Direct Comparison of Characteristics, Treatment, and Outcomes of Patients Enrolled Versus Patients Not Enrolled in a Clinical Trial at Centers Participating in the TIMI 9 Trial and TIMI 9 Registry

Maria Cecilia Bahit, MD, Christopher P. Cannon, MD, Elliott M. Antman, MD, Sabina A. Murphy, MPH, C. Michael Gibson, MD, Carolyn H. McCabe, BS, Eugene Braunwald, MD


Am Heart J. 2003;145(1) 

In This Article


Of the total of 3580 patients with STEMI or LBBB, 2992 patients were enrolled in the TIMI 9B trial and treated with fibrinolysis, 257 of whom were enrolled at TIMI 9 Registry hospitals. This group of 257 had baseline characteristics similar to the entire cohort in the TIMI 9B trial. Of the remaining non-TIMI 9B patients, there were 214 fibrinolytic-eligible patients who were treated with reperfusion therapy and 82 who were not treated. There were 107 patients who were considered ineligible for fibrinolysis on the basis of relative contraindications but who were nonetheless treated with reperfusion therapy, and 175 patients not eligible, not enrolled, and not treated (Figure 1).

Patient population and groups. Because only 10 patients enrolled in TIMI 9 trial did not receive reperfusion therapy, the group enrolled as not treated was not considered.

The baseline characteristics of the patients by group are shown in Table I . Across the 5 groups, a gradient of high-risk baseline characteristics was observed. Overall, patients not enrolled in the TIMI 9 trial tended to be older, female, and to have a history of MI and a worse Killip class on admission compared with those who were enrolled ( Table I ). For time from symptom onset to presentation, patients not eligible and not treated had the longest time (median 5.25 hours, interquartile range 1.34-21 hours) among the groups, and nearly half arrived after 12 hours (P < .001 compared with each other group of patients).

Reperfusion therapy was administered in almost all (99.6%) of the patients enrolled in the TIMI 9B trial. However, only 72% of patients deemed eligible for fibrinolysis but not enrolled in the trial received reperfusion therapy; 60% of these patients received fibrinolytic therapy and 12% underwent primary percutaneous coronary intervention (PCI). Conversely, 38% of those deemed ineligible by ACC/AHA criteria actually received reperfusion therapy, consisting of 25% who received fibrinolysis and 13% who received primary PCI.

A gradient was seen in the use of aspirin among the groups, with patients enrolled in the TIMI 9 trial having the highest inhospital rate of use of aspirin (96.6%) and those not eligible, not enrolled, not treated having the lowest (64.5%) (P < .001). Likewise, the use of ß-blockers also followed the same pattern (P < .001) (Figure 2).

Use of guideline-recommended medications and selected cardiac procedures across the 5 groups of patients.

Conversely, cardiac procedures including use of cardiac catheterization (P = .02), PCI (P = .002), and intra-aortic balloon pumping (P = .03) were used less often in patients in the TIMI 9 trial compared with patients not enrolled in the trial. Only the group ineligible for fibrinolysis and not treated had a lower rate of use of cardiac procedures ( Table II ).

Inhospital mortality rates are shown in Figure 3. Patients enrolled in the TIMI 9B trial had the lowest inhospital mortality rate at 5.0%, compared with 8.0% for those eligible, treated, 9.7% for those eligible, not treated, 13.1% for those ineligible, treated, and 24% for those ineligible, not treated (5-way P < .001). Patients not eligible, not treated had a 4-fold-higher rate of cardiogenic shock and severe heart failure (11.5% and 16.6%, respectively) compared with patients enrolled in the TIMI 9 trial (2.8% for both) ( Table III ).

Likewise, occurrence of the composite end point of death, repeat MI, new cardiogenic shock, and severe congestive heart failure was 3 times higher in patients not eligible (30%) compared with those enrolled in the TIMI 9B trial (9.0%).

Inhospital mortality rates across the 5 groups of patients.

On univariate Cox modeling, patient ineligibility was associated with inhospital death: not eligible, treated odds ratio (OR) was 2.8 (1.57-5.08, P < .001); not eligible, not treated OR was 5.9 (4.04-8.71, P < .001). On multivariate modeling including baseline characteristics and patient groups, only 5 features were found to be independently associated with inhospital death: increasing age, severity of Killip class, anterior MI, lower blood pressure, and status of not eligible and not treated (OR 2.8, 95% CI 1.80-4.54, P < .001) ( Table IV ).

In addition, various clinical features were independently associated with the occurrence of the composite of death, MI, new cardiogenic shock, and severe heart failure: increasing age, male sex, severity of Killip class, anterior MI, lower systolic blood pressure, eligible patients not enrolled, treated (OR 1.7, 95% CI 1.15-2.63, P < .001), and patients not eligible, either treated or not treated ( Table IV ).

We directly compared patients treated with fibrinolysis in the context of a clinical trial with eligible patients not enrolled treated with the same reperfusion therapy ( Table V ). Eligible patients treated within a clinical trial received aspirin more frequently both in the hospital and at discharge and ß-blockers and angiotensin-converting enzyme (ACE) inhibitors compared with those treated outside a clinical trial. Even though these 2 groups of patients received fibrinolysis as a reperfusion therapy, patients enrolled in TIMI 9 trial had a better outcome compared with those receiving fibrinolysis outside a clinical trial. Conversely, the use of cardiac procedures including cardiac catheterization, PCI, and coronary artery bypass grafting was statistically significantly higher among patients not in the TIMI 9 Trial compared with those enrolled in the TIMI 9 Trial (P = .01). Patients treated had a lower unadjusted mortality rate (5.0%) compared with those eligible but treated outside a clinical trial (8.5%, P = .047); however, when controlling for baseline characteristics, no difference was observed (OR 0.83, 95% CI 0.44-1.56, P = .563).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.