Decades of behavioral economics research and management science predict that framing quality of medical care as something that can be captured effectively by a simple metric, and then incentivized by financial rewards or punishment based on that metric, would misfire. The experience of a colleague suggests that this is not a theoretical concern. Here’s what she told me.
To meet the coming 2014 Meaningful Use quality goal of Future Fall Risk (NQF 0101) her well-meaning and high-quality institution proceeded as follows:
- They identified a screening tool for fall risk: Have you fallen twice in the last year, or once with an injury?
- They identified a threshold for a (+) screen result: Two falls or 1 fall with injury in the last 12 months.
- They identified a target group: all patients over 65. (Actually, the government defined the group.)
- They designed a screening process: the nurse asks all patients over 65 the two questions while preparing them for their office visit, and if the answer to either question is yes, “High risk of falls” is added to the patient’s problem list.
- FULL. STOP.
Can you, gentle reader, see any problems? Here is the list of concerns raised by my friend and her colleagues when they discovered ‘high risk of falls’ had been added to problem lists on their patients (without their knowledge):
- Positive screens were not followed by any assessment of the patient. (Two falls on black diamond trails at Sugarloaf would not confer a high risk for falls or the need for treatment. Two falls in the living room would. An appropriate assessment would include things like a timed up and go (TUG) task or other formal falls assessments, assessment of medical history, medications, living environment (stairs, alone), strength, posture, balance, gait and mental state, alcohol or drug use,
- The result of the screen was treated as a diagnosis. A positive screen does not mean the patient is at risk, it means the patient needs to be assessed for potential risk.
- The ‘diagnosis’ was added to the problem list by a staff member not credentialed to make diagnoses.
- The data was collected for audit and compensation purposes, but the provider not made aware that there was a potential high risk condition. Really? Does anyone, anywhere think it is reasonable to add a diagnosis of high risk for bad outcome to a patient problem list and not notify the responsible provider who is going to sign the note?
- Worse than not notifying the clinician was the fact that the patient was not made aware that there was a potential for a high risk condition. Does anyone, anywhere think it is reasonable to add a diagnosis of high risk for bad outcome to a patient problem list and not notify the patient who is the subject of the note and potentially at risk for the bad outcome?
- There was no thought about arranging a treatment plan. This is not surprising, as it is impossible to do so without first making an assessment of the problem, but it begs the question, how meaningful can a screening process be, if it does not include any attempt to improve outcomes for patients (as opposed to audit outcomes for the institution)?
- There was no system to track results and use them for PDSA and ongoing improvement. How would you know that you are doing as good a job as possible, if you do not monitor the results of your initiatives and self-assess?
The accepted requirements for ethical screening are:
- There must be a significant burden of disease. (One would not screen for blond hair or clinically insignificant skin lesions like freckles.)
- The screening instrument must be safe and easy with adequate sensitivity (finds those at risk) and specificity (avoids excess false positives).
- One must be able to target an appropriate population. (One would not screen a population that includes males for pregnancy.)
- There must be an available and affordable process to follow the screening with more formal assessment and diagnosis.
- There must be an available treatment that is safe and effective (prevents bad outcomes more often than it causes bad outcomes).
- Screening must not violate the ethical or cultural norms of the population screened. (One would not administer pork to Jewish or Muslim populations as part of a screening program.)
- Screening must be transparent and voluntary: the screened individuals must know they are being screened and have the option to decline.
I do not see this as a incorrect behavior or a personal failing by the individuals involved. On the contrary, I see this as the expected behavior of skilled and dedicated people functioning appropriately within a system that has lost its way. By report from my friend, the people involved are sincere, dedicated, bright, and hard working. They see themselves as hard at work on projects to improve quality. If they have a failing, it is neither ethical nor motivational. It is failure to recognize they are trying to improve something they do not do and therefore do not understand, compounded by failure to involve the actual people (patients and their clinicians) impacted in planning and implementing their initiatives.