[Difficulties of assessing postoperative pulmonary embolisms. Risks of meta-analysis from clinical trials]

During the 1970's, several large clinical trials have shown low dose subcutaneous heparin to be effective in the prevention of post-operative lethal pulmonary embolism. These claims have been widely accepted mainly because the selected main criterion was effectively pulmonary embolism. To-date, this criterion appears to be poorly reliable clinically and ethically unacceptable due to lethal risk. In practically all recent trials with low molecular weight heparin fractions in this indication, the main criterion has been prevention of deep vein thrombosis and not of pulmonary embolism. This is why, among other reasons exposed in this study, it would be hazardous to perform any meta-analysis related to pulmonary embolism from trials which have not been designed for this purpose.