Appendix D. Methodology Enhancements for 2007 Ratings Models
© Copyright 2007 Health Grades, Inc. All rights reserved.
Hospital Report CardsTM Methodology 39
May not be reprinted or reproduced without permission from Health Grades, Inc.
Cardiac
Cohorts Affected
2007 Ratings' Model Change
Rationale for Changes
Coronary Bypass
Surgery
Excluded patients with Carotid
Endarterectomy (38.12, 39.72) in the
same hospitalization
Low volume prohibited achieving the statistical
significance required to adequately risk adjust for
this population.
Coronary Bypass
Surgery
Valve Surgery
Excluded patients with aortic
dissection (441.0-441.03)
These patients were excluded because they are
not representative of the bypass population. This
rare group can represent the sickest, most acute
population and possibly be under risk adjusted.
Coronary Bypass
Surgery
Valve Surgery
Excluded patients with aortic
abdominal aneurysm, or thoracic
aneurysm (38.44, 38.64, 39.71,
441.2)
This situation is infrequent but when it does occur,
inadequate risk adjustment may occur due to the
low volume of this patient population (and not
reaching the statistical significance required to
remain in the final prediction model), thus
necessitating exclusion from the patient
population evaluated.
Valve Surgery
As in the 2006 model, included
combined codes for "other valve
repair" (35.10, 35.11, 35.13, 35.14)
and "mitral repair" (35.12) as potential
predictors (positive or negative) of
mortality
Although rates of valve repairs are increasing
across several institutions, there are still
insufficient numbers associated with each valve
repair procedure code to reach statistical
significance. As such, "like" valve repair codes
were combined to increase the likelihood of
reaching statistical significance to be included in
the final model.
