October 31th: Digital Medicine News Picked for You

Facebook launches a new tool to help you get health tests, Rock Health published a report on consumerization of health, and other Digital Health news.

Weekly Digital Dose
Share on
#AI#Clinical Trials#Digital Medicine

Dissecting racial bias in an algorithm used to manage the health of populations

The U.S. health care system uses commercial algorithms to guide health decisions. Obermeyer et al. find evidence of racial bias in one widely used algorithm, such that Black patients assigned the same level of risk by the algorithm are sicker than White patients. The authors estimated that this racial bias reduces the number of Black patients identified for extra care by more than half.

Our take: It has been well publicized across industries that algorithms perpetuate bias. For example, facial recognition technology used in police surveillance has significantly different levels of specificity and sensitivity for White vs Black individuals. We also know from landmark studies that racial bias in medical care can lead to different clinical decision making, and recent reviews indicate that these issues persist (see, A Systematic Review of the Impact of Physician Implicit Racial Bias on Clinical Decision Making, published by Academic Emergency Medicine in 2017). Designers of commercial algorithms used in large populations have a responsibility to test for racial bias and account for any findings before they are fully deployed. – An Advisor

Rock Health, Stanford report: Consumerization of health reshaping doctor-patient relationships, data conversations

Researchers found that over the last two years, more patients are using digital health tracking tools and sharing these data with their providers. Patients are also seeking health information online to inform discussions with their doctors — often in the form of self-diagnosis, developing treatment plans, and asking to add or stop a medication.

Our take: The amount of information patients consume from non-medical sources — from patient communities/online support groups to “ask your doctor” ads, WebMD and PubMed — is only going to increase. Providers are increasingly expected to begin their clinical encounter by responding to information that the patient has collected online. It may be time to offer clinicians training on how to respond to patient-generated questions and suggestions that are responsive to their concerns and not dismissive. Reframing this increased patient engagement as an asset with the potential to improve outcomes will be critical. – An Advisor

How Good is Your Webside Manner?

57 percent of respondents in a new Deloitte report expressed an interest in telemedicine and virtual health tools. But, the survey contained some bad news as well. Specifically, patients surveyed reported that their telehealth providers didn’t seem as knowledgeable or as professional as their in-person providers. Only one-third of respondents felt that they received enough information during their virtual care visit. The Deloitte report’s conclusion: clinicians need more training to bridge the differences between a telemedicine visit and an office visit.

Our take: The really interesting finding in this report is that, contrary to expectations, users are accessing healthcare through technology regardless of age, although for very different purposes. The majority of seniors access virtual health to refill their prescriptions, while younger patients are using telehealth services to monitor health and fitness goals. It is likely that the issue of “webside manner” may be due to such different use cases. To improve satisfaction ratings, there may have to be different protocols (and perhaps even different virtual providers) depending on the service the individual is trying to access. – An Advisor

Novartis-partnered TrialSpark pens ‘digital’ trial collab with Limbix

Limbix, which develops mobile apps and virtual reality (VR) focused on helping solve mental health issues, is collaborating with TrialSpark for the “Limbix Spark five-week program”. The program focuses on behavioral activation and guides teens through tasks that teach coping skills and mindfulness.

Our take: Virtualizing behavioral modification has the potential to improve access and adherence to behavioral interventions. Approaching teenage users who are “digital natives” through mobile and VR where they are already comfortable may increase utilization and, ultimately, improvement in outcomes. – An Advisor

Facebook launches new tool to help you get health tests and screenings

The tool, called Preventive Health, launched on Monday in the United States. It focuses specifically on cardiovascular health, cancer screening and seasonal flu, making recommendations such as when to get a cholesterol test or where to get a flu shot. Facebook partnered with the American Cancer Society, American College of Cardiology, American Heart Association and US Centers for Disease Control and Prevention to develop the tool.

Our take: Here’s a rare example of social media using its microtargetting capabilities for good. Rather than personalized ads, users will receive personalized screening recommendations through an interface they already use regularly and are familiar with. The questions will be whether these recommendations change behavior and whether, in turn, outcomes improve. – An Advisor

HealthMode News

Why deep-learning AIs are so easy to fool

Great article showing how introduction of simple noise to a visual signal renders almighty AI completely useless. It is food for thought to chew on while looking into various directions – whether it’s reliability of deep-learning models, expert oversight or data security.

Let's stay in touch

I consent to the use of my data for the purpose of receiving a weekly email. Please see our privacy policy for additional information on how we protect your data.