More Encouraging News
I have posted many articles discussing the problems encountered with guideline-centered versus patient-centered care as well as the distortions of clinical judgment associated with mindless incorporation of such guidelines into care pathways. So, it is no surprise that a recent opinion piece in the Annals of Internal Medicine caught my eye. The authors, all practiced in guideline development, posit the issue this way.
“Clinical guidelines seek to standardize care to increase adoption of proven, superior interventions and reduce ineffective, unsafe, or wasteful practices…Studies of guidelines unfortunately have reported a high proportion of recommendations—including strong recommendations—based on evidence that is low quality and hypothesis-generating rather than high quality and hypothesis confirming…”
By itself, this is not a terrible issue, as a lot of medical practice is done by consensus—what seems to be the best general approach to a common problem given the current state of the art. The problem, though, is articulated well by the authors.
“Once guidelines are published, they unfortunately tend to be treated ‘as if stone tablets descended from the heavens.’ They set medicolegal standards of care that drive lawsuits, government and private payer standards, subspecialty billing practices, and hospital order sets and regulatory compliance standards.”
The authors present a table of a dozen guideline recommendations that have resulted in harm or waste, and put special emphasis on those dealing with timely antibiotic administration and the Surviving Sepsis campaign, both of which were based on small, local, retrospective studies that should, in the authors’ opinion, have been considered hypothesis-generating. None have actually worked.
The timely antibiotic standard is an interesting one for me, because some years ago, I got involved in such an effort as an early hospital-based quality improvement project. The hospital’s CMO had a previous bad experience where a patient had been admitted to the hospital with pneumonia, the doctor had ordered antibiotics, and the patient had died some six hours later without ever receiving the first dose. Even though this is an example of an anchoring heuristic, we decided to find out how long it took to get antibiotics going. Now this was in the days of paper orders entered into the system by a clerk, but we found the average gap between order and administration was about six hours. Note we were timing from the decision to give antibiotics and were not trying to influence the decision to order. In other words, we wanted to work on internal processes.
We decided one of the problems was an antibiotic order was treated as routine, so it went into the queue along with every other medication order at the pharmacy, and even when filled, went back to the floor where the nurse administered it when she/he got the chance. He addressed those issues within the bureaucracy and the time delay got to where we were hitting the four hour mark or less with regularity.
When the community-acquired pneumonia standard was issued, the major difference was the clock started when the patient presented to the emergency department, so the four hours included all the usual processes involved in trying to make a diagnosis, triage, and patient flow management. And, of course, there were financial penalties for not making the four hour mark. So, what happened? Doctors started ordering antibiotics “just-in-case.” And, since there was often a change in physician coverage between the ordering and the return of culture results, there was delay in stopping antibiotics when confirmatory tests showed they were not needed. There was a strong temptation to defer a decision, so a “just-in-case” order turned into 10 days of therapy with associated expense, risk, and increased antibiotic resistance. The same logic was true in the Surviving Sepsis experience as well, with the additional example that administering antibiotics was done whenever the patient’s presenting lactate level was elevated. In effect, sepsis was being diagnosed by administrative fiat using a test not specific for the problem.
The authors note:
“It is critical that guidelines possess the humility of uncertainty in the absence of hypothesis-confirming evidence. Yet, some form of advice is needed for unsettled questions to inform clinicians providing care in fields where they are not expert. How can we address this desire to standardize care and provide thoughtful expert opinion on unsettled questions, balanced against the need to humbly admit uncertainty where it exists and avoid mandating incorrect care?”
The authors make five proposals. First, guideline committees should include a broad mix of both generalists and specialists to reduce specialty bias and to ensure the guidelines include both in-patient and out-patient considerations. Second, recommendations should be made only where confirmatory evidence is strong. Third, recommendations should not be made where such confirmatory evidence is lacking. Fourth, in commentary, including expert opinion, options and differences of opinion should be highlighted and should not be labeled as a recommendation. Lastly, guidelines should be living documents updated in real time.
As I have suggested previously, the problem of uncertainty is real for both physicians and patients as well as payers looking for “value”. Experience suggests it is a larger issue than either are willing to recognize.
I remember a patient years ago where I was trying to treat an rare problem using aggressive therapy. One day she asked me what I was going to do next. I said: “I don’t know. I’m making this up as I go along, and it depends on how it is working.” After I left, the nurse asked the patient if she believed me. She didn’t, saying she was convinced I knew what I was doing. While I knew what I was doing, I had also told her the truth that I was making it up as I went. Fortunately, it worked and she got over the illness. But I still have no idea if what I did was the “right” thing to do. Such is the reality of clinical practice. New data may move the boundary of uncertainty, but will not eliminate it.
17 January 2022
 Spellberg B, Wright WF, Shaneyfelt T, Centor RM. The Future of Medical Guidelines: Standardizing Clinical Care With the Humility of Uncertainty. Ann Intern Med 2021;174(12):1740-1742. doi:10.7326/M21-3034.
Beyond Evidence-Based Medicine
CQI - The Good, the Bad, and the Ugly
Are the uses and misuses of the continuous quality improvement method emblematic of our times?
More Data on the Value Proposition
Value-Based Purchasing" is a complex program designed to improve hospital quality and outcomes by using financial leverage. A recent study by Ryan and associates suggest it has had minimal effect.
More on Variation - Part 1
Variation is not peculiar to healthcare, but is a general issue with the way the people think, and occurs whenever judgment is needed and the data are fuzzy.
Dealing with uncertainty is at the core of practicing medicine. Have we tried to escape this reality?