To the Editor:
We thank Jiménez-Navarro et al for their letter and we can only agree with their comments. In relation to our study,1 the authors refer to the relevance of knowing the outcomes of healthcare in real clinical practice. While agreeing with this, we want to highlight some aspects regarding the interpretation of our study that illustrate the complexity of interpreting observational studies. These difficulties have to be added to those already mentioned by Jiménez-Navarro et al.
The MASCARA study shows that, during 2004-2005, invasive procedures in acute coronary syndromes in the participating hospitals were not associated with evident clinical benefit. While this datum seems barely debatable, its interpretation is open to speculation. Compared to previous Spanish registries (PRIAMHO II2 and DESCARTES3), the MASCARA study found a striking increase in the use of drugs and percutaneous coronary intervention, and it can be assumed that this represents an important change in healthcare practice over a short period. We could form the hypothesis that the process that would have enabled good outcomes (prehospital and hospital waiting periods, correct patient selection,4 etc) would have been highly complex; even with a certainly appropriate technical execution of the intervention, the whole process would not have been sufficiently well-developed by 2004-2005. The message of this interpretation would be that to implement the invasive procedures recommended in the guidelines would not only mean carrying these out, but also appropriately modifying the healthcare management process. Our study could be a useful reference point for each center to assess to what extent this is the case at present.
On the other hand, the quality requirements that a valid registry should have are far less established than those for clinical trials, which can further hinder correct interpretation. The difficulties involved in registries that accurately reflect the situation of the participating centers are usually underestimated and barely recognized in the studies. The MASCARA study necessarily involved complex quality control that excluded 18 centers to ensure the validity of the results obtained. It is far from easy to ensure consecutive and complete inclusion in the current conditions of hospital practice. And in its absence, the resulting biases can be surprisingly high.5
These observations serve to illustrate the complexity involved in conducting and interpreting observational studies, the need for which Jiménez-Navarro et al make very clear.