New California Data Helps Patients Evaluate Hospital Quality

California starts to report hospital mortality rates

California is breaking new ground by publishing eight hospital mortality rates as a start so patients can determine which hospitals have the best quality outcomes. The state is using "risk adjusted" measures to better compare how "sick" a patient is and how likely they would be to die from the condition.

The eight measures include death from three medical conditions; acute stroke, hip fracture and gastrointestinal bleeding as well as five surgical procedures: coronary angioplasty, carotid endarterectomy, craniotomy, esophageal resection and pancreatic resection.

The new data from of hospital mortality rates offers a snapshot of the quality of care provided by over 400 California hospitals.

"This is the first time for many hospitals that they have been able to benchmark their performance on these procedures against all California hospitals," said Joseph Parker, director of the health care outcomes center for the Office of Statewide Health Planning and Development.


Of the state's 384 such hospitals, 25 performed better than the state average on at least one of the procedures or conditions, and 94 did worse in 2007. In 2006, 33 hospitals scored better than average, and 98 rated worse on at least one of the indicators.

It is hoped that public reporting of quality data will give patients information to make health care decisions. However, it would be premature to think that these measures are anything other than a raw beginning. There are a number of problems with the mortality rate data that patients need to know:

1. The eight conditions it tracks cannot be generalized to looking at how a hospital performs in other aspects of patient care. They are very specific and reflect only that measure and cannot be used to conclude that a hospital is good or bad
2. A hospital can be above average on one measure and below average on another.
3. The data is old (2006,07) and many hospitals have made improvements since that time.
4. The data fails to properly account for patients with multiple conditions or those who do not wish to be resuscitated.
5. The state relied on data used for billing purposes and did not verify to make sure that the coding was correct and that it matched the patient record information.

Most patients don’t plan a stroke or a broken hip so this type of data might even be confusing and make them feel insecure about their community hospital if it scored average or lower on a certain measure.

And finally, not everything that matters can be measured and not everything that is measured matters.