Assessment of In-Hospital Mortality and the Need for Mechanical Ventilation in Acute Exacerbations of COPD: A 2-Years Prospective In-Hospital Observational Study

Abstract

PURPOSE: Hospitalization of acute exacerbations of COPD patients is sometimes associated with a high risk of inhospital mortality and need for mechanical ventilation. Meanwhile, physicians lack a risk-stratification instrument to easily predict the site of care and the need of hospitalization or ICU admission for high-risk patients. Objectives: to assess different COPD scoring systems to accurately predict in-hospital mortality of AECOPD and the need for MV, trying to optimize the treatment plan and stratify the patients' site of care. METHODS: We prospectively recruited 250 patients with AECOPD in chest department, Assiut university hospital from December 2012 to December 2014. This is a tertiary care hospital in upper Egypt. We compared 4 scoring systems in predicting in-hospital mortality and the need for MV. These were the CAUDA-70(Confusion, Acidosis less than 7.35, Urea >7mmol/l, MRC>4, Albumin less than 35g/l and age >70 years), BAP-65 (Urea >25 mg/dl, Alert mental status, Pulse rate >110/m, 65 Years of age), CURB-65 (Confusion, Urea >7mmol/l, RR >30/m, Hypotension, 65 Years age, and CAPS (The COPD and Asthma physiological score from 0-100 points) scores. RESULTS: the mean age ± SD for our recruited 250 patients was 64.59± 8.45 years. The majority were males (76.4%). The total in-hospital mortality was 17.2% and the need for MV was 54.4%. The mortality and the need for MV increase with escalating scores. The AUROC for in-hospital mortality were CAUDA-70=0.82, CURB-65=0.79, BAP-65=0.77 and CAPS = 0.73 while for need of MV were 0.79, 0.69, 0.75, 0.74, respectively. The cut off values for predicting in hospital mortality were > 4, >2, >3, >33 respectively while for predicting the need for MV was >3, >1, >2, and >27 respectively. CONCLUSIONS: The CAUDA-70 score outperforms all other scores in predicting both in-hospital mortality and the need for MV in AECOPD patients. Being a simple 6-point score based on clinical variables obtaied easily at presentation, the CAUDA-70 gives a very good option to be used by clinicians for risk stratification for AECOPD and the need for MV. the study limitation was that all the recruited patients were hospitalized by AECOPD and future studies should include less severe cases that could be managed at home. CLINICAL IMPLICATIONS: At a cut off value of more that 3 poins in the CAUDA-70 score, the physician can refere the AECOPD patient for MV as the mortality rate was significantly higher in this group of patients.