End-of-life care in ICU – what clinicians think

Abstract

Introduction: Globally Intensive Care Unit (ICU) mortality rates vary, ranging from 10-30% with Australian rates at 8%. End-of-life care (EOLC) in ICU is an under researched area despite ICU clinicians being at the forefront of transitioning critically ill patients from receiving life-sustaining treatment to commencing EOLC.

Objectives/Aims: To examine ICU clinicians’ perspective of EOLC in a tertiary referral Australian ICU.

Methods: A convergent parallel mixed-method design was conducted at a 28-bed tertiary adult ICU. ICU clinicians were invited to complete Hansen and colleagues’ 30 item, 5 sub-scale EOLC questionnaire examining knowledge and ability, work environment, support for patient/family, work stress, and staff support. The 4-point Likert scale ranged from 1 = very good to 4 = poor. Demographic data were analysed using descriptive statistics and mean scores were calculated for sub-scales. Focus groups were held with nurses to elicit their perceptions of EOLC. Field notes were made, and inductive thematic analysis was used to identify emerging themes. Ethical approval was obtained.

Results: 143 questionnaires were returned (response rate 64%). The mean overall score was 2.7 (SD 0.9), with scores > 2 indicating a need for improvement. Exploratory analysis identified that nurses rated all domains of EOLC as having potential for improvement, with support for staff 2.6 (SD 0.48), and work stress work stress 2.9 (SD 0.54) emerging as the prominent area requiring improvement. Focus groups identified themes of workload, ICU ward acuity, clinician inexperience, lack of education, role ambiguity, person-centered care, and effective interprofessional team work. Respondents suggested areas for improvement including increasing EOLC resources, education, an EOLC guideline and improved collaboration, including with EOLC specialists.

Conclusion: Nurses perceived that EOLC can be improved in this ICU with key barriers and enablers identified. Questionnaire results identified that interventions targeting all five domains is needed.