News center

Mar 09, 2020

Don’t Fear the Disruption – AI Is Here in Healthcare to Improve Productivity

Estimated reading time: 3-5 minutes

Artificial intelligence has taken the world by storm. From transportation to education to human resources, no matter the industry, ears perk up when AI is mentioned and is commonly referred to as the next solution that will greatly disrupt our everyday lives. AI has already shown its potential, with many success stories on how it can improve efficiencies. Now, it has finally made its way into the most complex industry yet – healthcare.
 

For years, there has been a lot of talk around AI’s role in healthcare – it’s currently a hype cycle. This hype has sparked an ongoing conversation around how we can leverage information in a way we never thought was possible. It also has subjected the industry to many bold predictions – from AI replacing physicians, to predicting when a person will die, to robots prescribing medications. The truth is we’re not entirely sure where AI is headed, but we do know AI’s impact is limitless, and we’re already seeing why it needs to have a seat at the healthcare table.
 

Recognizing the next wave of technology
 

As an industry, we realize that AI’s potential influence in healthcare is huge, but the reality is it has been rolled out in a slow and cautious way – as it should be in a field like healthcare where lives are on the line. Healthcare is filled with complexities that differ depending on the care setting and even the specific unit in the hospital. In addition to no two hospital settings being alike, the healthcare industry is slow to change, and providers often stick to the processes that work and have initial hesitation around new technology – especially when it can directly impact patients’ quality of life.

Christine Storm - Patient Care Analytics General Manager at Philips

Despite the shellshock of new and shiny AI technology and the hesitation it brings, providers shouldn’t fear this disruption, but rather embrace change management in this conservative industry. In fact, adoption of AI in healthcare is expected to grow rapidly due to its tremendous value in clinical, operational and financial applications, with McKinsey Global Institute estimating that 15-20 percent of the healthcare market has the potential to be impacted by AI [1].
 

Adopting AI to support proactive care
 

In this value-based care environment, there’s more pressure now than ever before on providers to deliver efficient and effective care at an affordable price. Yet, given the number of patients, influx of data – with most of this data incomplete and unstructured – and the inundation of sounding alarms, they often find themselves in a reactive state rather than practicing proactive, preventative care. It's not that physicians can't provide proactive care, but rather that they don't have the bandwidth to do it alone. In addition to the explosion of data and bandwidth personnel issues, hospitals have been plagued by labor-intensive, inefficient administrative tasks, which affect physicians’ overall ability to deliver high-quality care and increase the likelihood that a physician may consider seeing fewer patients in the future [2]. 
 

AI’s clinical decision support meets this concern head on by accurately reading patient data, predicting adverse events earlier and helping minimize human error – specifically in departments like the ICU that are often overcrowded. It is part of the reason we recently extended our AI capabilities for high acuity areas with our Sentry Score technology [3]. While Sentry Score does not address all of the ICU’s challenges, it is a predictive algorithm for the adult intensive care unit (ICU) that shows a patient’s probability of receiving an intervention within 60 minutes [4]. Sentry Score patient risk predictions are continuously updated as new vital sign data are received, providing deeper insights into the patient’s trajectory, which can provide the opportunity for earlier intervention. The addition of Sentry Score to eCareManager builds upon currently available algorithms that target early deterioration detection, patient discharge readiness scoring, clinician workflow productivity and acuity-based population health prioritization. The Sentry Score proprietary algorithm was developed with a regression machine learning model using patterns of vital sign data from the Philips eICU Research Institute associated with clinical interventions. Sentry Score has been developed in collaboration with current Philips eICU Program care customers who are part of the largest Tele-Critical Care network of leading academic medical centers and integrated delivery networks (IDNs) across the US.
 

Healthcare is turning to a new chapter of care and the title is AI. AI has already shown its ability to screen for cancer, review imaging scans and remind patients to take their pills. Now, AI is tackling healthcare productivity by improving workflow and clinician efficiencies, and moving toward alleviating other healthcare pain points. This is just the beginning, and I look forward to seeing what’s next.
 

As patients transition through the health system, conveying accurate patient information during patient transfer or hand-off is essential to ensuring safe, high-quality care. However, studies have shown that more than two-thirds of adverse patient events are related to communication errors [5].  AI tools across the industry could potentially help enhance the transfer process by connecting the dots across systems and departments to coordinate care around the patient, rather than point solutions that only focus on a specific disease or step in a clinician’s workflow. AI’s ability to interpret and integrate data, pick up on subtle changes in vital signs and provide actionable insights earlier will bring the industry closer to the coveted idea of a proactive healthcare system

 [1] https://www.mckinsey.com/~/media/McKinsey/Industries/Advanced%20Electronics/Our%20Insights/How%20artificial%20intelligence%20can%20deliver%20real%20value%20to%20companies/MGI-Artificial-Intelligence-Discussion-paper.ashx

[2] https://journals.lww.com/academicmedicine/FullText/2017/02000/The_Impact_of_Administrative_Burden_on_Academic.30.aspx

[3] AI used in development and is not actively learning.

[4] Sentry Score Development and Validation Among Patients Under Intensive Care

[5] https://www.americannursetoday.com/issues-up-close-22/  [Copyright in the Publisher’s Journals and their contents, including the abstracts, is vested in the Publisher, whose proprietary and publishing rights are protected under national and international law. The Subscribing Individual shall not have any proprietary right in any of the Journals by reason of their access to them.] HealthCom Media Permission Notice: https://www.myamericannurse.com/terms-of-use/

Share on social media

Topics

Author

Christine Storm

Author

Patient Care Analytics (PCA) General Manager, Philips
Christine Storm is the Patient Care Analytics General Manager at Philips. In this role, she leads the organization addressing the needs of healthcare providers in the acute care space with informatic solutions which touch point-of-care, individual units, and the broader enterprise through telehealth solutions. Christine joined Philips in 2005 and has over fifteen years of experience in Sales and Services within the healthcare management field.

Follow me on

More related news