to reflect patient demographics skew the results of expensive, time-consuming, clinical research. An example is when a disease may disproportionately affect a group who traditionally are difficult to recruit to studies. Hospitalization of people with Covid-19 with Black or Asian heritage in the UK and the US is higher than for white people, yet they are not recruited proportionally. 

In a recent simulated screening process for a pancreatic cancer trial in the US, Black patients were significantly more likely to be excluded than their white counterparts, with likely higher rates of exclusion on factors such as infection with hepatitis C. However, these factors do not warrant the exclusion of these groups, which may actually be the most likely to use the drug.

Why are certain demographics underrepresented? 

Clinical study recruitment is often undertaken by resource-stretched healthcare providers with limited time for research. This frequently results in failing to represent ethnic minority populations. A recent US paper has suggested the majority of enrolment into clinical studies continues to be white men, with the number of participants from ethnic minorities actually declining in the last 20 years.  

While proactive equality policies such as demographic-based marketing have a role to play, they are usually not sufficient to deliver representative patient cohorts. Traditional patient recruitment relies on cumbersome and time-intensive approaches. For example, in primary care settings, staff manually check a patient’s electronic health record to identify them as relevant for a study. They issue the patient with a standard ‘one-size-fits-all’ invitation to participate in the study. However, these standardized scripts resonate with some demographics far better than with others. 

The time needed to achieve a balanced and diverse patient population can be burdensome, even for a specialist hospital department, let alone for a small family medical practice. Other impeding factors include a patient’s distance from a health center, their educational and cultural background, their socio-economic status, employment patterns, or affluence. Disability, religion, language, sexual orientation, as well as access to digital technologies all come into play.  

How automation can help recruit the right patients into clinical trials

These recruitment challenges have led researchers to look for more individualized engagement. These personalized approaches potentially show how we can boost engagement and create trust with under-represented groups. Emerging patient recruitment technologies — sometimes called automated recruitment platforms (ARPs) — are able to tailor communications between the recruiting healthcare professional and the potential volunteer. 

ARPs are an emerging technology that can open the door to volunteers from traditionally ‘less heard’ groups. For example, an invitation from family doctors using ARP technology could use informative cultural or social imagery for a specific community: a language, ethnicity, gender, or other identity. This is a huge jump from traditional template approaches, creating a relationship of trust and potentially richer, safer, more accurate data to be registered in the patient’s electronic health record.

umed phone demo

ARPs can also extract deeper information on a patient’s experience through more substantive, ongoing interaction via digital channels. These platforms collect patient-reported outcomes through convenient and easy-to-use personalized questionnaires. For example, a question could be “Do you become breathless when getting dressed in the morning?” with closed answer options: ‘yes’, ‘no,’ or ‘sometimes’. This question could be tailored to a particular ethnic or age group who may have a higher risk factor for a disease. Depending on these answers, the patient could receive further questions to help researchers understand if the patient fits the study profile required.

This technology is already demonstrating how it is possible to improve diverse clinical trial representation. In a recent study undertaken by our team, over 25% of patients signing up were from ethnic minority backgrounds compared to 3% in a traditionally recruited study in 2020. The ability to use imagery and content that resonates with under-represented groups has been particularly effective in engaging new volunteers. Using the patient’s electronic health record as the ‘backbone’ of the data, ARPs are able to segment a patient population into multiple subgroups, which can then be engaged with targeted and bespoke content. 

Other advances, such as wearable devices and text communications are boosting engagement rates. All help to save clinician time. Because ARPs use tailored engagement and individualized approaches, they offer an opportunity to screen patients for symptom severity and engage patients with bespoke questionnaires; an example is the issuing of letters to older, less digitally savvy patients, explaining that they may receive a text message, or sending texts at different times for each demographic. We have found that some older people who have a mobile phone usually prefer to receive texts in the morning, while some younger people prefer to receive texts in the afternoon.  

We are working on a study with Queen Mary University of London in Parkinson’s disease, where diabetes is a potential background risk factor. With ARP technology, we are seeing astonishing levels of responses from ethnic minority patients, with an admission rate of around 8%. This means almost 7,000 patients engaging, and 20% of these clicking through and responding. 

In the study, we used recruitment imagery for some under-represented groups, but not with all patients. This revealed a click-through rate of 12% for those who received imagery against 8% for those who did not see an image. These high levels of conversion and engagement are very rare in traditional manual recruitment of patients where the typical conversion rate is around 1-3%.

How could this technology benefit pharma and biotech companies?

Life sciences companies face perennial challenges in recruiting the right mix of patients to clinical studies in the required time span. Besides the challenges in engaging with some demographic groups, there are also technical obstacles in recruiting patients more generally. These include initial failure at the screening stage; impractically small samples of patients from which to recruit through traditional methods as in some rare diseases; and disengagement by patients over the course of a study. All add costs and risks for the industry and make it even more challenging to recruit patient cohorts that are reflective of population diversity and less heard demographic groups.

ARP technology changes this landscape and can help the life science industry recruit the right patients into clinical trials, building evidence in days rather than the usual months or years. For example, in a recent UK-wide hypertension study, ARP technology recruited 150 patients in under four weeks with an average conversion rate of 10% from initial outreach.

No technology is a panacea for such a challenging issue as patient recruitment. However, automated recruitment platforms are already revolutionizing both patient engagement and the personalized, consented, and trust-based sharing of patient data. Used correctly, these platforms can play a vital early-stage role in helping biotech researchers tailor innovation in far more effective ways than in the past, ultimately to bring the right treatments to the right patients.  

Matt Wilson is a former accident and emergency doctor, anesthetist, and medical officer in the UK’s Royal Marines. He founded the uMed platform in 2018, recognizing the benefits to clinical research of automating key parts of the clinical study process by using electronic health record data. The platform is engaging with two million patients in approximately 200 primary care sites in the UK and will soon include some US health systems. 

Explore other topics: Clinical trialLab automation

Newsletter Signup - Under Article / In Page

"*" indicates required fields

Subscribe to our newsletter to get the latest biotech news!

This field is for validation purposes and should be left unchanged.
Labiotech.eu