Guest post: Using the Elixhauser model for PSIs

by James S. Kennedy, MD, CCS, CDIP, CCDS

Many hospitals find that their patient safety indicator (PSI) ratios remain high despite doing a spectacular job of addressing these events and exclusions. They may fail to realize that the Health and Human Services Agency for Healthcare Research and Quality (AHRQ) has a risk-adjustment methodology that predicts each of these PSIs and is dependent upon the documentation and coding of PSI-sensitive risk factors. These follow the Healthcare Cost and Utilization Project (HCUP) model using the Elixhauser Comorbidity software.

For example, in PSI 11 (Postoperative respiratory failure rate), the observed rate is compared to an expected rate determined by certain regression model coefficients.

The trick is knowing what codes drive the coefficients that predict the likelihood of PSI 11 (and others as described in the reference).

Note that many conditions not affecting the MS-DRG for payment, such as hyperkalemia or hypokalemia, do affect the expected PSI metric.

This exercise must be repeated for all the PSIs, keeping in mind that to be coded, they must be explicitly documented when present.

A list of all ICD-10-CM codes mapping to the Elixhauser model is available online. I suggest developing preoperative documentation templates that capture these conditions in your preoperative and postoperative workflows so that when the observed PSI occurs, it is in proportion to what is expected.

Editor’s note: This article originally appeared in JustCoding. Dr. Kennedy is a general internist and certified coder, specializing in clinical effectiveness, medical informatics, and clinical documentation and coding improvement strategies. Contact him at 615-479-7021 or at jkennedy@cdimd.com. Opinions expressed are that of the author and do not necessarily represent HCPro, ACDIS, or any of its subsidiaries.