It was easy to feel a bit sorry for the Centers for Medicare & Medicaid Services Wednesday. They've got a big and likely acrimonious job ahead of them.
They've been charged by the Affordable Care Act to create PhysicianCompare.hhs.gov by January 1—only 64 days away—and make it even better than HospitalCompare.hhs.gov.
Then, not later than Jan. 1, 2013, for reporting periods to begin no sooner than Jan. 1, 2012, the agency must post information on quality, performance and patient experience for physicians enrolled in the Medicare program on a public website.
The smart people at CMS will have to figure out how to fill this site—an adaptation of Medicare's current Physician Finder not just with information about doctors that patients can use to make smart decisions about where to get care. They also must make the data useful for providers to improve upon that care.
They'll have to figure this all out and propose a rule that tries to satisfy everyone—or maybe not everyone and suffer the political slings and arrows —and put it into play in a dry run before next fall in a way that is "statistically valid and reliable," risk adjusted, "robust and accurate," the legislation directs.
On Wednesday, the agency took the bull by the horns and held a five-hour-long Town Hall to ask doctors, hospitals, employers and anyone else: What information, exactly should the website contain?
They heard an earful.
"We realize this creates quite a few more issues than other (website) settings may not have, and we know there's a lot of interest and opportunity but also some anxiety about it as well and how it will come out," acknowledged Michael T. Rapp M.D., CMS' director of quality measurement and health assessment.
Physicians' sex, race, languages spoken, office hours, medical school degrees and board certifications, age, hospitals where they have staff privileges and for how many years were just a few things many people said should be posted. To which health plan networks do they belong and are they still accepting new patients?
Have they won honors or awards, and do they devote time to caring for the poor?
And what about what quality measures? A recurring theme among some of the Town Hall speakers was whether the information to be posted will be adequately risk-adjusted for doctors who treat sicker patients, so they won't look worse than their colleagues who treat comparatively healthier patients. Do scientists today have enough confidence that the formulas are fair?
What about the average lengths of hospital stay for patients of a certain physician, adjusted by disease or condition? Or the rates of readmission among patients in their care?
Will a physicians' charges be part of what's viewed? Or whether they ever make housecalls? Do they enroll patients in clinical trials, accept money from pharmaceutical companies, or spend a good portion of their day in the classroom?
Can the patients talk with them by e-mail? Do they have an electronic medical record system in their practice? Who covers for them when they go on vacation?
How many patients must a physician care for in order to qualify for inclusion, so the data has enough power to be right, and over how many months or years?
Should quality and patients satisfaction measures be included for non-Medicare/Medicaid patients, the privately insured or self-pay patients the doctor may treat? And should those scores be shown separately, so patients can see if the doctor treats all patients the same?
Should the website show how often the doctor updates his continuing medical education, and in which specialties he's overdue?
The American Medical Association has some very big concerns about how this all plays out, especially when talking about evaluating doctors' care provided in a hospital setting. "We want to have reliable information that a patient can use if they are trying to select a physician or facility to get elective care," AMA president James Rohack MD, told me in an interview Tuesday.
"But our concern is that individual doctor level data right now is not read for prime time, especially in complex situations. The attribution of who's really responsible for that care is not worked out."
Rohack added that risk adjustment is still a big problem as well. "Some elective claim submission forms are limited in the number of co-morbid conditions they'll accept. Some take 4, 6, but the patient has 10."
In the Town Hall, many of the speakers had similar questions. Should the physician's track record on numerous care processes, such as 15 things doctors should do for a patient with diabetes, be rated individually, or will a composite for certain diagnoses suffice?
"Composites for consumers are very, very important," said Steven Findlay of Consumers Union. "(Having) 7, 10, 15 diabetes (measures) is not something most people are going to go through."
What if the doctors work in a group practice? Should the entire group be scored as well as the individual practitioner?
Should individual performance be rated for doctors who work in hospitals? And what about physicians who aren't chosen by the patient, and may never see the patient, but provide essential care on demand nevertheless such as hospitalists, radiologists, emergency room physicians and surgeons, anesthesiologists and pathologists? Perhaps there needs to be another way to score them.
"I don't think it's appropriate to put up information that's not actionable for patients," said one speaker who said he represented the American Society of Anesthesiologists. "Comparing anesthesiologists to anesthesiologists is not necessarily the best way to do things."
Will the scoring system rank 90% of the doctors as average or "good," eliminating much of the variability?
And what about patient satisfaction scores? How will they be rated and what measures will be used? Is it important to show how long patients had to wait to get an appointment and whether the physician was always on time?
Some of the speakers argued that the more information that is put on the website, the better for all. Others said that if there is too much data, it will fail in its primary purpose.
"It seems to me that some research in the past has indicated that more information is not necessarily better for consumers," said Ed Mendoza of the California Office of Statewide Health Planning and Development, which operates the nation's largest hospital discharge database. "You can have so much information that you disengage people before they've had a chance to think about choice."
Mendoza advised the federal planners to make sure they do user testing to make sure the site can engage consumers.
And what if the doctors don't like what gets posted? Will they have the right to review the data with sufficient time to appeal or prevent it from going up? Will visitors to the site know see an alert that the provider has questioned the data or has filed an appeal?
Rohack told me on the phone that the doctors aren't afraid of being graded. "The reason we became doctors was because we got good grades," he says. "And we want to have valid reliable data that patients can use to select their physicians and we believe that accurate reliable data is important for quality improvement studies.
"But if someone is going to use current methodology, it may give patients misleading information."
I understand his view. But as many of the Town Hall speakers pointed out, the perfect should not be the enemy of the good. Once the site is operating, problems can be identified and worked out. They certainly have been for HospitalCompare, a dynamic process ongoing as we speak.
PhysicianCompare.hhs.gov is going to happen because it's now legally required. I hope that providers, consumers and risk adjusters can make it all work well.