External validity of ST elevation myocardial infarction trials: the Zwolle studies
Saman Rasoul1, Jan Paul Ottervanger, Jan-Henk E Dambrink
1Isala Klinieken, Department of Cardiology, Zwolle, The Netherlands.
Insights
Patients with ST-elevation myocardial infarction (STEMI) not enrolled in trials were older, had higher mortality, and differed significantly from enrolled patients. Guidelines should include observational data from unselected STEMI patients to reflect real-world practice.
Area of Science:
- Cardiology
- Clinical Research
- Public Health
Background:
- Clinical practice guidelines often rely on randomized controlled trials (RCTs) with limited external validity.
- Routine patient care may not fully align with RCT findings, necessitating broader data considerations.
- This study examines differences in ST-elevation myocardial infarction (STEMI) patients enrolled versus not enrolled in trials at a single hospital.
Purpose of the Study:
- To compare baseline characteristics, treatment, and long-term prognosis of STEMI patients enrolled in clinical trials versus those not enrolled.
- To identify predictors of non-enrollment in STEMI trials.
- To inform guideline development by highlighting the need to incorporate data from unselected patient populations.
Main Methods:
- Retrospective analysis of individual patient data for all STEMI discharges between January 2001 and December 2001.
- Follow-up data collected until December 2004.
- Multivariate analysis to identify independent predictors of non-enrollment and long-term mortality.
Main Results:
- Of 583 STEMI patients, 78% were enrolled in trials. Non-enrolled patients were older, had more prior myocardial infarctions (MI), and higher risk profiles.
- Higher age was the sole independent predictor for non-enrollment.
- Non-enrolled patients received more conservative treatment and had significantly higher long-term mortality (36% vs. 6%).
Conclusions:
- Significant differences exist in characteristics, treatment, and prognosis between STEMI patients enrolled and not enrolled in clinical trials.
- Non-enrollment, older age, and diabetes are independent predictors of long-term mortality in STEMI patients.
- Guidelines should incorporate findings from observational studies of unselected patients to better represent daily clinical practice and improve STEMI patient outcomes.
Background:
Guidelines are mainly based upon results of randomised controlled clinical trials. However, due to low external validity of these trials, their results can not reasonably be applied to all patients in routine practice. In our hospital, all patients with ST-elevation myocardial infarction (STEMI) are eligible for inclusion in one of our ongoing trials or registries. To asses differences between patients enrolled versus not-enrolled in a trial or registry, we evaluated all patients with a discharge diagnose of STEMI during the study period.
Methods:
Retrospectively, individual patient data from all patients with a discharge diagnosis of STEMI between Jan 2001 and Dec 2001 were evaluated. Follow-up data were obtained until Dec 2004.
Results:
A total of 583 patients were discharged with a diagnosis of STEMI. About 455 (78%) patients were enrolled in one of the ongoing clinical trials or registry and 128 were not. Not-enrolled patients were significantly older; more often had a history of previous MI and had higher risk profiles. Multivariate analysis revealed that higher age was the only independent predictor for non enrollment. Not-enrolled patients were more often treated conservatively and had a higher mortality rate (36% vs. 6%, P << 0.001). After multivariate analysis, non-enrollment (OR: 95% CI) 4.02 (1.98-8.16), age 1.07 (1.04-1.12), and diabetes 2.39 (1.17-4.89) were the only independent predictors of long term mortality.
Conclusions:
This study shows important differences in baseline characteristics, treatment, and prognosis between patients with STEMI who were enrolled or not in a trial. To better reflect daily clinical practice, guidelines should also consider results of observational studies of unselected patients.
