M H Ebell1, L L White, D Weismantel. 1. Department of Family Practice, Michigan State University, East Lansing 48824-1315, USA. ebell@msu.edu
Abstract
OBJECTIVE: The accuracy of the troponin T and I test as a prognostic tool for patients with chest pain varies considerably depending on the patient population, the cutoff for an abnormal test result, and other factors. The goal of our systematic review was to synthesize the best available evidence on this topic. SEARCH STRATEGY: We searched the MEDLINE database, bibliographies of identified articles, and articles identified from a previous meta-analysis of diagnosis. SELECTION CRITERIA: We included cohort studies that had at least 80% follow-up and reported useful data. DATA COLLECTION AND ANALYSIS: Data from each study were abstracted by 2 investigators. We calculated sensitivity and specificity for the prediction of death, fatal or nonfatal myocardial infarction (MI), or any cardiac event for each combination of patient population, troponin test, interval from admission to blood draw, and cutoff for an abnormal test result. MAIN RESULTS: For patients with chest pain and a normal electrocardiogram, the peak troponin I level drawn 6 or more hours after the onset of chest pain is useful for identifying patients at low risk of death or nonfatal MI at 30 days (negative likelihood ratio=0.07; probability of outcome=0.3% with a negative test, given a pretest probability of 4.4%). For patients with unstable angina, the sensitivity of troponin I for the identification of patients who die or have a nonfatal MI in the next 30 days is only 59%, and the specificity is only 79%. The sensitivity and specificity varied widely for patients with unstable angina or non-Q-wave MI depending on the inclusion criteria, cutoff used, timing of the blood draw, duration of follow-up, aim other factors. CONCLUSIONS: If the peak troponin T or I level measured at least 6 hours after the onset of chest pain symptoms is in the normal range in a patient with a normal electrocardiogram, it is very unlikely that the patient will die or have a nonfatal MI in the next 30 days (< or =1%). The initial troponin value is not as helpful as the peak value at least 6 hours after the onset of chest pain. An abnormal troponin test result for patients with unstable angina or non-Q-wave MI identifies a subset at greater risk of death.
OBJECTIVE: The accuracy of the troponin T and I test as a prognostic tool for patients with chest pain varies considerably depending on the patient population, the cutoff for an abnormal test result, and other factors. The goal of our systematic review was to synthesize the best available evidence on this topic. SEARCH STRATEGY: We searched the MEDLINE database, bibliographies of identified articles, and articles identified from a previous meta-analysis of diagnosis. SELECTION CRITERIA: We included cohort studies that had at least 80% follow-up and reported useful data. DATA COLLECTION AND ANALYSIS: Data from each study were abstracted by 2 investigators. We calculated sensitivity and specificity for the prediction of death, fatal or nonfatal myocardial infarction (MI), or any cardiac event for each combination of patient population, troponin test, interval from admission to blood draw, and cutoff for an abnormal test result. MAIN RESULTS: For patients with chest pain and a normal electrocardiogram, the peak troponin I level drawn 6 or more hours after the onset of chest pain is useful for identifying patients at low risk of death or nonfatal MI at 30 days (negative likelihood ratio=0.07; probability of outcome=0.3% with a negative test, given a pretest probability of 4.4%). For patients with unstable angina, the sensitivity of troponin I for the identification of patients who die or have a nonfatal MI in the next 30 days is only 59%, and the specificity is only 79%. The sensitivity and specificity varied widely for patients with unstable angina or non-Q-wave MI depending on the inclusion criteria, cutoff used, timing of the blood draw, duration of follow-up, aim other factors. CONCLUSIONS: If the peak troponin T or I level measured at least 6 hours after the onset of chest pain symptoms is in the normal range in a patient with a normal electrocardiogram, it is very unlikely that the patient will die or have a nonfatal MI in the next 30 days (< or =1%). The initial troponin value is not as helpful as the peak value at least 6 hours after the onset of chest pain. An abnormal troponin test result for patients with unstable angina or non-Q-wave MI identifies a subset at greater risk of death.