We've tried to make the Texas Hospital Checkup as accurate as possible. We used patient discharge data from the billing records of all acute care hospitals in Texas gathered by the Texas Health Care Information Council.

Billing records provide a lot of information about what each patient was sick with, what procedures were done and whether the patient died. But billing records don't provide other details that could help determine the expected outcome for each patient.

To be fair, we took into account the differences between each hospital's patients—including how sick they were and differences in age and sex. But it's not possible to account for all differences among patients—like smoking, diet and other lifestyle issues. We used the 3M All Patient Refined Diagnosis Related Group (APR-DRG)™ system to account for how sick patients were and how likely they were to die during their hospital stay for a particular procedure. We didn't show mortality rates for hospitals that did fewer than 75 of each type of procedure a year because there isn't enough information to calculate a statistically reliable score.

The comparative analysis presented here was prepared by the consulting division of Information Management Systems, Inc., IMS of Miami, Florida, using data collected by the Texas Health Care Information Council (THCIC). You can link to IMS here.

Here are some other things to remember about information on hospital performance:

	There are coding differences between hospitals. Some hospitals code more completely and accurately than other hospitals, making it tough to compare all hospitals fairly.
	It's sometimes unclear when a condition occurs. Billing records can't tell you if a condition was present at admission or if it happened during the hospital stay. For example, a hospital may have a high rate of complications because lots of physicians refer complex cases to it or because the care that the hospital delivered caused the complications.
	It's hard to classify disease accurately. The ICD-9 classification codes contained in the patient record may not be specific enough to adequately characterize a patient's condition, making it difficult to accurately perform risk and severity adjustment calculation.
	The billing record doesn't show things that happen after the hospital stay. For example, a patient may have been discharged alive but died two days later.
	The information is for a specific period of time—from April 2001 through March 2002. The report is a snapshot of that 12-month time period. Hospitals may be doing a better or worse job now.
	Charges shown are the hospital's billed rates. Actual amounts paid may be significantly less due to rate negotiations with payers. Physician charges aren't included.

I've read and understand the limits of the information in the Texas Hospital Checkup and want to move ahead.