We have read the article by Caravaca et al. entitled “Sudden death in patients with advanced chronic kidney disease”1 and we would first like to congratulate the authors.

It is a subject of great interest among both kidney failure patients and the general population, which is a cause for social concern and requires a complex approach.

We believe its complexity is based on several factors. The first is the definition of sudden death (SD) itself. The one used by the authors is the most widely accepted and includes natural death occurring within an hour of the onset of symptoms, although they also include unexplained death during sleep. The latter definition can be assimilated to those that propose other time intervals (2, 6 and 24h) between the onset of symptoms and death for specific circumstances such as unwitnessed death, which reduce the probability of heart-related death the longer the time that elapses between death and the onset of symptoms.2 This variability in the time period in question is one of the factors that contribute to the differences found in the prevalence of SD in different studies. Caravaca et al. found an annual incidence of 1.8%, not including that of death during sleep. In a study conducted at our hospital on a dialysis population, the annual incidence was 1.7% when we considered an interval of one hour, and 2.9% when the interval was extended to 24h.3 The lack of autopsies in the Caravaca et al. study, as in ours (something that was particularly inexplicable for the reviewers of the American journal in which it was published), poses many obstacles when it comes to establishing the cause of death. In a Japanese study in which autopsy was performed on 81.4% of patients who had SD, strokes and aortic dissection accounted for more than 45% of the causes of death.4

Another aspect to be taken into account when analysing SD is the need to search for those factors that identify the population at greatest risk in order to come up with targeted and profitable prevention strategies. Caravaca et al. found through Cox regression analysis that greater age and greater comorbidity were associated with a higher probability of SD, and the use of antiplatelet therapy had a protective effect. In our study,3 with the same analysis, we identified a combined variable that included a documented history of coronary disease and electrocardiographic abnormalities (abnormal Q waves, subendocardial lesion, negative T-wave and QRS >120ms) which clearly identified two risk groups. Patients who had none of the aforementioned findings had almost the same probability of SD as the general population; however, those who had one of them had a probability of SD similar to that shown by patients in the general population with left ventricular dysfunction and who have been included in studies demonstrating the efficacy of the automatic implantable defibrillator.5,6 The lack of availability of electrocardiography records in the Caravaca et al. study, which the authors point out as a limitation, makes it impossible to verify whether the findings of our study can be extrapolated to a renal insufficiency population that is not on dialysis.

The study by Caravaca et al. provides information and raises awareness of the importance of the problem and urges the scientific community and the health authorities to establish measures. The first of these is to give a precise definition of SD in all registers in order to reduce variability between centres. The need to perform autopsies in cases of unexpected, unexplained deaths, even in patients with severe chronic diseases, is undoubtedly an aspect that should be given greater attention.

Finally, the identification of risk subgroups does not necessarily mean that we can extrapolate the prevention strategies that have been shown to be effective in the general population. The automatic implantable defibrillator is less effective in patients with kidney failure,7 and even something as basic as having external defibrillators in dialysis units also has a limited efficacy.8

In conclusion, we again congratulate Caravaca et al. for their contributions to an important issue that has so far not been given the attention it deserves.