Adverse Cardiac Events in Emergency Department Patients with Chest Pain Six Months after a Negative Inpatient Evaluation for Acute Coronary Syndrome.
Reviewed by: Jennifer Tobin, MD
Reference: Manini AF, Gisondi MA, van der Vlugt TM, Schreiber, DH. Adverse Cardiac Events in Emergency Department Patients with Chest Pain Six Months after a Negative Inpatient Evaluation for Acute Coronary Syndrome. Acad Emerg Med. 2002; 9:896-902.
This study was performed at the Stanford-Kaiser Emergency Medicine Residency Program associated with the University of California at San Francisco School of Medicine and Stanford University School of Medicine. Results were presented at the Society for Academic Emergency Medicine Annual Meeting in May 2001.
The authors' objective was to evaluate the risk of adverse cardiac events (ACEs) in patients who had six months previously presented to the emergency department with chest pain, were then admitted, and subsequently underwent a negative cardiac work-up for acute coronary syndrome (ACS). They sought to determine this population's long-term probability of experiencing an ACE, the latter being defined as cardiac death, cardiac arrest, MI, emergent revascularization (PTCA or CABG), or unstable angina. The authors' motivation for such a study derived from the current Agency for Health Care Policy and Research guidelines, which recommend that patients with unstable angina who are judged to be at low short-term risk for adverse cardiac outcomes be evaluated as outpatients within 72 hours after initial presentation. They note that much of the data regarding outcomes for low-risk populations stem from the surveillance of patients treated in chest pain observation units. These patients are at relatively low risk (<7%) for cardiac events within 72 hours if they have a negative work-up based on history and ECG findings alone, and findings of negative serial cardiac markers can identify a population at even lower risk for cardiac events at 30 days (0.3-0.7%). Nevertheless, these "low risk" patients remain at a relatively high risk for subsequent cardiac morbidity within six months (5-23%), and it is this long-term outcome that interests the authors. Finally, while several recent prospective studies have investigated the six-month and even one-year prognosis of patients with negative diagnostic cardiac studies, none so far has evaluated the effect of diagnostic test setting--inpatient versus outpatient--on long-term risk of ACEs. The authors incorporate such an analysis into their results.
The study identified all ED patients with chest pain who were admitted to the hospital with suspected ACS, and who had a subsequent negative in-hospital ACS evaluation. Patients were excluded if they 1) had a positive diagnostic cardiac study either in-hospital or as an outpatient; 2) were transferred to an outside hospital during the study period; 3) had a heart transplant prior to the study period; 4) had no place of residence; or 5) resided outside California. At six months, patients were contacted by telephone to document any subsequent clinical cardiac events (ACEs as defined above). The presence or absence of cardiac risk factors, including any of the following--smoking, hyperlipidemia, hypertension, diabetes mellitus, prior MI, and prior CAD--were gleaned from the patients' charts. Finally, patients were classified according to whether they were scheduled to receive a cardiac diagnostic test, and according to the scheduled setting--inpatient (IP) or outpatient (OP)--of that test. Inpatient tests were conducted prior to discharge, and outpatient tests were scheduled as soon as possible after discharge and no later than two weeks following discharge. For all subjects in the OP group, no ACE occurred in the interval between discharge and the scheduled time of the study. Tests included any of the following: exercise treadmill testing, angiography, stress echocardiography, or stress thallium perfusion scans.
199 patients met the initial criteria, of which 26 were excluded for the reasons stated in the above paragraph, so that 157 were available for evaluation and follow-up. Of these patients, 40% (62/157) had pre-existing CAD and 26% (41/157) had had a previous MI. Overall, 78% (122/157) were scheduled to have a cardiac diagnostic test (i.e., were in either the IP or the OP group) while 22% were not scheduled for a test (were in the group designated "N"). Crucially, patients in the OP group who were non-compliant and were not tested as scheduled remained in the OP group. This had a significant impact on the authors' results.
At six months, the entire cohort of 157 patients had an ACE rate of 14% (22/157). The OP group had a rate of 23% (12/52), while the IP group had a rate of 10% (7/70), and the N group had a rate of 9% (3/35). The authors also calculated the ACE rate of the three groups, while excluding patients with prior CAD: 24% for the OP group; 5% for the IP group; and 0% for the N group. Logistic regression analysis was used to evaluate whether any cardiac risk factors (hyperlipidemia, hypertension, smoking, etc.) were independent predictors of ACEs in this population, and none was found to be.
The authors conclude that ED patients with subsequent negative inpatient ACS evaluations have a higher rate of ACEs (14%) than that predicted by previous reports (2-5%). Even excluding those with prior CAD, the ACE rate was 11%. The rather high numbers obtained by these authors is due in part to their inclusion of unstable angina in the category of ACE; they justify its inclusion because unstable angina patients have been shown to have a long-term prognosis similar to that of patients with MI. Finally, while no cardiac risk factor was found to be an independent predictor of an ACE at six months, the setting of scheduled diagnostic tests for chest pain patients seems important to long-term cardiac morbidity: the group scheduled for outpatient tests was at significantly higher risk for ACEs at six months than was the group scheduled for inpatient tests. The authors suggest that one explanation for this finding is the relatively high rate of noncompliance with diagnostic testing in the OP group, and they conclude that "in patients for whom compliance and follow-up are problematic, inpatient evaluation appears warranted and indicated."
Limitations of the study derive from the fact that it is retrospective: the setting of scheduled diagnostic testing was not randomized, but rather assigned by each patient's physician. Therefore, it is not clear that the setting of scheduled diagnostic testing per se affects cardiac morbidity. Perhaps those patients who are noncompliant with follow-up testing (all of whom were by necessity in the OP group) are also more prone to ACEs for other reasons. For example, they may exercise less--sedentary lifestyle is one of the cardiac risk factors which the authors did not eliminate as a predictor of ACE. Also, while the authors found that no one cardiac risk factor was an independent predictor of ACEs in their patients, it may be that certain combinations of these factors help to identify patients at risk. Finally, while the authors suggest that for patients in whom there is a risk of noncompliance, inpatient evaluation is warranted, this may not be the best use of resources. If these patients could be successfully convinced by telephone or mail (or cash incentives or lower insurance premiums) to go to their follow-up appointments, costly admission to the hospital would be unnecessary.
