目的：分析住院病人使用注射型類固醇引起高血糖之危險因子，建立風險管理觀念，提醒醫療人員對於存在類固醇性高血糖危險因子之病患，住院期間應該規律監測血糖。
方法：收納胸腔內科、風濕科、神經內外科、皮膚科及腎臟科等住院病人，分析使用注射型類固醇後血糖變化，以病例對照研究分為類固醇引起高血糖及沒有引起高血糖兩組，以獨立樣本t 檢定 (independent sample t-test) 與卡方檢定 (chi-square test) 分析兩組病人特徵，並以多變項分析找出獨立的危險因子。
結果： 研究共收納132 人， 使用類固醇引起高血糖者有63 位， 發生率為47.72%。年紀 ≥ 65 歲 (odds ratio [OR] = 3.168; 95% confidence interval [CI] = 1.13 ~8.91, p = 0.03)、糖化血色素 (hemoglobin A1c, HbA1c) ≥ 5.7% (OR = 8.31; 95% CI = 3.51~ 19.63, p < 0.01)、糖尿病前期 (OR = 9.92; 95% CI = 4.45 ~ 22.09, p < 0.01) 及末期腎臟疾病者 estimated glomerular filtration rate (eGFR) < 15 ml/min/1.73 m2 (OR = 10.62;95% CI = 2.85 ~ 39.48, p < 0.01) 是發生類固醇性高血糖的獨立危險因子，兩組使用類固醇劑量並無統計上的差異。
結論：年紀 ≥ 65 歲、HbA1c ≥ 5.7%、糖尿病前期，以及末期腎臟疾病 (eGFR < 15 ml/min/1.73 m2) 是發生類固醇性高血糖的獨立危險因子。存在這些危險因子的病人，住院中使用類固醇應規律血糖監測，提高醫療人員臨床警覺、早期診斷及處置。

Objective: Our study aims to investigate the risk factors for hyperglycemia caused by steroid injection in non-diabetic inpatients, and to establish a risk management program for medical staffs to regularly monitor patients with the risk factors.
Methods: This study enrolled inpatients from the department of pulmonary medicine, rheumatology, neurology, dermatology and nephrology. Factors associated with the occurrence of hyperglycemia after injection of steroid were analyzed. The case-control study was divided into two groups: the steroid-induced hyperglycemia (SIH) group and the non-SIH group. Patient characteristics were compared using the independent t-test and chi-square test. Multivariate analysis was performed to identify the independent risk factors.
Results: Of the 132 enrolled subjects, 63 (47.72%) developed SIH. The independent risk factors included age ≥ 65 years (odds ratio [OR] = 3.168; 95% confidence interval [CI] = 1.13–8.91, p = 0.03), hemoglobin A1c (HbA1c) ≥ 5.7% (OR = 8.31; 95% CI = 3.51–19.63, p < 0.01), pre-diabetes (OR = 9.92; 95% CI = 4.45–22.09, p < 0.01) and end-stage renal disease estimated glomerular filtration rate (eGFR) < 15 ml/ min/1.73 m2 (OR = 10.62; 95% CI = 2.85–39.48, p < 0.01). The dose of steroid was not statistically related to the development of SIH.
Conclusions: Age ≥ 65 years, HbA1c ≥ 5.7%, pre-diabetes and end-stage renal disease were identified as the independent risk factors for SIH, regardless of the dose of steroid administered. Patients who use steroids and have these risk factors should be regularly monitored. The results of this study allow medical staffs to predict SIH, raise clinical alertness and improve early diagnosis and treatment.