A National Survey on Moral Obligations in Critical Care

Carol L. Pavlish, PhD, RN; Katherine Brown-Saltzman, MA, RN; Joseph A. Raho, PhD; Belinda Chen, MPH


Am J Crit Care. 2019;28(3):183-192. 

In This Article


Data Collection

We created a questionnaire based on the results of our focus groups to explore teamwork in ethically complex situations (see Appendix). We tested the clarity and relevance of the questionnaire with 6 physicians and 6 nurses. After revisions, the final online questionnaire included (1) clinician demographics; (2) ranking the top 3 "most frequent and disturbing" ethical situations; (3) rating the degree to which nurses, physicians, and the health care team have moral responsibilities for implementing specific ethics-related actions while caring for critically ill patients; (4) rating agreement with statements about leadership, communication, and trust among nurses and physicians, and about the availability of ethics resources; and (5) final comments. Reliability measures using Cronbach α included 0.945 for participants' ratings of physicians' responsibilities, 0.895 for nurses' responsibilities, and 0.896 for team responsibilities.

After receiving approval from the institutional review board at the University of California, Los Angeles, we recruited participants from the American Association of Critical-Care Nurses and the Society of Critical Care Medicine. Both organizations distributed study information and a survey link in their respective e-newsletters. The survey link was open on Qualtrics (the online survey platform) for 30 days. No incentive was offered for completing the survey.

Data Analysis

Data were analyzed with IBM SPSS Statistics version 24. Baseline characteristics of clinicians were compared by using χ 2 analysis and the Student t test. Clinician groups and moral responsibility ratings were compared by using 1-way analysis of variance with the post hoc Tukey-Kramer test for significant comparisons. The Likert-style ratings in the questions for perceptions of trust, communication, and organizational support were categorized into dichotomous responses (agree/strongly agree and disagree/strongly disagree). Differences between clinician groups were assessed by using the χ 2 test. We used stepwise multivariate logistic regression to identify clinician factors associated with differences in perceptions of trust, communication, and organizational support. Variables that were significant in univariate analyses were entered into multivariate models. Significance was set at P less than .05 for all analyses.

We applied narrative analysis to examine the qualitative data provided by 144 nurses and 7 physicians (N = 151) in the free-text section of the survey. Final comments ranged from brief statements (3 words) to descriptive paragraphs (242 words). First, we inductively coded each response for "moral obligations," then we sorted codes across all responses. This sorting provided clusters of similar codes from which we created thematic categories to understand clinicians' experiences and obligations when working together during complex clinical situations.

While analyzing the narrative responses, we noted that most respondents were dissatisfied with team interactions and reported numerous disappointments, regrets, and failed expectations. As Hyvärinen[27] (p456) has argued, "changing, failing, or realized expectations" are embedded in almost every narrative. Tannen[28] studied linguistic expression of human expectations and found 9 indicators, among which are negative, evaluative, contrastive, and tentative statements. We returned to narrative comments to code for embedded expectations. First, we inductively coded responses for expectations. We then searched for indicator words such as should, must, and not, which are based on the linguistic studies by Tannen.[28] We identified categories of embedded expectations about ethical decision-making.