Using Predictive Analytics to Improve PT Outcomes, Lower Costs

February 18th, 2020
By: Content Staff

Executives at health-care payers and providers are betting big on the power of predictive analytics to decrease health-care costs and increase patient satisfaction. According to a 2019 survey conducted by the Society of Actuaries, 60 percent of health-care executives were using predictive analytics within their organizations, a 13 percent increase from 2018.

Similarly, 60 percent of survey respondents said they expected to dedicate at least 15 percent of their budgets to predictive analytics in 2019.

Predictive analytics can be used in a variety of ways in the health-care industry, according to a 2018 article by, including avoiding 30-day hospital readmissions, preventing appointment no-shows, predicting patient-utilization patterns, developing precision medicine and innovative therapies, and increasing patient engagement and satisfaction.

Jennifer Bresnick, a writer for, explained it this way in her article: “As health-care organizations develop more sophisticated big-data analytics capabilities, they are beginning to move from basic descriptive analytics toward the realm of predictive insights. Instead of simply presenting information about past events to a user, predictive analytics estimate the likelihood of a future outcome based on patterns in the historical data.”

Big data is starting to make a big impact on physical therapy practices as well, PT in Motion, a publication of the American Physical Therapy Association, reported in February. ATI Physical Therapy, for example, has been collecting patient data since 1996 on metrics such as range of motion, pain and activities of daily living and is now using it to optimize care for patients across its 900 locations.

In 2015, ATI Physical Therapy began using standardized patient-reported outcome measures (PROMs) to better measure patient outcomes and the performance of its 3,000 clinicians. That effort eventually led to the creation of a national patient registry approved by the National Institutes of Health’s Library of Medicine, according to PT in Motion.

The PROM data allow the company to identify the clinicians who consistently produce the best patient outcomes when dealing with certain diagnoses. Those PTs then can mentor ATI Physical Therapy’s other clinicians to improve their handling of these diagnoses, which leads to better patient outcomes, PT in Motion said. After clinicians are mentored, their patients’ PROM data in subsequent months can be used to determine whether the mentorship program yielded positive results.

The data from these reports allow ATI Physical Therapy to identify which topics to focus on in its continuing-education efforts and to create customized online courses to meet particular needs, PT in Motion reported. In the future, the company will introduce a new system that will allow every clinical director to sort the data easily and break it down, providing a clearer picture of a clinic’s performance in a wide variety of areas.

Smaller PT practices might not have enough data in their electronic health records (EHR) database to fully utilize predictive analytics because they don’t see enough patients. But there are efforts to help providers pool their data to increase its usefulness, including the APTA’s Physical Therapy Outcomes Registry. The registry collects and reports data for all of its participating practitioners, and as of last summer, it contained data on 35,000 patients, PT in Motion reported.

Having more accurate data allows PTs to advocate for their patients more effectively, PT in Motion said. For example, if an insurance company is paying for only six visits but the PT’s data show that patients with a particular diagnosis greatly benefit from eight visits, the PT can then make a data-supported case for extending treatment.

“We certainly can go to payers with this data,” Dr. Stephen Hunter of Intermountain Healthcare told PT in Motion. “Most payers know that if patients don’t improve in physical therapy, they’re going to require more expensive care.”

In the future, PTs may be able to use an algorithm to predict how long it will take a patient to recover from an injury, based on the records of thousands of other patients and data points from that current patient, including age, body mass index, level of fitness and health issues, PT in Motion said. Then, if the patient fails to make adequate progress within that timeframe, the team can decide how best to alter the treatment program.

“We can learn what to expect the course of care to be – maybe three weeks and six visits – and what outcome realistically can be achieved,” Dr. James Irrgang, chair of the Department of Physical Therapy at the University of Pittsburgh, told PT in Motion.

Search Jobs



Latest Career Advice