D25-13 Characterizing Patterns of Interaction Between Clinicians and Family Members in the Intensive Care Unit

Abstract Rationale Regular communication between clinicians, patients, and patient families in the ICU is essential; it preserves trust, supports the wellbeing of patients and their families, and improves understanding between clinicians and families. Prior interventional studies aimed at improving the content and quality of communication show reductions in family members’ anxiety and depression symptoms. Comparatively less is known about the typical patterns of these interactions, including frequency, which may be an important contributor to patient and family outcomes. The objective of this analysis was to characterize current patterns of clinician-family interaction in medical ICUs. Methods This study is a secondary analysis of a multicenter ethnographic study of time-limited trials in patients with acute respiratory failure receiving invasive mechanical ventilation. In this analysis, we analyzed electronic health record (EHR) documentation to identify in-person and remote (e.g., telephone) clinician-family interactions. Two coders identified interactions using a deductive codebook. Clinical and demographic data were abstracted from the EHR. Descriptive statistics were used to compare patients in higher- and lower-interaction groups, separated by median interactions per day. Results In this secondary analysis, we analyzed 30 parent study patient cases and identified 126 total clinician-family interactions. The mean number of interactions per case per day was 0.53 (±0.51), ranging from 0 to 2.5 (Figure). The higher-interaction group (0.39 interactions per day) had a higher proportion of male patients (73%) compared to 47% in the lower-interaction cases (p = 0.04). The mean age was similar between higher-interaction and lower-interaction groups (64 95% CI: 56—72 and 65 95% CI: 61—69 years, respectively; p = 0.94). We also found no differences between the higher- and lower-interaction groups when comparing patient racial distribution, Area Deprivation Index (a measure of socioeconomic disadvantage), and Charlson comorbidity index. Conclusions In this multicenter observational study of EHR documentation, we found that clinician-family interactions occur on average less than once per day for critically ill patients receiving invasive mechanical ventilation in an ICU. While the optimal frequency of interactions remains undefined, our findings suggest substantial patient-level variation and highlight that some families have sparse interaction. The only patient characteristic significantly different between higher- and lower-interaction groups was patient sex; further work is needed to elucidate the reason behind this difference and define what constitutes ideal clinician-family interaction frequency. The analysis presented here can contribute to novel and targeted interventions on clinician-family communication in ICUs. This abstract is funded by: NIH/NHLBI grant R01HL168474