The Archives of Internal Medicine published an article questioning the conventional wisdom regarding physician selection. In this article they tried to correlate a number of variable with a number of indicators of “quality care” (more on this, later). The authors did find that female sex, board certification (indicating completion of a residency and ability to pass an exam on set intervals), and graduation from a domestic medical school correlated with “significantly” better performance. Medical malpractice history, interestingly, did not indicate poor medical practices.
When looking more closely at this data, it shows some of the weaknesses of using statistics on numbers to describe behavior. The authors looked at 10,000 physicians and over a million patient encounters. They apparently analysed all of this information with a statistical package and then looked to see if any of the better performance was seemingly due to something other than chance. They found that all physicians did 62% of the activities that were expected. Female physicians for example were 1.6 percentage points higher than male physicians in regards to performance as a group (and the other differences were equally small). Because of sheer numbers this was seen to be “statistically significant” but given that based on this a random male doctor would be correct 61% of the time compared to the females 63% I think I would find a different way to pick a doctor.
This story was picked up in the lay press by the LA Times. In fairness to the writer, she did point out the weaknesses of the study, pointed out how limited information was for doctor selection, and pointed the reader to an AHRQ website with some pretty good advice:
Look for a doctor who:
[x] Is rated to give quality care.
[x] Has the training and background that meet your needs.
[x] Takes steps to prevent illness-for example, talks to you about quitting smoking.
[x] Has privileges at the hospital of your choice.
[x] Is part of your health plan, unless you can you afford to pay extra.
[x] Encourages you to ask questions.
[x] Listens to you.
[x] Explains things clearly.
[x] Treats you with respect.
The site goes on to explain quality care and offers a checklist to use when interviewing a physician to see if they meet your criteria.
This came up because Blue Cross/Blue Shield of Alabama has decided they are going to try to reward us primary care physicians for delivering “quality care” by increasing the visit fee by 5%. To find out whether I would qualify I decided to look on their Find-A-Doctor website to see where I stand. I was somewhat hesitant because I am an educator in a teaching practice. As such, all of the patients seen by learners while I am a teaching attending are counted against me in addition to those patients I am personally caring for. I would like to believe our learners are delivering excellent care but belief often is trumped by evidence.
Going beyond the user friendliness of the site (too many poorly categorized physicians with incorrect practice addresses) I am pleased to say that we are delivering good (and sometimes great) care. When measured on screening for certain types of cancer as well as diabetes care we were at or above the national and Alabama average. On the other hand, we do not perform as well as we could.
This measure is done on patients who have come to our office for care. They are not asked if they want for us to provide these services nor are we paid for providing such services. Although my staff, my residents, my colleagues, and myself are certainly motivated to deliver quality care, wouldn’t it be easier if we could dedicate staff to assure quality happened? Wouldn’t it be nice if we could enroll patients to incorporate their wishes and beliefs regarding their healthcare rather than assigning patients to me based on (perhaps) a single visit and assuming they want me to take responsibility for the care plan? Wouldn’t it be better if we were paid well for delivering care for acute illness and equally well for handling chronic illness and preventive services that don’t require a visit. Oh well, at least I’m liked (search for Perkins) by all 7 of the people who bothered to filled out the survey.