LAUSR

Multimorbidity, is certainly one of the most relevant and challenging issues in medicine, for both health-care professionals and policymakers [1]. Despite this, the research advances in multimorbidity are fraught with difficulties for many reasons, mainly related to the lack of knowledge about its aetiology, epidemiology, risk factors, and, most importantly, best treatment and care [2]. Ageing is certainly a key determinant of the multimorbidity pandemic, as more than 90% of individuals aged 65 years and greater suffer from this condition [3]. However, multimorbidity is not only confined to the elderly population, and many adult patients are multimorbid in terms of absolute numbers [4]. Most of the available evidence about multimorbidity derives from the primary care setting [2]. Hence, the paper by Aubert et al. [5] published in this issue of the journal, that focuses on the patterns of multimorbidity in 11 hospitals from three different countries, is more than welcome, considering that inpatients may be burdened by more intensive and resource-consuming needs. The aim of the study was to identify less obvious patterns of multimorbidity along with the relative proportions of chronic and acute conditions, given that these latter do not pertain to the classic definition of multimorbidity [1]. Starting from these premises, the authors performed a retrospective, administrative database-driven study in which they described and quantified the most common combinations of comorbidities. The data of a large cohort of 147,806 inpatients (median age 64 years, 52% males) were retrieved from electronic medical records, using International Classification of Diseases (ICD)-9 or -10 codes. It was found that 86% of patients suffered from multimorbidity, with a median of five chronic diseases per patient, and the most common comorbidities were chronic heart and kidney diseases. The Chronic Condition Indicator was used to categorise chronic vs acute conditions, classifying more than 14,000 ICD codes into 285 categories, considering the involved body systems [6, 7]. Chronic conditions turned out to be twice more common than the acute ones. The topic covered by this study is certainly of interest to clinicians, particularly for those involved with internal medicine, even though the patients included in the study were enrolled in various medical speciality wards, with the only exception of surgical, and this might have determined an overor underestimation of some disease patterns. Their demographic results did not differ from those of other studies focusing on multimorbidity in hospitalised patients [8–13], except for the median age that was unusually lower, with no patients older than 80 years. Regarding patterns of multimorbidity, the fact that the study was based on administrative data provides a fairly static and a bit unfocused picture. Studies on administrative data lead to a classification of too broad categories, and it is sometimes hard to exploit these data from the point of view of a clinician. When multicentric, administrative results may be undermined by differences in coding used in different hospitals. They do not provide any information on patients’ outcomes, such as inhospital and long-term mortality, they do not allow to evaluate the type of discharge (home, nursing house, others), to prioritise an index disease and, therefore, to verify causal associations. In other words, no disease outcome can be predicted with this information, despite that the database used in the study by Aubert et al. [5] was firstly set up for studying hospital readmissions. Although it was correct to differentiate acute from chronic conditions, the mere description of their proportion, instead of how their presence could have negatively impacted on final results, does not provide any meaningful clinical information. Similarly, risk factors were excluded, and this makes it difficult to ascertain possible * Gino Roberto Corazza [email protected]