Triple Growth and Four Major Challenges: Digital Therapeutics Clinical Trials Push Forward

In early March, the Hainan Provincial Health Commission held a promotion meeting for the Hainan Digital Therapeutics Trial Center and awarded plaques to the first batch of 20 provincial-level digital therapeutics clinical trial centers in the province. These clinical trial centers cover nine areas: chronic disease management, rehabilitation, oncology, sleep disorders, orthopedics, psychiatric, behavioral and cognitive disorders, ophthalmology, nutrition, and the Boao Lecheng International Medical Tourism Pilot Zone’s Clinical Research and Translation Base for Digital Therapeutics.

During the five-year construction period, these clinical trial centers are required to meet quantifiable targets to effectively advance clinical trials for digital therapeutics. For instance, each center must conduct more than five clinical studies and evaluations of digital therapeutics, thereby promoting the clinical translation of key technologies and products. Additionally, they must complete research on at least one high-quality clinical diagnosis and treatment guideline for digital therapeutics. Furthermore, each trial center must ensure that at least one digital therapeutic product completes regulatory filing within one year, entering the stage of outcome transformation.

Following the official announcement of the clinical trial center’s accreditation, the first clinical trial since its establishment was quickly implemented. On March 15, a clinical trial led by Hainan General Hospital, titled “Evaluation of Sodium-Control Management Software as a Digital Therapeutic for Improving 24-Hour Urinary Protein in CKD Patients,” was officially launched. The aim is to assess the impact of the latest generation of digital therapeutic technology for sodium control on 24-hour urinary sodium excretion in clinical CKD patients.

As digital therapeutics gradually transition from concept to practical application, a growing number of industry participants are recognizing that clinical trials play a pivotal role in unlocking the full potential of this field. However, it is evident that this is no easy task. What is the current state of digital therapeutics clinical trials in China, and what challenges do they face? VCBeat (WeChat ID: VCBeat) has investigated these questions.

In terms of the definition of digital therapeutics, it is widely recognized that digital therapeutics provide patients with treatment or intervention measures based on evidence-based medicine. These interventions are driven by high-quality software programs and are used for disease treatment (prevention, diagnosis, and control).

Among these, reliance on evidence-based medicine is the core feature that distinguishes digital therapeutics from digital health. Generally, evidence-based medicine requires validation through high-quality clinical trials to demonstrate the efficacy and safety of digital therapeutics in treatment or intervention. It is precisely for this reason thatVCBeat has always maintained that digital therapeutics, in their full sense, must be supported by clinical trials. This view has now largely become a consensus.

However, during the early stages of the emergence of the digital therapeutics concept in China, there was some divergence over whether digital therapeutics must undergo clinical validation. At that time, some viewpoints held that digital therapeutics primarily targeting consumers might not need to go through the medical device registration process, and thus would naturally not be required to complete clinical trials.

However, as a new product, it is clear that patients do not fully understand it and may still seek advice from their doctors. For physicians, recommending treatments with unproven efficacy typically presents significant challenges from various stakeholders. Increasingly savvy patients are also skeptical about the effectiveness of this novel form of intervention; they may view it as a nice-to-have trial if offered free of charge, but would find it difficult to accept if a fee were required.

“I believe digital therapeutics is a sector that integrates medicine and the internet, and its future commercialization will follow the same path. It does not necessarily mean products must be purchased in hospitals; however, the prerequisite is that they must first be backed by clinical evidence and gain professional recognition before they can be marketed and brands established. Otherwise, digital therapeutics would merely become digital health supplements, with companies making unsubstantiated claims—how could consumers possibly purchase them on their own under such circumstances?” This statement by Zhang Jing, CEO of Senmei Medical, may represent the prevailing view.

It is not only difficult to sell; from the perspective of industry development, the lack of robust evidentiary support is clearly detrimental to an emerging product, particularly one driving innovation in the serious medical sector.

“I believe clinical trials are of paramount importance. Therefore, I have always held myself to this standard. In my view, digital therapeutics are first and foremost ‘therapeutics,’ with ‘digital’ serving merely as a modifier. All healthcare professionals should adhere to high standards to prevent bad money from driving out good and to safeguard the industry. We must remain grounded and focus on building a solid foundation. Once this foundation is established, not only will doctors and patients benefit, but companies will also gain corresponding advantages.” This was the perspective shared by Qin Lu, CEO and Founder of Fanju Technology, in an interview with VCBeat.

“When I first started communicating with doctors many years ago and suggested that virtual reality (VR) could be used for the diagnosis and treatment of certain neurological disorders, their initial response was that VR was not yet viable. Do you know why? Because they had tried too many substandard VR devices, which led them to believe that the technology had not yet achieved the desired outcomes. Nowadays, digital therapeutics are increasingly coming into public view. If clinicians’ only exposure is to ‘digital health’ software products that lack validation through clinical efficacy evaluations, their initial reaction may similarly be that digital therapeutics are ineffective. Therefore, the absence of verification through clinical evidence can undermine trust in digital therapeutics and result in inconsistent therapeutic outcomes. I believe this is a critical issue affecting the future development of the digital therapeutics industry,” she added.

Moreover, the recent surge in popularity of digital therapeutics has attracted substantial investment. Naturally, investors impose corresponding requirements on portfolio companies, and the ability to successfully complete clinical trials and obtain medical device registration certificates is generally regarded as a critical factor in pre-investment due diligence.

Liu Xiao, Executive Director of YuanBio Venture Capital, who is well-versed in the digital healthcare sector, once stated in an interview with VCBeat: “We maintain very high standards for our portfolio companies. First, they must possess strong core competencies, including R&D capabilities, the ability to collaborate with clinical institutions, and proficiency in medical device registration. Second, they must demonstrate robust commercialization capabilities, ensuring effective market promotion within a reasonable timeframe.”

He stated,Digital therapeutics must return to the fundamentals of medical practice, adhere strictly to regulatory boundaries, and first demonstrate clinical value through clinical trials or real-world studies. Only on this basis can market potential be discussed and commercial value be proven.

Correspondingly, the ability to achieve corresponding clinical trial milestones is of great significance for enhancing a company’s valuation. Although this topic is often avoided in public discourse, it has indeed become a driving force for digital therapeutics companies to advance their clinical trials.

As time progresses,An increasing number of companies have gradually recognized the necessity of clinical trials for digital therapeutics and are allocating more resources to them.

VCBeat searched the Chinese Clinical Trial Registry website using keywords such as “digital therapeutics,” “software,” “digital,” “mobile phone,” and “virtual reality.” According to incomplete statistics, we found that 60 clinical trials related to digital therapeutics were registered between January 1, 2021, and March 31, 2023 (the statistical results may contain omissions or errors and are for reference only). This performance is considered moderate for an emerging field.

Partial Digital Therapeutics Clinical Trials as of March 31, 2023 (Chart by VCBeat)

Meanwhile,The Trend of Rapidly Increasing Clinical Trials Is Very ObviousAccording to statistical results, there were only 13 relevant clinical trials in 2021. By 2022, the number of clinical trials increased nearly threefold to 37. In the first quarter of this year, 10 relevant clinical trials had already been registered, representing a twofold increase compared to the same period last year. If the implementation of clinical trials across the nine key areas in Hainan proceeds as planned, the number of digital therapeutics clinical trials this year is expected to see significant growth compared to last year.

Another piece of good news is,In terms of the types of diseases involved, an increasing variety of digital therapeutics have begun to explore clinical trials., covering a wide range of disease areas including mental health and psychology, sleep disorders, sports rehabilitation, diabetes, hypertension, fatty liver disease, pain management, chronic kidney disease, strabismus, amblyopia and myopia, chronic respiratory diseases, speech delay, and addiction withdrawal.

While the rapid increase in the number of clinical trials is commendable, VCBeat’s extensive research has revealed that, beyond quantitative growth, there are still many areas for improvement in the current state of digital therapeutics clinical trials.

The most critical issue may be the lack of unified clinical trial standards that align with the characteristics of digital therapeutics.This is also an issue raised by nearly every company that has communicated with VCBeat.

There are currently no guidelines for clinical trials of digital therapeutics in China. Given the wide variety of products, companies are uncertain about how to set clinical trial objectives, which clinical software to use, and which clinical endpoints to select.

Fang Qiuxue, Marketing Director at SLANHEALTH, told VCBeat that there is currently no unified standard for clinical trials of digital therapeutics in China. As a result, even digital therapeutics targeting the same disease exhibit significant variations in trial design; some products are designed for randomized controlled trials, while others adopt single-arm trials. Meanwhile, the number of participants enrolled in these trials can vary considerably.

“Some trials may apply strict inclusion and exclusion criteria, resulting in a final enrolled cohort of only dozens or even fewer than twenty participants. Frankly speaking, this reflects a degree of artificial control. In contrast, other trials enroll thousands of participants, with some pre-registration investigative studies enrolling nearly two thousand subjects. The disparity between these scenarios is substantial. So, how should product efficacy be properly assessed?” she explained to VCBeat.

“Evidence-based standards for clinical trials targeting the same medical condition also vary, meaning that the guidelines selected differ. The endpoints based on evidence, i.e., the endpoints set in clinical trials, are also different. Furthermore, there is currently no unified standard for the implementation and management processes of these trials. Therefore, overall, I believe that clinical trials for digital therapeutics are currently rather disordered, resembling a ‘hodgepodge.’” she added.

“The digital therapeutics industry is still in its infancy. Consequently, there is a lack of experience and unified standards across all stakeholders, including R&D enterprises, hospitals, contract research organizations (CROs), and regulatory authorities.” Fanju Technology’s digital therapeutics product received certification as an innovative medical device in Zhejiang Province last year, granting it access to an expedited approval pathway. Qin Lu also acknowledges that the absence of standardized guidelines is one of the most significant challenges facing clinical trials for digital therapeutics. “Take CROs, for example; we approached many firms, but very few had prior experience with similar products.”

She stated that all forms of clinical trials require a baseline calculation to determine the sample size, thereby establishing an approximate number of enrolled cases needed to demonstrate product efficacy. However, as digital therapeutics are still in their early stages of development, there are few precedents to provide reference baselines. Coupled with the lack of unified standards, it is difficult to obtain definitive results in practical implementation.

“For clinical trials of pharmaceuticals, the traditional approach is to reference clinical trial data from similar drugs and use this as a baseline for statistical calculations. This way, researchers can estimate the number of participants required to demonstrate efficacy—whether it be 100, 500, or 1,000 cases—at least providing a preliminary projection. However, digital therapeutics are so novel that there is often no reference data available for many indications, making it difficult to calculate the appropriate sample size. Moreover, given their unique characteristics, direct comparison with pharmaceuticals is impractical. So, what should be done?”

Qin Lu told VCBeat that during the evaluation of innovative medical devices in Zhejiang Province, regulatory authorities attached great importance to the process, convening meetings with enterprises and experts to explore uncharted territory together. However, in the absence of uniform standards and regulations, case-by-case oversight may vary across regions, posing significant challenges for local regulatory bodies.

“The standards in Zhejiang are exceptionally high. From our initial application for innovative medical device designation to the clinical trial process, both the regulatory authorities and clinical centers in Zhejiang have maintained a particularly conscientious and rigorous approach. We held meetings with regulatory experts to jointly discuss and study international practices, exploring potential lessons we could adopt. Leaders from the drug administration encouraged us to excel, emphasizing that as the first innovative medical device in the digital therapeutics sector, it must uphold internationally consistent rigor and set a national benchmark. I find this highly commendable, as it fosters healthier development within the industry.” She strongly endorsed the meticulous and serious attitude of the local regulatory authorities.

Of course, the field of digital therapeutics is not entirely devoid of guidance documents, but there is still room for improvement in terms of detail. Liu Xuan, Vice President of Ningdong Medical Technology (Shenzhen) Co., Ltd., who has many years of industry experience, mentioned in an interview with VCBeat: “Regulatory authorities have issued numerous regulatory documents concerning digital health, essentially extending device regulations to the digital health sector. However, after reading them, you only come away with one piece of information—that you must meet the general requirements for medical devices.” Liu added, “Relying solely on reading regulatory documents does not clarify how to practically proceed; interpretation and collaboration from professionals specializing in registration and clinical affairs are still necessary.”

As digital therapeutics defined as medical devices, they inevitably encounter the common challenges faced in traditional clinical trials for pharmaceuticals and medical devices; patient recruitment is naturally one of these issues.

Chongqing Kaiqiao Information Technology Co., Ltd., a well-known enterprise in the healthcare IT sector, recently ventured into the research and development of digital therapeutics (DTx) products. Although the company had accumulated substantial experience in healthcare informatics, this marked its first foray into core medical interventions. In practice, patient recruitment for DTx clinical trials has proven more challenging than anticipated. As Song Xianlin, the CEO, stated, “Clinical trials for digital therapeutics require the recruitment of specific patient populations, which may be very limited in size, thereby posing significant recruitment challenges.”

Theoretically, the characteristics of digital therapeutics can digitize the informed consent process for patients in clinical trials, which not only reduces the burden on research efforts but also enables decentralized clinical trials to easily recruit patients globally at an astonishing speed. For instance, in the groundbreaking Apple Heart Study, 400,000 participants were recruited and enrolled via an iPhone app in just nine months.

However, eligibility for enrollment is a separate matter. Li Dai, Founder and CEO of Wangli Technology, stated that in clinical trials of digital therapeutics, patients recruited through self-assessment or non-professional evaluation during the recruitment phase undergo further screening by professional physicians, resulting in very few patients who actually meet the requirements of data algorithm models for enrollment: “If enrolled patients are not rigorously screened, the result will be a large volume of noisy data, making it difficult to generate valuable research findings. For instance, in clinical trials of digital therapeutics for depression, typically only 10–20% of applicants meet the enrollment criteria. From a cost perspective, screening presents a relatively high barrier, but it filters out noise in advance, helping to ensure the accuracy of data and labels.”

Liu Xuan stated that the difficulty of recruiting participants for digital therapeutics clinical trials depends on several factors. One key factor is the disease area. If the patient population is large, recruitment is relatively easier. “For example, with diabetes, there is a vast number of patients, making recruitment comparatively straightforward. However, the sample size must comply with statistical calculation rules and meet the minimum sample requirements set by regulatory authorities for that specific therapeutic area.”

Patient recruitment for clinical trials is closely tied to the commercialization progress of a company’s product pipeline. Ningdong Medical Technology (Shenzhen) Co., Ltd. obtained its first Class II medical device registration certificate in February 2022. This virtual reality-based digital therapeutic provides adjunctive treatment for acrophobia by offering virtual exposure scenarios. In fact, through discussions with the company’s senior management, VCBeat learned that its product pipeline targets various specific phobias. However, acrophobia was the first indication to complete clinical trial recruitment and ultimately gain approval, indirectly indicating a relatively larger patient population and more favorable market prospects.

Second, there is the cost of clinical trial recruitment. In fact, recruitment costs are closely tied to the disease area. For common diseases, the large patient pool, combined with the relatively low risk inherent in digital therapeutics, makes recruitment easier and less costly. “The cost of clinical trial recruitment cannot be avoided. Even if you are a small startup in the digital therapeutics space with a very tight budget, you still need to allocate funds sufficient to support at least a small-scale clinical trial for your product.”

Digital therapeutics companies need to seek assistance from professional organizations such as CROs during the registration phase. However,Due to the scarcity of CROs with experience in the field of digital therapeutics at this stage, both the explicit and implicit costs of clinical trials for digital therapeutics are exceedingly high.

First is the explicit cost, which is an economic issue. Fang Qiuxue stated that there is often a significant gap between the fees charged by traditional CROs and the budgets of digital therapeutics startups: “There are certain compatibility issues between the clinical trial protocols provided by traditional CRO companies and the protocols required for digital therapeutics products. Their quotes tend to be relatively high, failing to consider the economic suitability for digital therapeutics products. It is not feasible to conduct a clinical trial with an extremely large sample size in the early stages, as digital therapeutics companies cannot bear such costs.”

This is merely the tip of the iceberg, with even greater invisible costs lurking beneath. Currently, there is no unified standard tailored to the unique characteristics of digital therapeutics (DTx) for clinical trials, and many DTx companies lack the necessary in-house professional teams. Consequently, the design of clinical trial protocols and the validation of efficacy rely heavily on Contract Research Organizations (CROs). However, few institutions possess relevant experience in this field. Given the highly personalized nature of DTx products, any flaws in protocol design that prevent the achievement of expected outcomes could prove disastrous for DTx companies.

Qin Lu stated that the design of clinical trial protocols determines the outcomes of future trials: “This encompasses the overall methodological approach, such as choosing between single-arm and placebo-controlled designs, and deciding whether to implement blinding; it also involves designing measures to avoid or mitigate bias, for instance, how to isolate behavioral therapy from other concurrent training modalities to minimize potential sources of bias; furthermore, the inclusion criteria must comply with ethical requirements.”

“Taking ADHD as an example, its current clinical diagnostic criteria rely on subjective rating scales administered in hospital settings. However, these scales are highly subjective. Therefore, during our discussions with experts, we concluded that it is essential to adopt objective methods for evaluation.”

“A well-designed protocol in the early stages makes the results less susceptible to challenge. If there are flaws in the trial protocol design, you might be unable to answer a question posed by an expert reviewer, necessitating a repeat of the trial. The impact would be substantial,” she explained.

Furthermore, high dropout rates and poor data quality represent significant pain points in clinical trials for digital therapeutics. During the discussion, Song Xianlin stated, “If adherence is low, convenience is compromised, and the user experience is not sufficiently friendly, the dropout rate in clinical trials will be relatively high. Additionally, we have found that clinical trials for digital therapeutics generally require the collection of large volumes of data, which may suffer from quality issues such as data loss or errors. These factors can adversely affect the outcomes of clinical trials.”

Undoubtedly, these challenges are more or less present across all fields of clinical trials; however, there may be a significant reason for their particular concentration in the digital therapeutics industry. As digital therapeutics gradually gain popularity,An increasing number of software and internet companies with no prior experience in the healthcare industry are actively positioning themselves in the digital therapeutics sector, driven by optimistic projections of its future. However, these entrants may lack a sufficiently clear understanding of the medical field, let alone the capacity to navigate clinical trials, which remain a formidable challenge even for established industry insiders.In VCBeat’s previous exchanges, there was no shortage of similar “optimists.”

Years ago, Qin Lu also “crossed over” from the software sector into digital therapeutics, giving him deep insight into this trend: “Digital therapeutics are gaining significant momentum, attracting many enterprises with no prior experience in medical R&D. Some lack fundamental understanding of the healthcare industry and treat digital therapeutics as mere software. Take a digital therapeutic product I previously encountered as an example. It was named after a specific disease, yet its primary indication targeted anxiety and depression associated with that condition. This naming convention alone fails to meet medical standards and regulatory requirements. Furthermore, I later learned that what they conducted was not a clinical trial but rather user experience testing prior to product launch. Thus, they continued to develop the product following software logic. Personally, I believe this represents a significant issue.”

Fang Qiuxue has engaged with numerous digital therapeutics startups. She noted that clinical trials require specialized teams and talent, yet non-medical industries often lack sufficient understanding of healthcare and generally do not have professional medical teams. “Clinical trials must meet applicable regulatory requirements and be conducted within a compliant regulatory framework. However, I have observed that many digital therapeutics companies have failed to establish such compliant systems, which can result in clinical trials that do not meet registration requirements.”

It must be acknowledged that digital therapeutics possess unique characteristics compared to existing traditional medical devices. Consequently, clinical trials for digital therapeutics also have their distinct features when compared to those for pharmaceuticals and traditional medical devices.

First, there are significant differences in trial design. The primary delivery mechanism of digital therapeutics is software-based patient use, which makes it difficult to fully conceal treatment group assignments. Consequently, digital therapeutic trials currently typically employ open-label or single-blind designs, whereas traditional medical device and pharmaceutical trials usually adopt double-blind or even triple-blind designs.

“Blinding in clinical trials of pharmaceuticals can be implemented quite easily. I give you a pill, and you do not know whether it is a placebo or the actual drug. However, the content of digital therapeutics is directly presented, whether through visual or auditory information, making it easy to discern and difficult to achieve blinding. How to implement blinding in future clinical trials of digital therapeutics is therefore a subject worthy of further study,” explained Qin Lu to VCBeat.

It is worth noting that a small number of placebos suitable for use as controls in digital therapeutic trials have indeed emerged abroad. Unlike placebos used in traditional pharmaceutical trials, those in digital therapeutic trials may differ substantially from the actual control products. Generally, these placebos remove or alter the core features responsible for the therapeutic intervention, retaining only auxiliary functions. For example, Pear Therapeutics employed a placebo called HealthWatch in the clinical trial of Somryst. While HealthWatch retained certain secondary elements of Somryst, such as its interactive interface, it did not deliver the therapeutic effects for insomnia. Similarly, Akili Interactive and NightWare have also utilized comparable placebos.

In terms of trial categories, clinical trials for digital therapeutics have promoted the decentralization of clinical trials. Fang Qiuxue told VCBeat that a characteristic of digital therapeutic clinical trials can be understood as the use of digital tools. This enables researchers to assess patients' health conditions while evaluating their needs, thereby freeing clinical research from the geographical limitations and the traditional, clinic-centric constraints of randomized controlled trials, significantly reducing costs.

“Many clinical trials for traditional medical devices and pharmaceuticals need to be conducted within hospitals. However, digital therapeutics offer advantages in data collection and remote intervention, often without the need for real-time monitoring, thus providing greater flexibility in terms of time and space. Furthermore, they enable the collection of structured medical data from patients outside hospital settings, such as the automatic acquisition of digital biomarkers via wearable devices. Therefore, clinical trials for digital therapeutics can effectively address the gap in out-of-hospital data.”

Second, the metrics for evaluating treatment efficacy may differ. The evaluation of therapeutic outcomes for certain digital therapeutics is typically based on multi-source data, such as electronic health records and data collected from mobile sensors, whereas traditional medical device and pharmaceutical trials usually prioritize biological markers as the primary evaluation metrics.

“Traditional medications allow physicians to monitor adherence through direct supervision or by testing for drug components in the blood during follow-up visits. However, conventional methods such as blood tests cannot verify adherence to digital therapeutics. Since long-term rehabilitation via digital therapeutics is mostly conducted outside clinical settings, it is difficult for physicians to assess patient compliance even with supervision. For instance, it is hard to determine whether a patient strictly follows the prescribed training regimen or merely goes through the motions. This significantly impacts final outcomes. We need technical solutions to ensure quality,” Qin Lu explained to VCBeat, highlighting the distinction between the two approaches.

Currently, there is a growing number of efforts to develop digital biomarkers specifically targeting clinical endpoints for digital therapeutics. A notable trend is the increasing use of gait alterations as a digital biomarker for patients in the early stages of Alzheimer’s disease. Historically, obtaining real-time data on gait changes was challenging; however, with the maturation and widespread adoption of Internet of Things (IoT) devices and artificial intelligence technologies, the industry now has greater capacity to collect extensive data to validate the role of gait alterations as a digital biomarker. Furthermore, data on behavior, facial expressions, and eye movements are also poised to serve as clinical endpoints for digital therapeutics in the future, with similar developments increasingly emerging across the industry.

“Given the need to collect a large volume of accurate and reliable data, Li Dai stated that the complexity of clinical trial design for digital therapeutics is in a league of its own compared with conventional clinical trials. ‘Digital therapeutics may involve multiple components, such as hardware, sensors, and communication systems. Therefore, the trial design for digital therapeutics must account for a variety of factors, including device reliability, software stability, and data acquisition and analysis.’”

“Clinical trials for digital therapeutics must ensure the continuity of follow-up to meet the need for long-term follow-up and monitoring of enrolled patients, thereby evaluating the long-term efficacy and safety of digital therapeutics,” he added.

Third, digital therapeutics differ in their methods of data collection, storage, and analysis, and impose higher requirements for data security and privacy protection. Clinical trials for digital therapeutics often require the collection of large volumes of data, including patients’ biological data, daily behavioral data, and user experience feedback. These data need to be stored and analyzed using technologies such as cloud computing and artificial intelligence to obtain more accurate results.

“Digital therapeutics often require patients to input personally sensitive information, such as phone numbers and health records. These data are generally de-identified before being uploaded to the model engine. This imposes higher requirements on data security and privacy protection for digital therapeutics, necessitating the formulation of more stringent data security and privacy protection measures,” said Song Xianlin.

Fourth, clinical trials of digital therapeutics are also characterized by iterability and personalization.Therefore, Li Dai believes that clinical trials for digital therapeutics must also account for usage across different versions and diverse populations to progressively enhance their efficacy and safety.

Stay tuned for VCBeat’s article tomorrow on how to address these challenges in digital therapeutics clinical trials.

References:

Huh KY, Oh J, Lee S, Yu KS. Clinical Evaluation of Digital Therapeutics: Present and Future. Healthc Inform Res. 2022 Jul;28(3):188-197. doi: 10.4258/hir.2022.28.3.188. Epub 2022 Jul 31. PMID: 35982593; PMCID: PMC9388922. 

Harmon DM, Noseworthy PA, Yao X. The Digitization and Decentralization of Clinical Trials. Mayo Clin Proc. 2023 Jan 18:S0025-6196(22)00560-2. doi: 10.1016/j.mayocp.2022.10.001. Epub ahead of print. PMID: 36669937.