The Centers for Medicare and Medicaid Services sent out a confidential report to a small group of physicians to provide them feedback on their care of fee-for-service Medicare patients. The report is based on data from 2007.
One physician agreed to share this report. While we can argue about the quality of the data, inaccuracies, errors etc, it seemed at first glance to be a huge step forward in fixing the problem of variation in practice patterns. We are all familiar with the data showing how states that spend more money on Medicare patients have worse outcomes (first graph below). This report brings it down to the level of individual physicians.
Here is one graph that shows the variation in cost of care of Medicare patients with Diabetes amongst 317 physicians in North East Ohio. These costs were calculated only for physicians who had 30 or more Medicare patients with Diabetes. These cost were adjusted for risk from co-morbidities. The range was from 4,300 to 48,300 dollars!
One physician agreed to share this report. While we can argue about the quality of the data, inaccuracies, errors etc, it seemed at first glance to be a huge step forward in fixing the problem of variation in practice patterns. We are all familiar with the data showing how states that spend more money on Medicare patients have worse outcomes (first graph below). This report brings it down to the level of individual physicians.
Here is one graph that shows the variation in cost of care of Medicare patients with Diabetes amongst 317 physicians in North East Ohio. These costs were calculated only for physicians who had 30 or more Medicare patients with Diabetes. These cost were adjusted for risk from co-morbidities. The range was from 4,300 to 48,300 dollars!
The report also includes quality measures related to process of care (e.g. testing for HbA1c) but not even short term outcome measures like level of HbA1c.
Of course there will be legitimate quibbling about the quality and accuracy of the data in this report but the startling spread of the cost curve indicates that there is probably some room for improvement in practices at the high end. With the advent of EHRs and meaningful use, this data may be easier to collect for many more physicians. Will this thought scare off the physicians on the high end of the cost curve from implementing EHRs?
Can we make a substantial dent in the Medicare budget by impacting this variation? Atul Gawande described how presently there is no incentive for not doing more tests or spending more money. This may be the first step towards constructing some incentives.