Decisions in chest pain: how often do they change after the initial evaluation?

Abstract

Study objectives: We determine to what extent decisions in chest pain change from initial emergency department (ED) patient encounter to final ED diagnosis, disposition, and treatment. Methods: We designed a prospective, observational study of a convenience sample of chest pain patients triaged to the critical care area of an urban ED. Patients were excluded if they were diagnosed with ST-segment elevation myocardial in farction. Physician decisions about these patients were studied. The physicians involved in the study were attending emergency physicians and upper-level emergency medicine residents who staff 100% of cases with an attending emergency physician. Physicians were surveyed after initial patient evaluation. This initial evaluation was defined as initial medical history and physical examination and reading of the ECG. The survey elicited initial working diagnosis, treatment, and disposition. The treating physician was again surveyed after final ED decisions and disposition were made. The second survey elicited the same information as the first survey about decisions. The second survey also asked the treating physicians to cite the factors that led to changes in their decisions. The surveys were then compared, and decision changes were noted as upgrades or downgrades. Changes in decisions were correlated with cardiac marker results and physician-cited factors that led to those decision changes. Results: Of 83 physician/patient interactions studied, 20 (24%) had an upgrade in management, 24 (29%) a downgrade, and 39 (47%) had no change. Seventy-six patients had negative cardiac markers. Of the patients with negative markers, 14 (18%) had an upgrade, 24(32%) had a downgrade, and 37 (49%) had no change in decisions. The most common factor cited by the treating physician as affecting decisions was change in patient status (N=34, 41%). The second most common factor cited was cardiac marker results (N=30, 36%). From these 30 interactions, 4 had positive markers, all of which had an upgrade in management; 26 had negative markers. Of these 26, 6 were upgraded and 8 were downgraded; 12 had no change in management, even though the physician stated that the marker results affected their decisions. Finally, from these 30 interactions in which cardiac markers were cited as affecting decisions, 9 patients had negative markers and duration less than 6 hours. Follow-up was available for 81 of the 83 patients. No significant correlation between changes in decision and outcomes was found. Conclusion: Nearly 50% of chest pain management decisions did not change after the initial evaluation. Physicians may inappropriately rely on initial cardiac marker results in determining the etiology of a patient's chest pain.