Mrs. ZHANG Xiaoming

emlyon Global DBA 2018 Intake

ZHANG Xiaoming, Vice President of PPC Group, decided to give herself an unprecedented challenge.Already with double master's degrees, she chose to study at emlyon Global DBA Program and initially planned to focus her dissertation on big data and clinical trial.

Two years ago, an internationally renowned clinical trial company she had served was merged with a medical data company, which inspired her curiosity and thoughts. "How will big data change the industry in the future?"

The deeper the pain point, the greater opportunities AI will bring

Xiaoming chose Prof. Margherita Pagani at emlyon business school as her advisor. Prof. Pagani appreciated her attempt after a long conversation with her yet suggested that she might put big data aside and take a look at the applications of artificial intelligence (AI) in clinical trials.

At the beginning, Xiaoming did not take it seriously, because she felt that the topic was far away from practical applications. At the suggestion of her advisor, she did some homework on the applications of AI in clinical trials of new drugs and found that there were real-world applications at home and abroad.

Behind AI, there are big data and deep learning. Big data is a tool, AI is a vehicle and the combination of the two can create unexpected results.

Inspired by the advisor, Xiaoming began to re-examine the digital innovation in the field of clinical trials of new drugs.Industry insiders understand that the development of new drugs is a time-consuming and laborious project with high failure rate, huge cost and extremely long cycle. Usually, a new drug takes 10-15 years to market and US $1-2 billion to invest.

Due to the complexity of the R&D process, pharmaceutical companies hope to outsource some or all of new drug R&D to specialized companies in order to improve efficiency and cut the cycle. Therefore, many contract research organizations (CROs) have emerged in this field and PPC is one of them, which specializes in pharmaceutical clinical trial services.

In the clinical trials of a new drug, Phase III validates the efficacy of the treatment, which calls for a large number of patient samples. It helps to acquire more details on drug safety and efficacy and provide evidences for the approval of the drug.

Clinical trials are extremely complex processes. The design of the pre-trial plan involves experts in medicine, regulatory registration, statistics and drug safety. The implementation and data analysis after the completion of the design also need to hand over to a professional team. “In clinical trials of new drugs, the biggest challenge is in fact the recruitment of patients. Phase III usually needs to recruit hundreds or even thousands of patients, which takes two to three years to complete. 90% of clinical trial delays are caused by the lack of compatible patients." said Xiaoming.

There are deep pain points in this domain. On the one hand, pharmaceutical companies or CROs are unable to recruit compatible subjects. On the other hand, a large number of patients want to participate in clinical trials of new drugs in order to try the latest treatments or reduce the burden of treatment costs. However, with limited medical knowledge and resources, they can hardly identify the clinical trials suitable for their conditions.

The application of AI has provided Xiaoming and her peers a new alternative. "The application of AI in this field is similar to the internet revolution in the past. Although it is still early, the way of thinking is essential."

Emerging industry opportunities

In a clinical trial of an anti-tumor drug, there are often 20 or 30 filters on patient enrollment, each with a detailed medical description, which needs the supervision of medical professionals to match the appropriate patients.

To this end, some international AI startups have developed AI-based recruitment systems for clinical trial patients. A patient uploads his/her medical records to a system, which automatically searches the clinical trial database to recommend clinical trials of new drugs that may be suitable for the patient. "If it works well, a patient can be matched in a few minutes, which solves the most time-consuming and difficult part of clinical trials," claimed Xiaoming.

First of all, AI application needs to solve the challenge of data sources. Data in the medical industry are often extensive and inconclusive, with data of previous clinical trials, hospital electronic medical record (EMR) systems, medical insurance and drug sales, which are scattered across different systems. “My advisor suggested that instead of studying big data, it’s better to look at AI one step ahead, and the combination of the two disciplines will produce the desired results,” explained Xiaoming.

At present, PPC Group is carrying out a number of clinical trials of new cancer drugs. To this end, she began to collaborate with AI technology companies in China's medical field, preferring specialized companies with large enough medical big data, one of which has already partnered with hundreds of medical institutions in China to build a data network. “These companies have emerged over the past few years and provide professional medical data and AI solutions.”

In addition to patient recruitment, Xiaoming also saw more and more AI applications in clinical trials. Some companies have uploaded clinical trial protocols to relevant platforms in Word or PDF documents, using AI to guide and optimize the design of clinical trials.

Whether a patient who participates in a clinical trial takes or discards the drug upon receipt depends entirely on his/her report. If the patient fails to report it, the trial data will be inaccurate. "The combination of AI and face recognition will be a great solution. On the one hand, it is possible to remind the patient to take the drug via an App. On the other hand, it can also validate whether the patient him/herself is taking the drug, which will ensure the authenticity of the data collection," noted Xiaoming.

In addition to patient recruitment, Xiaoming also saw more and more AI applications in clinical trials. Some companies have uploaded clinical trial protocols to relevant platforms in Word or PDF documents, using AI to guide and optimize the design of clinical trials.

Whether a patient who participates in a clinical trial takes or discards the drug upon receipt depends entirely on his/her report. If the patient fails to report it, the trial data will be inaccurate. "The combination of AI and face recognition will be a great solution. On the one hand, it is possible to remind the patient to take the drug via an App. On the other hand, it can also validate whether the patient him/herself is taking the drug, which will ensure the authenticity of the data collection," noted Xiaoming.

The use of AI in the clinical trials of drugs for Parkinson's disease will also bring an opportunity to the industry.Clinical trials of drugs for Parkinson's disease require the collection of patient behavioral and cognitive data. The current method is that patients go to a physician for data collection, in either paper or electronic records, which is time-consuming and labor-intensive. A potential alternative is to provide patients with wearable devices supplemented by AI. These behavioral and cognitive data can be transmitted to the data collection system in real time, so that physicians get real-time feedback and patients do not have to visit hospitals frequently. "Parkinson patients usually can't describe their own conditions. With AI assistance, the judgment of physicians will be more accurate and timelier, and the clinical data will be more accurate." said Xiaoming.

“Whether AI shall be used to drive organizational change is a decision to be made by the strategic team, not just for the technical team.” Xiaoming has already practiced it on a small scale in her company. Although no one can foresee the future, studying at emlyon Global DBA Program and engaging in theoretical research and exploration helps Xiaoming accurately identify business opportunities. "We are a service provider. If we do not structurally change our business model and keep up with the trend of AI, we will miss this time window."