Background and importance Continuous infusion of high dose piperacillin/tazobactam (16/2 g in 264 mL NaCl 0.9%) has been included in the UZ Leuven outpatient parenteral antimicrobial therapy (OPAT) protocol. Elastomeric pumps (Infusor LV10, Baxter) were selected as the drug delivery device, as the patient’s mobility and comfort are maintained. Unfortunately, incomplete infusions after 24 hours were observed, related to a reduced flow rate. A mean daily residual volume of 50 mL, corresponding to a dose of 3/0.38 g piperacillin/tazobactam, was detected, resulting in substantial underdosing with the risk of treatment failure.

Aim and objectives To analyse two hypotheses: a reduced flow rate could be the result of particulate formation of piperacillin dimers due to the absence of stabilising excipients (hypothesis 1) or a result of high viscosity (hypothesis 2).

Material and methods Hypothesis 1: particulate formation was detected by comparing the flow rate of tazocillin (with stabilising excipients) versus generic piperacillin/tazobactam (without this excipients), by measuring light absorbance (600 nm) by spectrophotometry and by measuring total piperacillin content at different concentrations after storage for 24 hours at 33°C.

Hypothesis 2: the effect of concentration on the density and viscosity at 33°C was measured. Additionally, the relation between viscosity and flow rate was evaluated.

Results Hypothesis 1: no difference was observed in the flow rate between Tazocillin and generic piperacillin/tazobactam. No difference was observed in absorbance between Tazocillin and generic piperacillin/tazobactam, and no difference was observed in absorbance between piperacillin/tazobactam and a blank. Generic piperacillin/tazobactam seemed to be stable for 24 hours at 33°C.

Hypothesis 2: a linear relationship was observed between concentration and viscosity. An inverted linear relationship was observed between viscosity and flow rate of piperacillin/tazobactam solutions.

Conclusion and relevance The in vitro experiments suggest that the reduced flow rate is a result of high viscosity, related to the concentration of piperacillin/tazobactam. As it is impossible to lower the concentration, the final volume of the solution should be adjusted. Before being used in clinical practice for OPAT, this mode of administration will first be validated in five patients during hospitalisation. In general, healthcare teams need to be aware of factors which may lead to longer flow durations with these infusion devices.

References and/or acknowledgements No conflict of interest.