Hmm...I agree that poorly designed quality measures (antibiotics for pneumonia) are, on net, bad. I agree that all quality measures have negatives that need to be part of an cost/benefit evaluation. I agree that measures are probably net negative for the very best practitioners and systems. Incidentally, this is true for all metrics in all industries.
I don't think that it follows that all or most quality measures are bad on net, especially at the system/payer level. The A1c measure, for example, doesn't prescribe a treatment protocol (as the pneumonia one did.) Whether a patient's diabetes is uncontrolled due to lifestyle or medical factors, their health system (though maybe not an individual doctor) should try to help them improve it, and measuring success or failure is meaningful to incentivizing that.
I'm not totally clear on your argument yet though. One of the following?
1. All metrics are always or usually bad and healthcare quality metrics are no exception.
2. Quality measures in healthcare are always or usually bad for healthcare-specific reasons and should be eradicated.
3. Quality measures as they're currently implemented are bad, and they should be changed in xyz ways.
4. Quality measures are, on net, good, but not as uniformly beneficial as some VBC proponents think they are.
Thanks for reading the article and comment. Let me try to answer one by one:
1. All quality measures are not bad, especially at the system/payor level: At a 50,000-foot view, I agree with you. Health plans cannot directly control outcomes and will transfer the responsibility down the chain. The end of the chain is the individual doctor -- and this is where Goodheart's law comes into play.
2. Diabetes control does not prescribe a treatment protocol: I agree, but the only tools most doctors have at their disposal are meds and discharging patients from the practice. If you hold an individual doctor responsible for A1c control, they will use these tools, especially given the financial penalties and public humiliation from posting quality performance.
On the overall argument, all quality metrics in use today have not been tested using a randomized controlled trial, and almost all of them are bad when implemented at an individual doctor's level.
The current methodology uses guidelines as quality measures. This is a subtle but very important distinction--guidelines are not rules, but quality metrics turn them into rules--with unintended side effects and patient harm.
I have upcoming articles that will take individual quality measures and dive into how they cause harm.
Hmm...I agree that poorly designed quality measures (antibiotics for pneumonia) are, on net, bad. I agree that all quality measures have negatives that need to be part of an cost/benefit evaluation. I agree that measures are probably net negative for the very best practitioners and systems. Incidentally, this is true for all metrics in all industries.
I don't think that it follows that all or most quality measures are bad on net, especially at the system/payer level. The A1c measure, for example, doesn't prescribe a treatment protocol (as the pneumonia one did.) Whether a patient's diabetes is uncontrolled due to lifestyle or medical factors, their health system (though maybe not an individual doctor) should try to help them improve it, and measuring success or failure is meaningful to incentivizing that.
I'm not totally clear on your argument yet though. One of the following?
1. All metrics are always or usually bad and healthcare quality metrics are no exception.
2. Quality measures in healthcare are always or usually bad for healthcare-specific reasons and should be eradicated.
3. Quality measures as they're currently implemented are bad, and they should be changed in xyz ways.
4. Quality measures are, on net, good, but not as uniformly beneficial as some VBC proponents think they are.
Tyler,
Thanks for reading the article and comment. Let me try to answer one by one:
1. All quality measures are not bad, especially at the system/payor level: At a 50,000-foot view, I agree with you. Health plans cannot directly control outcomes and will transfer the responsibility down the chain. The end of the chain is the individual doctor -- and this is where Goodheart's law comes into play.
2. Diabetes control does not prescribe a treatment protocol: I agree, but the only tools most doctors have at their disposal are meds and discharging patients from the practice. If you hold an individual doctor responsible for A1c control, they will use these tools, especially given the financial penalties and public humiliation from posting quality performance.
On the overall argument, all quality metrics in use today have not been tested using a randomized controlled trial, and almost all of them are bad when implemented at an individual doctor's level.
The current methodology uses guidelines as quality measures. This is a subtle but very important distinction--guidelines are not rules, but quality metrics turn them into rules--with unintended side effects and patient harm.
I have upcoming articles that will take individual quality measures and dive into how they cause harm.