A clinician views AI-powered neuroimaging analytics resulting from a scan on a Philips MR system. [Photo courtesy of Philips]
For its latest Future Health Index report, the device developer and manufacturer surveyed 2,011 healthcare professionals (surgeons, other doctors, nurses and physician assistants) and 20,085 patients ages 18 and older in online surveys lasting 15 minutes and 10 minutes, respectively.
Healthcare professionals said they’re using AI to transcribe clinical notes (52%), to discuss work-related ideas with a chatbot like ChatGPT (46%), schedule patient appointments (20%) and speed up X-rays or scans (9%).
They’re also using AI to suggest diagnoses based on symptoms (45%), warn about dangerous drug combinations (44%), help analyze X-rays or scans (24%), monitor patients and flag risks or changes (24%), and triage patients based on the severity of their condition (18%).
Philips said 49% of clinicians reported that AI saved them at least 132 hours per year on average, while 57% of patients who use generative AI for health information say it also saves them time.
Though the report didn’t include financial figures for AI spending and return on investment, 30% of healthcare leaders reported budget savings from AI implementation and half said the benefits are meeting or exceeding costs.
As far as how they’ve experienced AI improving care, 58% of healthcare professionals said AI improves their decision-making confidence, and an equal share reported more efficient workflows. They also identified faster diagnostic decision-making (54%), less time spent on routine or administrative tasks (50%), better ability to access consolidated patient data across care teams (40%), faster diagnostic results (33%), quality of time spent with patients during interactions (33%), and quality of collaboration with colleagues (26%)
Related: Philips CMO Dr. Carla Goulart Peron on how medtech could empower patients
About 10% of the healthcare professionals surveyed are in the U.S. (the survey was approximately evenly split to also include Brazil, China, France, Germany, India, Indonesia, the Netherlands, Saudi Arabia and the U.K).
Of those in the U.S., 36% said AI has increased their capacity to see patients, with a median increase of five patients per week among those reporting increased capacity. And nearly 30% of U.S. healthcare professionals say AI helped them identify or prevent a potential medical error at least three times in the previous three months.
Nearly three-fourths of U.S. healthcare professionals said they use their own personal AI tools when workplace options don’t meet their needs.
A similar rate of clinicians said they have strong concerns about AI tools making errors (76%), inconsistent performance across clinical settings (70%) and a lack of transparency in how AI reaches its recommendations (67%).
Previously: Five tips from Philips for building trust in medtech AI
Clinicians overwhelmingly (93%) want to keep a human in the loop with AI to make sure professionals are ultimately making the decisions.
AI is also creating some issues for clinicians, with 65% reporting they’ve had to correct AI-generated misinformation and 33% saying they’ve seen patients lose trust in their care after learning AI was involved.
“How clinicians introduce, discuss and explain AI and its use may increasingly shape patient confidence and clinical interactions,” Philips said in the report. “As AI becomes more embedded in healthcare, transparency and human judgment will remain central to building trust.”
Of the U.S. clinicians, 67% said they’ve had patients show up to consultations with AI-generated health information.
“Patients are using tools such as generative AI to understand their own symptoms, explore treatment options and review medical information,” the report said, pointing to 45% of patients who say knowing how to use AI will be important for managing their health and 40% who said AI will help them be more active in their care.
“Patients say these tools help them feel more informed about their health (45%), ask their doctor better questions (47%) and make the most out of their appointment (41%),” the report said. “At the same time, more information does not necessarily lead to deeper understanding or better decisions, reinforcing the need for clinical context, guidance and support to help ensure the benefits of AI-supported care are accessible to all patients.”
