Improving patient care quality: 3 ways health organizations use predictive analytics

Technical Writer

Wouldn't it be great if your patients had signs over their heads when they walked into your practice or hospital that told what the future holds for their health over the next few hours, months or even potentially a few years down the road? You would then know exactly what patient care techniques to recommend. With predictive analytics, you can have the next best thing: real-time data that gives physicians insight into a patient's mental and physical health risks. Instead of reacting to symptoms, you can focus on preventing the illnesses or issues from occurring in the first place.

ScienceDaily recaps a study published in Health Affairs, titled Six Cases Where Big Data Can Reduce Healthcare Costs, that found six areas where hospitals can use data for early intervention and prevention. The study found that predictive analytics are especially helpful in the following situations:

  • High-cost patients.
  • Readmissions.
  • Triage.
  • Decompensation.
  • Adverse events.
  • Treatment optimization for diseases affecting multiple organs.

Here are three examples of healthcare systems currently using predictive analytics combined with health intervention strategies in one of these areas to improve operations by implementing health interventions.

Maine HealthInfoNet: Predicting and preventing ER visits

Often, when a provider contacts one of its patients through the statewide electronic health record (EHR) program to let them know that they are at high risk for a condition or disease, the patient had no idea of the risk factor beforehand. Data analytics is applied to 1.3 million patient health records, alerting providers if their patients are at risk of requiring an ER visit, according to Health IT Outcomes. When the software was testing using prior data, 74 percent of the patients flagged as being at risk actually did end up in the ER six months later. The article highlights how the most important part of the program is what happens after the software flags the patient: personal follow-up to work through an issue.

Carolinas HealthCare System: Using consumer data to learn about lifestyle habits

It's no secret that patients are often not honest with their doctor about lifestyle issues that can affect risk factors. Bloomberg reported that Carolinas HealthCare System, which has over 900 patient care centers in North Carolina and South Carolina, is currently working on a program to use predictive analytics on consumer data to spot health risk factors. Examples given in the article include buying cigarettes, living in a high-pollen-count area, canceling your gym membership and purchasing junk food. The information is also used to determine how likely someone is to have a heart attack or an asthma attack that warrants an emergency room visit, along with providing interventions such as gym recommendations, information about smoking cessation programs or a referral to a dietitian. Since some patients may not want their physician to access this information, Carolinas HealthCare System plans to allow the patient to opt out.

University of Iowa Hospitals and Clinics: Preventing post-op infections by identifying high-risk patients

After surgery, University of Iowa Hospitals and Clinics uses predictive analytics on data about each patient's history, medical conditions, vital signs during surgery and complications to determine which patients are at high risk for surgical complications. EMSWorld reported that using predictive analytics reduced infection in colon-surgery patients by 58 percent. By using this data, providers are able to provide a higher level of care to those high-risk patients and save money by not using excessive measures on low-risk patients.

By using predictive analytics to learn more about your patients and then intervening to reduce risk, you are both saving the hospital money and improving your patients' health. More importantly, you have the opportunity to change your patients' futures.

Improve the quality of patient care with analytics