Insights Thought Leadership

Reducing Patient Readmission Risk through Better Data Analytics

Readmission Risk-1



We’re all familiar with the “80/20 rule,” but in many cases this benchmark doesn’t go far enough. In healthcare, for instance, 5 percent of the U.S. patient population accounts for nearly half of overall costs. Some of this statistic is attributable to the sad reality that the sickest people require the most care. However, a meaningful share of expense concentration can be traced to frequent revisits and readmissions for conditions normally characterized as low or moderate risk. Medicare estimates its cost from the problem to be $26 billion, $17 billion of it avoidable.

The Affordable Care Act attempts to address this issue through the Hospital Readmissions Reduction Program (HRRP) by imposing penalties - in the form of lower Medicare reimbursements - on hospitals with high readmission rates. A Kaiser study found that HRRP penalties increased to 0.63% of reimbursements in 2018, up from 0.38% the prior year.

A recent ARGO white paper details healthcare technology that utilizes a data-driven readmission risk model that has delivered a 16% improvement for hospitals, compared with previous efforts to manage the process. Collaborating with Dr. Susan McBride of Texas Tech University, ARGO built a model incorporating 382 disease conditions, procedures and social factors, further honing its effectiveness by focusing on cases with moderate to low expectation of readmission.

The Elusive Leading Indicators

The healthcare technology tools available to providers today are reliant on existing segmented IT applications ranging from electronic medical records to add-on analytics to standalone case management platforms. These tools require significant time to gather information, toggling between applications and doubling or tripling documentation in the process. Obviously, this diverts resources from the true objective of delivering care. 

Reducing readmissions requires planning and communication across the entire chain of a patient’s providers and caregivers. One major challenge is identifying patients in need of intensive discharge planning, or those who are at risk of poor transition of care and therefore susceptible to readmission. For example, in addition to fairly intuitive factors such as age, gender and socioeconomic status (based on ZIP code), the model incorporates key social and demographic factors including:

  • Discharge location (e.g., home routine, home healthcare, short term care, other)
  • Number of emergency visits in patient’s history
  • Number of inpatient visits in patient’s history

The model has considered a total of 2.52 million patient visits, far more than could be incorporated absent a robust healthcare technology management tool, and is continuously updated. One key to its effectiveness is limiting focus to unexpected readmissions; by excluding chronic conditions and known risky cases from the sample, greater success has been achieved identifying and mitigating addressable risk cases.

Formulas Aren’t Everything

Hopefully it goes without saying that healthcare decisions cannot be reduced solely to numbers. For example, caregivers purposely pursue early interventions with maternity patients in order to avoid preterm labor. Hospitals may also require frequent visits for patients with chronic illnesses in order to explore alternate medicines and treatment options. No one wants to dissuade such actions.

Hospitals anticipate multiple planned and unplanned visits from patients with known high revisit health conditions requiring repeated and intensive treatment. Because of the (presumably) unavoidable nature of these revisits, hospitals typically do not track readmission activity from these cohorts. While outside the scope of this exercise, such cases are candidates for future data analysis.

By using healthcare technology that leverages a robust model like the one developed by ARGO based on Dr. McBride’s research, hospitals can readily gain insight into patients with the highest readmission risk and the greatest care gaps. Healthcare organizations can then allocate resources to the appropriate patients, generating benefits for both patients and the bottom line. The HRRP is intended to deliver better health outcomes with the help of predictive algorithms, hospitals can avoid reimbursement penalties while also providing true value-based care. 

For more information download our white paper, "Using Analytics to Reduce Preventable Readmission Risk."

Download White Paper Using Analytics to Reduce  Preventable Readmission risk