In a national first for China's digital healthcare market, Beijing Tongren Hospital has successfully completed the country's first legally compliant sale of a massive ophthalmic patient dataset to commercial pharmaceutical companies.

Known for its leading expertise in treating eye diseases, the hospital has formed an ophthalmic database built upon a population of 100,000 people "during more than a decade of in-depth follow-up research and long-term accumulated clinical data", according to the hospital.

It contains data of 30 million entries, with 873 professional quantitative parameters, 188,000 fundus images and 3.5 million OCT images.

The data are sourced from real-world clinical diagnosis and treatment scenarios and cover a wide range of eye diseases.

Included are highly prevalent ocular disorders such as corneal diseases, dry eye disease, myopia, glaucoma, cataract, age-related macular degeneration, diabetic retinopathy and retinal vein occlusion. The database enrolls participants across all age groups.

Relying on two authoritative compliance platforms — the National AI Application Pilot Base (Medical Sector) and the Beijing International Big Data Exchange — the hospital said all sensitive personal information has been anonymized.

"It has proven that the market-oriented allocation of medical data elements is feasible and highly efficient," said Yuan Jin, president of the hospital, adding that it will turn dormant medical records into valuable assets that empower the industry.

Jessica Liu, director of imaging and the autoimmune product line business unit of Jiangsu Hengrui Pharmaceuticals Co, said that it purchased a full-eye follow-up dataset for patients with fundus diseases receiving anti-VEGF therapy, which helps induce regression of abnormal blood vessels, reduce vascular leakage and alleviate edema.

Liu said patients with fundus diseases face a high risk of developing concomitant dry eye syndrome due to long-term anti-VEGF treatment as well as the intrinsic impact of the underlying disease.

However, standardized protocols for ocular surface and dry eye management are currently lacking for this specific patient subgroup in routine clinical practice.

"The acquisition of this dataset aims to scientifically and rigorously evaluate the efficacy and safety of our company's perfluorohexyloctane eye drops in real-world clinical settings, supplement clinical evidence, and provide high-level evidence-based support for clinical treatment decision-making that aligns with real-world practice," Liu said.

Liu added that the company has established protocols to grant data access exclusively to designated teams, with all access activities fully traceable, and that all data usage will be conducted in strict compliance with relevant regulations.

Kevin Wang, head of medical affairs of Bayer Pharmaceutical China, said the acquisition of high-quality real-world data has long been challenged by lengthy cycles, high organizational costs and poor sustainability. "Standardized, transparent and compliant data transactions can meet enterprises' data demands for medical research, while fully protecting patient privacy and data security," Wang said.

The company purchased a longitudinal real-world dataset of fundus diseases in patients treated with anti-VEGF therapies.

It will support real-world studies by enabling in-depth analysis of disease characteristics, treatment pathways, clinical outcomes, and treatment effectiveness.

In addition, it will facilitate the development of artificial intelligence models and research tools based on fundus imaging and clinical data provided by Tongren.

"The transactions marked a pivotal exploratory step toward enabling standardized circulation and value release of medical data elements within the institutional framework," Wang said.

"In the long run, it will help medical big data deliver greater value under compliance constraints, facilitate positive interaction among scientific research, the pharmaceutical industry and the healthcare system, and play a positive role in empowering the development of precision medicine and high-quality development in the healthcare sector."

Yuan, from the hospital, said its data will mainly be targeted toward pharmaceutical enterprises for R&D of innovative drugs, medical technology companies for AI algorithm training, universities and research institutes, as well as public health and epidemiological research institutions.

The access threshold is stringent. Applicants must have legitimate and compliant purposes for data use. They shall also pass compliance review conducted by authorized designated data trading institutions approved by relevant regulatory authorities.

Within the scope of compliance with laws and regulations, a special data trading agreement shall be signed.

Based on cost estimates, data costs vary significantly depending on labeling depth. In addition, the terms of use will be clearly stipulated in the special agreement, and data usage must be strictly limited to the agreed scenarios, according to Yuan.

He added the dataset will expand to more regions and will extend from pure imaging data to multi-modal integration of imaging, genomics and molecular samples, covering more rare diseases and subdivided disease types.

"Other tertiary hospitals with superior specialized departments can also draw on this pathway to translate their respective specialty advantages into data assets, thereby activating China's massive medical data resources as a whole," he added.

chenmeiling@chinadaily.com.cn