Variation in hospital results – how to use the data

It has long been known that there’s wide variation in the type, quantity, and outcomes of medical care across providers.  A new report – research done by Dartmouth, and funded by the Robert Wood Johnson Foundation – looks at variations in re-admission rates among and between hospitals, and provides some striking insights.

The researchers used 2010 Medicare data; the overall results indicate one of eight surgical patients were readmitted within 30 days of discharge.  Non-surgical patients were readmitted more often; one out of six was back in within 30 days.

According to the report, the “issue of patients being readmitted to the hospital is considered important because many are avoidable and, as the report notes, can occur because of differences in patient health status; the quality of inpatient care, discharge planning, and care coordination; the availability and effectiveness of local primary care; and the threshold for admission in the area.” [emphasis added]

CMS recently began reducing reimbursement to hospitals with high levels of readmissions – which will make it really important for those hospitals.

So that’s kinda interesting, but not really. Here’s what’s really interesting.

The good folks at Dartmouth have published the re-admit rates for all hospitals, and you can download the spreadsheets.  Now before we go picking the best hospitals based only on their numbers, let’s look a little deeper.

Looking at two hospitals in my home state of CT, one can see the readmit rate for St Vincent’s in Bridgeport is much higher than Middlesex Hospital’s.  One answer may lie in the population; Bridgeport is a lower-income area than Middlesex, and likely has a much higher proportion of patients without adequate primary care and/or insurance.  Dartmouth provides some insight into this – 82% of patients discharged from Middlesex after congestive heart failure treatment saw a primary care provider compared to only 60 percent at St Vincent’s.

A couple other stats looked interesting; the data for surgical re-admits for UPMC facilities indicates they do a pretty good job keeping readmits down – and therefore overall quality is likely better than most (again this is just one data point).  Similarly, patients discharged after a heart attack from Geisinger’s Wyoming Valley facility have a high incidence of primary care follow up – compared to other facilities in PA (58 percent v 48 percent.  However, they’d be just above average in Wisconsin (54.4 percent).

What does this mean for you?

While there’s a LOT to digest here, I’d suggest one use would be for network direction.  Identify the hospitals with statistically better results, assess them for confounding factors, and think about how best you can direct patients/injured workers to those better-performing facilities.

 

 

Comments are closed.