Presented in part to the UK Society for Intravenous Anaesthesia, Chepstow, UK, 2006.
Patient-maintained propofol sedation using reaction time monitoring: a volunteer safety study*
Article first published online: 15 NOV 2012
Anaesthesia © 2012 The Association of Anaesthetists of Great Britain and Ireland
Volume 68, Issue 2, pages 154–158, February 2013
How to Cite
Allam, S., Anderson, K. J., O’Brien, C., Macpherson, J. A., Gambhir, S., Leitch, J. A. and Kenny, G. N. C. (2013), Patient-maintained propofol sedation using reaction time monitoring: a volunteer safety study. Anaesthesia, 68: 154–158. doi: 10.1111/anae.12036
You can respond to this article at http://www.anaesthesiacorrespondence.com
- Issue published online: 9 JAN 2013
- Article first published online: 15 NOV 2012
- Accepted: 9 September 2012
Previous volunteer studies of an effect-site controlled, patient-maintained sedation system using propofol have demonstrated a risk of over-sedation. We have incorporated a reaction-time monitor into the handset of the patient-maintained sedation system to add an individualised patient-feedback mechanism. This study assessed if such reaction-time feedback modification would reduce the risk of over-sedation in 20 healthy volunteers deliberately attempting to over-administer themselves propofol. All the volunteers successfully sedated themselves without reaching any unsafe endpoints. All volunteers maintained verbal contact throughout, in accordance with the definition of conscious sedation. The mean (SD) lowest SpO2 was 97 (1.7) % when breathing room air and no volunteer required supplementary oxygen. The mean (SD) maximum effect-site propofol concentration reached was 1.7 (0.4) μg.ml−1. The present system was found to be safer than its predecessors, allowing conscious sedation, but preventing over-sedation.
Effect-site controlled, patient-maintained sedation with propofol allows the patient to control their level of sedation using a hand-held demand button. The system sets a target effect-site concentration  at a low level that can then be increased when the patient presses the handset button twice within 1 s. The system increases the plasma concentration to twice that of the target effect-site concentration to ‘force’ the drug rapidly into the effect-site, and then reduces the plasma concentration as the target effect-site concentration is approached. Instead of a traditional lockout period, the patient cannot initiate a further increment until the plasma concentration decreases to within 10% of the target effect-site concentration. As the brain is the site of action of propofol, this paradigm allows equilibration between the blood and brain to occur. We have found this technique to be safe and effective in patients undergoing dental surgery , but before such systems can be used safely in the absence of an anaesthetist, it is necessary to demonstrate that patients are unable to become over-sedated, even when intentionally trying to do so.
Previous volunteer safety studies, which have tested the safety of the patient-maintained sedation system when healthy volunteers were deliberately attempting to over-sedate themselves, have shown that some volunteers manage to reach potentially unsafe levels of sedation (with minor arterial oxygen desaturation or loss of verbal contact). This is despite altering the lockout time on a plasma concentration-driven patient-maintained sedation system [3, 4], or delivering propofol to a target effect-site concentration. Some volunteers seem able to double-press the drug-demand button successfully despite being over-sedated and this system alone does not appear to provide adequate patient feedback .
We decided to investigate reaction-time monitoring as a potential way to improve individual patient feedback and reduce the risk of over-sedation. We have shown that as the patient’s effect-site propofol concentration and level of sedation increase, their reaction time increases also, and this becomes more marked before verbal contact is lost . We have now incorporated a reaction-time monitor into our sedation system’s handset. The handset vibrates at regular intervals and reaction time is measured as the time taken for the patient to press the sedation-demand button in response to the vibration. The reaction-time information is assessed as a lockout, either preventing further increases or reducing the target effect-site concentration of propofol if the reaction time becomes too slow. The aim of this study was to assess whether the addition of reaction-time monitoring would reduce the risk of over-sedation noted in previous volunteer studies.
We recruited 20 healthy volunteers (ASA physical status 1–2) following local Research Ethics Committee approval, Clinical Trial Authorisation and informed written consent. All volunteers were fasted for 6 h before sedation, which was supervised by an anaesthetist with full access to resuscitation equipment and drugs. An intravenous cannula was sited and volunteers were monitored throughout using ECG, pulse oximetry and non-invasive blood pressure measurements. Supplementary oxygen was not administered routinely, but was available if the volunteer’s peripheral oxygen saturation dropped below 90%.
Hardware and software modifications were made to the patient-maintained sedation system (Graseby 3500; Smiths Medical International Ltd, Ashford, UK), which was connected to a laptop computer via a serial port, to allow measurement of reaction time. The latter was logged and displayed along with target and calculated effect-site concentrations, and the activation of either of the system’s two lockouts, if appropriate. Reaction time was measured to the nearest millisecond, as the time taken to press the patient control handset button after it had vibrated. Before sedation started a minimum of three baseline reaction times were recorded, of which the last two measurements were averaged and used as that patient’s baseline reaction time. The effect-site-controlled patient-maintained sedation system was based on the Marsh pharmacokinetic model with a t½ ke0 (half-life of the equilibration constant) of 2.6 min .
Propofol sedation was started at a target effect-site concentration of 1 μg.ml−1 and the handset vibrated at intervals of 1 min during sedation to monitor the reaction time. The system’s existing pharmacokinetic lockout required the calculated effect-site concentration to equilibrate to within 10% of the calculated plasma-site concentration to permit an increment. In addition, the volunteer’s reaction time was used as an additional pharmacodynamic lockout; their measured reaction time was compared with their baseline measurement in an algorithm: reaction time ≤150% of baseline – increment allowed; reaction time > 150% and < 200%– target effect-site concentration maintained for 2 min; reaction time > 200% and < 300%– target effect-site concentration decreased by 0.2 μg.ml−1; and reaction time > 300%– target effect-site concentration decreased by 0.4 μg.ml−1. A double-press of the handset button within 1 s increased the calculated target effect-site concentration in increments of 0.2 μg.ml−1 up to a maximum concentration of 3 μg.ml−1, provided neither the pharmacokinetic nor reaction time lockouts were enabled. Finally, a reaction time of < 30% of baseline was considered ‘too quick’ and might be due to the volunteer's coincidentally double-pressing the handset (to request more sedation) immediately before or during a reaction-time measurement. The reaction time was rechecked after 20 s. An improvement in the reaction time to between 30% and 100% of the previous baseline measurement led to the resetting of the baseline reaction time to the new value.
Blood pressure, heart rate, peripheral oxygen saturation (SpO2) and respiratory rate were recorded before sedation and every 5 min during sedation. The volunteer’s sedation score (using the Observer’s Assessment of Alertness/Sedation score; OAAS ) was recorded with each reaction-time measurement. Volunteers were also given simple words to remember at intervals of 10 min during sedation. Any safety interventions made by the anaesthetist such as airway-support manoeuvres, use of airway adjuncts or supplementary oxygen were recorded.
Volunteers were continuously encouraged to use the handset button to make themselves as sedated as possible, particularly when neither lockout was enabled. Endpoints were defined as loss of verbal contact (OAAS score ≤2), SpO2 < 90%, requirement for airway intervention, apnoea, three consecutive reductions in target effect-site concentration due to slow reaction time, or the completion of 30 min of sedation. Volunteers were closely observed during the recovery period until they reached standard discharge criteria, when they were discharged into the care of a responsible adult. A short questionnaire was completed before discharge to assess recall and satisfaction.
We recruited 20 volunteers, 12 males and 8 females of mean (SD) age 36 (5.3) years and weight 74.4 (15.3) kg. All 20 volunteers successfully completed the study and in no volunteer was an unsafe endpoint noted; 10 volunteers ended the study after three consecutive reductions in target effect-site concentration and 10 after 30 min of safe sedation. The maximum effect-site concentration, lowest sedation score, lowest SpO2, recall of words and length of sedation for volunteers with these two endpoints are shown in Table 1. The mean (SD) maximum propofol effect-site concentration reached by the complete group of 20 volunteers was 1.7 (0.4) μg.ml−1 (range 1.2–2.4 μg.ml−1).
|Endpoint of study|
|Three consecutive decrements in effect-site concentration||30-min sedation completed|
|Number of volunteers||10 (50%)||10 (50%)|
|Maximum Ce; μg.ml−1||1.66 (0.4)||1.8 (0.4)|
|Lowest OAAS score||4.5 (3–5 [3–5])||3 (3–4 [3–5])|
|Lowest SpO2; %||97 (1.9)||97 (1.6)|
|Full recall of words at all times||5 (50%)||1 (10%)|
|Duration of sedation; min||21 (5.1 [12–29])||n/a|
All volunteers maintained verbal contact and obeyed commands throughout in accordance with the definition of conscious sedation. The lowest OAAS score was 3, reached by 10 volunteers. Five volunteers reached a sedation score of 4 and the remaining five volunteers maintained a score of 5. The mean (SD) minimum SpO2 was 97 (1.7)% while breathing room air. Supplementary oxygen or airway intervention was not required for any volunteer. Mean arterial pressure and heart rate remained within 30% of baseline measurements in all volunteers.
Volunteer satisfaction with this technique was high, 19 out of the 20 volunteers indicating that they were happy with the sedation they received and would use this technique if they required sedation in the future.
This study demonstrates that the addition of reaction-time monitoring and feedback control improves the safety of effect-site controlled patient-maintained sedation with propofol. For the first time, volunteers deliberately attempting to over-sedate themselves were unable to do so. All volunteers maintained verbal contact throughout in accordance with the definition of conscious sedation and maintained adequate levels of oxygenation breathing room air.
In clinical practice, patients receiving patient-maintained sedation would not be instructed, and would be unlikely to attempt to over-sedate themselves deliberately. However, for this system to be used without direct anaesthetic supervision, it is necessary to ensure that potentially unsafe levels of sedation cannot be reached. In our previous studies, volunteers succeeded in deliberately over-sedating themselves using systems where a successful button press would increase their plasma concentration [3, 4], or more recently, their effect-site concentration of propofol . This was despite previous attempts to make the system ‘unbeatable’ by prolonging the lockout time for increasing plasma concentration (to allow the brain to equilibrate with an increasing plasma concentration of propofol), and delivering propofol to a predicted effect-site concentration (which has been derived from a measure of clinical effect (auditory-evoked potentials ). With this latter system, it was noted that a number of volunteers retained the ability to press the demand button successfully, despite being significantly sedated, including one volunteer who had reached an unsafe level of sedation with oxygen desaturation. It was decided, therefore, that relying on a patient’s inability to press the button when becoming over-sedated was an insufficient control on its own.
In this study, we have investigated the addition of reaction-time monitoring as a method of providing individualised feedback within the patient-maintained sedation system and consequently its potential to reduce the problem of over-sedation. In a previous study by our group, it was found that increasing propofol effect-site concentration increases simple reaction time, as measured by the system’s vibrating handset. Even though there is a wide variation in reaction time during sedation in any population, it becomes more reproducible when it is considered as a percentage of an individual’s own baseline. It was also found that as a patient’s effect-site propofol concentration increases during sedation (before the induction of anaesthesia), individual reaction time increases more markedly just before loss of response to the vibrating handset, and this point is soon followed by loss of verbal contact. Reaction-time monitoring, therefore, may be a useful way of detecting when the depth of sedation is approaching an unsafe level .
In this study, reaction time was regularly monitored throughout sedation and significant slowing automatically prevented the volunteer being able to make further successful demands, either maintaining or actively reducing their effect-site concentration of propofol depending on the degree to which their reaction time had slowed. This allowed the volunteers to maintain their sedation within a safe level (sedation score 3 or above), avoiding loss of verbal contact.
Currently, the majority of patients in the UK receive sedation administered by non-anaesthetists for a variety of procedures including dental, endoscopic and radiological procedures. Midazolam is the agent most commonly used, but over recent years, issues regarding its safety have arisen [9–12], prompting the development of alternative systems using propofol, that have been studied in a range of clinical settings [13–17]. Patient-maintained sedation with propofol allows more rapid titration of and recovery from sedation (including faster discharge), better anxiolysis, less psychomotor impairment and fewer adverse effects including oxygen desaturation when compared with operator-administered midazolam . Other medical specialists seem to recognise its desirable effects [18, 19]; however, there is some concern about its administration by non-anaesthetists [20, 21]. We are aware that a much larger database of safety would be required to convince our peers and regulators that propofol sedation is safe without an anaesthetist present.
Attempts by volunteers to over-sedate themselves deliberately is an extremely artificial situation, designed to stress the safety of the patient-maintained sedation system. Our experience throughout development of this system has been that patients do not abuse or reach unsafe levels of sedation in clinical practice [13–17]. It is encouraging that in this pilot safety study, we have been able to show that over-sedation can be prevented when volunteers deliberately try to abuse the system. We are hopeful, therefore, that in the future it may be possible for such a system to be used in the absence of an anaesthetist. This would allow many more patients undergoing unpleasant procedures to benefit from the advantages that propofol sedation offers. Further work is now required to assess its safety and efficacy in large numbers of different patient populations.
We thank Martyn Gray of Anaesthesia Technology Ltd for the modification to the target-controlled infusion system to enable reaction time monitoring and feedback control.
No external funding and no competing interests declared.