“He took a turn for the worse, but the good news is that he is stable now,” the doctor explained to the patient’s son.

Three weeks earlier, our patient had been admitted to the intensive care unit (ICU) after suffering a myocardial infarction. As his condition worsened, his hospital room gradually accumulated life-support machines to support him—a BIPAP machine, followed by a ventilator, and then a dialysis circuit. When his heart abruptly stopped a few days later, he was placed on extracorporeal membrane oxygenation (ECMO). Standing amid a jungle of machinery, the patient’s son looked exasperated. The doctors were cautiously optimistic about his recovery on ECMO.

“Stable?” his son asked with confusion, his eyes darting between the plethora of machines, flashing monitors, and hoses circulating blood through the air.

How many life-support machines can a patient have and still be described as stable? In the fourth-century BCE, Greek philosopher Eubulides of Miletus considered an equally nebulous question in the context of a heap of sand. Eubulides imagined a heap of one million grains of sand, from which 1 grain at a time was removed.1 At what point was the sand no longer a heap? After 1 grain was removed? After a thousand? The question, which has come to be known as the Sorites Paradox or the Paradox of the Heap, is premised on the concept of vagueness, the idea that some descriptions lack boundaries and are subject to interpretation. In the case of the sand, there is no clear boundary between heap and no heap. In modern critical care, we posit upon this philosophical challenge as we choose our words and craft our vernacular. At what point does a patient transition from stable to unstable? From unstable to critical? And most importantly, how can we reframe vague terminology as objective information so our patients and their families can meaningfully participate in their own care?

In a modern context, philosopher Rory Collins considers in his essay, “On the Borders of Vagueness and the Vagueness of Borders,” how people’s understanding of language evolves as technology changes.2 Considering the smartphone as an example, years ago, the word may have conjured an image of a basic device with email. Now when we say “smartphone,” people understand we are describing a device with far more advanced features, such as GPS or Bluetooth. In the setting of advanced life support, while surgeons and critical care physicians’ understanding of commonly used words such as “stable” or “improving” have evolved alongside the technology, these words may be variably interpreted by patients and their families who are unfamiliar with modern critical care practice. “Stable,” a word traditionally used in the setting of a patient in the recovery room after an uneventful surgery, for example, may now describe a patient on level doses of vasopressors with septic shock or a patient who has just had a ventricular assist device placed after receiving cardiopulmonary resuscitation only hours earlier. In each context, “stable” connotes something different.

Such vagueness of language can impact our patients and their families’ understanding of their condition and ability to maintain autonomy, as was the case with our patients in the ICU. Although from the surgeon’s perspective, he was physiologically stable on ECMO, his son’s exasperation with this description was justified, as he struggled to contextualize this traditionally positive word with the unmistakably grave accumulation of life-support machines in his father’s room. A surgeon’s understanding of the word “stable,” based on experiences garnered over a career of treating critical illness, may vastly differ from a layperson’s more traditional understanding of the word. Such differences in perspective impact not only the language we choose but help explain why patients and health care providers may variably interpret a situation. In the case of our patient, the disconnect contributed to a growing sense of frustration that he did not understand his father’s condition and was unable to meaningfully act as his surrogate.

The value of words such as “stable” or “recovering” can be improved when we intentionally decrease their ambiguity. This is achievable through a 2-tiered approach that emphasizes a heightened awareness of the vagueness of commonly used words in the medical setting, combined with a strategy of consensus seeking. Identifying terminology that is prone to variable interpretation encourages the medical team to contextualize and provide clarification, before misunderstandings may develop. We know from the structured handoff literature, and specifically from handoff practices in high-risk areas such as space shuttle operations, that failures in communication may result in an “incorrect or incomplete model of a system’s state, failing to anticipate future events, and unwarranted shift in goals, decisions, priorities, or plans.”3 If we consider the transfer of information between provider and patient as analogous to a handoff between 2 providers and a patient’s accurate understanding of their condition as a measure of quality of care, we can appreciate the importance of eliminating ambiguity and preemptively clarifying information to improve patient satisfaction and guide shared decision-making. What does it mean to tell our patient’s son that his father is stable on ECMO? Does this mean that he is no longer at risk of dying, or simply not imminently? Does it mean that he is likely to improve? We could have mitigated his son’s frustration and empowered him as his father’s surrogate by more objectively explaining, “your father is stable, in terms of having reached a stable level of flow on ECMO with a survivable blood pressure, but overall, he remains dependent on multiple life support machines.” In the pediatric critical care population, this ability to responsibly manage parents’ hope regarding their child’s prognosis has been identified as a critical element in communication between the patient (parent) and the provider.4 While we may be limited in our prognostic abilities for patients receiving advanced mechanical life support, the importance of clear, effective communication in this area cannot be overstated.

Consensus seeking, which is one resolution to the Sorites Paradox, proposes that the definition of a vague state, or condition, be determined by stakeholders. In the context of the original paradox, the number of grains of sand that define a heap is determined by stakeholder consensus, and this definition may vary between different groups. In medicine, consensus seeking is already used to set goals of care, allowing patients and their medical providers to mutually define what is an acceptable clinical outcome, and when treatment has failed.5,6 In the context of improving language ambiguity, consensus-seeking is an effective means to ensure we share an understanding of critical terminology, not only with our patients but also among different specialty providers within the team. While it may seem counterintuitive to ask a patient or their family “What does stable, or improving, mean to you?”, questions such as these often yield important, difficult-to-ascertain information about the less tangible aspects of a patient’s condition, such as mindset, motivation, or sense of hopelessness or despair. In this way, consensus seeking allows a more complete, objective understanding of a patient’s condition among stakeholders, improving patient autonomy and allowing us to progress together toward a patient’s goals.

Consensus seeking may be particularly effective when used in conjunction with a strategy such as Best Case/Worst Case ICU, a graphic tool that allows patients or their surrogates to visually track their progress toward a best or worst case outcome through their ICU stay.7 Consensus-seeking allows us to improve our communal definition of what defines a best or worst case, ensuring that decisions are made, and care targeted, toward an individual patient’s goals of care.

Our success in treating critical illness in a surgical setting has been fueled by scientific ingenuity and innovation, but our vernacular and word consciousness as we communicate with our patients have not developed in conjunction. We have become increasingly comfortable filling our patient’s rooms with life-support machines but remain challenged to use the right words in a context that will provide objective, useful information to our patients. After several weeks on ECMO, our patient’s son chose to pursue palliative measures for his father despite our descriptions that he was “making progress” and “improving.” When his nephrologist predicted that he would have a life-long dependence on dialysis, his son knew it was time to stop. His father had never wanted to be dependent on life-support machines such as dialysis or the ventilator, and it became clear to him that the medical team did not share his understanding of his father’s condition and progress.

As we guide our patients through the complexity of modern surgery and critical care, moving in concert with them is commensurately more challenging. Socio-biologist E. O. Wilson once noted, “We have Paleolithic emotions, medieval institutions, and godlike technology.” As we carefully select our words to describe the technology that fills our rooms, when do we pause to ask again, at what point is sand no longer a heap?

1. Oms S, Zardini E,. An Introduction to the Sorites Paradox. In: The Sorites Paradox. Classic Philosophical Arguments. Cambridge University Press; 2019:3-18. 2. Collins R. On the borders of vagueness and the vagueness of borders. Vassar College Journal of Philosophy. 2018;5:30–44. 3. Patterson ES, Roth EM, Woods DD, et al. Handoff strategies in settings with high consequences for failure: lessons for health care operations. Int J Qual Health Care. 2004;16:125–132. 4. Gordon C, Barton E, Meert KL, et al. Eunice Kennedy Shriver National Institute of Child Health and Human Development Collaborative Pediatric Critical Care Research Network. Accounting for medical communication: parents’ perceptions of communicative roles and responsibilities in the pediatric intensive care unit. Commun Med. 2009;6:177–188. 5. Hendricks Sloan D, BrintzenhofeSzoc K, Mistretta E, et al. What does the word healing mean to you? Perceptions of patients with life-limiting illness. Palliat Support Care. 2022:1 1–5. 6. Taylor LJ, Nabozny MJ, Steffens NM, et al. A framework to improve surgeon communication in high-stakes surgical decisions: best case/worst case. JAMA Surg. 2017;152:531–538. 7. The Patient Preferences Project. Wisconsin Surgical Outcomes Research Project. Accessed January 10, 2024. https://patientpreferences.org/bcwc-icu/