Predictive Information Helps Hospital Case Managers

by Stefani Daniels, Managing Partner
Published on Jul 28, 2015

We've long taken the position that effective case management in the hospital setting is too expensive for all and not needed by most.  And we're talkng about professional case management practice - not the activities related to utilization review, discharge planning or discharge arranging. Basically, if the small number of patients who are destined to be high cost or at risk for clinical or financial burdens, can be prospectively identified and helped, a lot of costs could be saved by preventing readmissions, reducing resource utilization, and minimizing complications.  By focusing resources on a selected population of patients in need of advocacy to facilitate their progression-of-care through the entire continuum, the costs for the entire population of patients can be decreased.  In the meantime, more modest transitional care programs can be implemented for the remaining patients including payer sponsored community care management programs. 

How are the small numbers of patients selected and who are they?

Generally, unless the hospital has predictive analytics capabilities, the hospital case manager has to rely on the initial patient assessment that the nurse completes for each patient upon admission. That assessment - whether paper or embedded within the EMR - must be able to capture:

  1. patients with two or more chronic illnesses that are being treated
  2. patients with advanced illness who are nearing end of life who may benefit from palliative care
  3. patients with high use of hospital/community resources including ED visits or past admissions and, depending upon your EMR interoperabilities, physician office visits. 
  4. patients with polypharmacy or specific categories of pharmaceuticals like anti-coags
  5. patients with socioeconomic challenges who may benefit from a non professional community health worker.

The science behind the use of these inputs, whether technologically generated or not, is imperfect but getting better, and the more inputs, the better.  As Voltaire said, "Don't let the perfect be the enemy of the good."  Using the initial patient assessment until the hospital can generate real-time objective information based on a combination of hospital and payer data, using readily available information from changes to the initial patient assessment  will get you and the patient's healthcare team started.  The cost of care management/care coordination is variable.  It can require a small investment by a small hospital using just the information captured by the admitting nurse and then can be flexed up as the expertise grows and the benefits of care management across the continuum is evident.  Or it can be a large scale investment with resources positioned across the continuum using a single best practice algorithm to keep the patients' progression-of-care on track.