Differences between nurse- and physician-assessed ICU characteristics using a standardized survey

Objective: Surveys are often used to assess intensive care unit (ICU) organizational characteristics for quality improvement. Typically these surveys target ICU nurse managers and/or physician directors. However, it is unclear whether these providers' assessments differ. We sought to determine whether differences existed in nurse-and physician-assessed ICU characteristics using a standardized survey. Design: We administered a previously developed survey to nurse managers and medical directors in adult ICUs within a single healthcare system in 2013. The survey asked about interprofessional staffing and evidence-based protocols. We examined differences between nurse managers' and medical directors' responses using McNemar's test and assessed concordance using the kappa statistic. Setting: Twenty-three ICUs in 10 hospitals in Southwestern Pennsylvania. Results: Sixteen (69%) were specialty ICUs. The median number of ICU beds was 34. Concordance was moderate for high-vs. low-intensity physician staffing (kappa = 0.60) and almost perfect on questions related to interprofessional staffing (kappa = 0.83 nurse practitioners/physician assistants; 1.0 respiratory therapists; 0.83 physical therapists). However, concordance was slight to fair with regard to the presence of these providers on rounds (kappa = 0.20-0.21) and poor to slight for protocols for liberation from mechanical ventilation (kappa = 0.19), sedation (kappa = -0.03) and central line insertion (kappa = -0.03). Conclusions: Despite a standardized survey, we found substantial disagreement on ICU characteristics when assessed by the nurse manager or physician director. This study raises questions about the use of surveys to examine ICU organizational characteristics and suggests that differences in nurse managers' and medical directors' assessments could be helpful in guiding future ICU quality improvement projects.