Reducing Readmission with AI
Lowering readmission rates has always been a priority for healthcare organizations, and in the midst of a pandemic it's an even more pressing concern. Yet, many organizations are neglecting the opportunity to use artificial intelligence (AI) and predictive analysis to lower readmission rates in post-acute care management.
Tackling High Readmission Rates
Readmission is a $25-billion problem for hospital systems. What's more, it increases patient risk for additional hospital-acquired infections and COVID-19 exposure. This is especially problematic for certain patient populations, such as those with chronic conditions and patients with socioeconomic challenges.
The Hospital Readmission Reduction Program (HRRP), which started in 2012, aimed to improve the quality of healthcare and make hospitals more accountable for their readmission rates. With reduced reimbursements as the penalty for high readmission rates, it provided a strong incentive to make the prevention of readmission a priority. However, a study by CMS Patient Readmission found that 2,599 hospitals—or 82% of the participating organizations—received reduced reimbursements in 2019 due to readmission rates that were too high.
The data also shows that hospital and healthcare systems that have used artificial intelligence to reduce readmissions reported lower rates and increased patient satisfaction. One healthcare system managed to cut its readmission rates from 21% to 14% within a year.
Following are some ways that AI can help with readmissions in the post-care continuum:
- Monitoring readmission data and trends
- Helping to create customized solutions to keep readmission rates down
- Ensuring plan of care compliance across patient populations
- Implementing innovative strategies to improve readmission rates.
How Artificial Intelligence and Predictive Analysis Can Help Healthcare Organizations
AI platforms can use data to determine which patients are at the greatest risk of readmission, and then assign a risk score. Hospitals can use those scores to determine which patients to prioritize follow-up with after discharge.
AI and predictive analysis can also help determine the best time for reaching out; most healthcare systems find that checking in with patients within 5 to 10 days after discharge is optimal.
A follow-up by mobile unit, phone, or text can help evaluate whether the patient:
- Is taking all medication as directed
- Has someone helping take care of them
- Has food and meal plans in place
- Scheduled follow-up appointments as directed
- Has a sounding board to allay any questions, problems, or challenges.
Without AI, the LACE score is the most common way to identify patients at risk of readmission. This score identifies patients that are at-risk for readmission within 30 days of discharge after an acute hospital stay. The score assesses four variables:
- The length of stay
- Whether the patient was Admitted through the Emergency Department (ED) or voluntarily
- Whether the patient has more than one Condition
- The number of previous ED visits in the past 6-month period.
Patients with scores over 10 are at higher risk of a readmission, but the LACE score has a high rate of false positive and false negative results.
With AI, a predictive readmission calculator yields a higher rate of accuracy in readmission predictions based on patient's electronic health records (EHR).
One study of heart failure patients found that AI and deep learning techniques did far better at 30-day prediction of readmission rates than traditional methods, enabling targeted interventions and better outcomes.
Artificial Intelligence is Poised to Play an Important Role in Post-Acute Care
AI may be the best way to reduce readmission rates, improve outcomes, and target the right patients for post-acute care coordination and follow-up. The ROI of AI can be easily evaluated by comparing the number of predicted readmissions with the actual numbers of patient readmissions.
Using artificial intelligence to lower readmission rates will likely become the standard in patient care across the continuum. It can play a critical role at a time when hospitals across the country are seeking to reduce the flow of incoming patients so they can focus on handling COVID-19 patients.