Effective resource management using a clinical and laboratory algorithm for chest pain triage

Abstract

Background: Bridgeport Hospital recently compared the use of a clinical algorithm with unaided physicians for the triage of 200 emergency department (ED) patients with chest pain for coronary care unit (CCU) admission.

Hypothesis: Cardiac troponin-T may be an effective adjunct to an established clinical algorithm for accurate selection of patients who are admitted to the CCU for suspected acute myocardial infarction (AMI). It is elevated 6 hours earlier than lactate dehydrogenase fraction 1 (LD1) and may be the most effective test for confirming non Q-wave AMI (usually with ST-T changes and LD1/lactase dehydrogenase > 35%). It is elevated in class 3 angina with a high risk of fatal arrhythmia.

Study design: Clinical and electrocardiogram criteria for triage are as defined by Goldman et al. (see references 6 and 11). The exclusion of non-AMI within 4 hours of ED visit using troponin-T (or an alternative test) should allow more rapid treatment of non-AMI patients at a lower intensity of hospital service (e.g., telemetry versus ICU). This should translate into shorter length of stay for non-AMI patients and facilitate thrombolytic therapy with significant resultant medical and cost benefits.

Laboratory data: Creatine kinase and its MB isoenzyme (CK-MB) are measured at the time of clinical triage and 4, 8, and 12 hours later. LD1 and total lactase dehydrogenase are measured 12 hours after initial sampling. Cardiac troponin-T is measured at the time of ED arrival and 4 hours later. ANALYSIS OF BENEFIT: The first thing to consider is reducing the cost from stress thalliums for low risk evaluation. The second is the evaluation of active ischemia--unstable angina and diagnosis of missed AMI. The Goldman algorithm eliminates unnecessary admissions to the CCU for chest pain, including unstable angina. There is a difference between AMI and triage for CCU admission (200 patients). CK-MBs allowed for the elimination of 37 non-AMIs. However, we missed eight cases of AMI that would have been LD1 positive (troponin positive) and gained 10 cases of unstable angina that should have been assigned to CCU or a monitored bed. Troponin found six more cases of unstable angina that were CK-MB negative but should have been class 3 unstable angina that could be assigned to at least a telemetry bed. This makes 14 cases of AMI or unstable angina unaccounted for by CK-MB (5.6%). All of the cases were at risk of ventricular arrhythmia within 36 hours. The only question was whether to assign them to CCU or to monitored beds. The third point in the analysis is to examine the savings in operating costs by substitution. The cost model, in selecting a strategy, takes test costs, clinical outcomes, and the cost of clinical algorithms into consideration.