BACKGROUND: The prevalence of glucose abnormalities has been reported to be greater in hospitalized patients receiving gatifloxacin than in patients receiving ceftriaxone. OBJECTIVE: The purpose of the current study was to identify risk factors for hyperglycemia in hospitalized patients receiving gatifloxacin. METHODS: This retrospective, nested case-controlled study was conducted in 69 patients and 69 controls in 2 nonteaching hospitals in Houston, Texas. Classification and regression tree (CART) analysis was conducted to identify break points in continuous variables. Univariate and multivariate logistic regression analyses were also performed. RESULTS: A total of 138 patients (mean [SD] age, 68.3 [17.3] years; weight, 77.4 [21.1] kg; sex, women/ men, 69.6%/30.4%; race [Hispanic, 58.3%; white, 28.5%; black, 9.2%; other, 4.0%]) were identified. Patients who developed hyperglycemia (blood glucose level > 200 mg/dL) within 72 hours of gatifloxacin administration (cases) were matched 1:1 with patients who received gatifloxacin for > or = 5 days and did not experience hyperglycemia during treatment and for 3 days after the end of treatment (controls). In univariate analysis, cases were more likely than controls to have diabetes (89.9% vs 13.0%; P < 0.001), have a decreased mean (SD) creatinine clearance (CrCl) (53.8 [36.8] vs 69.2 [45.3] mL/min; P = 0.025), be located in the intensive care unit at the time of the event (20.3% vs 7.2%; P = 0.02), have received gatifloxacin doses higher than recommended based on renal function (> 200 mg/d when CrCl < 40 mL/min) (25/69 [36.2%] vs 9/69 [13.0%]), have leukocytosis (mean [SD] white blood cell [WBC] count, 15.5 [7.4] vs 12.8 [6.2] cells/mm3), and have received concomitant total parenteral nutrition (7/69 [10.1%] vs 0 [0%]; P = 0.02) and/or steroids (22/69 [31.9%] vs 8/69 [11.6%]; P = 0.008). CART analysis identified age > 65 years, albumin < 3.2 g/dL, WBC count > 14.5 cells/mm3, and CrCl < 40 mL/min to be more common in hyperglycemic patients. In logistic regression, diabetes (odds ratio [OR] = 233; 95% CI, 19-2784) and excessive dose for renal function (OR = 42; 95% CI, 2-770) were identified as independent predictors for hyperglycemia in patients receiving gatifloxacin. After controlling for diabetes, concomitant steroids (OR = 5.7; 95% CI, 1.5-21.1) and excessive dose for renal function (OR = 4.9; 95% CI, 1.2-20.4) were associated with hyperglycemia in patients receiving gatifloxacin. CONCLUSIONS: This retrospective, nested case-controlled study found that the risk of hyperglycemia increased in diabetic patients who received concomitant steroids or higher gatifloxacin doses than recommended by the manufacturer's prescribing information. The application of these data should be assessed for other fluoroquinolones used at higher doses.
BACKGROUND: The prevalence of glucose abnormalities has been reported to be greater in hospitalized patients receiving gatifloxacin than in patients receiving ceftriaxone. OBJECTIVE: The purpose of the current study was to identify risk factors for hyperglycemia in hospitalized patients receiving gatifloxacin. METHODS: This retrospective, nested case-controlled study was conducted in 69 patients and 69 controls in 2 nonteaching hospitals in Houston, Texas. Classification and regression tree (CART) analysis was conducted to identify break points in continuous variables. Univariate and multivariate logistic regression analyses were also performed. RESULTS: A total of 138 patients (mean [SD] age, 68.3 [17.3] years; weight, 77.4 [21.1] kg; sex, women/ men, 69.6%/30.4%; race [Hispanic, 58.3%; white, 28.5%; black, 9.2%; other, 4.0%]) were identified. Patients who developed hyperglycemia (blood glucose level > 200 mg/dL) within 72 hours of gatifloxacin administration (cases) were matched 1:1 with patients who received gatifloxacin for > or = 5 days and did not experience hyperglycemia during treatment and for 3 days after the end of treatment (controls). In univariate analysis, cases were more likely than controls to have diabetes (89.9% vs 13.0%; P < 0.001), have a decreased mean (SD) creatinine clearance (CrCl) (53.8 [36.8] vs 69.2 [45.3] mL/min; P = 0.025), be located in the intensive care unit at the time of the event (20.3% vs 7.2%; P = 0.02), have received gatifloxacin doses higher than recommended based on renal function (> 200 mg/d when CrCl < 40 mL/min) (25/69 [36.2%] vs 9/69 [13.0%]), have leukocytosis (mean [SD] white blood cell [WBC] count, 15.5 [7.4] vs 12.8 [6.2] cells/mm3), and have received concomitant total parenteral nutrition (7/69 [10.1%] vs 0 [0%]; P = 0.02) and/or steroids (22/69 [31.9%] vs 8/69 [11.6%]; P = 0.008). CART analysis identified age > 65 years, albumin < 3.2 g/dL, WBC count > 14.5 cells/mm3, and CrCl < 40 mL/min to be more common in hyperglycemicpatients. In logistic regression, diabetes (odds ratio [OR] = 233; 95% CI, 19-2784) and excessive dose for renal function (OR = 42; 95% CI, 2-770) were identified as independent predictors for hyperglycemia in patients receiving gatifloxacin. After controlling for diabetes, concomitant steroids (OR = 5.7; 95% CI, 1.5-21.1) and excessive dose for renal function (OR = 4.9; 95% CI, 1.2-20.4) were associated with hyperglycemia in patients receiving gatifloxacin. CONCLUSIONS: This retrospective, nested case-controlled study found that the risk of hyperglycemia increased in diabeticpatients who received concomitant steroids or higher gatifloxacin doses than recommended by the manufacturer's prescribing information. The application of these data should be assessed for other fluoroquinolones used at higher doses.