Apple's ResearchKit at a Crossroads: Questions Mount Over Clinical Validity and Purpose

Apple released its groundbreaking digital health platform, ResearchKit, in 2015 with the aim of revolutionizing the clinical trial market. However, it now faces intense competition from numerous companies in the mobile health sector, with many arguing that ResearchKit has “lost its original purpose.”

When it was first released, ResearchKit was hailed as “a revolutionary medical tool.” But just as the first studies conducted on the ResearchKit platform began to yield results, competitors and digital health analysts stepped forward to question its utility.

AppleResearchKit (image source fromiSmartLiving.net）

ResearchKit was introduced to bridge medical research with consumers in the digital health sector, leveraging health data from smartphones, laptops, and Apple Watch to collect valuable medical information. However, some argue that clinical research requires more specialized data, and the accuracy of such data remains to be verified.

“ResearchKit is well-suited for usability and feasibility studies,” said Ryan Rossier, Vice Chairman of Medullan, an e-health consulting firm. “A closer examination reveals that ResearchKit is merely a data repository rather than a medical platform. Due to insufficient data reliability and professional rigor, Apple has thus far only scratched the surface, whereas the medical community requires more in-depth solutions. If Apple truly intends to enter the healthcare sector, it must advance ResearchKit significantly further.”

In the market where ResearchKit operates, companies such as ePatientFinder, ProofPilot, KlickHealth, and Medidata have flocked in. They all proclaim to the market that they represent a new wave of mobile health clinical trial platforms, capable of providing healthcare providers with medical-grade consumer data.

According to Tom Dorsett, CEO of ePatientFinder, Apple’s ResearchKit offers only “very simple observational tools” and “very basic data.” He stated that Apple’s clinical research platform falls far short of professional medical standards and fails to meet the most urgent needs of the healthcare and pharmaceutical industries. In contrast, ePatientFinder’s clinical trial platform truly excels at accurately matching patients with clinical trials by conducting in-depth analysis of first-hand electronic health records, rather than hastily sifting through large volumes of consumer data to identify potential participants. Leveraging medical-grade data enhances the reliability of trial results, while also necessitating collaborations with healthcare providers such as hospitals.

Another company, ProofPilot, also takes a different approach from Apple’s ResearchKit. The company’s research platform is a SaaS (Software-as-a-Service) solution that enables medical researchers to independently launch targeted studies by accessing appropriate tools and templates through the platform. “As long as researchers know how they want to conduct their studies, they can leverage our expertise to use these tools with ease,” said Matthew Amsden, CEO of ProofPilot, in an interview. “But in reality, they often struggle to translate their conceptualized experiments into automated, efficient workflows. We frequently feel that ProofPilot is essentially a consulting firm wearing the mask of a technology company.”

There is also a company called Medidata that sells medical device data platforms. Its business involves aggregating data from hundreds of thousands of medical devices and apps, and uploading it to the “Medidata Medical Cloud.” The company’s data platform was developed in partnership with Validic. At the platform’s launch, Chairman Glen de Vries declared that this move represented “a major step forward in the mobilization of clinical research.”

Even more threatening is the competition ResearchKit faces from Android. Earlier, a group of researchers at Cornell Tech released ResearchStack, which is somewhat akin to an Android version of ResearchKit. Within months, their first app—MoleMapper—appeared on the Google Play Store.

ResearchStack, an Android-based medical research app platform (Image source: eHealthNews)

Supporters of ResearchStack argue that the platform holds greater potential than Apple’s ResearchKit, given that 70% of smartphone users worldwide are on Android. Clinical trials aim to recruit a broad base of middle- and low-income Android users, who typically have relatively limited access to healthcare services.

It is commendable that the developers of ResearchStack have stated their goal is to create a clinical trial platform accessible to both Apple and Android users. “The top priority for ResearchStack at present is to enable clinical research apps already developed for iOS to be ported to the Android system. Although the same app may not have an exact one-to-one correspondence across the two platforms, the resulting increase in the data pool is substantial.”

A large wave of followers on Apple’s ResearchKit has been drawn in part by Apple’s brand effect, and in part by the fact that many prominent medical organizations have also conducted research through the platform. They view Apple’s platform as an entry point for launching their first population health studies, but they have also raised questions about its practicality.

Although experts believe that Apple’s ResearchKit still requires further development, a study launched on the platform has already yielded promising results.

An article published last December in JAMA Cardiology indicates that the MyHeart Counts app, developed by Stanford University School of Medicine, has demonstrated that mobile apps can be utilized for large-scale clinical medical research. MyHeart Counts was one of the first apps released on ResearchKit for clinical research purposes.

The MyHeart Counts study aims to identify, from real-world evidence, the exercise patterns and frequencies most beneficial to health, as well as the most effective strategies for helping people maintain consistent exercise habits. The advantages of Apple’s ResearchKit platform are demonstrated in this study: traditionally, experiments on exercise and cardiovascular health have relied on self-reported data recalled and recorded by participants, which is prone to high error rates; in contrast, mobile devices can directly measure physical activity throughout the day.

Just recently, Stanford announced the integration of artificial intelligence and data analytics tools into the latest version of its MyHeart Counts 2.0 app, offering users a more personalized health management experience. “Artificial intelligence will help us determine which types of information are most useful or motivating for different population groups,” said a professor of medicine and health research at Stanford School of Medicine who was involved in designing the app.

Euan Ashley, principal investigator of the MyHeart Counts project, added that the most distinctive feature of the new version is its ability to randomize patients and their corresponding experimental interventions. This approach better highlights whether a given intervention has a positive impact on individual health behaviors, thereby enhancing the real-world validity of the study.