Unpacking resident-led code status discussions: results from a mixed methods study.

backgroundThe quality of code status discussions (CSDs) is suboptimal as physicians often fail to discuss patients' goals of care and resuscitation outcomes. We previously demonstrated that internal medicine residents randomized to a communication skills intervention scored higher than controls on a CSD checklist using a standardized patient. However, the impact of this training on CSD content is unknown.

objectiveCompare CSD content between intervention and control residents.

designWe conducted qualitative analysis of simulated CSDs. Augmenting a priori codes with constant comparative analysis, we identified key themes associated with resident determination of code status. We dichotomized each theme as present or absent. We used chi-square tests to evaluate the association between training and presence of each theme.

participantsFifty-six residents rotating on the internal medicine service in July 2010 were randomized to intervention (n = 25) or control (n = 31).

interventionIntervention residents completed CSD skills training (lectures, deliberate practice, and self-study). Six months later, all 56 residents completed a simulated CSD. MAIN MEASURE: Comparison of key themes identified in CSDs among intervention and controls. KEY

resultsFifty-one transcripts were recorded and reviewed. Themes identified included: exploration of patient values/goals, framing code status as a patient decision, discussion of resuscitation outcomes and quality of life, and making a recommendation regarding code status. Intervention residents were more likely than controls to explore patient values/goals (p = 0.002) and make a recommendation (p < 0.001); and less likely to frame the decision as one solely to be made by the patient (p = 0.01). Less than one-third of residents discussed resuscitation outcomes or quality of life.

conclusionTraining positively influenced CSD content in key domains, including exploration of patient values/goals, making a recommendation regarding code status, and not framing code status as solely a patient decision. However, despite the intervention, residents infrequently discussed resuscitation outcomes and quality of life.