Sedatives, analgesics and antipsychotics in tracheostomised ICU patients - Is less more?

BACKGROUND

Sedation and anaesthesia are used universally to facilitate mechanical ventilation - with larger cumulative doses being used in those with prolonged ventilation. Transitioning from an endotracheal to a tracheostomy tube enables the depth of sedation to be reduced. Early use of speaking valves with tracheostomised patients has become routine in some intensive care units (ICU). The return of verbal communication has been observed to improve ease of patient care and increase patient and family engagement, with a perceived reduction in patient agitation.

OBJECTIVES

To investigate the potential impact of speaking valve (SV) use on requirements of sedatives, analgesics and antipsychotics in ICU patients with a tracheostomy.

METHODS

A retrospective data audit was undertaken for all tracheostomised patients in a cardio-respiratory ICU from 2011 to 2014. Use of sedative, analgesic and antipsychotic drugs was captured for endotracheal tube, tracheostomy and SV periods, including patient demographics, disease specifics and severity. Stratified Cox regression analysis was performed to determine the effects of SV on drug dosage.

RESULTS

Of 257 patients, 144 (56%) received an SV. Use of an SV was associated with reduced risk of being in the upper quartile of daily dosage of analgesics (HR: 0.6; 95% CI: 0.5-0.8; p < 0.001). In the final adjusted multivariable model, analgesic dose was additionally associated with age, and attendance to operating theatre during ICU. Sedative dose was associated with age, gender and SOFA score. Antipsychotic dose was associated with gender (less likely in females: HR 0.6, 95% CI: 0.4-0.8), age and APACHE score.

CONCLUSIONS

There was significantly less analgesic used in patients with an SV compared to those without. However, SV use in patients with tracheostomy was not found to be associated with reduced dose of sedatives or antipsychotics, despite the clinical impression. Future prospective studies are needed to more adequately investigate the association between drugs and patients' ability to verbally participate in their care.

Incremental research approach to describing the pharmacokinetics of ciprofloxacin during extracorporeal membrane oxygenation

BACKGROUND

Significant interactions between drugs, extracorporeal membrane oxygenation (ECMO) circuits and critical illness may affect the pharmacokinetic properties of antibiotics in critically ill patients receiving ECMO.

OBJECTIVE

To describe the pharmacokinetic properties of ciprofloxacin during ECMO by integrating pre-clinical findings (ie, ex vivo and in vivo ovine models) to a critically ill patient.

DESIGN, PARTICIPANTS AND INTERVENTION

An ex vivo model of an ECMO circuit was used to describe ciprofloxacin concentration changes over 24 hours. An in vivo ovine model of ECMO was used to describe the population pharmacokinetic properties of ciprofloxacin in three different groups of sheep, and to investigate sources of pharmacokinetic variability. In the final phase, data from a 39-year-old critically ill man was used to validate the findings from the ovine pharmacokinetic model.

RESULTS

In the ex vivo model of ECMO circuits, the median concentrations of ciprofloxacin at baseline and at 24 hours after ciprofloxacin infusion were similar. The time course of ciprofloxacin in the in vivo ovine on ECMO model was adequately described by a two-compartment model. The final population primary parameter mean estimates were: clearance (CL), 0.21 L/kg/h (SD, 0.09 L/kg/h) and volume of distribution (V_d), 0.84 L/kg (SD, 0.12 L/kg). In the critically ill ECMO patient, the primary pharmacokinetic parameter estimates were: CL, 0.15 L/kg/h and V_d, 0.99 L/kg.

CONCLUSIONS

We provide preliminary evidence that ciprofloxacin dosing in ECMO patients should remain in line with the recommended dosing strategies for critically ill patients not receiving ECMO.

Sequestration of drugs in the circuit may lead to therapeutic failure during extracorporeal membrane oxygenation

Introduction

Extracorporeal membrane oxygenation (ECMO) is a supportive therapy, with its success dependent on effective drug therapy that reverses the pathology and/or normalizes physiology. However, the circuit that sustains life can also sequester life-saving drugs, thereby compromising the role of ECMO as a temporary support device. This ex vivo study was designed to determine the degree of sequestration of commonly used antibiotics, sedatives and analgesics in ECMO circuits.

Methods

Four identical ECMO circuits were set up as per the standard protocol for adult patients on ECMO. The circuits were primed with crystalloid and albumin, followed by fresh human whole blood, and were maintained at a physiological pH and temperature for 24 hours. After baseline sampling, fentanyl, morphine, midazolam, meropenem and vancomycin were injected into the circuit at therapeutic concentrations. Equivalent doses of these drugs were also injected into four polyvinylchloride jars containing fresh human whole blood for drug stability testing. Serial blood samples were collected from the ECMO circuits and the controls over 24 hours and the concentrations of the study drugs were quantified using validated assays.

Results

Four hundred samples were analyzed. All study drugs, except meropenem, were chemically stable. The average drug recoveries from the ECMO circuits and the controls at 24 hours relative to baseline, respectively, were fentanyl 3% and 82%, morphine 103% and 97%, midazolam 13% and 100%, meropenem 20% and 42%, vancomycin 90% and 99%. There was a significant loss of fentanyl (p = 0.0005), midazolam (p = 0.01) and meropenem (p = 0.006) in the ECMO circuit at 24 hours. There was no significant circuit loss of vancomycin at 24 hours (p = 0.26).

Conclusions

Sequestration of drugs in the circuit has implications on both the choice and dosing of some drugs prescribed during ECMO. Sequestration of lipophilic drugs such as fentanyl and midazolam appears significant and may in part explain the increased dosing requirements of these drugs during ECMO. Meropenem sequestration is also problematic and these data support a more frequent administration during ECMO.

Increased sedation requirements in patients receiving extracorporeal membrane oxygenation for respiratory and cardiorespiratory failure

Critically ill patients receiving extracorporeal membrane oxygenation (ECMO) are often noted to have increased sedation requirements. However, data related to sedation in this complex group of patients is limited. The aim of our study was to characterise the sedation requirements in adult patients receiving ECMO for cardiorespiratory failure. A retrospective chart review was performed to collect sedation data for 30 consecutive patients who received venovenous or venoarterial ECMO between April 2009 and March 2011. To test for a difference in doses over time we used a regression model. The dose of midazolam received on ECMO support increased by an average of 18 mg per day (95% confidence interval 8, 29 mg, P=0.001), while the dose of morphine increased by 29 mg per day (95% confidence interval 4, 53 mg, P=0.021) The venovenous group received a daily midazolam dose that was 157 mg higher than the venoarterial group (95% confidence interval 53, 261 mg, P=0.005). We did not observe any significant increase in fentanyl doses over time (95% confidence interval 1269, 4337 µg, P=0.94). There is a significant increase in dose requirement for morphine and midazolam during ECMO. Patients on venovenous ECMO received higher sedative doses as compared to patients on venoarterial ECMO. Future research should focus on mechanisms behind these changes and also identify drugs that are most suitable for sedation during ECMO.

Development of simulated and ovine models of extracorporeal life support to improve understanding of circuit-host interactions

BACKGROUND

Extracorporeal life support (ECLS) is a lifesaving technology that is being increasingly used in patients with severe cardiorespiratory failure. However, ECLS is not without risks. The biosynthetic interface between the patient and the circuit can significantly alter inflammation, coagulation, pharmacokinetics and disposition of trace elements. The relative contributions of the pump, disease and patient in propagating these alterations are difficult to quantify in critically ill patients with multiple organ failure.

OBJECTIVE

To design a model where the relevance of individual components could be assessed, in isolation and in combination.

DESIGN AND SUBJECTS

Four ECLS models were developed and tested - an in-vitro simulated ECLS circuit; and ECLS in healthy sheep, sheep with acute lung injury (ALI), and sheep with ALI together with transfusion of old or new blood.

MAIN OUTCOME MEASURES

Successful design of in-vitro and in-vivo models.

RESULTS

We successfully conducted multiple experiments in the simulated circuits and ECLS runs in healthy and ALI sheep. We obtained preliminary data on inflammation, coagulation, histology, pharmacokinetics and trace element disposition during ECLS.

CONCLUSIONS

The establishment of in-vitro and in-vivo models provides a powerful means for enhancing knowledge of the pathophysiology associated with ECLS and identification of key factors likely to influence patient outcomes. A clearer description of the contribution of disease and therapeutic interventions may allow improved design of equipment, membranes, medicines and physiological goals for improved patient care.