Background Accurate prediction for prognosis is important for hospitalized patients with chronic obstructivepulmonary disease(COPD)requiring invasive mechanic ventilation(IMV)and for their family members to makeend-of-life decisions.The response to therapy in such a patient population has rarely been investigated.The aim of thestudy was to evaluate the risk factors in these patients and investigate their response to IMV and the relationshipbetween their responses and prognosis.Methods A cohort of 138 patients with COPD requiring IMV≥12 hours for acute respiratory failure of diverseetiological factors during a 4-year period were retrospectively studied using prospectively gathered data.All variablespotentially related to hospital mortality were evaluated by univariate and multiple stepwise logistic regression analysis.Results The mean age of all patients investigated was(65.7±11.6)years and the hospital mortality was 39.9%(31.1% with COPD exacerbation).Correction of acidosis(pH≥7.30)was seen in 58 patients(69.9%)in survivors butonly 12 patients(21.8%)in nonsurvivors(P<0.05)after ventilation.Using multivariate logistic analysis,the variablesindependently associated with hospital mortality were a higher acute physiology score before intubation,lower pH valuemeasured 24 hours after the onset of ventilation and development of multiorgan dysfunction syndrome(MODS).Conclusions In COPD patients requiring IMV,the postintubation pH value can not only reflect patients’ response totreatment,but also serve as an independent determinant of hospital mortality apart from other risk factors such as ahigher preintubation APACHE Ⅱ score and development of MODS.A close correlation between the response to IMV andprognosis was proved in these patients.Chin Med J 2007;120(4):287-293 Background Accurate prediction for prognosis is important for hospitalized patients with chronic obstructive pulmonary disease (COPD) requiring invasive mechanic ventilation (IMV) and for their family members to make end-of-life decisions. The response to therapy in such a patient population has rarely been investigated The aim of the study was to evaluate the risk factors in these patients and investigate their response to IMV and the relationshipbetween their responses and prognosis. Methods A cohort of 138 patients with COPD requiring IMV ≧ 12 hours for acute respiratory failure of diverseetiological factors during a 4-year period were retrospectively studied using prospectively gathered data. All variablespotentially related to hospitalized were evaluated as univariate and multiple stepwise logistic regression analysis. Results of the mean age of all patients investigated was (65.7 ± 11.6) years and the hospital mortality was 39.9 % (31.1% with COPD exacerbation). Correction of acidosis (pH> 7.3 0) was seen in 58 patients (69.9%) in survivors butonly 12 patients (21.8%) in nonsurvivors (P <0.05) after ventilation. Using multivariate logistic analysis, the variables in dependently associated with hospital mortality were a higher acute physiology score before intubation, lower pH valuemeasured 24 hours after the onset of ventilation and development of multiorgan dysfunction syndrome (MODS) .Conclusions In COPD patients requiring IMV, the postintubation pH value can not only reflect patients’ response totreatment, but also serve as an independent determinant of hospital mortality apart from other risk factors such as ahigher preintubation APACHE II score and development of MODS. A close correlation between the response to IMV and prognosis was proved in these patients. Chin Med J 2007; 120 (4): 287-293