Patient-controlled administration of sedation (PCS) is safe and acceptable to patients, surgeons and anaesthetists alike, but its clinical use remains limited. This review examines the development of PCS, with a brief explanation of the possible psychological benefits.
PCS followed naturally from the acceptance of Patient Controlled Analgesia (PCA). Self-administration of analgesia was first described in the hospital context by Sechzer in 1968 , but it did not become a practicable proposition until the development of the Cardiff Palliator in 1972. By 1979, the desirability of PCA was widely accepted and industry responded by manufacturing a range of suitable devices, so that PCA is now offered routinely in most of the major hospitals in Western Europe.
Though there are several reports of the use of PCS, that use is not widespread, despite the fact that PCA devices can easily be utilized for PCS. There is unanimous agreement that PCS appears to be a safe and effective way to make potentially unpleasant procedures acceptable to the patient. Sedation has been used for decades on its own or as an adjunct to local or regional anaesthesia. Its main benefits are to allay the patients' anxiety and reduce the difficulties of the procedure by enabling their co-operation. Sedation administered by persons other than the patient has well-known disadvantages arising from either under-sedation or over-sedation - the latter being potentially more catastrophic. These can usually be attributed either to a lack of specialist knowledge of sedative drugs on the part of the administrator, or to the normal variation between patients in susceptibility to sedative drugs. The potential advantage of PCS is that, if a suitable dose and lock-out interval are selected to allow a safe but adequate upper-limit to the dosing, patients can generally titrate themselves to an appropriate level of sedation without risk of over-sedation.
The conceptual development of PCS
The first report of patient-controlled administration of drugs other than analgesics was by Loper et al., describing the use of midazolam for anxiolysis in the Intensive Care Unit (ICU) in two patients requiring mechanical ventilation. A Lifecare PCS Infusor was used to give doses of midazolam 0.25 mg with a lock-out interval of 8 min. There was a reduction in the patients' subjective anxiety, as measured on a linear analogue scale, and a reduction in daily requirement for morphine. Neither patient suffered any adverse cardiovascular or respiratory sequelae and Loper et al. concluded that patient-controlled anxiolysis was a promising method of moderating patients' anxiety in the setting of Intensive Therapy.
The second report of patient-controlled anxiolysis was by Galletly et al. in 1989  and described the effectiveness of self-administered diazepam, in 2 mg doses with a 3-min lock-out interval, as an alternative to standard premedication. Fifty patients awaiting a wide range of surgical operations were allowed to give themselves diazepam 20 mg over 30 min. Anxiety and sedation were assessed by both the patients and the anaesthetists, the patients by a linear analogue score while the anaesthetists used a system which gave a separate score for anxiety and sedation. The patients who rated themselves the most anxious gave themselves the largest doses of diazepam and obtained the most benefit. Worryingly, the anaesthetists' assessments of patients' anxiety bore no relation to those of the patients, and their predictions of the patients' requirements for premedication were inaccurate. The paper concluded that, though this method of premedication was effective, it was also costly in both time and equipment.
Ure et al. presented another study of self-administered premedication to the Anaesthetic Research Society in 1991. It was a double-blinded comparison of the anxiolytic and sedative effects of doses of propofol 10 mg against those of a placebo consisting of Intralipid 10%. Assessments were made both by the patient and an observer. The patients and observer reported significantly reduced levels of anxiety in the propofol group but, interestingly, the patients' reports of significantly increased sedation with the propofol were not reflected in the observer's assessments.
The term Patient Controlled Sedation (as opposed to anxiolysis) has come to be applied when the self-administered sedative is intended by itself to enable some surgical procedure to be carried out (and not simply as a premedication for general anaesthesia). PCS received an impetus after the advent of propofol, whose rapid onset and rapid offset make it easier to titrate for per-operative sedation than the agents that preceded it. The first report of anaesthetist-administered sub-anaesthetic doses of propofol as an adjunct to regional anaesthesia was by Mackenzie and Grant in 1987 .
The first paper describing PCS using propofol was by Rudkin et al., in 23 patients undergoing surgical extraction of third molars as day-cases in Adelaide, Australia. This was a landmark paper which has defined the direction the development of PCS has taken. Their choice of agent was influenced by earlier favourable comparisons of the rate of recovery and incidence of post-operative amnesia after propofol sedation with those after sedation with diazepam  or midazolam . After an initial loading-dose of propofol 0.7 mg kg−1 over 15 s at the start of the procedure, patients in the Adelaide study received further doses on demand from a Graseby PCA device (Graseby Medical Ltd, Watford, UK) modified to deliver propofol 0.7 mg kg−1 at a rate of 16.7 mg min−1 with a lock-out interval of 1 min. The results were very encouraging: haemodynamic responses were stable and oxygenation was well maintained. There was a 70% incidence of intra-operative amnesia without any significant post-operative amnesia and the patients found PCS highly acceptable.
The same group followed up with a study  in a similar group of patients, that compared an identical dosing scheme with propofol PCS with anaesthetist-administered sedation using repeated doses of midazolam and fentanyl. In addition to the level of sedation, duration of procedure, and degree of surgical difficulty, the study examined the degree of intra- and post-operative amnesia, post-operative cognitive function and the patients' overall satisfaction. Both methods of sedation provided satisfactory with safe operating conditions, but the propofol group indicated a greater degree of satisfaction, with amnesia that was more selectively intra-operative and good recall of the recovery period and the journey home.
The study's most important technical aspect was in modifying an Ohmeda 9000 pump (Ohmeda, West Yorkshire, UK) to infuse the demand dose of 0.7 mg kg−1 at 1200 mL hr−1. At their previously used infusion rate of 100 mL hr−1, it took over 2 min to deliver a demand dose to a 70 kg patient and this was followed by a lock-out interval of only 1 min. When demand rates were high towards the beginning of procedures, the long delivery times and the short lock-out intervals meant that the patients were effectively receiving an interrupted infusion rather than an intermittent dose of sedative. The higher infusion rates increased the ratio of successful to unsuccessful demands for sedation because fewer demands were being made during the lock-out interval, and, compared with a previous study, the mean total propofol consumption per patient was significantly reduced, from 260 mg to 160 mg.
A highly significant paper from Cook et al. introduced the concept of a zero lock-out interval . Patients, who were undergoing transvaginal oocyte collection, were randomly assigned to receive either propofol or midazolam via a Grase-by PCA pump (Graseby Medical Ltd, Watford, UK) with an infusion rate of 200 mL hr−1 and a zero lock-out interval. The items that were investigated included the speed of onset of sedation, sedation levels, patient satisfaction and psychometric testing. The aim of the study was to provide an infusion system that could deliver enough of the drug at the start of the procedure for the patients to be able to establish their own level of sedation, thus eliminating the requirement for an initial loading-dose administered by an anaesthetist. This, according to the authors, was true patient-controlled sedation. The challenge was to be able to produce a fast-onset sedation, but also to minimize the possibility of over-sedation. As mentioned above, patients' and anaesthetists' judgements of levels of anxiety and requirements for sedation may differ markedly, and this type of sedation system would remove the anaesthetist from the equation.
The studies by Osborne et al. and Cook et al. introduced the concept of effective lock-out time (the time taken to infuse the dose and the succeeding set lock-out time), and moved PCS towards explicit clarification of operational objectives (which are reminiscent of the operational objectives for PCA). The goal with any PCS system is to be able to establish rapid and effective sedation (which requires capability for a fast initial rate of drug delivery to an adequate total loading dose), to maintain a required level of sedation and to re-establish a deeper level of sedation quickly if it has been allowed to decay (which also requires the ability to deliver small incremental doses of drug quite rapidly): all this has to be carried out in the face of wide individual and temporal variation in requirements for sedation. Safety is, of course, the overriding factor, and the situation to avoid is that in which a mildly drowsy and unco-operative patient is given that additional increment of sedative that will push him or her from sedation to anaesthesia, with the attendant risks of loss of airway or apnoea. Cook and Whitwam  discuss the concept of maximum possible overdose that could arise from the number of additional increments of drug that could be self-administered before the previous increment has had time to exert its full effect (dose-stacking). If the time from administration to effect just exceeds the effective lock-out time and the effect of the most recent dose is just enough (to cause sedation rather than anaesthesia) the patient could self-administer one further dose, which would cause anaesthesia rather than sedation. The excess over the just adequate dose would be in units of the incremental dose set for self-administration, and the number would be the whole number of effective lock-out times that can be fitted into the time from administration to effect. The excess could be limited to one dose by matching the effective lock-out time to the time from administration to effect and its size could be limited by reducing the size of the dose (though smaller doses reduce the maximum loading rate and a small dose for one patient may be a large dose for another).
Although most authors have adapted electronic PCA pumps for use in PCS, a recent paper evaluated a large capacity, high-flow mechanical infusor, the Baxter Intermate LV250 infusor (Baxter Healthcare Ltd, Compton, UK) . This is a development of the Baxter mechanical PCA devices. The addition of a patient-control module (PCM) is claimed to make this suitable for use as a PCS pump, and the authors assessed the pump in this capacity on adult dental patients with good results. This mechanical system differs from electronic PCS devices in a number of key ways. First, there is no true lock-out interval, the maximum rate of flow is dictated by a flow restrictor within the infusor which, in the case of the LV250, permits a maximum flow rate of 250 mL hr−1. Second, the bolus dose is not fixed as activating the system via the wrist button empties a 0.5 mL bladder of the sedation mixture into the intravenous (i.v.) cannula, which then takes 7.2 s to refill. Pressing the button again within this period results in a proportion of the 0.5 mL being delivered. With an electronic PCS pump, pressing the button to activate the pump within the equivalent period would not result in any drug being delivered. The authors also found that, as would be expected from Poisseuille's Law, the viscosity of the solution in use affected the flow-rate of the device so that the standard propofol emulsion had a mean flow-rate 65% of that of 0.9% NaCl solution. Also, as the distensible bladder which forms the main reservoir of the infusor decreases in volume, the radius decreases at a faster rate than the tension in the bladder wall, which (in accordance with LaPlace's Law) results in a higher pressure and therefore a faster flow. Despite these concerns, the authors conclude that this style of pump is a feasible alternative to the electronic variety.
Cook and Whitwam  also discuss the roles of analgesics, notably opioids, co-administered with sedative agents. The unpredictable additive effect of opioids and sedatives may be minimized by giving a combination in which the overall sedative effect is greater so that there will be fewer demands for additional increments, or by giving the analgesic early so that it will have had its full effect by the time sedation is started, which would allow the effect of the analgesics to be taken into account. This obviously relies on using an analgesic with a fast onset of action and ideally a short elimination half-life. The combination of propofol and alfentanil has been advocated  and the very short-acting opioid remifentanil may be useful for this purpose. Alfentanil has also been successfully used alone in a PCS system . It may be that the prior administration of a non-steroidal anti-inflammatory drug may diminish the analgesic requirements, or indeed the amount of sedation required. Further work is necessary to examine this.
Although propofol seems to be the drug of choice for PCS, an abstract by Hamid et al. compared two PCS regimens, one using an incremental dose of methohexitone 2.5 mg and the other propofol 5 mg in 40 adult dental patients . The rationale for using methohexitone was that it is cheaper than propofol but has a similarly rapid clearance from the body. Sedation scores were similar for both groups of patients and psychometric testing demonstrated that methohexitone was only mildly inferior in terms of post-operative recovery. No data were provided on the incidence of nausea and vomiting. The authors concluded that methohexitone is a reasonable alternative to propofol.
Single-agent PCS as a technique has been established for midazolam during dental extractions  and for propofol during dental  or general surgical and orthopaedic procedures [18,19].
Psychological aspects of PCS
A midazolam/fentanyl mixture featured in an important study by Park and Watkins , in which patients undergoing lower limb or pelvic surgery under epidural anaesthesia were randomized to either self-administer the sedative mixture or receive incremental sedation from the anaesthetist. The PCS dose consisted of midazolam 0.2 mg mixed with fentanyl 10 μg and the lock-out time was 5 min. There were no significant differences between the two groups in terms of duration of anaesthesia and surgery, the deepest level of sedation attained and the total dose given. However, linear analogue assessments of comfort during surgery were significantly higher in the PCS group than in the patients being sedated by the anaesthetist. The authors discussed why this might be the case, noting that patients vary in how sedated they wish to be, and display widely varying pharmacodynamic responses to sedative drugs. Over- and under-sedation is therefore not infrequent, whether in absolute terms, or in terms of what the individual patient might want. The authors' general discussion introduced the concept to PCS that there may be appreciable psychological benefits from allowing patients to exert a substantial degree of control over an aversive stimulus.
Lefcourt  reviewed a number of studies of the stress-reducing effects of control over aversive stimuli, and stated that 'the perception of control would seem to be a common predictor of the response to aversive events regardless of species .. the sense of control, the illusion that one can exercise personal choice, has a definite and a positive role in sustaining life'. Averill, reviewing 'Personal control over aversive stimuli and its relationship to stress'  described three types of control:
- behavioural control, the ability to modify the characteristics of a threatening event;
- cognitive control, where the subject may diminish the psychological stress of an event by being able to interpret the available information; and
- decisional control, the opportunity to choose between various courses of action.
All patients should receive information on the nature of any intended operative procedure and on the options for sedation or anaesthesia that will be available to them. It has been repeatedly shown that this sort of pre-operative preparation is beneficial in terms of lower post-operative pain scores and analgesic requirements, and in shorter stays in hospital . A degree of cognitive control should therefore already be at work in patients being offered PCS, although it should be noted that individuals may display a large degree of variation in their response to information on a forthcoming aversive event . In addition, PCS confers a degree of behavioural and decisional control over the depth of sedation and therefore the responses, both psychological and autonomic, to painful or otherwise stressful events. Patients frequently comment, after a procedure performed under PCS, that they 'liked having some control over what was going on.'
There may be an additional pharmacodynamic reason, peculiar to propofol, why PCS is so popular with patients. Several papers have commented on the amorous advances made towards medical staff by some patients recovering from proprofol anaesthesia [25-28]. Others have reported anecdotally on the mild euphoria that may be associated with recovery from proprofol [29,30], although strong evidence of this is lacking. In a well-designed study, Whitehead et al. measured six psychometric variables during low-dose proprofol sedation that gave rise to plasma concentrations of between 0.3 μg mL−1 to 0.9 μg mL−1, and failed to find evidence of mood elevation.
Patient controlled sedation is a relatively new technique but has been used successfully for a broad range of surgical procedures. The technique is easily learned. It uses equipment that is available and familiar to anaesthetists in most European hospitals. PCS has displayed advantages over conventional forms of sedation, not only in terms of the quality and safety of sedation, but also in terms of acceptability to the patient. Its most notable advantage may well be the psychological benefit it confers on its users, from the sense of control over procedures that are inherently stressful.
1 Sechzer PH. Objective measurement of pain. Anesthesiology
2 Loper KA, Ready LB, Brody M. Patient-controlled anxiolysis with midazolam. Anesth Analg
3 Galletly DC, Short TG, Forrest P. Patient-administered anxiolysis-a pilot study. Anaesth Intensive Care
4 Ure RW, Dwyer SJ, Blogg CE, White AP. Patient-controlled anxiolysis with propofol. Br J Anaesth
5 MacKenzie N, Grant IS. Propofol for intravenous sedation. Anaesthesia
6 Rudkin GE, Osborne GA, Curtis NJ. Intra-operative patient-controlled sedation. Anaesthesia
7 Valtonen M, Salonen M, Fossell H, Scheinin M, Viinamatii O. Propofol infusion in outpatient oral surgery. A comparison with diazepam. Anaesthesia
8 Wilson E, David A, MacKenzie N, Grant IS. Sedation during spinal anaesthesia; comparison of propofol and midazolam. Br J Anaesth
9 Osborne GA, Rudkin GE, Curtis NJ, Vickers D, Craker AJ. Intra-operative patient-controlled sedation. Anaesthesia
10 Cook LB, Lockwood GG, Moore CM, Whitwam JG. True patient-controlled sedation. Anaesthesia
11 Cook LB, Whitwam JG. Patient controlled sedation. In: Whitwam JG, ed. Day-Case Anaesthesia and Sedation.
London: Blackwell Scientific Publications, 1994: 275-288.
12 Hamid SK, Wong PK, Carmichael K, White K, Asbury AJ. A novel device for patient-controlled sedation: laboratory and clinical evaluation of the Baxter Intermate LV250 infusor and patient-control module. Anaesthesia
13 Sherry E. Admixture of propofol and alfentanil. Anaesthesia
14 Zelcer J, White PF, Chester S, Paull JD, Molnar R. Intraoperative Patient-Controlled Analgesia: An Alternative to Physician Administration During Outpatient Monitored Anesthesia Care. Anesth Analg
15 Hamid SK, McCann N, McArdle L, Asbury AJ. Comparison of patient-controlled sedation with either methohexitone or propofol. Br J Anaesth
(Suppl. 1): 4.
16 Rodrigo MRC, Tong CKA. A comparison of patient and anaesthetist controlled midazolam sedation for dental surgery. Anaesthesia
17 Osborne GA, Rudkin GE, Jarvis DA, Young IG, Barlow J, Leppard PI. Intra-operative patient-controlled sedation and patient attitude to control. Anaesthesia
18 Grattidge P. Patient-controlled sedation using propofol in day surgery. Anaesthesia
19 Pizzirani E, Pigato P, Fornasier L, Lacamera F, Giron GP. Patient-controlled sedation during spinal anaesthesia. Br J Anaesth
(Suppl. 1): 4.
20 Park WY, Watkins PA. Patient-Controlled Sedation During Epidural Anesthesia. Anesth Analg
21 Lefcourt HM. The functions of the illusions of control and freedom. Am Psychol
22 Averill JR. Personal control over aversive stimuli and its relationship to stress. Psychol Bull
23 Egbert LD, Battit GE, Welch CE, Bartlett MK. Reduction of post-operative pain by encouragement and instruction of patients. New Engl J Med
24 Andrew JM. Recovery from surgery, with and without preparatory instruction, for three coping styles. J Pers Soc Psychol.
25 Hunter DN, Thonnling A, Whitburn R. Arousal from propofol. Anaesthesia
26 Young PN. Hallucinations after propofol. Anaesthesia
27 Bricker SRW. Hallucinations after propofol. Anaesthesia
28 Smith DG, Collins-Howgill PJ. Hallucinations after propofol. Anaesthesia
29 O'Toole DP, Milligan KR, Howe JP, McCollum JSC, Dundee SW. A comparison of propofol and methohexitone as induction agents for day case isoflurane anaesthesia. Anaesthesia
30 McDonald NJ, Mannion D, Lee P, O'Toole PD, O'Boyce C, Keane PK. Mood evaluation and outpatient anaesthesia. Anaesthesia
31 Whitehead C, Sanders LD, Oldroyd G et al.
The subjective effects of low-dose propofol. Anaesthesia