Family physicians and others have for years complained about ridiculous quality measures such as the meaningless use program, HEDIS, MACRA, etc. To its credit, the Agency for Health Care Research and Quality was willing to take a step back and ask, “What does high quality look like?”
Researchers led by Rebecca Etz, PhD were funded to go out and ask a variety of stakeholders their answers to this question. I’m sure this comes as no surprise to family physicians, but patients did not answer blood pressure numbers, hemoglobin A1c levels, mammogram rates, etc. In fact, when told how current quality measures often work, many responded, “Wait, you mean the doctor advised the patient to take their blood pressure medicine and the patient didn’t, but you blame the doctor? That’s stupid.” From the mouth of babes!
The researchers did a bunch of fancy qualitative work with layers of analysis. They identified 11 separate factors:
Accessibility, advocacy, community context, comprehensiveness, continuity, coordination, family context, goal-oriented care, health promotion, integration, and relationship.
Notice that none of these are easily measured with EMR data. They are too complex and too grey. They reflect the complexity and humanness of family medicine. One of the specific responses under these 11 was, “My family doctor and I have been through a lot together.” This becomes harder when the patient is incentivized to visit urgent care centers and tele-doc strangers.
Unfortunately, I’m afraid the suits — the MBAs, CFOs, MHAs, and CEOs who control the purse strings — simply won’t get this. They want measurable markers, and these aren’t, at least not from data sources that already exist.
What we do in family medicine is simply too complex for computers to keep up with or understand. We now have some evidence that patients understand this. When will the money changers?
Image credit: Shutterstock.com
Original Article Posted at : https://www.kevinmd.com/blog/2019/08/medicine-is-too-complex-for-computers-to-keep-up-with-or-understand.html