How AI Can Help Reduce Hospital Readmissions
There is an all-too-common problem faced by patients and the hospitals that serve them: readmissions. Besides the health risks, reduced quality of life for patients, and anguish for family members, the problem costs money, and lots of it. Hospital readmissions cost an estimated $41.3 billion a year in the United States. The average readmission costs $14,400, and Medicare penalizes hospitals deemed to have excessive readmissions.
Given the toll in human health and the financial cost, it seems clear that something must be done to reduce hospital readmissions. The central challenge is predicting just who, and with what conditions, is most at risk of failing to recover after an initial hospital discharge.
Data provides an important key to solving this issue. Data on the outcomes of specific procedures, on the people who return to the hospital, and on how long it takes them to recover — both in the hospital and after discharge. Sorting through all that data and drawing meaningful conclusions has historically been a daunting challenge all of its own. But now, artificial intelligence (AI) systems can allow hospitals to use their data to make better treatment decisions and to reduce readmissions.
AI for Reducing Readmissions
AI-powered analysis of data can help hospitals pinpoint which patients are most at risk for readmission and help hospitals identify proactive intervention strategies, such as visiting patients within a critical window following a hospital discharge.
Of course, getting a handle on big data is a challenge. Hospitals can quickly become overwhelmed when dealing with thousands of patients, diverse populations, illnesses, injuries of various kinds, socioeconomics, and all the other factors that influence readmissions.
AI offers a way to process and analyze large quantities of data about patients, illnesses and conditions, and treatments. All of which can help providers predict who is most at risk for a readmission — and do something about it proactively. What's more, easy-to-use AI analytics tools now let the business user or analyst make sense of data — without having to rely on higher-level data science teams, saving time and reducing overhead.
Hospital executives are starting to embrace AI solutions, and the healthcare industry's use of data is growing at the rate of 48 percent a year. This is helping improve outcomes and reduce costs in areas such as diagnostics and patient monitoring.
Preserving Health and Saving Money
Using data such as medical records, diagnostic output, and medical claims information can help hospitals predict and prevent readmissions, leading to more efficient use of resources while improving patient care. For example, an anonymized dataset of 16 million patients and AI software trained to recognize 50 different potentially avoidable risks helped Mercy Medical Center, based in Canton, Ohio, reduce readmissions by 20 percent.
Atlanta's Grady Health used the same system to help identify patients at risk for readmission, showing that 382 patients did not come back within the 30 days / readmission. At Grady, AI helps target particularly high-risk patient populations, based on data from millions of other patients and the outcomes of their diseases and conditions, for in-home visits. The purpose: to head off readmissions by intervening in a patient's care before he or she experiences a downturn.
Hospital directors at Grady have found that following up with home visits to at-risk patients within five to 10 days following a hospital visit works best. Patients get seen by off-duty EMS workers who form a mobile integrated health team.
Since a home visit costs $200 versus $11,000 for a readmission, the hospital has achieved significant savings with its new program, as well as helped patients feel better and get better sooner. And while it's not always possible to tell whether a given intervention prevents a readmission, the members of the mobile integrated health team are sure they're making a difference.
While the results of automated data analysis are promising, there are roadblocks to using AI tools in healthcare, such as the non-negotiable need to protect patient data and preserve privacy. That's one reason many organizations prefer AI systems that can run locally, on-premise at a hospital, rather than in the cloud. Healthcare providers prefer to crunch potentially sensitive data on private servers rather than on more public networks.
Powerful software tools, such as Ki, can run locally as well as in the cloud. That helps to keep data private, speeds processing time, and helps medical providers identify common denominators in aggregated patient data that might be missed by human analysts. These patterns can help predict outcomes, such as the likelihood of a hospital readmission.
Increased costs for insurers, providers and patients — including missed time from work and time with family and reduced quality of life — all result from hospital readmissions. Fortunately, the power of AI-driven automated analytics software produces actionable insights that can help reduce costs while improving patient outcomes.
To see how an automated analytics software can help you reduce hospital readmissions and turn your data into action, request a demo of Ki.
LATEST POST : Data Management Best Practices: Six Ways to Improve Analytics Impact