This white paper explores the importance of choosing an EDC with an API before starting a clinical trial. We discuss the benefits of automated data capture for clinical trials, API-facilitated research opportunities, and the importance of an API in an increasingly interoperable research environment.

Click here to download the white paper as a PDF

How clinical research differs from tech development

Digital therapeutics (DTx) manufacturers may be surprised by how slowly clinical research moves compared with technology development. For example, pharmaceutical or medical device clinical trials can take ten or more years from conception to product commercialization.¹ The clinical world operates by different conventions and practices. Everything from planning to recruitment, to data collection and validation, to regulatory approval takes significant amounts of time.

In addition to the time clinical research requires, it doesn’t always include the connectivity DTx developers may expect. Not all digital research tools, such as electronic data capture (EDC) systems, automatically communicate with information sources or other software tools. Clinical research often includes slowly-moving, siloed components. For example, when a participant visits their local clinic for tests, clinic staff enter results into the participant’s EMR. Research staff may need to manually re-enter the data into the EDC for the clinical trial. Re-entry duplicates time and effort, which impacts trial efficiency.

Yet DTx developers need to enter the world of clinical research to establish their products. Not only are clinical trials often necessary for regulatory body approval or clearance, but they also help validate DTx in the eyes of the medical community, payers, and the public. Clinical trials also provide DTx developers with a wealth of data to improve and expand their products. Despite these benefits, app producers are not taking advantage of the opportunities available through clinical trials. According to a recent review, clinical research supports a meager two percent (2%) of mental health wellness apps.²

App developers may be reluctant to conduct clinical trials because they can be time-consuming and expensive. They may also underestimate the importance of clinical trials in establishing efficacy. Choosing the right trial technology, such as an EDC with an application programming interface (API), takes some of the most time-consuming elements out of clinical trials—making them a more approachable investment.

Effective use of an API can help data systems automatically capture data from varied sources, such as smartphones and consumer wearables. Eliminating a human intermediary improves trial efficiency, data quality, and security.

Automate your data capture

API-facilitated automation revolutionizes data capture during clinical trials. An API enables machine-to-machine data sharing, meaning that researchers can step out of the process of requesting, sending, entering, and processing data. Automation can significantly improve data quality and trial efficiency and can allow researchers to focus on trial elements other than data acquisition and entry.³ This table compares manual and automated data capture.

Improving data quality

API-facilitated data capture improves data quality by eliminating transcription errors and supporting an increased quantity of collected data. Data gathered manually naturally includes chances for error, such as disparities in how parameters are measured or recorded, and typos. For instance, a digital monitoring device can record many data points at once and automatically enter them into the patient’s EMR, whereas the analog counterpart must be read and transcribed manually. Consider also administrative staff, inputting assessment results late into the night, may transpose digits by mistake. In comparison, software-enabled data capture ensures consistency, helping to generate reliable, objective data.

To that point, a 2020 Mayo Clinic study comparing manual to electronic data capture for a catheter device found electronic counts more reliable. Not only were errors in entry mainly manual (84.7% vs. 15.3%), but electronic errors ceased entirely after the initial reporting period while manual errors remained consistent throughout the study.⁴

An API can facilitate automatic capture of large amounts of data from disparate sources—everything from consumer wearables to clinical-grade biosensors. For example, a dairy farm used their analytical software’s API to gather data from various machinery and services, uncovering hidden insights. The results helped the farmers make highly informed decisions about maintenance, operations, and animal care.⁵ Although dairy farming might appear far from DTx clinical trials, the same principles apply.

Besides gathering data from disparate sources, an API facilitates data capture by helping different systems work together. For example, a cohort of medical centers struggled to obtain patient-reported outcomes (PROs) from computer-assisted surveys in a separate system from their EMR. The centers reliably implemented an API to communicate between computer-adaptive tests and patient EMRs. Not only did the API-facilitated assessments achieve a high rate of completion, but the centers also minimized workflow disruptions by eliminating the need for clinicians to log in to two different systems.⁶ The centers could get the data they needed without overtaxing staff and other resources.

APIs can improve data entry time and simplify monitoring and documenting clinical trials. API-facilitated automation and remote data capture benefit decentralized and hybrid clinical trials. Since clinical trials often occur in multiple physical locations, having an EDC with a robust API helps researchers collect, store, and analyze data efficiently.

Regular flows of reliable data also contribute to scientific integrity. Dr. Dominik Meller, product owner at Castor responsible for Tech Projects and Integrations, explains the significance of API-driven data capture:

“[It’s the] first time in the history of medicine we can measure, with such diligence, how the therapeutics interact with the patient. This has never been done before,” said Meller. “API is pushing this industry further. Not a single trial or a single product, but the whole industry.”

The sheer amount of good quality data generated through API-facilitated processes makes for robust studies, creating the evidence base DTx manufacturers need to prove their product’s efficacy. Effective usage of APIs must become standard for clinical trials utilizing decentralized methods in the future.

A pragmatic investment

After the initial investment in setup, an API can continue to operate for a long time with minimal administrative costs. Workflows developed with the API can also be used in future clinical trials, meaning that money and time put into an API during development will pay off down the line. For a better idea of how automated data entry makes clinical studies more cost-effective, here are two examples:

• • • The ADAPTABLE clinical trial: If operated in a traditional model without automated data capture, it would have cost more than $150 million. Utilizing novel tech reduced the actual cost by almost 10x the initial figure.⁷
    • The SWEDEHEART national registry: Automated data capture for participant baseline characteristics and end-point data cost a fraction of the cost of traditional trials.⁷

Security made better

Automated data entry also helps with security. An API enables more secure data collection with the benefit of encryption and precise data access control.^5,6 According to Meller, security is one of the primary concerns and not an afterthought for developers designing an API.

Use case: API for remote monitoring services

One Castor client, RSP Systems, needed to bring their non-invasive, convenient glucose monitoring tool to market for patients with diabetes. Moving a medical device from development to market is never easy or quick. RSP Systems needed to capture the measurements from devices in the patients’ homes, collect survey data via electronic patient-reported outcomes (ePRO), and gather data during in-clinic sessions through electronic clinical outcome assessments (eCOA). This data had to be securely captured in real-time from multiple devices from homes in different locations.

Given the amount of data to be captured, the short timeline, and the varied data sources, RSP Systems needed a robust solution to capture data automatically into a single and secure system to save time and personnel.

Capturing multiple measurements from every device was the primary, most complicated challenge. Several ideas were discussed, including manually copying the data from the device into a database, developing a bespoke solution that could pull the data from the device and enter it into a database, and other complex solutions. But then they found a simple solution that required very minimal programming from their team—the Castor API integration.

The Castor EDC API supports authentication and authorization of API calls through the industry-standard OAuth2. With the device and EDC linked together, RSP Systems automatically pushed the data captured on every patient’s device directly into the correct patient record in Castor EDC. The ability to send data directly from a patient’s home into the Castor EDC database enables RSP Systems to accomplish a lot more in a short period and provides them with controlled access to their study data.

Learn how an EDC with integrated API can save you time and resources in your next DTx trial. Get in touch

APIs enhance user experience

In addition to helping with the research and administrative sides of a DTx clinical trial, APIs can also enhance the experience of users or patients. Without API-facilitated data capture, users may have to log in separately to multiple systems to access and contribute to their data. Multiple logins may make the trial process burdensome for users.

APIs can allow users to access multiple systems simultaneously with a single login. Single logins provide a seamless user experience, helping ensure users can easily participate and researchers can get critical patient data.

Research opportunities

Not only does an API facilitate efficient, high-quality data capture, but it also improves feasibility for long-term studies. Studies lasting many years are often necessary to determine a product’s safety and efficacy. However, maintaining funding and research infrastructure over long periods can be difficult.⁸ This table lists how an API can solve many of these challenges.

Clinical studies utilizing research tools with API enablement can avoid common pitfalls of long-term studies. Successful long-term studies have exciting implications for DTx manufacturers, helping researchers see how their products can affect participants during an extended period, and comply with post-marketing surveillance regulations.

Interoperability is the future

Standardization in data sharing is becoming the norm. For example, Fast Healthcare Interoperability Resources (FHIR), a commonly-used standard for exchanging healthcare information electronically, is used in almost 96% of US hospital EMR systems.^{9, 10} Research software without a good API to facilitate data sharing with EMRs will be critically deficient soon.

Choosing an EDC with an API-forward strategy helps DTx manufacturers react flexibly to technological advancements. EDCs with APIs can easily switch out or add new software systems to their ecosystem. Manufacturers can respond quickly and efficiently to the rapidly changing world of technology.

Also, choosing an EDC with an API prepares DTx manufacturers to fully use available patient history data (with proper consent, of course). In 2020, the Office of the National Coordinator for Health Information Technology (ONC) added a final rule to the 21st Century Cures Act to standardize API access to EMRs.¹¹ The vision behind the rule is an ecosystem of apps that can work with any hospital system in the US—providing patients with access to their health records and apps (or medical treatments) with the information they need to deliver personalized care.¹²

In theory, the 2020 final rule looks promising for DTx manufacturers using EDCs with APIs. Unfortunately, implementation poses challenges. EMR vendors and healthcare systems fight to control patients’ records, and hospital systems question who will pay to update archaic technology.¹² Advocates for APIs must continue to push for governmental regulations that help APIs communicate with EMRs.

Automate data capture successfully with APIs in digital therapeutics

The clinical research world moves at a different pace than the tech world. Technology that enables interoperability, such as an API, may not be a given. That’s why it’s essential to consider research software with a strong API when looking for the tools you need to make your trial a success.

APIs help automate data capture—increasing the quantity and quality of trial data while streamlining processes. Built properly, they are cost-effective, secure, and free up researchers to focus on the trial itself. Automated data capture also creates new research opportunities and is a must to keep pace with rapid increases in interoperability and long-term monitoring.

Getting ready to choose an EDC for your next DTx trial? See how Castor’s EDC with integrated API can be a great fit!

Talk to sales

References

1. How long does the FDA take to approve a drug? US Department of Veterans Affairs. https://www.hiv.va.gov/patient/clinical-trials/drug-approval-process.asp. Accessed November 9, 2021.
2. Torous J, Bucci S, Bell IH, et al. The growing field of digital psychiatry: Current evidence and the future of apps, social media, chatbots, and virtual reality. World Psychiatry. 2021;20:318-335.
3. Zhang J, Sun L, Liu Y, et al. Mobile device-based electronic data capture system used in a clinical randomized controlled trial: Advantages and challenges. J Med Internet Res. 2017;19(3):e66.
4. Stone EC, Miller V, Shedenhelm HJ, et al. The validity of validation: A practical assessment. Infect Control Hosp Epidemiol. 2020;41:400-403.
5. Ferris MC, Christensen A and Wangen SR. Symposium review: Dairy Brain—informing decisions on dairy farms using data analytics. J Dairy Sci. 2019;103:3874–3881.
6. Bass M, Oncken C, McIntyre A, et al. Implementing an application programming interface for PROMIS measures at three medical centers. Appl Clin Inform. 2021;12(5):979-983.
7. Marquis-Gravel G, Roe MT, Turakhia MP et al. Technologyenabled clinical trials: Transforming medical evidence generation. Circulation. 2019;140:1426–1436.
8. Arslan RC, Walther MP and Tata CS. formr: A study framework allowing for automated feedback generation and complex longitudinal experience-sampling studies using R. Behav Res Methods. 2020;52:376-387.
9. FHIR overview. HL7. https://www.hl7.org/fhir/overview.html. Accessed November 17, 2021.
10. Digital Quality Summit: FHIR for dummies (or the forgetful). NCQA Communications. https://blog.ncqa.org/digital-quality-summit-fhir-for-dummies-or-the-forgetful/. Published July 15, 2021. Accessed January 13, 2022.
11. 21st Century Cures Act. FDA. https://www.fda.gov/regulatory-information/selected-amendments-fdcact/21st-century-cures-act. Updated January 31, 2020. Accessed November 17, 2021.
12. Gordon WJ and Mandl KD. The 21st century cures act: A competitive apps market and the risk of innovation blocking. J Med Internet Res. 2020;22(12):e24824.