Both
patients and clinicians would agree that the healthcare system has an abundance
of inefficiencies and frustrations. Alyssa Jaffee, partner at 7wireVentures,
said it well last month at MedCity News’ INVEST conference — “it’s not that
hard to figure out what the problems are — they’re everywhere.”
As
healthcare leaders work to transform the industry, there are a few key problems
that are in dire need of innovation, such as the siloed nature of care, poor
care access, the burden of manual workflows and the scarcity of personalized
care. All of these problems are further exacerbated by healthcare’s workforce shortage
and burnout crisis — the industry is projected to be short 10 million workers
globally by 2030.
Many
experts think technology will help mitigate these sweeping challenges, but the
healthcare industry still has a lot to figure out when it comes to choosing
which tools to deploy and getting its workers on board with these new tools,
according to a report released Tuesday by GE HealthCare.
For
healthcare leaders to achieve any success in their efforts to improve and
modernize the field, they will need to champion a culture shift to see the
healthcare workforce as an asset, the report argued.
For
its report, GE HealthCare surveyed 5,500 patients and their families as well as
2,000 clinicians across eight countries. Of the clinician respondents, 42% said
they are actively considering leaving healthcare. Clinicians cited poor
work-life balance, exorbitant workloads and inadequate compensation as some of
their top reasons for having these thoughts.
Another
reason that healthcare workers often experience poor job satisfaction is that
many feel they are not operating at the top of their licenses, the report
revealed. To remedy this, healthcare leaders must deploy technology that
delivers on its promises to reduce administrative tasks, better allocate
resources and alleviate burnout, the report said.
Generative
AI — which includes large language models like ChatGPT, holds great potential
to eliminate workers’ mundane and time-consuming tasks, said Taha Kass-Hout, GE
HealthCare’s chief technology officer, in a recent interview.
“Generative AI introduces a different element that’s going
to be super important for healthcare — where data is natively multi-modal and
there is no need for synthetic data. But curating can be daunting, so
generative AI could help tremendously here, where clinicians provide prompts or
examples to orient the model — which in healthcare, could be transformational,”
he declared.
The
use of these AI models is still pretty nascent in healthcare, so the industry
hasn’t yet seen the true impact of these technologies. But with the appropriate
human supervision, generative AI will reduce clinicians’ burden when it comes
to data query and analysis. This way, they can be focused on what really matters
— improving patients’ health.
Nearly
all of the surveyed clinicians said they want to see patients and their care
teams linked together via easy-to-use, effective technology. In order to
improve healthcare workers’ job satisfaction and keep them from leaving the
industry in droves, this demand must be met, the report said.
Machine
learning holds great potential to connect patients to their care teams in a
less siloed, more accessible way, Kass-Hout pointed out.
“Leveraging a tool like machine learning — combined with
clinical expertise and mindfully improving the integrity of the data — can help
us make sense of it all to create a 360-view of a patient’s entire medical
history. We can manage data across populations and securely share data to
develop more predicative, preventative care responses — and ultimately help
clinicians improve patient outcomes,” he explained.
The
report showed that clinicians are still on the fence about using machine
learning and AI in medical care — especially clinicians in the U.S. Kass-Hout
argued that it’s crucial these technologies are integrated in clinicians’
existing workflows to help them understand the data when it’s presented.
“This helps ensure clinicians understand AI’s role in
augmenting their work. AI is a utility, a tool – an intelligent assistant,” he
said.
To
effectively harness the power of big data through AI, the healthcare field
needs to “break the black box of AI,” Kass-Hout declared. This means that
clinicians must understand what data comprises the AI model they’re using. They
have to know about the data points — such as age, gender, lab results, remote
monitoring vitals, genetic variants and images of lesion progression — so they
can better understand what is influencing the AI output.
Transparency
about the data that influences the AI model, as well as how it can be adjusted
is critical to building clinicians’ confidence in AI, Kass-Hout declared.
“As an industry, we need to build clinician understanding
of where and how to use AI and when it can be trusted fully versus leaning on
other tools and human expertise,” he said.
