End-of-life decisions in the intensive care unit: clinical, ethical, and collaborative challenges

Background
When making end-of-life decisions in intensive care units, the different staff groups have different roles in the decision-making process and may not always assess the situation identically. Practice recommendations for withholding or withdrawing therapy state that decisions should be interdisciplinary, but the literature shows that this is not always the case. Research on end-of-life issues in Danish ICUs is limited.
Aim
The aims of this thesis were to
• Examine Danish practices regarding end-of-life decisions in the ICU.
• Examine the opinions of nurses and physicians who work in Danish ICUs regarding
considerations of:
o What should induce withholding or withdrawing therapy.
o The multidisciplinary collaboration regarding end-of-life decisions.
o Interventions that may improve collaboration and decision-making.
• Examine and evaluate different methods to improve interdisciplinary collaboration and decision-making.
Hypotheses
• Nurses, intensivists, and primary physicians have different experiences of interdisciplinary collaboration regarding end-of-life decision-making in the ICU.
• Specific interventions targeting end-of-life decision-making in the ICU, such as interdisciplinary audits and guidelines, can improve both interdisciplinary collaboration and patient care.
Methods
A multi-method approach was used, including five sub-projects:
Subproject 1. Hospital record review: The review included all patients who had either died in two regional ICUs in 2008, or were discharged with treatment withheld or withdrawn (264 patients). The basic characteristics of the patients who were discharged from the units with full therapy were also collected (1401 patients).
Subproject 2. Interviews: Mono-professional focus-group interviews with 11 nurses and 10 intensivists, and individual interviews with 8 primary physicians were conducted in two ICUs.
Subproject 3. Questionnaire survey: A questionnaire regarding different aspects of end-of-life practices was developed based on literature and the interviews. After pilot testing the questionnaire, it was used in a survey among nurses (495) and intensivists (135) from 10 ICUs in the Region of Southern Denmark. Additionally the survey included primary physicians (146) from two regional ICUs.
Subproject 4. Audit: Three interdisciplinary audits with the participation of 8 primary care physicians, 9 intensivists, and 12 nurses were conducted. Form and profit of the audits were evaluated by a short questionnaire at the end of the sessions and again three months later.
Subproject 5. Guidelines: The guidelines for withholding and withdrawing therapy were developed based on prior projects. These guidelines were implemented in two regional ICUs in May 2011 and were evaluated after 6 months with a questionnaire survey and a hospital record review. Prior projects provided baseline data.
Results
Hospital record review: Out of 1665 patients admitted to the ICUs, 176 patients (10.6%) died; of these 34 (19.3%) died while still receiving full active therapy, 25 (14.2%) died after therapy was withheld and 117 (66.5%) died after therapy was withdrawn. An additional 88 patients (5.3%) were discharged alive with therapy either withheld or withdrawn.
Interviews: The participants identified the main challenges regarding end-of-life decision-making as different assessments of recovery potential, unnecessary changes and postponements of withholding or withdrawing therapy orders, and how and when to identify the patient’s wishes.
Questionnaire survey: The unified response rate was 84%. “Futile therapy” and “Patient’s wish” were the main reasons for considering withholding or withdrawing therapy for all of the participants. Out of the primary physicians, 63% found their general experience of collaboration to be very or extremely satisfactory compared to 36% of the intensivists and 27% of the nurses. Forty-three percent of the nurses, 29% of the intensivists, and 2% of the primary physicians found that decisions regarding withdrawal of therapy were often, very often or always unnecessarily postponed.
Audit: Both immediately after and after three months most of the participants (97% and 89%, respectively) found that this type of audit (to some or great extent) was usable to improve interdisciplinary collaboration regarding end-of-life decisions. All of the participants emphasised the interdisciplinarity as one of the benefits of the discussions. After three months 35% - 45% found that both collaboration and their own practice (to some or less extent) had changed.
Guidelines: The guidelines were evaluated with a questionnaire and a hospital record review.
Questionnaire survey: The unified response rate was 81%. No significant changes were found in satisfaction with the interdisciplinary collaboration regarding end-of-life decision-making or in experiences of withholding or withdrawing decisions being unnecessarily postponed.
Hospital record review: For patients, who died after their therapy had been withdrawn, the median time from admission to the first consideration on the level of therapy decreased from 1.1 to 0.4 days (p=0.03), and the median time from admission to the decision to withdraw therapy decreased from 3.1 to 1.1 days (p=0.02). Total length of stay at the ICU decreased from 3.1 to 1.7 days (median) (p=0.06).
Conclusion
Withholding or withdrawing therapy is common in Danish ICUs. Nurses, intensivists, and primary physicians agreed in principle on what should induce considerations on the level of therapy, but they differed in their perceptions of collaboration and other aspects of withholding and withdrawing therapy practices at the ICU. Nurses were the least satisfied and primary physicians were the most satisfied with the collaboration. The studies suggest that the use of interdisciplinary audits and guidelines for withholding and withdrawing therapy may facilitate improvements in interdisciplinary collaboration, end-of-life decision-making and patient care.