Moving cardiology to the front of the hospital.

The major factor influencing the outcome of thrombolytic therapy for acute myocardial infarction is the " door to needle " time. For every hour that thrombolytic treatment is delayed after the onset of symptoms, 1.6 lives are lost per 1000 patients treated. 1 In many institutions the door to needle time exceeds one hour. Various attempts have been made to reduce unacceptable treatment delays. The most successful approach has been to shift the administration of thrombolytic treatment from the coronary care unit (CCU) to the emergency department. 2 In the study by Hourigan and colleagues which appeared in a recent issue of Heart, 3 the door to needle times of 100 patients treated by emergency physicians in the emergency department were prospectively compared with those of a historical group of 89 non-consecutive patients treated in the CCU between three years and 16 months previously. Not surprisingly, the door to needle times in the emergency department were much shorter than those in the CCU (37 v 80 minutes). This is an admirable achievement, and patient outcomes must have been improved as a result. However, according to current guidelines , the door to needle delay should have been no longer than 30 minutes, 4 5 and the usage of thrombolytic treatment, which was given to only 20% of patients with a subsequent discharge diagnosis of myocardial infarction, was rather low. 6 To evaluate the appropriateness of the treatment given in the emergency department and the CCU, one would need to know how well the ECG criteria and contraindications against thrombolytic treatment were adhered to. Judging by their discharge diagnoses, thrombolytic treatment was inappropriately given to one patient in the CCU and three in the emergency department who did not have myocardial infarction. There was a significantly higher mortality rate among the patients who received treatment in the emergency department (12% v 3.4% in the CCU, p = 0.02), including two deaths that were attributed to errors of judgement. In this selected population, it is possible that some patients receiving thrombolytic treatment in the CCU initially had non-diagnostic ECGs but subsequently fulfilled the ECG criteria for thrombolytic treatment, thus biasing the analysis against this group. A temporal ascertainment bias could also have occurred—that is, sicker patients may have died before reaching the CCU, aVecting both the door to needle times and clinical outcome measurements. The patients given thrombolytic treatment in the CCU may …