Bicarbonate as a predictor of successful insulin transition in critically ill patients with diabetic ketoacidosis: A retrospective cohort study.

BACKGROUND: Treatment of diabetic ketoacidosis with intravenous insulin is effective but resource intensive. Treatment guidelines recommend transitioning to subcutaneous insulin when the anion gap closes, but transition failures due to recrudescent ketoacidosis are common despite adherence to treatment protocols following such guidance. STUDY OBJECTIVE: The primary objective of our study was to evaluate the ability of serum bicarbonate levels of ≤16 mEq/L to predict intravenous to subcutaneous transition failures among those with a normal anion gap at the time of transition. DESIGN AND SETTING: This retrospective cohort study evaluated critically ill adult patients with a primary diagnosis of diabetic ketoacidosis. Historical patient data were obtained by manual chart review. The primary outcome was transition failure, defined as the re-initiation of intravenous insulin within 24 h of transitioning to subcutaneous insulin. Odds ratios were calculated using generalized estimating equations with a logit link and weighted by standardized inverse probability weights to assess the predictive value of serum bicarbonate levels. MAIN RESULTS: The primary analysis included 93 patients with a total of 118 distinct transitions. In the adjusted analysis, patients whose anion gap had normalized but had a serum bicarbonate of ≤16 mEq/L were significantly more likely to experience a transition failure (odds ratio = 4.74, 95% confidence interval: 1.24-18.1, p = 0.02). The results of the unadjusted analysis were similar. CONCLUSIONS: In patients with a normal anion gap at the time of insulin transition, serum bicarbonate levels of ≤16 mEq/L were associated with significantly increased odds of transition failure.

Decreased serum linezolid levels in a critically ill patient receiving concomitant linezolid and rifampin.

Serious gram-positive infections present an increasingly common therapeutic dilemma. Combination antimicrobial regimens (e.g., linezolid with rifampin) aimed at improving bacterial eradication and preventing resistance are often used; however, most data supporting this treatment strategy are not from randomized controlled trials. We describe a patient with disseminated community-acquired methicillin-resistant Staphylococcus aureus infection who experienced a possible drug interaction between linezolid and rifampin that resulted in decreased serum linezolid levels. To our knowledge, this is the first published report of a possible drug interaction in a critically ill patient receiving concomitant linezolid and rifampin. Although we hypothesize that the reaction was caused by P-glycoprotein expression, further study is warranted.