Within the history of healthcare innovation lies the profound lessons surrounding the adoption of revolutionary technologies. Whether it be unleashing genomic sequencing to personalize oncologic treatments or developing an electronic medical records system, the healthcare industry as a whole abides by the highest standards of testing and evaluation prior to welcoming novel solutions that may impact patient lives. Certainly, this level of prudence makes sense: we all want the technologies we utilize to be safe and effective, especially those in the field of medicine. In that regard, we’re finding ourselves at an incredible juncture for healthcare AI as we enter the new year. Backed by a stockpile of scientific research, clinical trials and critical feedback, healthcare AI has reached that inflection point where it is now primed for widespread adoption.
It’s safe to say that the healthcare community is turning a page in 2023 as far as acknowledging the value of AI. Scientific research and development will continue to catalyze its growth, reception, and market access, but 2023 brings an undeniable truth to the forefront of every health system in America: healthcare AI is here to stay.
As clinical AI continues its maturation, I predict it will touch up to 50% of providers in 2023. AI will impact the patient throughout their clinical journey, whether they know it or not, both within and outside the walls of the hospital.
According to a survey conducted in 2021, nearly 40% of respondents were engaged in the AI market or actively assessing the role AI would play in their health system. The major driving forces for this continued market growth is the constant necessity to reduce overall care expenditures for both patients and health systems alike.
This means that the potential of the AI market will continue to grow. Projections indicate that healthcare AI can grow at a compound annual growth rate of 46.2% from now until 2027, reaching up to $67.4 billion. In 2021, the AI in healthcare market was worth around 11 billion U.S. dollars worldwide. It has been forecast that the global healthcare AI market would be worth almost 188 billion U.S. dollars by 2030, increasing at a compound annual growth rate of 37 percent from 2022 to 2030.
One of the most promising starting points for healthcare AI has been radiology thanks in part to the department’s high volume output of structured data, which alone makes it a prime AI candidate. On top of that, radiology is a key entry point for countless hospital workflows for both in- and out-patient procedures. Thus, AI that can help flag suspected pathologies at the beginning of these workflows has greater potential for downstream impact.
Despite its humble beginnings at the radiologist’s workstation, however, the power of AI is now felt across the healthcare spectrum. For instance, medical oncology is at the forefront of implementing AI into the development of personalized care management solutions by leveraging machine learning to properly understand the complex nature of each patient. Within the field of gastroenterology, AI is leveraged for endoscopic analysis of lesions, in the detection of cancer and to analyze inflammatory lesions or gastrointestinal bleeding during wireless capsule endoscopy. Predictive analytics and deep learning tools enable physicians to highlight patients at risk of developing comorbidities, and even diagnose patients’ underlying disease states at earlier stages. Outside the walls of the hospital, AI and machine learning augment the world of clinical trials by better matching patients with novel and nascent therapies. Well beyond radiology, AI is currently being deployed in nearly every subspecialty of medicine, and will continue offering value to patients, physicians, and administrators across the spectrum of healthcare.
The increasing AI adoption across health systems in the US and Europe will continue leading us down the path of more advanced, critically examined AI strategies throughout hospitals.
Perspectives on AI have matured beyond “is it even worth it” toward questions like “how do we materialize the full ROI of AI” and “how do we build a strategy to adopt 20-30 solutions over the course of the next couple years.”
This strategy is not focused on scaling alone, but also on expansion between service lines. AI originated as a single service line, mainly due to workflow limitations. For example, helping radiologists read exams, ED physicians identify sepsis, etc. However, there is a lot of value in AI helping dissolve the silos between departments. At the end of the day, you don’t treat the data, you treat the patient, and patients often live and receive care at the intersection between multiple service lines.
One shining example of healthcare systems benefiting from AI in the care coordination realm is a recent study at Cedars-Sinai, which associated adoption of an AI worklist triage system with a decreased length of stay (LoS) for both pulmonary embolism (PE) and intracranial hemorrhage (ICH) patients by 26.3% and 11.9%, respectively.
The United States has one of the highest costs of healthcare in the world. In 2020, U.S. healthcare spending reached $4.1 trillion, which averages to over $12,500 per person. With a population of increasing age, experts estimate the cost of care to continuously expand over the next decade. Consequently, medical innovation need not exclusively focus on developing products that improve patient outcomes, but also accounts for cost containment while providing added revenue opportunities. A medical innovation that adds an unhealthy expense to a hospital’s bottom line will suffer from a lack of adoption in the market.
The ROI for AI algorithms is robust and proven. Beyond the tangible clinical benefits of healthcare AI, hospitals have realized an economic ROI that incentivizes health systems to rapidly adopt AI solutions. AI has proven to solve financial pain points of hospitals tied to clinical outcomes. For instance, overburdened emergency departments enjoy a decreased ED LoS by using AI, while inpatient decreases in LoS have also been realized across the healthcare continuum. Prompt treatment and management of patients suffering from pulmonary embolisms or brain bleeds has led to more timely interventions offered per year and better patient outcomes. The backbone of great AI solutions offering robust ROIs are well thought through clinical use cases that truly impact patient lives. As the ROI continues to shine through for AI in healthcare, especially in this challenging healthcare environment, the incentives for AI adoption will only grow.
The stresses on the healthcare system come from two enduring hospital issues that may at first seem separate from AI: labor shortage and revenue leakage.
Labor shortage has primarily led to staff burnout, resulting in many surveyed healthcare providers denoting that they either would like to or are planning on leaving the profession. Moreover, the labor shortage has created negative residual effects on the patient experience, such as an increase in the length of stay, both in the ED and post-acute inpatient settings, which raise operational costs and cut margins. With regard to leakage, as patient volumes decrease and ambulatory surgery centers and telehealth programs become more prevalent and viable alternatives to large hospitals, hospital EDs—considered important sources of institutional revenue—are likely to suffer.
In terms of labor shortage, we’ve already mentioned the impact AI can have on LoS and physician efficiency. On the revenue side, AI-driven care pathways can ensure that suspected pathologies receive proper follow-up, avoiding patient leakage while improving outcomes.
As the ROI evidence continues to grow, and the magnitude of AI becomes more apparent, I predict AI will become a strategic priority. Furthermore, I would estimate today that a well defined AI strategy can have about $25-30M EBITA impact for a 1500-bed health system, and this number will grow over the next few years, with an increasing number of AI-driven solutions gaining clearance and hitting the market.
The burden on the healthcare system to provide quality and effective care to every patient is enormous. Doctors work tirelessly to diagnose and manage patients through their journey of regaining health with a complex set of tools at their disposal: lab tests, physical examinations, historical medical records, blood biomarkers, imaging and more. In the imminent future, including AI in the arsenal of a physician’s toolset will be the standard of care. Every subspecialty is inching towards a future in which AI enables them to manage their ever-expanding patient load. The uniqueness of the healthcare AI revolution is its impact on every dimension of providing care: patients benefit from AI through better diagnostics and interventions, physicians benefit from AI through an increased ability to manage their patient panels, and health systems benefit from an efficiency and productivity perspective. The effect of AI solutions in medicine are undeniably positive, and will continue to grow as a vital component of the future of better care.