Data collection from medical charts was conducted at Far Eastern Memorial Hospital, a tertiary-care hospital with around 900 beds. Patients who were admitted from January 2008 to December 2008 and diagnosed as having chronic obstructive pulmonary disease with acute exacerbation (code ICD-9: 491.21) were included in this study. The study was approved by the Institutional Review Board of the Far Eastern Memorial Hospital (FEMH:098097-3).
Demographic data, medical histories and medicine prescriptions at outpatient clinics, emergency and inpatient departments were collected. The initial presentations of acute exacerbation and laboratory data were also summarized from the records of the emergency department. The reports of pulmonary function tests within 6 months before or after the exacerbations were included while reviewing the charts. The probability of concomitant pneumonia (pAECOPD) was determined by attending physicians and reviewed by the investigator according to the radiologic findings. If there were incoherent opinions about the probability of pneumonia between the physicians and the investigator, it was determined by another independent pulmonologist. The highest daily doses and total doses of systemic steroids were calculated and recorded as the equivalent doses of prednisolone (1 mg methylprednisolone is equal to 1.25 mg prednisolone; 1 mg hydrocortisone is equal to 0.25 mg prednisolone). The subjects were divided into 3 groups according to the highest daily dose of steroids as follows: patients receiving more than 60 mg/day were designated to the HD group; those receiving less than or equal to 60 mg/day to the LD group; and those who did not receive any steroids during admission to the NIL group. The length of stay (LOS) was defined as the duration from admission to the emergency department to discharge. Hyperglycemia, psychosis, nosocomial infections and active peptic ulcer diseases were regarded as acute side effects of systemic steroids. Hyperglycemia was defined as a new blood sugar level of more than 200 mg/dl in patients without a history of diabetic mellitus, or an increment in diabetes medications. The relapse rate within 30 days was assessed by outpatient follow-up records.
