- © 2004 Canadian Medical Association or its licensors
The study by Steve Goodacre and his colleagues1 addresses the issues of safe and efficient discharge decisions for patients with chest pain. Their data suggest that a chest pain unit is cost-effective compared with routine care in a British emergency department. However, the population in that study appears very different from that of most US chest pain units, especially in terms of the high admission rate after evaluation in the chest pain unit, and it would be difficult to compare the patient population in our study2 with the population in the British study.1 The Canadian health care system is substantially different from both the US and the UK models, and we maintain that a combination of sensible and safe early discharge combined with an efficient rule-out protocol would be more efficient than mandating that a large number of low-risk patients enter a US-style chest pain program, including stress testing before discharge.
Brian Steinhart has concerns about the proportion of missed patients who had anti-ischemic treatment or follow-up diagnostic testing (or both). Our definition of missed patients included only patients who were discharged without any anti-ischemic treatment and without specific follow-up evaluations or testing booked. In only 1 of the 21 cases did the patient end up, many days later, in a cardiology clinic, but we could not confirm any predischarge planning for this appointment. The definition of clinically significant adverse outcomes is an interesting one. The single patient who died had significant comorbidity, and the death was not unexpected. However, 10 of the 21 “missed” patients had a 30-day diagnosis of acute myocardial infarction (AMI). One had an elevated troponin level known by the treating physician but discounted as a false positive. The others re-presented with evidence of myocardial necrosis, and it is likely that the index presentation was unstable angina that could have been treated more appropriately (and the AMI potentially averted). There appears to be no consensus on whether this should be considered inappropriate management; however, our position is that the diagnosis of acute coronary syndrome should be made with the greatest possible accuracy on initial presentation and that each missed case is inappropriate.
We agree that there were differences in methods between our study and that of Pope and colleagues.3 We did not include some critically ill patients, but these patients by definition would not be missed. They may have had a small impact by increasing the denominator modestly. Steinhart contends that if we had used the methods outlined by Pope and colleagues, our rate of missed cases would have been lower than 5.3%. Although this is probably true, the question is which method is more appropriate. We prespecified detailed definitions for AMI and definite unstable angina and followed up patients very carefully and therefore are confident in underscoring our rate of missed acute coronary syndrome.
We encourage others to measure outcomes in patients with chest pain and challenge all to develop consensus on a more appropriate definition of clinically significant missed acute coronary syndrome.
Jim Christenson Grant Innes Barb Boychuk Ken Gin Department of Emergency Medicine Providence Health Care St. Paul's Hospital Vancouver, BC
Footnotes
-
Competing interests: None declared.