Letter: the burden of co‐morbidity in the population contributes more to upper gastrointestinal bleeding than aspirin – authors' reply

Abstract

We thank Dr Crooks and colleagues for their critical appraisal regarding our recently published article.1, 2 We agree that our tool will need to be updated with new risk factors and estimations for upper gastrointestinal (GI) bleeding in aspirin users. Obviously, the new risk factors to incorporate in the calculator will need to be consistently confirmed in different studies. Those we have used now are well known, widely accepted and validated in different studies, including those based on the General Practice Research Database from the UK.3-6 Crooks and colleagues suggest that co-morbidity is an independent risk factor for upper GI bleeding and that its contribution is greater than other factors. We do not deny the importance of co-morbidity, but to incorporate this risk factor into a calculator like the one we have developed requires precise scores, in conjunction with other well-known risk factors. Furthermore, it must be understood that the well-known and widely accepted risk factors used in our calculator are applied to baseline upper GI complication rates that change with the inclusion of them and that, when estimated, have already incorporated the contribution of the co-morbidity (e.g. age). The index described by Charlson et al.7 was a measure of co-morbidity to predict death, and not complications from ASA or NSAID users. To conclude, the goal of our calculator was to provide indirect estimates of the risks and benefits for patients who take aspirin, and to provide recommendations for clinicians. As we mention in our article, those recommendations are in agreement with current guidelines for both cardiovascular (CV) prevention and GI risk reduction, as it was shown in the various patient's profiles provided. We agree that future work with this tool is needed to validate and improve its capacity to provide accurate prognostic value at both GI and CV levels. The data we have reported here are the first step. The authors' declarations of personal and financial interests are unchanged from those in the original article.2