Building a Bundle: Setting the Baseline

(Post Four in a series)

By Joanne Conroy, MD

Last week, we launched into constructing a new bundle. We selected DRGs (Diagnosis Related Groups) 216-221, which cover valve replacement with and without complications. It makes sense because many AMCs actually do as many valves as they perform CABGs (Coronary Artery Bypass Grafts). Valve replacements can be trickier, can require postoperative anticoagulation if a mechanical valve is used, and can have a stormy postoperative course if left ventricular function has been significantly compromised prior to diagnosis.

When we looked at the principal diagnoses under these DRGS, the vast majority are open replacement of the aortic or mitral valves. There were a sprinkling of valvuloplasties, a few tricuspid valve procedures, and few of the new percutaneous procedures also included in our sample cohort. I also noticed a few admissions with strange principal diagnoses that obviously came in for something else and required a valve replacement.  

So last week was dedicated to plowing through the CABG episode grouper skeleton, trying to determine which of the exclusions and relevant readmissions were applicable to the valve population. This is where the involvement of clinicians in the discussion sped the process along immeasurably. We had a surgeon on the line from one of our member institutions, who in five minutes outlined the common perioperative concerns including recurrent pleural effusion, anticoagulation management, atrial fibrillation, infection, altered mental status, stroke, etc.  In order to be complete, we are going to cross-reference our list of relevant services with the Society of Thoracic Surgeons’ national database to ensure we exclude the catastrophic and rare events.

Although the inclusion or exclusion of a service or readmission in the bundle appears a bit random, it is all about setting your baseline.  If you have a wider net and are more inclusive, you may have greater opportunities for savings as you standardize care and focus on care coordination post-discharge.

There was a long discussion about the minimally invasive value procedures, which have lower discharge to post-acute care (in some studies up to 90 percent discharge to home) and could dramatically reduce episode costs. However, the device costs are quite high and sites are concerned that they have inadequate historical data to accurately construct a baseline. There are five new hospital inpatient ICD-9-CM procedure codes for valves via either an endovascular or transapical approach that, effective October 1, 2011 (FY2012), are grouped with the open valve procedure DRGs with a reimbursement range of $24,500 to $54,500. We will need to query CMMI about excluding these because of inadequate historical information. We should be examining these (from a clinical effectiveness perspective) to see if the total cost of care is less across a 90- to 180-day episode; we introduce much of this new technology with the hope that there will be decreased morbidity and improved outcome at a lower cost.

After slogging through the inclusions and exclusions for over an hour, we had consensus on the parameters and the episode logic code can be written and applied to the data.

Next week: on to quality and patient engagement measures…

—Joanne Conroy, MD, is Chief Health Care Officer at the Association of American Medical Colleges. She can be reached at Follow her on Twitter @joanneconroymd.