Background: Small differences in the language we use can have a large impact on how patients/families perceive decision-making. Consensus guidelines about discussing death recommend clear, direct language and avoidance of euphemisms; nonetheless, studies suggest that clinicians use words like “die” and “death” sparingly in documentation and recorded encounters (both real and simulated). There is much less data about death-related language in pediatric conversations. Most children who die do so in a medical facility, and many who die in ICUs do so after a decision to withhold or withdraw sustaining treatments. This study sought to characterize how death is discussed during family meetings about critically ill infants and explore difference in language between clinicians and families.

Design and Participants: Longitudinal qualitative study at a single academic hospital in the southeast United States, conducted from 2018-2020. Participants were parents of infants <1 year old with a neurologic condition in an ICU with a planned family meeting to discuss neurologic prognosis or withholding/withdrawing LST. Parents were >18 years old and English-speaking. Family meetings were audio recorded, and transcripts were screened for discussions about death; those that included discussions about death were qualitatively analyzed with focus on discussions about death and language used (direct references to death, euphemisms, etc.)

Results: 68 family meetings (FM) including 36 parents of 24 infants were recorded and screened; 33 of these involving 20 parents of 13 infants included discussions about death and were analyzed in detail. 65% of parents were mothers, and 60% of parents were Black, 40% White, 10% Asian. The median parent age was 28.5, most had completed high school or some college, and most reported low household income. The median infant LOS was 86 days, with median gestational age 37 weeks. Most (92%) received mechanical ventilation, 46% received chest compressions, 38% had DNR order, 15% died during admission. 
Throughout the 33 FMs, death was referenced 406 times, including 275 times by clinicians and 131 times by families. Direct language (“death”, “die”, “stillborn”) was used in 45% of meetings and accounted for only 8% of all references to death, including 5% of clinician references and 15% of family references.
Four types of euphemisms were identified, including survival framing (e.g. “not live”, “don’t survive”), colloquialisms (e.g. “pass away”, “not make it”), medical jargon (e.g. “code”, “stop breathing”), and pronouns without an antecedent or a vague noun (e.g. “when it happens”, “something might happen to her”). Families most used colloquialisms (34% of references) while clinicians most commonly used medical jargon (43% of references).
Other ways of referencing death were also described, including decision-focused language, hoping/planning for a good outcome, generalization, and burying discussion in a monologue – all these behaviors were more common in clinicians than families.

Commentary: Clinicians referenced death about twice as often as families, but used direct language about death less often than families did. Among the various types of euphemisms, families commonly used mutually understood colloquialisms while clinicians commonly used medical jargon. One major limitation of this study is that it did not evaluate relative clarity of euphemisms about death. For example, phrases like “pass away” or “not survive” are probably commonly understood to mean “die”, whereas “to code” or “to have an irrecoverable heart rate drop” are not likely widely understood. However, there is a lot of other evidence to suggest that medical jargon is problematic for clear communication and clinicians overestimate patient/family ability to understand it. This study also did not examine family preferences about language. There is evidence that families want clear communication, but it’s uncertain whether this necessitates using the word “death”/” die” itself. This study also was not able to discern the effect of language on decision-making or therapeutic alliance or listener understanding.  

Bottom Line: Discussion about death during family meetings about critically ill infants more commonly include euphemistic references to death rather than the word “die/death/dying” itself. In lieu of the word itself, families primarily use colloquialisms, whereas clinicians primarily use medical jargon. While this is a single-center pediatric study, language choice in discussions about death across all populations and settings is important and deserves further study aimed at better understanding patient/family preferences and the impact of language on decision-making.

Source: Barlet MH, Barks MC, Ubel PA, et al. Characterizing the language used to discuss death in family meetings for critically ill infants. JAMA Netw Open. 2022;5(10):e2233722. doi:10.1001/jamanetworkopen.2022.33722