The first post discussed the problem of medication overuse and the promises of medication discontinuation. It is a brave act to discontinue a medication in the hospital, and to help patients to leave the hospital with fewer medications than when they entered. It is brave because it runs against what is often requested of us by worried patients and their legitimately concerned families, and by the medical industrial complex as we understand it. And it is brave because it is often the right thing to do for those patients.
In this post, the focus will shift towards avoiding diagnostic testing without a clear indication. The competing demands of the medical industrial complex push physicians, physician assistants, and nurse practitioners towards ordering diagnostic testing. When you add the remaining ingredients of guideline-based decision making by well-intentioned medical societies, and a caustic legal climate against medical professionals, the baked product that comes out of that oven insists on a test with every bite. The easiest way forward is to identify whether the outcome of the test will truly change the trajectory of the patient’s life in a meaningfully positive way.
There are many examples to demonstrate this process, but a reasonable example is stress testing. In patients who present to the hospital with chest pain, especially in patients over 65, it is almost as sure as the rising sun that this patient will be admitted to a hospital and will undergo stress testing. Troponins will be drawn in addition to of course the requisite EKG. The stress test will be treated and interpreted and understood by patients as Truth itself. If the patient is unfortunate enough to have evidence of active coronary artery disease, this patient will be taken for cardiac catheterization, and in an even smaller group of patients, stents will be placed. This process has its merits for patients with intervenable targets. But what of the older, more fragile patients? Perhaps in their 80s, with a slightly elevated troponin to begin with? A touch of renal disease, a dash of COPD, and all of a sudden, the risk-benefit ratio shifts. In these patients, now testing can be problematic, because checking a troponin level can potentially lead to a cascade of events that may actually offer no benefit or even might harm the patient. Even absolute recommendations such as aspirin initiation may not be appropriate if the patient won’t live long enough to appreciate its benefit but will certainly live long enough to appreciate its harm. In certain populations, the testing inevitably leads to an undesirable outcome for the patient, the family, as well as the clinician caring for the patient.
The world of oncology has offered reasonable solutions to the problem of a small “time horizon”. The time horizon refers to the time patients have left before they will most likely die. While we may not have perfect knowledge, often we are able to predict, to some degree, a patient’s time horizon. For patients with time horizons of less than ten years, there is an armamentarium of non-intervention from which they benefit. Avoiding colonoscopies, avoiding mammograms are two of the oncologic outcomes we can now firmly offer. Liberalizing A1c is another well-known means of harm reduction in this group.
There are times when an MRI will offer no meaningful benefit to a patient with a suspected stroke if they will not tolerate a change in anticoagulation. A recent study in the New England Journal of Medicine, which looked at dual-antiplatelet therapy in stroke and TIA patients, excluded patients with severe renal or hepatic comorbidities, which exist in many hospitalized patients. There are times when a carotid ultrasound may not be appropriate for patients with a known stroke because they may not tolerate anesthesia or carotid endarterectomy. It is by no means exhaustive to identify these examples, but they require clinicians to spend more time considering and explaining to patients and their families, that sometimes, less is truly more
