Treatments Fail Patients, Patients Don’t Fail Treatments

When reviewing the chart for a patient with a complex diagnosis or constellation of diagnoses, it is not uncommon to see, in the section labeled “Past Medical History”, a remark that the patient has “failed” a trial of other interventions. A patient with pain failed a trial of NSAIDs. A patient with Crohn disease failed infliximab and is now on vedolizumab. A patient with cancer failed first-line treatment and now needs a bone marrow transplant. The intent of these statements is certainly not to place the blame on the patient, but that is the inevitable implication of ordering the phrasing in this way. Consider how differently the above sentences sit with an inversion of the word order: NSAIDs failed a patient with pain. Infliximab failed a patient with Crohn disease, who is now on vedolizumab. First-line treatment failed a patient with cancer, who now needs a bone marrow transplant. Patients are not taking a test that they must pass in order to heal – yet the construction of our medical lingo makes this the subtle implication. When a given linguistic construction is highly prevalent, it gives me pause to consider what purpose it serves, subconscious or not, to the writer. If patients fail treatments, rather than treatments failing patients, does that, in some subliminal way, help the doctor feel less complicit in the treatment failure? The prevalent structure implies that the treatment itself is blameless, while the patient is culpable – when in reality, treatments are useless in the absence of patient healing.

While word choice such as this in medical records may seem like a relatively minor issue in the broader context of clinical medicine, with all its complexities, emerging evidence suggests quite the contrary. Indeed, a 2018 study by Anna Goddu and colleagues examined the impact of stigmatizing language used in a clinical vignette of a patient with sickle cell anemia presenting with a vaso-occlusive pain episode and found that trainees who read a version of the vignette that used stigmatizing language were less likely to indicate they would provide opioid pain medication to this hypothetical patient than were those who read a version of the vignette that eliminated such stigmatizing language for neutral, medically accurate language. This is concerning not merely because undertreating pain is inhumane, but particularly because in the case of sickle cell disease, an acute vaso-occlusive pain episode needs treatment with opioid pain medications post-haste, as the vasodilatory effects of opioids like morphine help reverse the underlying pathophysiology of a vaso-occlusive episode and thereby treat the disease directly, not merely its symptoms. All it took was changing a few phrases, such as “sickle cell patient” to  “a patient with sickle cell disease”, “narcotic medications” to “home medication regimen”, “hanging out with friends outside McDonald’s” to “spent yesterday with friends”, and “has been cussing at nurse” to “seems distressed” to change the degree to which medical students and residents were willing to treat this patient’s disease. That language alone could create such a disparity in our treatment of sickle cell disease in particular is even more concerning considering the racism already inherent in our systemic undertreatment and under-research of sickle cell disease. Our words matter, and we must wield them wisely.  

The importance of phrasing – that is, the implication of one phrase over another – has become all the more relevant in light of the stipulations of the 21^st Century Cures Act, which requires hospitals to make most medical records available to patients immediately once they are signed (with exceptions in the case of certain sensitive records pertaining to mental health, suspected abuse or domestic violence,  adolescents’ desire for confidentiality from their parents, active legal proceedings, and so on). This Act has been met with mixed reactions by the medical community, with some praising this move towards transparency and patient advocacy, with others bemoaning a perceived implied need to write notes in patient-centered language, or the increased administrative burden of responding to patient questions regarding the content of clinical notes. Personally, I think this is a move in absolutely the right direction – it has always struck me as strange that previously if patients wanted access to their records, they had to go through an arcane and bureaucratic maneuvering process through the hospital’s medical records department, rather than me just printing out a copy of their admission history/physical and discharge summary, which we send to other medical providers without a second thought.

It does bear reflecting if and how I will change my documentation in this open access era. I do not plan to stop using medically appropriate terminology to describe my thought process, nor will I cease from including less savory considerations that were on my differential diagnosis (e.g., cancer) when documenting – in the end, medical records are as much records for ourselves and other providers who make take care of the patient than they are patient-facing documents, and accurate documentation of my thought process when writing a medical assessment remains essential. It is possible that this will result in increased questions from patients and parents who would like clarification of what I’ve documented, but after all, that’s my job – doctor comes from the Latin docco, to teach – after all. Indeed, I find that sometimes it is not until I take a moment to write my thoughts down in the medical record that I fully consider my assessment, and it is possible that I would have done a better job explaining my thought process to the patient at hand if I’d had a chance to foresee the future and write my note in advance. It seems not unreasonable that then a patient who reads my more fully processed reasoning could have a follow-up question, which could lead to a more thorough and nuanced discussion than what we’d been able to have in the context of our first encounter – and if part of that discussion means that I have to define “histiocytosis” or “SIADH”, then so be it!

What I will change – or, speaking more optimistically, rededicate myself to – is intentionality my phrasing. To reiterate my earlier example, I will avoid saying “patient tried and failed NSAIDS” – instead, NSAIDs did not relieve the pain. Antibiotics were unsuccessful in treating the infection. Steroids failed to induce remission. The patient is not to blame here. Moving beyond this, I will try more actively to adopt the patient perspective when structuring sentences in general – of course rededicating myself to patient-centered language (patient with autism spectrum disorder not autistic, patient with Type I diabetes not a Type I diabetic, etc) but even going farther to consider why a patient or family chose not to pursue a given intervention. There’s a part of our electronic medical record system that always irks me – if a patient or family declines a medication, it shows up in our medication administration record as “patient/family refused”. Refusal is potentially rather judgmental – while there are exceptions of course (due to lack of understanding or denial), patients and families tend not to “refuse” interventions that they’re finding beneficial. Perhaps it would be more empathic of me to phrase things as “after an informed consent discussion of the risks and benefits, the patient decided to pursue non-operative treatment” rather than “the patient refused surgery” – because of course there are any number of reasons why a patient might not want surgery even if recommended, including fear of complications from the surgery itself or from anesthesia, fear of debilitating medical bills, incomplete understanding of the need for the procedure, belief that non-operative management is equally or more effective, and so on. In addition, I will work to favor phrases such as “did not adhere to the originally prescribed plan” or simply “did not take all medicine as prescribed” rather than “did not comply with treatment regimen” – doctors offer suggestions and treatment options, rarely (generally in the case of a critically ill child) mandates of what must happen. We should all work on baking our respect of patient autonomy into the narrative summaries of our patient encounters.

To those in healthcare – how will you change your documentation given the Cares act? To all of us, who are patients at some point or another – how do you feel about this change in record accessibility? Will you read your own clinical notes?

February 17, 2021