Scientific and technological advances in drug development and discovery have significantly changed the pharmaceutical industry. And while drugs are becoming more targeted and more effective, the average length of time to execute the necessary clinical trials and gain approval for an experimental drug still takes about a decade and costs billions of dollars. Technological advances, such as artificial intelligence (AI), hold excellent promise to lower costs and create efficient solutions that will facilitate getting drugs to the market more quickly. Examples of cost and time-saving functions currently being performed by smart programs are; auto-generation of Case Report Forms, expediting target patient outreach, screening and enrollment, data capture, and protocol review. Matching eligible patients to a trial is a critical step in the successful conduct of drug evaluation and clinical trials. Whitehouse.gov statistics report that only 3% of cancer patients, who could benefit from clinical trials, are enrolled. While this type of technology has clear benefits for patients, as well as other stakeholder, researchers have only scratched the surface of understanding what intelligent algorithms can provide to enhance the speed, safety, and efficacy of clinical trials execution.
Artificial intelligence is defined in the dictionary as “the capability of a machine to imitate intelligent human behavior”. While the essence of this statement is true, more accurately, AI leverages vast amounts of data to refine underlying algorithms to produce a system that provides solutions through recognition of learned patterns and experiences. This “learning” can provide researchers with an invaluable tool to enhance clinical trial planning and execution and significantly decrease human error and protocol omissions. During a recent AI software demonstration that members of our company attended, the programmers were able to show that the AI-based clinical trials system they created could augment human intelligence during FDA protocol development by highlighting deficiencies in the inclusion/exclusion criteria and indicating “missed” steps in the data collection timeline. In addition, during conduct of the trial, the system prompts and reminds researchers and patients of data collection points and missed data entries. Missing data and patient drops represent a historically challenging problems in conducting clinical trials. Removing some of the human element in tracking and completing forms and protocols would eliminate some of these issues and greatly enhance trial accuracy and completeness.
Although technological advancements including artificial intelligence provide the opportunity to personalize and optimize clinical trials, in many cases, research centers are still using spreadsheets, paper, and phone calls to execute trials. This is somewhat surprising as recent reports have indicated that the use of an artificial intelligent program can reduce the number of required personnel and time to conduct the study by 30-50%. By incorporating AI, highly-skilled personnel can focus their attention on trial specific research tasks and not on more routine tasks such as phone call or text reminders or reviewing email or a website to ensure forms were returned and processed. AI programs have been used to match patients to the appropriate trial, consolidate medical history and lab results in one place, predict which patients have a higher chance of dropping out or not being fully compliant, and can utilize historical cases and data from your previous studies to improve current and future trials. An example of how AI can speed the trial process, a researcher that needs a small sample group of patients with a highly specific disease state will probably need six months to identify and enroll those patients, with AI, identification of eligible patients would take minutes and not weeks. According to Cognizant, 80% of trials fail to meet enrollment timelines and 1/3 of Phase III studies are terminated due to enrollment difficulties. Incorporating AI systems into clinical trials would mean enrolling and running more patients in a given time period, enhancing patient recruitment and retention, and decreasing the opportunities for protocol deviations and reports.
A common question asked by clinical trials researchers is whether these technologies are 21 CFR Part 11 compliant, which would determine if the FDA would accept forms, reports, and data produced from the system. During the course of our research, we have found that many of these platforms have gained full 21 CFR Part 11 compliance. As such, creating CFRs, using the technology to improve the conduct and data capture during the trial, and preparing final reports for submission would be acceptable to the FDA and sponsors. Moreover, recent comments by FDA Commissioner Scott Gottlieb, MD., show the FDA’s current belief that incorporating technological advancements are good for patients by delivering effective, cost conscious, and safe drugs, faster. “Our longstanding goal for medical care is to ensure that the right drug or device is delivered to the right patient at the right time. This vision is increasingly possible with the innovative products that are becoming available. These new technologies offer transformative opportunities, but they also challenge the US Food and Drug Administration (FDA) to modernize its approach to evaluating new innovations.” The efficiencies, and therefore, cost and time saving, that can be created by incorporation of new technologies are vast. Researchers can access patient data in real-time, send automated alerts through the system, remotely administer and collect data, track adherence, report patient health outcomes, and effectively communicate with medical directors, sponsors, and regulators, greatly simplifying research compliance and decreasing the time it takes to complete the trial.
After completion of data collection comes the critical process of reviewing and cleaning patient data in compliance with regulatory standards. This process is necessary to get to the final data validation run and “data lock”. This process can take a considerable amount of time and can be complicated by missing data. Incorporating AI into data capture and monitoring during the trial, and during the data review, quality check, and assurance portions of the study, can greatly improve data accuracy and completeness and assist in getting to “data lock” more quickly. Using algorithm-based programs, having all data consolidated one location, and the ability to have real-time linkages to multiple sources can decrease the time, and therefore cost, of certifying a complete data set. Speeding time to data lock can enable biostatisticians to quickly get to work analyzing the data and getting all the results compiled in preparation for FDA submission, the end game of conducting clinical trials. The incorporation of AI can expedite development of a final package, ensuring regulatory requirements have been satisfied and that all required documents are complete and accurate.
Artificial Intelligence has drastically reshaped the world that we live in. Not just in healthcare, but in our everyday lives. There will be more impactful changes in clinical research, hopefully in the near future; however, we must be willing to put aside our pre-conceived notions. We may not be able to completely rely on technology to do all research —we still need the bright minds of clinical innovators—but we will be able to rely on technology to streamline the process and move life-saving drugs to market faster. Developing best practices that combines state-of-the-science technology with innovative-minded medical researchers may be the ideal solution to ensure efficacious, novel, lower-cost medications get to patients in their safest form -------significantly faster.
Rita Simmons is the founder and lead consultant of Novelle, where she provides business and research consulting to companies across a variety of industries. Dr. Simmons leverages her drive for innovation and excellence along with her extensive executive and military experience to help business owners grow their business, drive revenue, and achieve strategic goals. When you’re ready to take your business to the next level, contact Dr. Simmons at firstname.lastname@example.org or connect with her on LinkedIn.