Can Artificial Intelligence improve medicine and reduce the cost of care?

Artificial Intelligence (AI) is a prime example of a horizontal innovation that is reverberating across all domains of our lives. Healthcare, in particular, is experiencing sweeping changes that are intended to improve medical care while at the same time reducing healthcare expenditures.

American healthcare is finding it increasingly hard to cater to the needs of over 327 million people. Despite the paradigm shift towards value-based care, where providers are offered incentives for giving quality patient care, the American people still spend a hefty $3.65 trillion – a cost that needs to be brought under control. AI presents a revolutionary opportunity to improve medical care while bringing substantial cost reductions in medical practices and hospitals. Accenture has termed AI ‘Healthcare’s New Nervous System,’ and it estimates that effective AI implementation can save the US Government a substantial $150 billion by 2025.^[1] Alongside this, AI will be a key player in ensuring that patients also enjoy a smarter and safer healthcare experience.

Catering to the KenSci has developed an AI solution designed to analyze clinical data and perform predictive analysis for smarter resource allocation. KenSci claims a straight 50% hard Return on Investment (ROI), along with other benefits, when implementing their AI solutions.

Following are some ways in which AI can improve medical care while reducing hospital costs:

Forecasting patient admission

AI can forecast when a patient may need to be admitted to a hospital. It can do so by keeping track of a patients’ clinical history, comparing the readings to a set of benchmarks. Any unusual outlier could require hospitalization. AI can do this for the whole community and generate a cumulative report with the number of patients a hospital could anticipate in the coming days.

Predicting cases of overstay

The predictive analytics in AI could calculate the expected duration of stay for each patient with regards to the treatment administered and the progress made by the patient. If the patient’s health doesn’t get better at a specific rate, the automated systems will let the provider know about that case. AI systems could go as far as calculating potential complications and may warn the provider if a particular treatment/drug does not suit a patient. Such a patient may need to be kept in the hospital premises until the vitals readings show significant improvement.

Building an administrative prediction model

Management and allocation of hospital resources such as beds, drugs, and machines can be a daunting challenge – particularly for a hospital with an influx of hundreds of patients. Manually managing a hospital’s resources may not offer the same efficiency as that of an AI system. AI could account for the entire patient flow (incoming, staying, and outgoing) and designates resources according to the specific needs of each case.

Predicting disease by intervening early

Diseases, as a result of pathogens, usually do not happen abruptly. Upon inculcation, they start replicating to a certain point when they finally overpower the natural defense system. AI can sense the slightest change in the vital signs and check for similar patterns in its clinical knowledge base. It also takes into account the psychosocial characteristics of an individual. By doing so, AI can predict a disease much earlier than a physician can, and will allow him to take evasive action much earlier.

Managing claims smarter

A major reason for physician burnout is the unrewarding nature of billing and claim management. It takes a significant toll on the provider (and his staff) to accurately gather the requisite information for claim submission. With AI, not only would this be an easier task, but any anomalies and errors would also be highlighted.

Uncover any attempts of data breaching

The one major problem that the current healthcare ecosystem faces is that of data breaching. The average cost of a single protected health information (PHI) breach is $15 million. As of now, healthcare companies are paying huge sums to counter this predicament. With AI, breaching clinical data  

Assisting in Integrated operating rooms

Integrated Operating Rooms (OR) is a giant leap in embracing AI and long-distance care. It is estimated that around 310 million patients (globally) are operated on each year. More than 50 million patients suffer post-surgery complications, and 1.5 million patients die from these complications. AI could potentially minimize complications due to human error, making operation theatres safer.

Performing Smarter bed management (to improve patient outcomes)

In a hospital filled with thousands of beds, manually managing each one can be a cumbersome task. AI could be used to ensure that resources are managed smartly. By predicting the estimated stay of a patient, AI can effectively allocate the beds to their respective patients.

Assisting Caregivers

1. An AI-enabled EHR system can assist caregivers to make the right decision. Proving additional insights and advanced data mining capabilities, AI could make better and more comprehensive predictive analysis, helping improve the clinical decision-making of a provider.

Calculating ROI

Providers can be indecisive about whether to invest in a certain commodity or technology with regards to its (Return on Investment (ROI). With advanced algorithms, AI could foresee the implications of each change in an already functioning hospital and recommend the healthcare strategists if a certain investment is worth it.

Conclusion

In a survey conducted by OptimIQ, 91% of the healthcare leaders have shown their interest in investing in AI. Growth in the AI health market is forecasted to reach $6.6 billion by 2021, from $600 million in 2014 by Accenture. With potential capabilities of improving medical care, as well as potential cost savings, AI is a technology the healthcare industry landscape could find beneficial.