How Physicians and Oncologists View AI Assistance
The progressive integration of AI within medicine is transforming the way physicians evaluate and monitor disease, enabling earlier detection and improving outcomes and quality of care. Medical practices are increasingly leveraging datasets, algorithms, and other machine learning techniques to identify subtle patterns that might not be discernable to physicians without this technology.
These AI tools are especially useful in detecting conditions like early-stage cancer, cardiovascular diseases, and neurological disorders. For example, AI can analyze imaging data, such as CT and MRI scans and mammograms, to flag abnormalities with greater accuracy and efficiency than current methods, reducing the risk of incorrect diagnoses.
While AI’s early achievements in diagnostic accuracy have already demonstrated its value, we wanted to know how physicians and oncologists are currently using this technology—and how they expect to use it in the future. In September 2024, the MMIT Oncology Index and Biologics & Injectables Index teams conducted qualitative research on physician and oncologist perspectives on AI and ChatGPT.
Reducing Administrative Burden
According to the MMIT Indices results, only 6% of physicians and 10% of oncologists are currently using AI in their practices. However, 22% of physicians and 11% of oncologists are actively researching/implementing these tools, and an additional 50% of physicians and 32% of oncologists are engaged in preliminary discussions on the use of AI/ChatGPT.
Physicians anticipate the primary uses of AI in their practices to be patient scheduling, patient education, clinical notetaking, data retrieval, and prior authorizations. AI can assist in drafting clinical notes by transcribing patient interactions, summarizing key points, and organizing information, saving valuable time. Physicians who already use for data retrieval report that it helps them find clinical study data with one simple search, as it immediately pulls and summarizes information from electronic records. AI assistance with prior authorization requests also quickly provides the relevant information, helping physicians communicate more efficiently with payers.
Oncologists also expect AI to be used in their practices to improve note-taking and chart evaluation, as well as data retrieval and claims processing. Oncologists anticipate that AI-assisted claims processing would provide their teams with fast access to the appropriate information, speeding up the process and improving accuracy. Overall, two-thirds of physicians—and slightly less than one-third of oncologists—believe that AI is very to extremely likely to reduce or eliminate administrative burden within their practices.
Notably, the majority of physicians (52%) expected that AI tools might be very or extremely useful for clinical diagnosis. While less than one-quarter of oncologists anticipated using AI for diagnostic purposes, several oncologists said AI tools might be useful in providing “differential diagnosis and likely diagnosis based on clinical factors.”
Ensuring Data Privacy, Security and Accuracy
Both physicians and oncologists who have implemented ChatGPT or similar AI tools agree that ensuring the accuracy of AI-generated data has been their biggest hurdle. Some physicians have also struggled to ensure that protected health information (PHI) is not stored in AI tools.
For those who have not yet implemented AI, the top areas of concern for both physicians and oncologists are data security, data privacy, and data inaccuracy. Oncologists are also concerned about data validity, while physicians are worried about their ability to integrate AI into their clinical practices and the overall investment in training. Both groups are somewhat concerned about whether or not AI will be cost-effective for their practices.
Although the majority of physicians and oncologists see the potential for using AI within their practices, many survey respondents highlighted the need for regulation to ensure appropriate use. Several oncologists stressed that such tools must be standardized and amended to ensure accuracy. As one respondent noted, the “lack of accountability if incorrect information [is] provided, incorrect authorization received, etc.” might present a challenge for practices.
While the continued integration of AI in medicine will undoubtedly create new challenges to address, AI has the potential to modernize how healthcare is delivered, improving patient outcomes and the overall quality of care. By helping to automate time-intensive tasks like insurance verification, charting, scheduling, authorizations and billing, AI can free physicians, oncologists and their teams to focus on patient care. In turn, greater workflow efficiencies are likely to also improve the patient experience by reducing wait times and eliminating administrative delays.
Clinical Applications of AI
In the clinical arena, AI is already making strides when it comes to the diagnostic process and personalized treatment options, as AI algorithms can now analyze medical scans, patient data, and other sources of information with greater accuracy than a human physician.
On the other end of the pipeline, AI could potentially reshape the landscape of drug development by accelerating the pace of molecular discovery. By analyzing large volumes of molecular data, AI can quickly identify promising drug candidates far more efficiently than traditional methods, which often take years. In the future, AI-assisted drug discovery could significantly lower drug development costs and enable faster patient access to new therapies.
AI also shows great potential in improving the safety and efficiency of clinical trials. For instance, AI is already being used to optimize clinical trial planning and site selection. AI could also be used to streamline the participant selection process, which typically requires considerable time and resources, making trials faster and more cost-effective. Additionally, AI tools could be leveraged to track patient data in real time, flagging any potential adverse reactions before they escalate into something more serious. This real-time tracking would not only enhance the safety of trials, but also boost their efficiency.
The role of AI in healthcare is just beginning, and its future looks incredibly promising. As AI becomes more integrated into healthcare systems, we expect this technology will ultimately expand access to healthcare, improving care quality, enabling more personalized treatments, and reducing overall healthcare expenses. Manufacturers must consider these applications when considering their clinical trial development, marketing tactics, and overarching brand strategies.
To track payer and HCP perspectives on current trends, learn more about our Oncology Index and Biologics & Injectables Index solutions.
