Abstract
Background: The fragility index (FI) is calculated by iteratively changing one outcome “event” to a “non-event” within a trial until the associated p-value exceeds 0.05. Purpose: To investigate the FI and fragility quotient (FQ) of trial endpoints referenced in the ACCF/AHA/SCAI guidelines in the management of ST-elevation myocardial infarctions. Secondarily, we assess the post-hoc power and risk of bias for these specific outcomes and whether differences exist between adequately and inadequately powered studies on fragility measures. Basic procedures: All citations referenced in the guideline were screened for inclusion criteria. The FI and FQ for all included trials were then calculated. The Cochrane ‘risk of bias’ Tool 2.0 was used to evaluate the likelihood and sources of bias in the included trials. Main findings: Forty-two randomized controlled trials were included for assessment. The median FI was 10 with a FQ of 0.0055. Seven trials were at a high risk of bias, all due to bias in the randomization process. Fifteen trials were found to be underpowered. Adequately powered studies had higher FIs and FQs compared to underpowered studies. Principal conclusions: Our findings support the use of FI and FQ analyses with power analyses in future methodology of randomized control trials. With understanding and reporting of FI and FQ, evidence of studies can be readily available and quickly eliminate some readers' concern for possible study limitations.
Original language | English |
---|---|
Pages (from-to) | 2229-2238 |
Number of pages | 10 |
Journal | American Journal of Emergency Medicine |
Volume | 37 |
Issue number | 12 |
DOIs | |
State | Published - Dec 2019 |
Keywords
- Practice guideline
- Randomized controlled trials
- Research methodology
- ST-elevation myocardial infarction
- STEMI
- Statistics