Through analyzing large patient data sets, AI is poised to significantly change the way physicians monitor and care for patients.
The increasing use of artificial intelligence (AI) to analyze large patient data sets promises to change the face of population health management in a way that will be far reaching across the industry and a game changer to the way physicians monitor and care for their patients.
AI’s ability to raise the level of evidence-based medicine can help primary care physicians make better decisions in several areas. These include the ability to determine appropriate treatments for their patients and to how best to monitor their care during and after hospitalization, to improving efficiency and productivity in care team workflows and finding better ways to reduce overall costs associated with patient care.
Supercomputers that compare and analyze large groups of patients’ clinical data, diagnostic images and claims data, are capable of identifying subtle patterns and changes in health and wellness that can foreshadow the start of an illness, monitor the effectiveness of drug treatments and identify patients’ health risks.
For example, when researchers at the MRC London Institute of Medical Sciences, wanted to find out which patients with pulmonary hypertension had the greatest risk of heart failure, they used AI software to analyze images of patients’ hearts and constructed a smart 3D heart that predicts patient survival rates.
The study, published in the journal Radiology, noted that researchers used the software to analyze MRI scans of 256 patients’ hearts along with blood tests to identify when the heart was about to fail.
With each heartbeat, the software examined the movement of 30,000 different points in the heart’s structure, and combined this information with historic patient health records to learn which abnormalities would predict when patients would die.
The study found that the software could see around five years into the future and correctly predict which patients would still be alive after one year about 80% of the time. This information is critical for physicians to determine when to intervene to save a patient’s life and what course of treatment should be taken, including drugs, injections straight to the blood vessel or a lung transplant.
On the cancer front, data presented at last year’s American Society of Clinical Oncology conference showed that a drug developed by biotechnology company Berg LLC using AI can slow the growth of cancer in clinical trials.
AI is also being used in other areas of healthcare including analyzing data to help patients take their medications on time, manage diabetes and improve clinical practices in mental health.
“Artificial intelligence can help providers gain insights about their patients, their treatments and their workflow that they could never have known on their own” said Harpreet Singh Buttar, transformational health industry analyst at Mountain View, California-based, research firm Frost & Sullivan.
He added that by using this technology, doctors will achieve their population health management goals as they make better decisions at a much faster rate. Furthermore, computational computing can be applied to many data sets associated with population health management including data registries, electronic health records, data generated from team care, as well as claims data, information gleaned from hospital readmissions and data related to social determinants that have a significant impact on patient populations.
Frost & Sullivan estimates that the market for AI will grow from earned revenues of $634 million in 2014 to $6.6 billion in 2021 at a compound annual growth rate of 40%. The company also predicts that overall, AI has the potential to improve outcomes by 30% to 40% while cutting treatment costs by as much as 50%.
Buttar told Medical Economics that the need for data mining and decision-making in healthcare, as well as the shift toward value-based care that demands greater efficiency and cost effective practices has put AI at the forefront of healthcare.
Buttar also said most hospital executives he has discussed the topic with are trying to implement an AI structure into their current workflows. As hospitals move toward value-based care providers are hoping machine learning can help their efforts to introduce cost cutting measures, improve patient outcomes and reduce hospital readmissions over time. He also warns that the ability to interpret the data is as important as analyzing it.
“Artificial intelligence can change every facet and aspect of population health management, but to get the best results, researchers will have to understand which data to select, what data to use for comparisons and how to correctly use the system to analyze the data before you can make the right decision,” Buttar said.