Wednesday, December 11, 2013
There Is A Real Sting In The Tail In These Great Reported Results From The Use Of Analytics In Healthcare.
Two very interesting reports appeared a week or two ago.
December 5, 2013, 7:12 PM ET
REDWOOD CITY, CALIF. — Kaiser Permanente’s use of data analytics is helping it lower hospital mortality rates and look for ways to diagnose illnesses earlier. John Mattison, chief medical information officer at Kaiser spoke, Thursday, at VentureBeat’s Data Science Summit in Silicon Valley. Dr. Mattison predicts that by the year 2020, ten times more medical research will be generated by analyzing vast quantities of medical data than by conventional models of clinical research.
Over the past several years, Kaiser Permanente’s hospitals in southern California – the region with the most members — have enjoyed a lower mortality rate than the national average, according to data from the Centers for Medicare and Medicaid Services. “Today you have a 26% lower chance of dying in one of our hospitals than you do in other hospitals,” said Dr. Mattison, adding that Kaiser is starting to lower its mortality rate much faster than the national average. “A lot of this is directly rated to how we use data and integrate data,” he said.
Kaiser Permanente has some advantages in data collection over other medical providers because it provides physician, hospital and pharmacy services as well as health insurance to patients. All of those records are electronic. When a patient visits a Kaiser hospital, their entire health record, including doctor visits and medications, is immediately available. Kaiser can easily track patient outcomes after hospital procedures because patients see their doctors within the Kaiser system for follow-up visits. It’s a closed loop and all of that information resides in one place.
The informatics department at Kaiser, which is growing, looks at medical studies as well as information from its anonymized pool of information about patient outcomes to make implementable recommendations that it sends to physicians and hospitals through information alerts. One of the most high profile examples of this happened about a decade ago when Kaiser looked at its database of 1.4 million members and discovered that patients who took Vioxx were more likely to suffer a heart attack or sudden cardiac death than those who took a competing medication. Physicians were resistant to these alerts in the early years but the culture has changed and the informatics department continues to get requests for more of these alerts, said Dr. Mattison.
We also had this appear on the very same day.
December 5, 2013, 10:32 AM ET
University of Pittsburgh Medical Center has slashed readmission rates by 37% since it began using analytics to predict which patients were more likely to be readmitted to the hospital within 30 days.
That represents considerable savings for the hospital in terms of providing urgent care, let alone saving the hospital from potential penalties levied by the Centers for Medicare and Medicaid Services for failing to lower those rates.
The trouble for most hospitals is that they’re geared up for the “average patient,” whereas no one is actually an average patient. The role of analytics at UPMC is to determine most precisely which course of treatment will be most effective for each individual.
“Analytics helps you determine who you should focus on,” said Dr. Pamela Peele, chief analytics officer for the UPMC Insurance Services Division during a visit to CIO Journal offices.
According to Dr. Peele, the factors that hospitals should pay attention to are “jaw-dropping.” Far from the actual health of the patient, those factors have to do with how patients used care in the past – what services they’ve received over time and whether the use of the services has been “lumpy or smooth” over time.
Lots more here:
What we have here are very positive reports of the value of analytics in improving hospital and health system performance at the level of the most important measure - improved clinical outcomes.
The sting in the tail is that both the organisations involved are very strategic users of Health IT and have been evolving and improving their Health IT infrastructures over decades. They also have integrated environments where EHR data from both hospitals and ambulatory systems is easily accessible as well as the billing / insurance information and all that can be used for analysis.
For Australian Hospitals they have no access to the GP records and Medicare Payment records - so it now becomes very tricky to obtain such benefits.
It is really only those organisations that hold relevant ambulatory, hospital and insurance information which is easily accessible, and that also have a very advanced IT infrastructure that can replicate this. I wonder are the gurus and NEHTA and DoH working out how these sorts of benefits can be replicated in Australia or is the plan to mine the PCEHR to do a very second best effort?
Time will tell I guess.
Posted by Dr David G More MB PhD at Wednesday, December 11, 2013