Creating an accurate system for evaluating hospital quality for payments, as well as spotting areas where savings could be achieved and deaths prevented, is becoming increasingly important to health system leaders. However, the current list of hospital-acquired conditions (HACs) that Medicare uses to set payments for hospitals, may be too narrow finds an analysis of discharge codes conducted by the Premier Inc. healthcare alliance.

The Charlton, North Carolina-based Premier mined more than 5.5 million de-identified ICD-9 coded discharge abstracts from approximately 530 inpatient facilities in its database of federal fiscal year 2013 claims. The facilities that produced the data ranged from small critical access hospitals to large academic medical centers, and represented facilities in 47 states.

Premier then used this data to develop a larger list of conditions that could be hospital-acquired and conducted further analysis to see which of its potential inpatient conditions (PICs) had the highest impact on cost, mortality and length of stay, as well as being addressable through the use of evidence-based best practices.

When Premier listed its top 10 addressable conditions that have the greatest negative impact cost, mortality and length of stay, it noted that only two of the 10 overlapped with Federal policies.

“We believe the methodology presented here represents a significant improvement over currently available methods. Moreover, in identifying high impact conditions, the Premier method for identifying complications, once refined with causal modeling, might suggest a broader set of measures for use in moving a range of quality indicators. Measures that are broad in scope, that include plausible causal inferences, and that are associated with solid evidence-based practices to reduce their incidence may ultimately gain traction and become the next-generation measures used in both CMS and private payer pay-for-performance programs,” the Premier report authors concluded.

The Premier top 10 list of possibly preventable potential inpatient conditions with the greatest impact on cost, mortality and hospital length of stay were:

1. Acute Myocardial Infarction
2. Acute Renal Failure
3. Aspiration Pneumonia
4. Cerebral Infarction
5. Gastrointestinal (GI) Ulceration and Hemorrhage
6. Hypotension
7. Respiratory Failure
8. Sepsis/Bacteremia
9. Pulmonary Embolism
10. Ventilator Associated Pneumonia

If half of these cases in just the Premier analysis of a single year of data could have been prevented, it would have saved more $1.2 billion in costs and nearly 12,000 lives.