Five Key Drivers Accelerating AI Adoption in Clinical Decision Support in 2019

2019 is set to be a year of breakthrough development for medical AI. Building on the positive feedback from artificial intelligence research and clinical practice in the healthcare sector over the past two years, doctors, hospitals, insurance companies, and patients are adopting a more open attitude toward AI. According to data from Tractica, a renowned U.S. market research firm, healthcare organizations are projected to spend over $34 billion on AI technologies by 2025, up from $2.1 billion in 2018.

The realization of artificial intelligence’s vast potential must be premised on “transforming how physicians deliver high-quality, cost-effective diagnostic and therapeutic services,” thereby maximizing the quality of clinical medical decision-making. In 2019, the application of medical AI in diagnosis, medical imaging, disease prediction and analytics, and healthcare management will continue to grow. The demand for AI in the healthcare industry stems from the following five aspects:

Irregular medical practices are a realistic issue in clinical settings. Non-compliant diagnosis and treatment may lead to a series of consequences, including doctor-patient conflicts, medical compensation claims, and challenges for health insurance coverage. Among these, drug abuse—such as the misuse or inappropriate use of methadone and morphine—has caused direct economic losses and behavioral health impairments to patients.

Introducing medical AI tools to “remind” and urge physicians to issue medical orders in accordance with clinical practice guidelines during the diagnosis and treatment process is a proven, effective approach to standardized clinical management, which will significantly reduce safety threats to patients posed by non-compliant practices. It is reported that the AI-based Huimei Clinical Decision Support System (CDSS) has been integrated into the electronic medical record (EMR) system at Xuanwu Hospital for stroke medical record quality control. The system automatically checks for deficiencies in physicians’ diagnostic and treatment behaviors while they document medical records and provides timely reminders. As a result, the compliance rate with acute cerebral infarction treatment protocols in the department increased by 33.7% within one month, with multiple quality control targets achieving 100% compliance.

Supporting healthcare operations management is the fastest-growing and largest area for artificial intelligence. By the end of 2019, AI’s market share in the field of healthcare information technology construction alone is projected to exceed $1.7 billion, potentially helping to save over $3 trillion in healthcare waste.

In China, as hospitals gradually implement the Diagnosis-Related Groups (DRG) payment system, they must focus on diagnosis and treatment to fundamentally reduce health insurance costs. AI- and big data-driven scientific decision-making can effectively assist physicians in making accurate diagnoses, facilitating correct DRG classification and reimbursement. Furthermore, by integrating comprehensive clinical patient data, AI can predict disease outcomes as well as the likelihood and severity of complications, thereby helping physicians determine the most effective treatment for each individual patient.

Since President Obama proposed the “Precision Medicine Initiative” in 2015, the term “precision” has arguably defined the future trajectory of disease diagnosis and treatment. This is particularly evident in oncology, where continuous advancements in genomics and immuno-oncology have instilled significant confidence in the industry and delivered tangible clinical benefits. However, achieving true “precision” in medicine is far from straightforward. A study by Scipher Medicine found that 65% of patients treated with the top five best-selling drugs worldwide showed no response to therapy. This underscores a critical reality: relying solely on a limited set of genetic markers, novel pharmaceuticals, or new medical devices is still insufficient to realize truly individualized, “tailor-made” diagnosis and treatment.

In 2019, artificial intelligence will become an accelerator for the development of precision medicine. Deep learning diagnostic and treatment models trained on patient data can rapidly identify drugs and therapeutic regimens suitable for patients, as well as track and predict the progression of individual diseases.

Over the past four to five years, there has been a growing trend of patients undergoing surgical procedures in outpatient settings rather than being hospitalized. Patients are continuously educated on the care philosophy of “the right care, in the right place, at the right time,” encouraging them to seek timely treatment at professional healthcare institutions. Insurance companies have taken the lead in shifting traditional inpatient surgeries to outpatient settings through incentives and patient education initiatives, aiming to improve bed turnover rates and control health insurance expenditures.

Consequently, Ambulatory Surgery Centers (ASCs) equipped for outpatient surgical procedures must maintain high standards of medical and nursing care to support the performance of complex surgical operations. In this context, in addition to offering competitive salaries to recruit talent and meet healthcare demands, the adoption of artificial intelligence to assist in scientific decision-making for diagnosis, treatment, and perioperative nursing represents a rapid, feasible, and reliable approach. Studies have shown that high-quality perioperative care provided by ASCs can help patients achieve lower postoperative infection rates and better prognoses, thereby saving millions of dollars in healthcare costs annually.

The impact of artificial intelligence extends beyond clinical diagnosis and treatment within hospitals; it will also enhance the accessibility of out-of-hospital services. Once patients leave the hospital, physicians may struggle to ensure adherence to prescribed treatment plans or monitor chronic health conditions. AI can bridge temporal and spatial gaps, enabling efficient transmission and sharing of information. Patient data collected via mobile applications will help physicians stay informed about patients’ health status in a timely manner, translating into “insights” that improve clinical decision-making.

In 2019, artificial intelligence demonstrated immense potential to optimize physicians’ workflows and enhance the efficiency and confidence of patient care decision-making. With the deep integration of “AI + Healthcare,” AI applications are becoming an indispensable tool in the modern physician’s arsenal, akin to the stethoscope for internists and the scalpel for surgeons.