• Salvatore Corrao, Giuseppe Natoli, Alessandro Nobili, Pier Mannuccio Mannucci, Antonello Pietrangelo, Francesco Perticone, Christiano Argano, and RePoSI Investigators.
• Dipartimento di Promozione della Salute, Materno Infantile, Medicina Interna e Specialistica di Eccellenza "G. D'Alessandro", PROMISE, University of Palermo, 90133, Palermo, Italy. s.corrao@tiscali.it.
• Intern Emerg Med. 2020 Jun 1; 15 (4): 621-628.

AbstractIn the last 2-3 decades internists have confronted dramatic changes in the pattern of patients acutely admitted to hospital wards. Internists observed a shift from younger subjects affected by a single organ disease to more complex patients, usually older, with multiple chronic conditions, attended by different specialists, with poor integration and treated with multiple drugs. In this regard, the concept of complex patients is addressed daily in clinical practice even if there is no agreed definition of patient complexity. To try to evaluate clinical complexity different instruments have been proposed. Among these, the number of comorbidities (NoC) was considered a marker of clinical complexity. However, this instrument would not give information about the clinical relevance of each condition. On the contrary, cumulative illness rating scale (CIRS) addresses the problem calculating both CIRS severity index (CIRS-SI) and CIRS comorbidity index (CIRS-CI). In light of this, 4714 patients from the RePoSI register were retrospectively analyzed to show if CIRS assessment of comorbidity burden is different from the simple count of comorbidities in predicting the length of hospital stay (LOS) and all-cause of mortality in hospitalized elderly patients and if NoC could be a valid tool to measure patient's complexity. CIRS-SI resulted the best predictor of all-cause in-hospital mortality [OR: 2.66 (1.88-3.77)] in comparison with NoC that did not result statistically significant (p = 0.551). CIRS-SI was also the best predictor of all-cause of post-discharge mortality corrected for age and sex [OR: 2.12 (1.53-2.95)]. CIRS-SI (coefficient ± standard error: 1.23 ± 0.59; p < 0.0381) and CIRS-CI (coefficient ± standard error: 0.27 ± 0.10; p < 0.011) were strong predictors of LOS in comparison with NoC that did not result statistically significant (coefficient ± standard error: 0.04 ± 0.06 p < 0.0561). In conclusion, CIRS assessment of comorbidity burden is a better clinical tool in comparison with the simple count of comorbidities especially considering the length of hospital stay and all-cause mortality in hospitalized elderly patients.

20 keywords...

hide…