Kutcher said those improvements likely aren’t reflected in the mortality rate figures in the current report because the information used is nearly 2 years old.

In the end, he said, patients should look at all available data on a hospital and see how open they are about sharing information and instituting programs for improvement.

The mortality rates on the Web site are adjusted to take into account how sick a patient was when they were admitted to the hospital.

While more telling than in the past, the figures used are still imprecise, particularly for smaller hospitals. That’s because the smaller the number of patients treated, the less certain the hospital’s death rate. It’s easier, for example, to get a more precise rate for a hospital that treats thousands of heart attacks than for a hospital that treats just a few a year.

The Web site compensates for that by including an “interval estimate” — a horizontal bar showing the range into which the actual rate could fall. ISJ for example has a death rate of 16.2 percent for heart attack patients. Their range of death rate is listed at 12 percent to 20 percent.

The data counts deaths if they occur within 30 days of the time the patient was released from the hospital, under the assumption that poor care could result in death weeks after being released.