Medical Decision-Making and Bereavement Experiences After Cardiac Arrest: Qualitative Insights From Surrogates

Abstract

Background: Surrogates of incapacitated patients in the intensive care unit (ICU) face decisions related to life-sustaining treatments. Decisional conflict is understudied.

Objectives: To compare experiences of ICU surrogates by reported level of decisional conflict related to treatment decisions after a patient's cardiac arrest preceding death.

Methods: Convergent mixed methods were used. Bereaved surrogates recruited from a single northeastern US academic medical center completed surveys including the low-literacy Decisional Conflict Scale (moderate-to-high cut point >25) and individual interviews about 1 month after the patient's death. Interview data were analyzed by directed and conventional content analysis. Surrogates were stratified by median total survey score, and interview findings were compared by decisional conflict level.

Results: Of 16 surrogates, 7 reported some decisional conflict (median survey score, 0; range, 0-25). About two-thirds decided to withdraw treatments. Three themes emerged from interviews: 2 reflecting decision-making experiences ("the ultimate act"; "the legacy of clinician communication") and 1 reflecting bereavement experiences ("I wish there was a handbook"). Surrogates reporting decisional conflict included those who first pursued but later withdrew treatments after a patient's in-hospital cardiac arrest. Surrogates with decisional conflict described suboptimal support, poor medical understanding, and lack of clarity about patients' treatment preferences.

Conclusions: These findings provide insight into bereaved ICU surrogates' experiences. The low overall survey scores may reflect retrospective measurement. Surrogates who pursued treatment were underrepresented. Novel approaches to support bereaved surrogates are warranted.

Figure 1

Mixed methods analytic approach. Interview data were first analyzed via directed, then conventional, content analysis. Next, descriptive statistics were calculated for decisional conflict survey scores. Interview data were then compared by level of decisional conflict (ie, above vs at or below median survey score).

Figure 2

Decisional Conflict Scale–Low Literacy (DCS-LL) scores. Dashed line indicates the generally accepted cut point of >25 for moderate to high decisional conflict.

Figure 3

Sample concordance/discordance analysis. Codes were reviewed and compared across decisional conflict groups. Codes reflective of similar ideas, thoughts, or messages were considered concordant; codes reflecting contrasting ideas, thoughts, or messages were considered discordant.