Predictors of emergency department visit disposition for patients with chronic obstructive pulmonary disease

Study objectivesThe study objective was to validate among patients with chronic obstructive pulmonary disease (COPD) clinical predictors of emergency department (ED) disposition (admission vs discharge) found in a previous study. We hypothesized that the initial heart rate (HR) at presentation and the number of nebulizer treatments would predict visit disposition adequately and that existing triage criteria would not.MethodsIn this correlational study, all ED visits to a university medical center related to a diagnosis of COPD over a 2-year period were identified and reviewed retrospectively (N = 114 patients and 226 visits). Associations of clinical predictors with visit disposition was by contingency table or receiver operating characteristic curve analysis.ResultsThe overall admission rate was 60.2%. An initial HR slightly greater than normal limits (HR ≥ 106 vs HR ≤ 105 beats/min) predicted visit disposition; receiver operating characteristic area under curve (95% confidence interval) = .67 (.60-.73); sensitivity = .61 (.52-.69); and specificity = .68 (.57-.77). Likelihood ratios for positive and negative tests were 1.93 and 0.57, corresponding to posterior probabilities of admission = .74 and .46, respectively. Triage classification did not adequately predict visit disposition. Patients receiving three or more nebulized bronchodilator treatments were more likely to be admitted than those receiving fewer treatments, but this predictor had low sensitivity and was no more discriminating than the initial HR. Women were more likely than men to be admitted, odds ratio = 2.66 (1.54-4.61), but there were few clinically meaningful differences between sexes.ConclusionsThe initial HR on arrival predicted disposition of ED visits for patients with COPD with operating characteristics that were superior to the existing triage criteria. Triage of patients with COPD could be improved simply, and at no additional expense, by recognizing that almost any degree of tachycardia in a patient with COPD substantially increases the likelihood of admission.