USA Today recently published a story about ProPublica, a nonprofit news organization that has developed a metric to score surgeons’ performance, comparing them with their peers. The study is intended as a tool for consumers, but it has generated concern among surgeons, who feel they are being treated unfairly.
What the article neglects to mention is that rating doctors and hospitals is not new in the general health world. Scoring medical providers has been a practice for decades. The Leapfrog Group, which scores hospitals, has been in business much more than 20 years. Doctor Scorecard scores medical doctors, and a Google search will offer more.
What is different about the ProPublica analysis is that it is based entirely on data and singles out surgeons treating the Medicare population. It also uses an adjustment score for the difficulty of cases analyzed called an adjusted complication rate.
The ProPublica study includes 17,000 doctors performing what are called low-risk, elective surgical procedures derived from Medicare data. The adjusted complication rate selects cases that are considered low risk, such as gall bladder removal or hip replacement. The study looks for complications such as infection or blood clots that require post-operative care, in this case re-hospitalization.
The cost of post-operative care requiring hospital readmission amounted to $645 million, which was billed to taxpayers for 66,000 Medicare patients from 2009 to 2013. Logic says that if surgical complications requiring hospitalization are so costly for Medicare patients, the costs must translate to astounding rates in workers’ compensation, as well. However, the study does not directly apply to work comp doctors.
The ProPublica study does not directly translate to workers’ compensation because the study examines Medicare patients only. While some injured workers qualify for Medicare, the majority are healthy, working adults under Medicare age.
What does translate from the study is that evaluating and rating medical doctor performance based on the data is do-able and important. However, it should not be limited to surgeons. The analysis of doctor performance must be comprehensive, accurate and fair.
Rather than using the limited measure of adjusted complication rate following surgery, a broader view of the claim and claimant is appropriate for workers’ compensation. Analysis is not limited to those cases with complications. Instead, all claims are analyzed. Results are adjusted by the claimant’s age, general health (indicated by co-morbidities), and the type and severity of the injury itself. Administrative management analyses are also important in workers’ compensation such as direct medical costs, indemnity costs, return to work, and case duration, among others.
Case complexity, sometimes presented as case mix adjustment, is important to fairness in rating doctors in workers’ compensation. Also, analyzing a broad scope of data elements smoothes the variability, leading to more accuracy. Fortunately, in workers’ compensation, claims have a very wide range of revealing data elements that can be drawn from a payer’s multiple data silos.
The ProPublica study has created pushback from the physician community for several reasons. For one, gall bladder surgery is often performed in an outpatient setting, so re-hospitalization is a meaningless metric. The same is also true for others of the so-called low-risk surgery category. Moreover, the study names names.
Published provider ratings from a national survey caused much of the angst noted in the USA article. Names were even published in local papers, naming physicians well-known in their communities. Doctors cried foul!
Expecting the general population of patients to understand what the ratings mean, regardless of their accuracy, is naive. Ratings listed as 2.5 or 1.6 have obscure meanings to the uninitiated. Fortunately, workers’ compensation providers do not face that level of exposure. Doctor ratings in workers’ compensation are not published for the general public or made available for consumer interpretation.