Twelve hours before New Year's Eve, California health officials released inpatient mortality rates for each of the state's 335 hospitals for each of a dozen care categories.
The report came in two parts, one for 2008 and one for 2009. Lo and behold, many hospitals that scored "better" than other hospitals in 2008 also scored "better" in 2009. Likewise, many hospitals that were "worse" in 2008 were also "worse" in 2009.
"This was a surprise," says Joe Parker, Director of the Healthcare Outcomes Center at the California Office of Statewide Health Planning and Development and the primary author of the report. "We had not seen that kind of consistency before across the 12 measures."
Wake up and smell the coffee. This is the dawning of the age of transparency in hospital mortality. (With apologies to the Fifth Dimension.)
Specifically, Parker explains, "Hospitals in 2008 that were outliers, that is worse, were five times as likely to be worse in the same way the next year." Hospitals that were better in 2008 were six times more likely to perform better in 2009.
Also, if a hospital appears worse in one category, it is likely to be either average or worse in another category. Likewise, if a facility shows up better in one category, it's more likely to be either average or better in another category.
Only 12 of the 335 hospitals in either year had a mixture of being better and worse in multiple categories, and these were all risk-adjusted according to criteria set by the Agency for Healthcare Research and Quality.
There wasn't the typical up/down variation as is usually seen across such large expanse of categories either. Perhaps, this is just the kind of information that will shine a light on those hospitals that are doing the right things with their patients, something employers, payers, patients and other providers most certainly want to know.
However, this report also begs the question, are poor hospitals just condemned to poor performance because of “extenuating circumstances,” with little to nothing they can do about it?
Going back to similar reports issued for 2006 and 2007, some of the same trends persist, although only eight procedures or conditions were measured and the AHRQ risk adjustment metrics, which have been revised three times since, were significantly different.
Two examples of this are St. Agnes Medical Center in Fresno and Presbyterian Intercommunity Hospital in Whittier.
St. Agnes Medical Center was "worse" in gastrointestinal hemorrhage and pneumonia death in 2008. Then in 2009, St. Agnes again scored "worse," but now in three categories, percutaneous coronary angioplasty, congestive heart failure and pneumonia.
Presbyterian Intercommunity Hospital was better in three categories, acute myocardial infarction, acute stroke and pneumonia, in 2008. But in 2009 it was better in six areas, having achieved additional lower mortality scores in gastrointestinal hemorrhage, congestive heart failure and craniotomy.
Parker says the rankings give clues as to what might be going on with a hospital's process of care. "In the case of stroke, for example, we know there's evidence that a timely CT scan and a swallow assessment can help guide the correct treatment decisions, and may prevent death or extend life. Likewise, timely administration of the correct antibiotic to patients with pneumonia has been shown to increase survival rates," he says.
He urges hospitals that received poor scores – or those simply wanting to improve – to go back to the patients' charts "to see whether there were issues with the patient's care."
Hospitals should review to see if there were extenuating circumstances or conditions that may be contributing to certain hospitals having higher death rates.
But isn't it possible that some of this trend is partially explained by patient payer mix and socioeconomic status? The unchangeable "extenuating circumstance?"
That's unlikely, Parker says. But it could be true that hospitals serving areas that lack palliative care resources might be forced to keep those patients until the end. That would be counted as an inpatient death, while hospitals in areas with the ability to discharge patients to palliative care services would avoid another mortality statistic.
Likewise, hospitals without transfer agreements to accept certain types of patients, for example, those on ventilators, might see themselves at a disadvantage. But that doesn't explain poor performers for all hospitals or in all care categories, Parker says.
There's evidence to think that hospitals are at least starting to pay closer attention to these publicly available numbers. Between 2008 and 2009, mortality in eight of the 12 categories went down. For example mortality per 100 cases for craniotomy was reduced from 7.5 to 6.7, while inpatient mortality for patients with heart attack went from 7.5 to 7.1.
Jim Lott, executive vice president of the Hospital Association of Southern California, which includes more than 150 hospitals in the state, says the reports are critically important. "In this day and age, hospitals can't afford to look the other way, because they know they are now transparent, although some hospitals obviously will try to do more (with the data) than others."
He adds that any hospital with a "worse" score, or which is unhappy with average scores, should first "look to the physicians practicing in that area of care. If a hospital has a high mortality rate for craniotomy, well you might want to look to see who the neurosurgeons are on staff. There's no secret that I would not go to certain hospitals (for surgery) not because of the hospital itself, but because of the members of its surgical staff."
After all, Lott says, payers are looking at reputation as well as the hospital's cost. "At the end of the day, they're looking at network placement and price. And the bottom line is if the hospital has a bad reputation, they'll pay attention to that and ask, do I need this hospital in my network?" he says.
Yes, this is the dawning of the age of mortality transparency. And as the lyrics from the 1969 song say, "Let the sun shine in."
California's report may be seen here.