BACKGROUND: Gatifloxacin, until recently one of the most commonly prescribed antibiotics, has been shown to produce hypoglycemia. METHODS: To further examine the effects of Gatifloxicin (G) on blood glucose (BS), we conducted a retrospective chart review on 264 inpatients, examining for both hypoglycemia and hyperglycemia, comparing G with another quinolone, Ciproflaxin (C), and nonquinolone, Ceftriaxone (R). RESULTS: We found that of 292 patient encounters, 28 hypoglycemia and 48 hyperglycemic events occurred. Patients given G were 5 times as likely to become hypoglycemic as C (P < 0.01) and 9 times as likely as those given R (P < 0.02). Patients given G were 5.6 times more likely to develop hypoglycemia (P < 0.001) than the combined group, R+C. Conversely, patients treated with G were 3.8 times as likely to become hyperglycemic as those give C (P < 0.01) and 9.8 times as those given R (P < 0.01). With C and R combined, those given G were 5.2 times as likely to develop hyperglycemia (P < 0.01). Looking at patient encounters where G was given, we found that having preexisting diabetes mellitus (DM) was positively associated with hypoglycemia (21/144, P < 0.001). Steroid use (P < 0.05) and being in the ICU (P < 0.01) were also positively associated with hyperglycemia (38/144, P < 0.01). CONCLUSIONS: In summary, G was clearly associated with both hypoglycemia and hyperglycemia compared with C and R. The risk of hyperglycemia increased in the presence of DM, steroid use, and "sick enough" to be in the intensive care unit.
BACKGROUND:Gatifloxacin, until recently one of the most commonly prescribed antibiotics, has been shown to produce hypoglycemia. METHODS: To further examine the effects of Gatifloxicin (G) on blood glucose (BS), we conducted a retrospective chart review on 264 inpatients, examining for both hypoglycemia and hyperglycemia, comparing G with another quinolone, Ciproflaxin (C), and nonquinolone, Ceftriaxone (R). RESULTS: We found that of 292 patient encounters, 28 hypoglycemia and 48 hyperglycemic events occurred. Patients given G were 5 times as likely to become hypoglycemic as C (P < 0.01) and 9 times as likely as those given R (P < 0.02). Patients given G were 5.6 times more likely to develop hypoglycemia (P < 0.001) than the combined group, R+C. Conversely, patients treated with G were 3.8 times as likely to become hyperglycemic as those give C (P < 0.01) and 9.8 times as those given R (P < 0.01). With C and R combined, those given G were 5.2 times as likely to develop hyperglycemia (P < 0.01). Looking at patient encounters where G was given, we found that having preexisting diabetes mellitus (DM) was positively associated with hypoglycemia (21/144, P < 0.001). Steroid use (P < 0.05) and being in the ICU (P < 0.01) were also positively associated with hyperglycemia (38/144, P < 0.01). CONCLUSIONS: In summary, G was clearly associated with both hypoglycemia and hyperglycemia compared with C and R. The risk of hyperglycemia increased in the presence of DM, steroid use, and "sick enough" to be in the intensive care unit.