#Faculty of Medicine, The University of British Columbia, Vancouver, BC, Canada. How low .... Thank you to the Royal Columbian Hospital operating room staff.
How low can you go?
The Design of a Respiratory Rate Mobile Examining pharmacokinetically defined minimum safety bounds for propofol Application during closed-loop control of anesthesia *, Walter Karlen†, Heng Gan*, Michelle Chiu#, Christian L Petersen*, Guy A Dustin Dunsmuir Sonia M Brodie*, Klaske van Heusden†, Nicholas West*, †Matthias Görges†, Richard*Merchant*, Guy A Dumont†, J Mark Ansermino* Dumont , J Mark Ansermino *Anesthesiology, Pharmacology & Therapeutics, and †Electrical & Computer Engineering, * † Departments of Departments of Anesthesiology, Pharmacology & Therapeutics and Electrical and Computer Engineering, and #Faculty of Medicine, The University of British Columbia, Canada BC, Canada The University of British Vancouver, Columbia, BC, Vancouver,
INTRODUCTION »
In closed-loop control of anesthesia • drug infusion rates are automatically adjusted using continuous feedback from a measure of clinical effect1 • safety constraints based on the drug’s known therapeutic window are required to avoid under/overdosing2
»
For closed-loop propofol administration, we limit the predicted effect site concentration (Ce) to [1.5 mcg/ml - 8 mcg/ml], based on the Schnider pharmacokinetic (PK) model3
»
A recently published general purpose PK model for propofol4 suggests the Schnider model underestimates drug concentrations
»
Objective: We examine the incidence of reaching the lower safety bound for propofol with our current system, and compare the predicted plasma concentrations (Cp) of the Schnider model3 to those of the general purpose PK model4
Figure 2: Overview of the first 100 minutes of all 82 cases. Representative cases shown from Phase 1 (red) and Phase 2 (blue). Shown are: depth of hypnosis (top), and predicted plasma concentration (Cp) for propofol (bottom) [this calculated using the Schnider model (solid), and the general purpose model (dashed)] Figure 3: Total distributions of predicted plasma concentration (Cp) for propofol calculated using the Schnider model (black), and the general purpose model (gold). Medians (solid) and 5th & 95th percentiles (dashed) are shown.
METHODS »
Our closed-loop control system, iControl-RP • receives processed electroencephalography feedback from the NeuroSENSE Depth of Hypnosis measure (WAVCNS) [NeuroWave, Cleveland, OH5] • controls two infusion pumps for propofol and remifentanil during both induction and maintenance of anesthesia (Figure 1)
» With Health Canada/REB approval
and written informed consent, data were collected from 82 patients recruited from a population of ASA IIII adults undergoing routine, elective surgery
DISCUSSION »
Considerable intra/interpatient variability in propofol requirements emphasizes the utility of a closed-loop system which can administer low doses of propofol when needed
»
Using the Schnider model3, the minimum bound for propofol was reached in the majority of our cases • Median propofol dose was relatively low • WAVCNS was adequate, vital signs within normal clinical range
»
If our safety bounds were calculated using the general purpose model • minimum bound may not have been reached as frequently • propofol infusion may have gone lower in some cases This highlights challenges of using population-based prediction models for safety bounds, and questions how the bounds should be formulated: how low is too low?
»
To reduce the number of threshold violations we propose to • make it easy to manually adjust the safety thresholds to reflect individual requirements • consider using the stability in WAVCNS to permit automated adjustments to the threshold
» All cases were conducted with
propofol in closed-loop • 51 cases with remifentanil administered by target controlled infusion [Phase 1: complete] • 31 cases with remifentanil administered in closed-loop [Phase 2: ongoing] Figure 1: The iControl-RP system
RESULTS » Minimum bound for propofol reached at least once in 51 cases (62%) » Table 1 shows summary of data while at lower bound and WAVCNS setpoint ≤55
Table 1.
Median propofol dose (mcg/kg/min)
65 [56-85]
Median remifentanil dose (mcg/kg/min)
0.10 [0.083-0.12]
% of case at minimum bound WAVCNS MAP (mmHg) HR (bpm)
17 45 87 67
[5.8-33] [43-48] [74-94] [55-73]
Minimum bound for propofol defined when predicted Ce
