Prediction of in-hospital mortality of patients with SARS-CoV-2 infection by comorbidity indexes: an Italian internal medicine single center study

A. De Giorgi, F. Fabbian, S. Greco, E. Di Simone, R. De Giorgio, A. Passaro, G. Zuliani, R. Manfredini; the OUTcome and COMorbidity Evaluation of INTernal MEDicine COVID19 (OUTCOME-INTMED-COV19) Study Collaborators

COVID-19 Internal Medicine Units (Clinica Medica, Medicina Interna Universitaria, Medicina Ospedaliera 2), Azienda Ospedaliero-Universitaria “S. Anna”, Ferrara, Italy. f.fabbian@ospfe.it

---

• Infectious Diseases

OBJECTIVE: Clinical outcomes in patients hospitalized for severe acute respiratory syndrome due to coronavirus (SARS-CoV-2) infection seems to be closely related with burden of comorbidities. A comorbidity score could help in clinical stratification of patients admitted to internal medicine units. Our aim was to assess a novel modified Elixhauser index (mEi) and the Charlson Comorbidity Index (CCI) for predicting in-hospital mortality (IHM) in internal medicine patients with SARS-CoV-2 infection.

PATIENTS AND METHODS: This single-center retrospective study enrolled all consecutive patients discharged from internal medicine unit with confirmed SARS-CoV-2 infection. Both the mEi and CCI were easily calculated from administrative data. Comorbidity scores were tested using receiver operating characteristic (ROC) analysis, and the respective area under the curve (AUC).

RESULTS: The total sample consisted of 151 individuals, and 30 (19.9%) died during their hospital stay. Deceased subjects were older (82.8±10.8 vs. 63.3±18.1 years; p<0.001) and had a higher burden of comorbidities: the mEi and CCI were 29.9±11 vs. 8.8±9.2 and 4.6±2.6 vs. 1.2±2 (p<0.001), respectively. Only the mEi was independently associated with IHM (OR 1.173), and ROC curves analysis showed that the AUCs were 0.863 and 0.918 for the CCI and for mEi, respectively.

CONCLUSIONS: In patients admitted to internal medicine wards with SARS-CoV-2 infection, the mEi showed a better performance in predicting IHM than CCI.

Free PDF Download