[Observer variation in diagnostic procedures related to cardiothoracic surgery].

The results reliability is a predominant condition in the validity of clinical research. The observer agreement determination is a useful procedure in the confirmation of the obtained results in clinical research. The present work is a presentation of the basis about observer agreement and a review of national publications in cardiology, pneumology and cardiothoracic surgery between january 1980 and july 1993, where we found just two publications with the application of observer agreement determination in echocardiography, electrocardiography, microscopy and hemodynamics. It was concluded that is necessary the increase in determination of observer agreement and inform this results in our publications, excluding the variability attributed to patients, diagnostics equipment or the circumstances.