Healthcare News & Insights

Prediction model assesses patients at risk for 30-day readmissions

Just think how much you could lower your readmission rate if you had a tool to predict those patients at greatest risk of being readmitted after discharge. Even better, what if that tool used administrative and clinical data readily available at no cost to your facility? Sound like a dream come true? 

The good news is researchers from Brigham and Women’s Hospital have done just that.

According to a study published in JAMA Internal Medicine, the researchers collected data between 2009 and 2010 for all patient discharges from any medical services, which ended up being approximately 10,731 discharges. Of those discharges, 2,398 (22.3%) were followed by a 30-day readmission. And of those readmissions, 879 (8,5%) were deemed as potentially avoidable.

Readmission predictors

From the avoidable readmissions, a multivariable logistic regression was used to identified seven independent factors that best predict whether a patient will be readmitted. They are:

  1. hemoglobin levels at discharge
  2. whether the patient was discharged from an oncology service
  3. sodium levels at discharge
  4. number of procedures during the first admission
  5. whether the admission was elective or non-elective
  6. number of admissions within the prior year, and
  7. length of hospital stay.

Simple but effective

Each variable is assigned a value, with the more heavily weighted predictors being worth two points, and the rest being worth one point. Physicians can easily go through the seven independent variables at a patient’s bedside before discharge to determine if he or she is at a high-risk for readmission.

The higher the score, the more likely the patient is to be readmitted. So if a patient scores high, then the physicians knows that patient needs intensive transitional care interventions to make sure he or she doesn’t end up back in the hospital.

These can include:

  • reviewing discharge instructions with the patient and a family member
  • home visits from a nurse or physician
  • follow-up calls from a provider to check on how the patient is doing and answer any questions, and
  • scheduling follow-up appointments for the patient and providing reminder calls.

With hospital readmissions being such a costly problem for the industry, preventing avoidable readmissions would be a huge accomplishment. Studies have shown that nearly 20% of Medicare patients experience 30-day readmissions at an cost of $17 billion a year. So preventing them would not only save our healthcare system a ton of money, but it would also improve patient care. And this simple prediction models gives hospitals a no-cost way to identify those patients needing intensive transitional care interventions.

 

 

 

Subscribe Today

Get the latest and greatest healthcare news and insights delivered to your inbox.

css.php