From a doc who just received a report card:
You're absolutely right on several levels.
The "Small N" problem:
Judging your effectiveness and quality as a physician based on infinitesimal biometrics is silly, potentially dangerous, certainly adds expense with little if any value. Any good researcher can tell you the risk of assuming your overall performance based on a handful of metrics applied to a very minor sample of patients.
Treating organ system or treating people?
The paradigm of measuring quality by organ system and bio metric is based on the notion that overall quality of care is the sum of the parts and that humans are sums of organ systems. This is a demonstrably false notion that leads us down into the weeds of measurement. This path is very seductive because of the obvious logic behind the micrometric: If having a mammogram is a good thing, then counting the percent of mammography is a good thing and is a good indicator of quality of care.
The problem with this path is the lack of replicability outside the research environment. In the real world of primary care we manage patients with multiple conditions, multiple mixed motivation, multiple needs. Nailing the perfect guideline driven care for one system often creates unintended consequences for another evidence based intervention.
Take diabetes and depression.
40% of people with diabetes have depression.
The guideline for diabetes care instructs patients to get a glucometer and check, document & report readings regularly.
Ignoring for a minute the lack of evidence that this intervention works for Type II DM, there are some recent studies that tell us the results tend to increase depression in patients with diabetes.
We know that treating depression helps patients with diabetes do a better job of managing their diabetes (linked to more exercise, losing weight and all sorts of good things linked to better outcomes).
Depression is bad for diabetes outcomes but is a consequence for some patients who strictly adhere to the guideline.
Guideline A is in direct conflict with Guideline B.
What's the best course for managing diabetes and depression? Only you and your patient can know the answer.
Shared decision making and patient choice
Really good studies tell us that we can increase the probability of our patients following through with certain things (mammography, colorectal cancer screening, smoking cessation, etc) if we work with them in a certain way. We can help more people achieve better clinical outcomes when we take the time to listen and have learned certain skills around shared decision making, motivational interviewing, self-management support, and when we have the resources to engage in follow-up.
This is some of what got the great outcomes in the IMP project. It is work we can do if we had the time & tools & funding to engage in a broader scope of work with our patients. It is nearly impossible to invest in this work in a hamster wheel practice. This is why we need to change the paradigm of how we're paid - so that we have the resources we need to engage in the full spectrum of our profession.
That being said, there will always be some patients who say 'No' despite our every effort. There should be some opt-out mechanism that the patient can trigger that alerts those who would judge us that we've truly done our best but can control only some of the variables that lead to the desired outcome.
We need a different system of measurement.
Comprehensive primary care is not the sum of disease management bio metrics. We work in the real world of complexity, ambiguity, mixed incentives, mixed patient motivation. I'm not asking to be let off the hook, but I want a new hook - one that is consistent with comprehensive primary care.
Remember that comprehensive primary care is based on four cardinal features:
"I can get care when and how I need it"
"I have a PCP who knows me as a person"
"My PCP takes care of the bulk of my health care needs"
"My PCP coordinates all my care"
We can (and did in the IMP project) successfully use aggregate patient experience data as a better proxy for what we do and who we are as professionals. These data meet the criteria set above:
Large N - all our patients
Person not organ system
Global and not in the weeds
Insurers who have trouble getting their heads past HEDIS and other irrelevant data rivers are urged to consider the following:
If excellence in disease management is the means to the end and we've been doing that for more than a decade, why have we yet to see global improvement in quality & cost?
Patients who receive care in practices that excel on these metrics report improved clinical process (A1c, SBP, DBP, etc) and outcome (ED use, hospitalization) measures.
This approach achieves the stated goal. The current approach (unfortunately embedded in all the "medical home" models) does not and never will achieve the goals.
Time to measure what really matters and stop swimming in a river of irrelevant data.
Gordon
This thing will be much worst in the current health care reform that they are planning!
Posted by: buy soma | September 01, 2009 at 03:43 PM
As we tend to reach the goal of Health Care System Like Canada and Britain, we are healpless regarding the quantity of people wanting this benefits.
Posted by: buy soma | August 23, 2009 at 11:25 PM