Industry Voices—Beyond the hype: Here's how oncologists view the future impact of AI

machine learning in healthcare
In a survey earlier this year, 57% of participating oncologists said they think AI will help improve clinical outcomes. (monsitj/Getty Images)

Artificial intelligence (AI) is among the shiniest new objects in healthcare today. Leveraging machine learning and data algorithms, AI emulates human decision-making—but at a speed and scale that exceeds human capabilities. By streamlining the collection and analysis of vast amounts of data, AI-enabled technologies may help find connections that point to new ways to prevent, diagnose and treat disease.

photo of man
Bruce Feinberg, D.O. (Courtesy of
Cardinal Health Specialty Solutions)
Researchers are already exploring AI’s potential to improve care in diverse areas such as diabetes and Alzheimer’s disease. The work in cancer detection is especially exciting. Consider these developments:

RELATED: Verily’s India project reveals the work of applying AI to real-world clinical care

Reflecting the growing excitement, the Centers for Medicare & Medicaid Services (CMS) recently announced the “AI Health Outcomes Challenge,” a $1.65 million contest to develop an AI application that can predict patient outcomes. A recent Accenture study predicts annual savings of a staggering $150 billion by 2026 through the use of AI.

Innovation Awards

Submit your nominations for the FierceHealthcare Innovation Awards

The FierceHealthcare Innovation Awards showcases outstanding innovation that is driving improvements and transforming the industry. Our expert panel of judges will determine which companies demonstrate innovative solutions that have the greatest potential to save money, engage patients, or revolutionize the industry. Deadline for submissions is this Friday, October 18th.

However, the ever-increasing enthusiasm is balanced with healthy skepticism. There have been some early disappointments, such as the failure of IBM Watson to detect cancer in a trial with MD Anderson Cancer Center. Some experts wonder if the reality will live up to the predictions.

Perspectives from the front lines

Beyond the speculation, what do clinicians delivering care really think about AI’s potential? To better understand the perspective of healthcare providers, Cardinal Health Specialty Solutions surveyed more than 180 oncologists earlier this year from community- and hospital-based practices across the United States.

RELATED: Harvard, Dana-Farber AI challenge uses crowdsourcing to improve cancer care

The results indicate oncologists are open to embracing these advanced technologies. Overall, oncologists are optimistic that AI will enhance the quality of patient care and outcomes.

More than half of participating oncologists said they are “excited” when asked about the future impact of AI on the oncology industry. Looking beyond three years, most participating oncologists said they think AI will help enhance the quality of care (53%), improve clinical outcomes (57%) and drive operational efficiencies (58%). Nearly half (47%) also expect it to lower the cost of care.

Nearly four in 10 participating oncologists rated AI’s ability to automate administrative tasks and enable them to focus more on patients as the most valuable potential benefit. Reflecting urgency in addressing the increasing complexity of clinical decision-making, the top-ranked opportunities to improve care included “helping to determine the best treatment paths,” “identifying patients most likely to develop complications” and “improving accuracy of diagnostics.”

The findings also highlight some interesting generational differences. For instance, participating oncologists under the age of 40 are more optimistic that AI will lower healthcare costs compared to those over the age of 50 (58% vs. 44%, respectively). This may reflect a higher level of skepticism among older physicians who have been disappointed by other technologies, such as electronic medical records, that promised to lower costs but did not deliver.

While survey participants are optimistic about AI’s overall potential, they also expressed concerns about the lack of clinical evidence (29%), potential bias within AI algorithms (27%) and providers’ lack of familiarity with AI tools (27%). To address these issues, oncologists will need more exposure to AI—further defining the role that AI should play in patient care.

Overall, the research provides a valuable look beyond the hype at how practicing oncologists view AI’s potential. It is evident that oncologists are open to adopting new technology and working in new ways. This positive outlook has important implications for the future of cancer care and, more broadly, for the pace at which digital medicine will be fully embraced.

To learn more about how oncologists view AI’s future role in oncology, download the full report at cardinalhealth.com/oncologyinsights.

Bruce Feinberg, D.O., is vice president and chief medical officer for Cardinal Health Specialty Solutions.

Suggested Articles

In a letter, 111 physician organizations weighed in on surprise billing, urging Congress not to turn more power over to health insurers.

Even when taking into account increased resources, general and vascular procedures performed in teaching hospitals are better for high-risk patients.

As members of Congress wrangle over the best way to stop surprise medical bills, one senator predicts Washington will pass a new law soon.