Elsevier documents progress in healthcare technology - can AI save the sector?

George Lawton Profile picture for user George Lawton September 11, 2023
Summary:
Elsevier looks at the global trend of a reduction in healthcare workers and discusses how technologies could play a role in supporting the healthcare sector in the future

Medic using healthcare technology on tablet © metamorworks - shutterstock
(© metamorworks - shutterstock)

A new report commissioned by Elsevier Health highlights the opportunities and challenges of emerging medical technologies, including generative AI, patient experience, data integration, and social determinants of health. The new report summarized the experience of over 2,600 doctors and nurses globally. The top concern is treatment backlogs, exacerbated by a shortage of doctors and nurses. 

This is driving increased interest in the role that generative AI could play in improving efficiency and training. Doctors and nurses are warming up to the potential for generative AI to improve health outcomes with appropriate guardrails. 

Another interesting trend is increasing interest in transitioning to value-based care to lower healthcare costs while improving outcomes. Interest is greatest in countries in Europe that already offer universal healthcare. However, this transition requires significant digital transformation to shift existing processes to the newer approach. 

Clinicians are burning out.

One early takeaway is that doctors and nurses are leaving jobs and the profession entirely at a much faster rate. Many clinicians are burning out against a backdrop of record waiting lists in the UK and elsewhere. The widespread adoption of digital technologies and electronic medical records has still not panned out regarding the efficiencies and effectiveness imagined. 

Globally, 37% are considering leaving their job in the next two to three years, up from 33% last year. In addition, 13% are considering leaving healthcare entirely. In the US and Europe, the proportion considering leaving their job is even higher, with 50% considering leaving in the US and 24% planning to leave healthcare or retire. 

This global trend suggests there will not be enough healthcare workers to provide sufficient levels of care using existing technologies, processes, and practices. Deloitte has estimated we will need 80 million more healthcare workers by 2030 globally. The World Health Organization has a more optimistic need for only 10 million more health workers in the same time frame. In the UK, in particular, the NHS estimates the current gap of 100,000 clinicians will rise to 179,000 by 2024. 

The rise of digital technologies is also driving new upskilling challenges for doctors and nurses to keep pace with both medical advances and how to use the latest digital technology. 

Growing demand for data and AI

The big question is - could recent innovations in AI and large language models help make up the difference both in training and medical practice? Opinions on this are mixed. China is the most optimistic, while other countries are more concerned about mitigating hallucinations and bad medical advice. 

There is a growing call for using AI tools and improved data integration and analytics to address time pressures, create efficiencies and enhance clinical practice. About 11% are using AI tools today, while 48% find it desirable to use AI in clinical decision-making in five years. However, the use of AI in clinical decision-making differs significantly across regions at 6% in Europe, 7% in North America, 19% in the Asia Pacific and 28% for nurses. 

On the educational aspect of new AI, there are concerns about the dangers of replacing hands-on experience. AI may also teach inaccurate or erroneous information and hinder students' critical thinking and decision-making. Support for AI training is highest in China, with support from 72% of doctors, compared with 41% in Europe and 36% in North America. 

Various institutions and organizations are ramping up efforts to mitigate these challenges, including the US General Services Administration Applied AI Healthcare Challenge, US healthcare giant Kaiser Permanente’s AIM-HI Program, and the UK National Health Services AI Lab. China is promoting Internet+Healthcare as a unified medical platform. 

Patient Empowerment

Another major trend is the growth of different facets of patient empowerment supported by metrics correlating the use of monitoring tools, social determinants of health (SDOH) knowledge, and health literacy. Clinicians predict these will increase in the next five years. 

Monitoring tools - the use of monitoring tools is expected to grow from 27% today to 45% in five years. This means more people using smart watches, rings, and special monitoring devices. Hospitals are still trying to sort out how to use this data as part of clinical practice. Doctors must also upskill their expertise to work with these new data streams. An additional issue is that many of these products only meet consumer requirements with lower accuracy and fidelity than more expensive medical equipment with special certifications. Hospitals will have to figure out how to flag and appropriately label data and its provenance to strike a balance between efficiency and risk. 

SDOH -social determinants of health knowledge is expected to grow from 35% today to 48% in five years. These provide the metrics to build digital twins for correlating behavioral, environmental, infrastructural, and institutional metrics with health outcomes. These can guide a person to make the most impactful health decisions, prioritize new programs or development for the community, and help doctors and hospital program planners make better decisions. 

Proactive health management  -health literacy measures a patient’s ability to interpret and act on medical information from the doctor, trusted third-party sources, and other information delivery mechanisms. This is expected to grow from 41% to 49% in the next five years. 

Need for integrated systems 

Clinicians highlighted the need for more efficient, integrated IT systems to reduce administrative burdens, enable data sharing, and give them more time with patients. Interoperability and patient data integration are priorities.

Existing approaches to medical records management started with a strong focus on billing and automating administrative practices. New healthcare standards and initiatives are also beginning to bring more attention to improving patient outcomes. For example, the UK NHS recently rolled out a new information standard to improve medication and allergy/intolerance information across all healthcare services in England that went live across all NHS and social care organizations last March. 

Meanwhile, healthcare organizations are still struggling with what to do with the tsunami of fitness information flowing out of low-cost consumer gadgets. Although 68% of clinicians feel this is desirable, opinions are mixed worldwide about new problems it may create. While 75% of Chinese doctors and 74% of US nurses like the idea, only 48% of Chinese nurses do. The biggest concerns are the inaccuracy of data from wearables and the risk of misdiagnosis. 

 

Value-based care (VBC)

One of the biggest innovations in healthcare has been finding ways to align the most efficient care with the best patient outcomes. This looks for prevention and risk reduction opportunities. This is a difficult proposition technologically, organizationally, and market-wise. First, most of the back-end systems underlying medical care are optimized for billing. Organizationally, hospitals would need to develop entirely new care practices and workflows. With regard to the market, established medical vendors and service providers are incentivized to maintain the current market structure and established billing practices. 

One of the biggest changes lies in transitioning payment processes from a fee-for-service approach to one that favors smart spending based on outcomes rather than services. A separate 2022 Survey of America’s Physicians found that 63% of physicians thought that measuring pay-for-performance was a challenge. 

Clinicians have mixed opinions on whether VBC will lower costs, with only 51% believing it will, at least in the short run. Even if the care costs less per patient, there is the enormous cost of putting in the new processes and supporting IT infrastructure. But a majority see many benefits - 67% believe the quality of consultations will be higher, and 69% expect hospital stays could be shorter. 

Essential innovations in IT systems will be required to enable population health management, support personalized care, and improve preventative health. Some early pioneers include:

  • NHS Rightcare is developing data packs for specific conditions like cardiovascular and respiratory disease management. 

  • The Netherlands is completing a 5-year Plan for Outcome-Based Healthcare to transform the healthcare system.

  • The Healthy China 2030 program is working on a new delivery model for universal health coverage. 

  • Singapore’s Three Beyonds looks at going beyond healthcare to health, beyond hospital to community, and beyond quality to value.

  • Kaiser Permanente is working on a preventative program for cardiovascular diseases. 

  • The US Centers for Medicare and Medicaid Services (CMS) is restructuring its first foray into value-based to build accountable care relationships for all Medicare patients by 2030.

My Take

With the recent onslaught of progress and the associated hype in generative AI, it's tempting to imagine that healthcare will soon get better. But that glosses over the fact that hallucinations and mistakes can kill. Not to mention, they can get us worked up about problems that are not significant while ignoring the real killers. 

However, I am not sure what to make of all the health gadgets and how they may improve future outcomes. The various health trackers I have experimented with over the last few years only sometimes and sort of agree. And then they all have strange metrics like body battery, readiness level, and energy that seem sort of worthless from a medical perspective. 

In the short run, progress in medical AI, and especially some of the automation for capturing doctors and nurses in the flow of work Diginomica previously covered, will free up a lot of time for patient-doctor interactions. In the long run, companies like Elsevier Health, which funded this survey and sells a lot of information services to support some of these innovations, will continue to adapt to address criticisms and build trust in the technology and information they provide. 

Loading
A grey colored placeholder image