A Little More on those Three Letters -- PSA

Sensible Medicine readers and writers seem unable to get enough of thinking about the PSA. I think this will be the last for a while. I appreciated this essay by Dr. Trump, in which he considers whether some of the ideas expressed in a previous post could be extrapolated.

Adam Cifu

I have followed the ongoing discussion on Sensible Medicine regarding PSA testing with interest, as it is something I struggle with in my own clinical practice as an internist. I encourage everyone to read Owen Stratton’s recent post. I enjoyed this post, but thought it might have limitations when applied more broadly to clinical practice. Time constraints will always lead to epistemic discordance between patient and clinician, regardless of how information is communicated. We have to rethink how we practically define epistemic justice as we strive to achieve it in clinical decision-making – regarding PSA testing and beyond. All quotes below are from Stratton’s piece.

In an era when primary care clinicians are limited to briefer appointments for sicker patients while handling ballooning administrative tasks, time constraints are real. Every decision to take time discussing a certain topic has an opportunity cost. We absolutely should present evidence behind recommendations in a more patient-centered way, largely shifting away from qualitative to more tangible, quantitative statements. But as much as we’d like to distill the evidence for and against something like PSA testing into a quick (hopefully) digestible soundbite, it often takes more time than that to reach a point where the patient feels fully empowered with all relevant information to make an informed decision.

It is one thing to say we should still take a few minutes to go through this process when discussing PSAs. But who’s to say we shouldn’t extrapolate this framework to more facets of our clinical decision-making? Don’t we want patients to make fully informed decisions about all aspects of their care in the name of Platonic justice? After all, “when reason is withheld, medicine moves from the pursuit of justice to the practice of persuasion.” Obviously, we can’t spend all the time it would take to explain the evidence and quantitative rationale behind every test we do or do not order. We are constantly “persuading” patients in the questions we ask or decide not to ask, and in the way that we frame our rationale when placing each order.

When I evaluate a patient for lower-extremity edema and ask whether they have had any recent periods of prolonged immobilization, I don’t always explain why I am asking (the reason being that it helps shape my pretest probability of a deep vein thrombosis, or DVT). Although I do try my best to talk through my differential diagnosis with my patients when discussing our plan, if my suspicion for DVT is sufficiently low enough that I do not feel the need to order a d-dimer blood test or an ultrasound, I will not necessarily give a full explanation of the quantitative data behind these tests (or maybe won’t even bring these tests up as options) for the sake of time. Is this epistemic injustice?

It would be impossible to spend extended time explaining every decision made. If we acknowledge that this would be untenable and that there will have to be situations in which the patient functions as a “passive recipient of medical authority,” one could argue that we limit the comprehensive application of this epistemic justice framework only to situations with enough nuance or complexity that “shared decision-making” is required.

Ultimately, however, someone – likely the clinician – would still need to decide which situations meet that requirement. Where would this line be drawn to determine whether a situation qualifies for shared decision-making? There are some obvious cases, like deciding on overall treatment goals for newly diagnosed metastatic cancer. But beyond that, do we base this on whether guidelines explicitly encourage us to engage in shared decision-making for things like PSAs? Guidelines are imperfect and do not comprehensively capture every scenario in which a “shared” decision might be warranted. Shouldn’t this framework also be employed any time that there is not necessarily an objectively correct answer to a question – e.g., whether or not to start anticoagulation in an elderly patient with atrial fibrillation at high risk of falls? Undoubtedly, these are all situations in which we should strive for epistemic justice. Still, if one argues that “the clinician’s task is not to impose a decision, but to support reasoning and understanding,” the list of topics requiring extensive counseling to fully achieve this goal begins to grow unachievably long within the confines of a fifteen-minute clinic visit.

Given this constraint, we must acknowledge that, even in the information age, epistemic asymmetry in the patient-clinician relationship will almost always exist and cannot be rectified in the moment for every clinical decision. There will be situations when we simply have to employ a certain degree of “paternalism,” though that is a terrible term – I prefer Dr. Cifu’s idea of “parentalism.” We won’t be able to provide a thorough quantitative explanation at every juncture of care. Such an omission of decision-making rationale can still fall within the confines of Platonic justice, however, if the clinician is truly practicing with as much empathy, respect, and knowledge of the patient’s health priorities as possible, along with humility and transparency when facing uncertainty. (This is admittedly a high bar and one that we too often do not reach in these moments, but developing strong patient relationships over long periods of time helps to build the trust that is required to meet this standard.)

Broadening our conception of Platonic justice can also allow us to include a patient’s emotions and anecdotal experience, even if these factors are not grounded in empirical “facts”. Ultimately, I agree with Stratton that we should always strive for clear communication about our decisions, but this may not always be practical. Carefully modifying this framework to account for appropriate and judicious use of parentalism allows us to practice in a way that maintains epistemic justice in clinical care.

Samuel Trump (no, he’s not related) is a general internist at the University of Wisconsin-Madison and serves as medical director of UW’s MEDiC Southside student-run free clinic.