The purpose of this paper is to present a critique on a systemic review article on the topic of emergency department crowding and patient outcome (Carter, Pouch & Larson, 2014). The problem addressed by the article is the global increase in Emergency Department. Crowding in the Emergency Department is well-known to be associated with poor care processes such as timeliness of care and diversion of ambulances (Johnson & Winkelman, 2011; Kennebeck, Timm, Kurowski, Byczkowski, & Reeves, 2011). The review article sought to focus on another facet of Emergency Room crowding and that was patient outcome.
The research question of the relationship of Emergency Department crowding and patient outcome was clearly stated, well-defined, and had an appropriate scope. The concept of patient outcome was defined as patient satisfaction, morbidity, mortality, and exiting the Emergency Department without being seen. Carter, Pouch and Larson (2014) defined crowding in terms of Emergency Department capacity and volume, and Emergency Department length of stay, which was appropriate. The results of the systematic review were integrated into a table with appropriate headings: design, sampling, crowding measure, outcome measure, and results, and thus provided a convenient means of assessing the body of literature.
The research strategy was clearly described, adequately covered the topic, and covered the period from January 2002 to January 2012. The authors used a standard method for identifying primary articles and included a table that showed the Boolean combinations of medical subject headings and key words that were used to search OVID Medline and OVID Medline in Process and search engines for non-indexed citations. The supplementary methods used to identify articles were an electronic search of the tables of content of six emergency medicine journals using the same key words and a hand search of the list of cited references of relevant articles locate by the search. A four-tier PRIMA flow chart was included to show the flow of 196 initially identified articles through to the finally selected 11 articles to be included in the review.
All the articles included in the final sample of 11 used a comprehensive range of acceptable research designs [case-crossovers (1), prospective observational (1), retrospective observational (2), correlational study (1), retrospective cohort (3), population-based retrospective cohort (1), retrospective stratified cohort (1), and prospective cross-sectional (1)]. All the 11 studies had robust sample sizes, which ranged from 213 to 13,934,542. The inclusion criteria were articles measure Emergency Department crowding or a proxy of crowding and at least one of the patient outcomes of interest. Articles were excluded if the study described crowding interventions, care processes related to crowding, and instruments to predict or measure crowding. Also excluded were studies without abstracts, commentaries, or that were editorials, or articles in a language other than English. Using an iterative method, the authors discussed each article for inclusion or exclusion. Disagreements were settled by discussion and consensus. The authors pointed out that only one search engine was used and gray literature was not examined for potential inclusion. Although the articles were discussed, ultimately the decision to include an article was subjective. No authors were contacted to obtain additional information.
The article indicated a comprehensive appraisal process. The quality of the articles included in the sample was appraised with the criteria being full descriptions of study question and study population, results, statistical analyses, exposure measurements, and outcome measurements. The authors stated that the majority of the articles satisfied the criteria, but there were some deficits in the descriptions of the sample population and in the research questions. Although most of the articles were appraised by one researcher, an acceptable interrater reliability was obtained on a subset of the articles. The authors included a table that summarized the appraisal information for each article.
The data extract was briefly but adequately described and covered the categories of sample size, methods, and sample characteristics. The data extraction tool used by the study was adapted from a recent study (Uchida, Pogorzelska-Maziarz, Smith, & Larson, 2013). All of the study’s authors piloted two articles using the data extraction instrument and a high level of interrater reliability was obtained.
A narrative integration rather than a formal meta-analysis was performed, which was described in the Results section of the article. As the number of articles to be analyzed was limited (n = 11) and the categories for comparison were well-defined and also limited, and the outcome measures were heterogeneous, a narrative integration is an appropriate method. Overall, the analysis process was adequate, clearly described, and credible. The comprehensive table describing the article content across relevant categories was sufficient to supply the information needed by the reader to access the integrity of the article.
The conclusions were drawn based on the research question, which was the relationship of crowding to patient outcome. Not all the studies in the sample focused on patient outcome, but those that did found that Emergency Department crowding was associated with higher rates of mortality for those admitted and discharged from the Emergency Department. The authors’ noted unevenness with regard to the rigour of some of the studies and the heterogeneity of the outcome measures. A consistent finding across the studies was that Emergency Department crowding was related to higher rates of people leaving the Emergency Department without having been seen.
The limitations of the study were described in thorough detail. The authors stated that the exclusion criterion of restricting the studies to those that used a direct or proxy measure of crowding may have led to missing some relevant articles. As mentioned previously, another limitation was that the study abstraction and assessments of the quality of the articles was conducted for the most part by only one researcher.
Although two of the three authors of the research are RNs, no implications for nursing practice were included in the article. The recommendation for further research was restricted to a single sentence, which called for policies to address crowding as it constituted a significant safety concern for patients. The research question was focussed on the relationship of overcrowding to patient outcomes, which might be regarded by the authors as more of an administration than a nursing issue in the Emergency Department, and the lack of any recommendations might be justified on that basis. Also, the sample was small and had inconsistencies in quality that would limit the authors’ ability to formulate any recommendations.
Given the limited number of studies that fit the inclusionary criteria, the researchers might have expanded the criteria to include qualitative research on the association on crowding in Emergency Department and patient outcome. The narrative synthesis method of systematic reviews has been used effectively to synthesize results of quantitative and qualitative studies (Elder, Johnston & Crilly, 2015; Wheeler, et al., 2015).
The major result of this systematic review was a consistency of the results concerning a higher rate of people leaving without being seen being associated with increased crowding in the Emergency Department. However, this finding is already well-established (Hsia, et al., 2011). Strict adherence to the inclusionary criteria is laudable in systematic reviews with large numbers of articles, but in this case, it makes the results less useful.
References
Carter, E. J., Pouch, S. M. & Larson, E. L., (2014). The Relationship Between Emergency Department Crowding and Patient Outcomes: A Systematic Review. Journal of Nursing Scholarship, 46(2), 106–115.
Elder, E., Johnston, A. N, & Crilly, J. ( 2015). Review article: systematic review of three key strategies designed to improve patient flow through the emergency department. Emergency Medicine Australasia, 27(5), 394-404. doi: 10.1111/1742-6723.12446
Hsia, R. Y., Asch, S. M., Weiss, R.E., Zingmond, D., Liang, L-J., Han, W.Sun, B. C. (2011). Hospital Determinants of Emergency Department Left Without Being Seen Rates. Annals of Emergency Medicine, 58, 24–32.
Johnson, K. D., & Winkelman, C. (2011). The effect of emergency department crowding on patient outcomes: A literature review. Advanced Emergency Nursing Journal, 33(1), 39–54. doi:10.1097/TME.0b013e318207e86a
Kennebeck, S. S., Timm, N. L., Kurowski, E. M., Byczkowski, T. L., & Reeves, S. D. (2011). The association of emergency department crowding and time to antibiotics in febrile neonates. Academic Emergency Medicine, 18(12), 1380–1385. doi:10.1111/j.1553-2712.2011.01221.x
Uchida, M., Pogorzelska-Maziarz, M., Smith, P. W., & Larson, E. (2013). Infection prevention in long-term care: A systematic review of randomized and nonrandomized trials. Journal of the American Geriatric Society, 61(4), 602–614. doi:10.1111/jgs.12175
Wheeler, C., Lloyd-Evans, B., Churchard, A., Fitzgerald, C. Fullarton, K., Mosse, L., Johnson, S. (2015). Implementation of the Crisis Resolution Team model in adult mental health settings: a systematic review. BMC Psychiatry,15, 74. 10.1186/s12888-015-0441-x
