WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT: The methodological and ethical value of equivalence and non-inferiority trials is questioned. These studies are still increasingly used in drug evaluation and accepted by the scientific community and the regulatory authorities. WHAT THIS STUDY ADDS: By applying the hypothesis of an equivalence trial of saruplase and streptokinase (SK) we proved that the no-thrombolysis approach in the GISSI trial was equivalent to thrombolysis with SK, i.e. death rates in patients given SK or not were similar enough to consider the no-thrombolysis regimen equivalent to thrombolytic treatment. These data illustrate the unreliability of equivalence trials, which can even disprove consolidated clinical evidence such as the efficacy of thrombolysis in acute myocardial infarction. Equivalence trials should not be considered an option by the scientific community and should not be accepted as a basis for marketing authorization by the regulatory authorities. AIMS To assess the reliability of equivalence trials we tested whether the no-thrombolysis approach was equivalent to thrombolysis with streptokinase (SK) in acute myocardial infarction. METHODS: We applied the hypothesis of an equivalence trial of a recombinant plasminogen activator and SK to the GISSI-1 control group. RESULTS: In at least one of three subsets randomly extracted from the GISSI database the equivalence criterion was satisfied, i.e. death rates in patients given SK or not were similar enough to consider the no-thrombolysis regimen equivalent to thrombolytic treatment. Two-thirds of 100 replications of the sampling gave this result. CONCLUSIONS: These findings suggest the unreliability of equivalence trials, which should neither be adopted by the scientific community nor accepted by the regulatory authorities.
WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT: The methodological and ethical value of equivalence and non-inferiority trials is questioned. These studies are still increasingly used in drug evaluation and accepted by the scientific community and the regulatory authorities. WHAT THIS STUDY ADDS: By applying the hypothesis of an equivalence trial of saruplase and streptokinase (SK) we proved that the no-thrombolysis approach in the GISSI trial was equivalent to thrombolysis with SK, i.e. death rates in patients given SK or not were similar enough to consider the no-thrombolysis regimen equivalent to thrombolytic treatment. These data illustrate the unreliability of equivalence trials, which can even disprove consolidated clinical evidence such as the efficacy of thrombolysis in acute myocardial infarction. Equivalence trials should not be considered an option by the scientific community and should not be accepted as a basis for marketing authorization by the regulatory authorities. AIMS To assess the reliability of equivalence trials we tested whether the no-thrombolysis approach was equivalent to thrombolysis with streptokinase (SK) in acute myocardial infarction. METHODS: We applied the hypothesis of an equivalence trial of a recombinant plasminogen activator and SK to the GISSI-1 control group. RESULTS: In at least one of three subsets randomly extracted from the GISSI database the equivalence criterion was satisfied, i.e. death rates in patients given SK or not were similar enough to consider the no-thrombolysis regimen equivalent to thrombolytic treatment. Two-thirds of 100 replications of the sampling gave this result. CONCLUSIONS: These findings suggest the unreliability of equivalence trials, which should neither be adopted by the scientific community nor accepted by the regulatory authorities.
Authors: U Tebbe; R Michels; J Adgey; J Boland; A Caspi; B Charbonnier; J Windeler; H Barth; R Groves; G R Hopkins; W Fennell; A Betriu; M Ruda; J Mlczoch Journal: J Am Coll Cardiol Date: 1998-03-01 Impact factor: 24.094