Automated systems that predict hospital readmissions are in the works – but are not yet ready for deployment, according to several recent studies. Some systems show promise, as the U.S. healthcare system refines progressively more precise metrics to predict what causes readmissions.
Authors of a study appearing in BioMed Central manually reviewed 459 readmissions and then repeated the process using software from 3M, one vendor with a readmissions application in development.
Researchers found some overlap — both approaches agreed that 56% of the readmissions were preventable — but concluded that “not enough concordance currently exists between methods to use automated classification to replace manual review for quality improvement initiatives.”
Hospitals should have more than improving care quality as motivation to keep patients from making multiple visits. Facilities with high readmission rates are subject to fines under the Affordable Care Act. One estimate says preventable readmissions cost Medicare more than $17 billion every year. Without the use of an automated readmission prediction system, some hospitals have turned to building their own internal programs. One medical center relies on an application that collects patient data specifically to measure their likelihood of readmission. Another sends out text messages to remind patients to attend heart rehabilitation classes.
Reducing readmission rates can be a byproduct of improving the overall quality of care, something that is a goal of facilities that move to the ACO model. One hospital that deployed ACO support software from Lumeris documented an 18% reduction in readmission rates after moving to the pay-for-performance model. They also reported a 16% increase in follow-up visits by high-risk patients when offering incentives to physicians.
Technology may not be the surest way to prevent or predict readmission rates. The answer is much simpler, according to a study performed by Health Services Research. The study found that 58% of variation in hospital readmission rates is determined by a hospital’s geographic location and not based on the facility’s performance. That percentage was calculated by measuring the variation in 30-day readmission rates at the county level, as noted by the Health Behavior News Service.