Methods: Retrospective cohort study of pts receiving 12g/day (d) of NAF or OXA for at least 24h between 4/2012 and 7/2013. Pts were followed until treatment discontinuation (DC) or discharge from the hospital. Those with pre-existing hypokalemia (K <3.4 mmol/L), liver disease (LFTs > 3x normal), or end-stage renal disease (requiring HD) were excluded.
Results: Of 224 pts included, 64% were male; median age was 64 yrs (range: 19 – 90). 71% and 29% received NAF and OXA, respectively. Charlson comorbidity scores, baseline serum creatinine (SCr), and receipt of concomitant medications did not differ between groups. 42% developed hypokalemia (defined as K ≤3.3) at a median of 4d (1 – 15) from start of therapy. Rates of hypokalemia were 51% and 17% for pts receiving NAF and OXA, respectively (P<0.0001). When defined as ≥ 0.5 mmol/L decrease in K, rates were 56% and 34%, respectively (P=0.005). Severe hypokalemia (K ≤ 2.9) was also more common among pts receiving NAF (20%) compared to OXA (3%, P=0.0008); by multivariable logistic regression, NAF was a predictor of K ≤2.9 (OR=6.74, 95% CI=1.45-31.2, P=0.02). There were no differences in rates of abnormal LFTs (P=0.45), hypersensitivity reactions (P=0.28), or leukopenia (P=0.49); however, 18% and 6% of pts developed acute kidney injury (AKI; increased SCr ≥ 1.5x baseline) after receipt of NAF and OXA, respectively (P=0.03). Adverse events led to drug DC in 18% receiving NAF vs 2% receiving OXA (P=0.0004). DC was secondary to AKI (50%), hypokalemia (40%), leukopenia (7%), and hypernatremia (3%). DC due to AKI and hypokalemia occurred exclusively among pts receiving NAF. Hypokalemia resolved in 86% of pts switched from NAF to OXA. Hospital re-admission within 30d did not vary by drug (28% vs 22%; P=0.38).
Conclusion: Comparing similar patient populations, those receiving NAF (compared to OXA) experienced higher rates of hypokalemia and AKI, which lead to treatment DC in 18%. Prospective studies validating these findings are warranted. Until such time, we recommend OXA as the preferred anti-staphylococcal penicillin.
J. A. Viehman,
K. Sheridan, None
K. Byers, None
P. He, None
B. Falcione, Hospira: Investigator , Co-investigator for an unrelated research study for which financial support was provided to the University of Pittsburgh
R. K. Shields, Merck: Investigator , Research grant
Astellas: Investigator , Research grant