Sub-optimal serum gentamicin concentrations in sickle cell disease patients utilizing the Hartford protocol

Several studies have reported that use of the Hartford nomogram in different patients' population was associated with low serum gentamicin concentrations (SGC) at different intervals or midpoints. This study was intended to determine the prevalence and predictors of SGC in patients with sickle cell disease (SCD) as another population representing low SGC while utilizing the Hartford protocol.

This retrospective observational study was carried out in a University-teaching hospital in Oman. The study was conducted from January 2005 through May 2008 and included all adult patients with SCD admitted during that time. Four-hundred and seven SGC representing 248 SCD patients were evaluated. The serum gentamicin concentration was considered sub-optimal if it was <2μg/mL (baseline of Hartford nomogram). Analyses were performed using univariate and multivariate statistical techniques.

Eighty-three percent (n=339) of SGC were sub-optimal. Multivariate analysis using logistic regression revealed that sub-optimal SGC were associated with younger patients with higher creatinine clearance. Specifically, patients who were ≤23years old were twice more likely to have sub-optimal SGC compared with those who were >23years of age (95% CI: 1·14-3·45; P=0·015). Patients with creatinine clearance of ≥200mL/min were 5·20 times more likely to have sub-optimal SGC compared with those with creatinine clearance <200mL/min (95% CI: 1·81-14·49; P=0·002). Furthermore, the logistic model also demonstrated that higher serum urea was associated with low SGC, with each one unit increase in serum urea, patients were 17% less likely to have sub-optimal SGC (95% CI: 0·72-0·96; P=0·011). Additionally, patients who were on piperacillin±tazobactam therapy given concurrently with gentamicin were 53% less likely to have sub-optimal SGC (95% CI: 0·28-0·83; P=0·009).

A majority of patients with SCD had sub-optimal SGC. The pharmacokinetic profile of such patients is apparently too variable to fit the existing Hartford protocol. The Hartford nomogram should be modified to address this issue. Otherwise, clinicians should revert to multiple daily dosing.