I’ve been thinking a lot about standards of care recently.
A “medical standard of care” is actually a legal term, typically defined as the level and type of care that a reasonably competent and skilled health-care professional, with a similar background and in the same medical community, would have provided under the circumstances that led to the alleged malpractice.
Caught your attention with that last word, didn’t I?
In my world, discussions of standards of care arise outside of malpractice considerations (fortunately, and I am knocking wood as I write this), and with some regularity, in two specific instances: defining standard therapies for patients with hematologic conditions or cancer, as opposed to what should be considered research; and identifying standard testing, office visits, and therapies for patients on clinical trials, distinguished from those that are considered “research specific.”
The first instance comes up frequently when we discuss drugs that comprise our formulary at Cleveland Clinic. I chair the Medical Staff Pharmacy and Therapeutics Committee for the entire enterprise, which is the group that decides what makes it onto our formulary and thus what is stocked in our pharmacy. Our purpose is to promote the rational and appropriate use of drugs within our health system, with the goal of improving safety and patient care.
By implication, the drugs that we approve also are getting the nod from our staff physicians as standards for the medical care of our patients. That imprimatur has substantial consequences for pharmaceutical companies: Our organization is large, and we may be the top prescriber of a medication in the world.
Nine separate pharmacy and therapeutic subcommittees (such as Hematology/Oncology, Neurosciences, Cardiovascular, etc.) report to our parent committee with recommendations for drugs to be added to the formulary within their specialties. Our discussions may be more far-ranging than those that the U.S. Food and Drug Administration (FDA) or its Oncologic Drugs Advisory Committee have when recommending regulatory approval of a drug. In contrast to the limitations placed on their considerations, we can discuss issues such as drug cost and we aren’t under the same pressures of political expediency. Thus, we may vote against a drug being added to formulary if we feel its relative balance of safety and efficacy is not favorable, despite its regulatory approval.
We also may place restrictions on drug use to encourage appropriate use, and limit overuse (particularly with expensive drugs) – even if these guidelines don’t align with the FDA label. This gets into standards of care.
What happens when someone at our institution uses a drug off-label? Most of the therapeutic approaches we recommend in hematology and oncology are actually off-label, so restricting use only to populations included in registration studies would cripple our practice of medicine. So, we allow it in our guidelines.
But how do we know if drug use is a bit too far off-label?
For example, the FDA approved enasidenib for the treatment of patients with relapsed or refractory acute myeloid leukemia who harbor an IDH2 molecular abnormality. Is it acceptable to use the drug in patients with myelodysplastic syndromes (MDS) who harbor the same mutation, as one clinical trial did include 17 patients with MDS? This is where we borrow a bit from the lawyers.
Our policy is to allow off-label use if a physician can identify at least one well-designed phase I or phase II study with an acceptable balance of safety and efficacy in a similar patient population, and two colleagues from comparable institutions can confirm that this approach is considered a standard (what we call our “1+2” policy). For molecular-driven therapy, we also have an off valve for recommendations coming from one of our genomics tumor boards.
Thus, for the patient with MDS and an IDH2 mutation, we could consider enasidenib a standard based on the clinical-trial data, and if two colleagues at comparable institutions agreed it was a standard. If a physician can’t meet that bar, we consider use of the drug for the proposed indication to be research and recommend a single-patient investigational new drug application, or an alternative approach.
We also identify standards of care every time we conduct a clinical trial. In clinical research, it is our responsibility to go meticulously through a trial to determine which tests, visits, or treatments are considered a standard part of medical care, and which are not.
Why do we go through this onerous process? Well, to be honest, to avoid Medicare fraud.
The law requires health insurance companies to pay for the “usual care” of medical conditions in those covered by the insurance. That makes sense, as it doesn’t seem fair that a patient with acute leukemia who sees her hematologist every two weeks would have those visits denied by insurance just because the biweekly visit occurs as part of a clinical trial. On the flipside, it wouldn’t be fair to force insurance to pay for EKGs that must be obtained within 10 minutes of each other in triplicate in an asymptomatic patient. We would identify the latter as being 1) purely for research purposes, and 2) a really dumb requirement.
But I digress.
Based on our determination, we bill insurance for some aspects of a patient’s care through a clinical trial and bill a study sponsor for the research aspects of that care.
The tricky part is how to decide what is “standard.” It’s easy enough to say that a patient with higher-risk MDS would be treated initially with the drug azacitidine, whether she has enrolled in a clinical trial or is being treated off-study. It’s much harder to say that we would usually perform bone-marrow biopsies in that patient every … three months? Six months? Two months?
If that patient were younger, we might agree that it should be de rigueur to perform a pregnancy test prior to her initiating chemotherapy. But is it standard to perform a pregnancy test prior to every cycle of therapy? Many clinical trials require it.
To answer questions about treatment standards, we often rely on treatment guidelines, acknowledging that not all of those adhere to rigorous guideline-development guidelines. And no, it is not considered a syntactic standard to use the word “guideline” three times in one sentence.
But where do we turn for guidelines on testing frequency or standard tests? I propose that guideline panels take this on – in addition to making recommendations about therapies – or that members of disease communities establish those standards. Otherwise, we may be leaving that decision to someone working at an insurance company who is having a particularly bad day.
And when we wrongly consider an aspect of a trial as a standard when it may not be, our patients may suffer the consequences by having to pay out of pocket at a time when they can least afford it.