There are a few themes that permeate this blog – the impact of new technology on medical practice, evidence-based care, health care financing, and a patient-centered approach to care delivery. The recent dust-up over the release of surgeon-specific outcome data touches almost all of them.
ProPublica, a not-for-profit organization devoted to investigative “journalism in the public interest” got the ball rolling last week with the publication of their “surgeon scorecard.” They compiled 5 years of Medicare data (2009-2013) on 8 generally elective surgical procedures: Knee and hip replacement, laparoscopic cholecystectomy, lumbar spinal fusion (broken out by anterior and posterior approach), “complete” prostatectomy, TURP and cervical spinal fusion. For each one, they identified a list of principal diagnosis codes associated with a hospital re-admission within 30 days of the surgery that could reasonably be interpreted as complications of the index surgery. For example, if a patient had undergone knee replacement and was admitted within 30 days with a principal diagnosis of “infection due to prosthesis” then that “counted” as a complication of surgery. Details of the methodology were provided online. The complication rates were adjusted by patient age, gender, and a few other variables, and their user-friendly tool allows for easy look-up of complication rates by surgeon or hospital.
The site also contains highly favorable testimonials supporting the public release of these data and the methodology used. Not surprisingly, lots of other online sources bubbled over with objections, some of which were pretty thoughtful and some of which were pretty angry. The major criticisms were mostly about the inadequacy of the risk-adjustment and the potentially misleading conclusions that result from limited sample sizes. This of course, could have serious implications for patients and surgeons. No one wants to mislead patients or mislabel physicians.
After wading through the methodology and a lot of the critiques, I ended up siding with others, like Ashish Jha who made a pretty compelling case that – from the perspective of a patient – the scorecard is a helpful advance. Not perfect, but better than the “data-free zone” that most patients experience when they are faced with choosing a surgeon. Furthermore, providing better tools for patients will require more investment in the collection of reliable clinical data (not just billing claims), and a commitment to transparency about patient outcomes.
What do you think?