We have read with interest the letter published regarding our article entitled The Importance of Dependence in the Global Assessment of Hospitalized Patients.1,2 We thank the authors for their comments on our research and agree with the main message of their letter, as it serves to reinforce the importance of dependence in performing activities (both basic and instrumental) in the prediction of mortality in patients with chronic diseases. We have observed that, independently or in addition to the many other clinical, demographic, or functional variables usually included in the scores used in clinical practice, variables related to dependence improve these scores and provide prognostic information.1–3

We would like to make a couple of comments on this communication. First, Briongos-Figuero et al.1 comment on our surprise that dependence for performing activities was an independent factor for mortality, with a higher capacity for prediction than other clinical variables in patients admitted for COPD exacerbation. However, in the introduction to the article we said that what surprised us was the scant evidence available in this field in COPD, unlike in other chronic diseases, since the results of our study were far from a surprise. In the SocioCOPD cohort itself, on which we carried out the mortality study, dependence for performing basic activities determined by non-specific COPD indices, such as Barthel, not only influences this event, but is also one of the variables that best predicts short- and long-term readmissions for any cause after a severe exacerbation of COPD4. Indeed, this is one of the alternative hypotheses of the initial research project. What surprised us, and still does, is that so little emphasis is placed on this area in the clinical practice guidelines on COPD.

Secondly, the authors refer to an interesting study that they published recently, and point out the limited value of comorbidities (determined by the Charlson index) when adjusted for disability in the prediction of mortality,3 an outcome that differs from our findings. In all the predictive models explored in the SocioCOPD cohort, the number of comorbidities measured by this index did have a predictive capacity, although the power was less than for dependence for both basic and instrumental activities.2 However, we would like to emphasize that these 2 studies are not at all comparable, as the mean upper age of our correspondents’ series was almost 20 years older than ours, they had many more comorbidities, and, in particular, less than 20% of their patients were identified as having “chronic respiratory disease”.3 In a large COPD population5 with a high comorbidity burden and age closer to that included in the study of Briongos-Figuero et al.,3 reported by a working group of the Spanish Society of Internal Medicine, the 2 factors that carried the greatest weight for predicting mortality were comorbidity and disability.

Therefore, we also believe that it is important to continue research in this line, as it does not appear to be exclusive to elderly patients with multiple diseases, but also affects to a large extent younger subjects with specific predominant diseases, such as COPD in this case.