Developing a Measure of Facilitators and Barriers to Rapid Response Team Activation

Purpose:   The purpose of this study was to develop and test an instrument that identifies specific facilitators and barriers to rapid response team (RRT) activation; this will ultimately provide a foundation for interventional studies designed to promote RRT utilization.

 RRTs, comprised of critical care experts, respond to the bedside of deteriorating patients in order to prevent cardiopulmonary arrest (IHI, 2009; VHA, 2007). Poor patient outcomes and increased costs may be associated with underutilization of RRTs (Morse, Warshawsky, Moore, & Pecora, 2007; Stolldorf, 2008; Winters et al., 2007). Little is known about facilitators and barriers to nurse activation of RRTs. Data from a previous qualitative study indicated facilitators and barriers are related to nursing unit culture, RRT member characteristics, and continuing RRT education (Astroth, Woith, Stapleton, Degitz, & Jenkins, 2013).

 Social judgment theory, as illustrated by the Lens Model of Cognition (Hammond, Hursch, & Todd, 1964), will serve as the conceptual framework for the study. According to the Lens Model, people make decisions based on cues, the significance they attach to those cues, input from colleagues, and available resources, such as time to process information (Hammond et al., 1964; Thompson, Foster, Cole, & Dowding, 2005). In the study, cues correspond to patients’ clinical signs and symptoms of impending CPA and the significance nurses attach to them. Input from colleagues corresponds to nursing unit culture and RRT member characteristics. Available resources correspond to RRT member characteristics and continuing RRT education.

Methods: We first developed survey items from a review of the literature and qualitative data; this data was obtained from a study of 15 acute care RNs exploring facilitators and barriers to activation of rapid response teams. We used the Lens model as a framework for identifying instrument subscales. Using an exploratory design, we surveyed a convenience sample of registered nurses (RNs) employed in five Illinois hospitals.

Participants completed the online RRT Facilitators and Barriers Survey, a 32 item survey developed by the researchers. The survey contains two subscales, facilitators and barriers. Items in both subscales describe nursing unit culture, continuing RRT education, and RRT member characteristics. Participants rated each item using a five-point, Likert-type scale; item responses range from strongly disagree (1) to strongly agree (5). We established face validity with 3 nurses and a non-nurse with no rapid response experience and content validity through a review of survey items by 4 nurses with expertise in staff development and rapid response teams. We piloted this newly developed instrument by surveying 50 RNs employed in one hospital using Select Survey, an online survey system. In order to fully describe the sample, we also collected demographic data: age, gender, nursing education, years of nursing experience, number of RRTs activated, and most recent RRT education program attended. Item analyses were conducted through exploratory factor analyses; internal consistency estimates were obtained. We utilized hierarchical cluster analysis to supplement factor analyses to identify homogeneous subsets of items, to refine items, and to eliminate redundant items. Due to low reliability of the education subscale and items clustered into other subscales, we developed additional items to the educational subscales.

We further tested the revised online 36 item instrument by surveying 250 nurses from the remaining hospitals, using Survey Monkey. We used confirmatory factor analysis to analyze these data using LISREL VIII and employing a multifaceted model-fit assessment strategy. Descriptive statistics were conducted on the demographic data to describe sample and setting characteristics.

Results: The final sample consisted of 194 nurses from four hospitals. The sample was predominantly female (74.8%), with a mean age of 39.16 (SD 12.04) years and 13.83 (SD 12.43) years of experience.  Most (48.4%) of the participants had a baccalaureate degree in nursing. Regarding the number of RRTs called, participants reported a range of none to too many to count. When asked about their most recent RRT continuing education, they reported a range of two months to four years. Some indicated they had never attended RRT continuing education.  

The full scale alpha is .73. Cronbach’s alphas for subscales measuring facilitators were: unit culture .83, team characteristics .83, and knowledge was .81. Cronbach’s alphas for subscales measuring barriers were: ranged from unit culture .81, team characteristics was .92, and education was .13. We analyzed item correlations within and between subscales.

We used LISREL 8.80 to conduct CFA on the proposed factor model for the 36 items.  The chi-square (579, N=194) = 812.80, the Root-Mean-Square-Error-of-Approximation (RMSEA) of 0.48, the Non-Normed Fit Index (NNFI) of 0.49, the Comparative Fit Index (CFI) of 0.53, and the Standardized Root-Mean-Residual of 0.71 all suggest poor model fit as none of these goodness of fit indices approach acceptable levels.

Conclusions: This scale shows promise for use in identifying facilitators and barriers to nurses’ use of rapid response teams. These facilitators and barriers may vary across institutions. Internal consistency of all subscales except education barriers reflects good reliability, especially for a new instrument (Nunally & Bernstein, 1994). The barriers associated with RRT education may not necessarily be highly correlated. The lack of internal consistency, in fact, is expected when considering items that reflect risk factors. Thus, experiencing one barrier to RRT education does not necessarily mean that one would experience the others. Further work to identify micro-structures within each factor and breaking factors apart using hierarchical cluster analysis and other item analysis techniques is warranted.