Patient-controlled sedation for cataract surgery ... - Oxford Academic

14 downloads 70898 Views 122KB Size Report
they were no longer anxious and “felt good”, and ... of a portable notebook computer (Compaq Contura, ... surgeon and was considered good if the patient.
British Journal of Anaesthesia 1996;77:370–374

Patient-controlled sedation for cataract surgery using peribulbar block C. K. PAC-SOO, S. DEACOCK, G. LOCKWOOD, C. CARR AND J. G. WHITWAM

Summary Patients undergoing cataract surgery using peribulbar block were allocated randomly to self-administer doses of either midazolam 0.1 mg or propofol 3.3 mg without a lock-out facility; in the control group the syringe was charged with saline, not as a placebo, but to “blind” the surgeon and the nurse observer. For midazolam and propofol, median doses were 2.54 (0.1–6.0) mg and 87.4 (0–145) mg, respectively. Patientcontrolled sedation significantly reduced the level of anxiety, with median visual analogue anxiety scores in the midazolam, propofol and saline groups of 5 (0–38) mm, 5 (0–25) mm and 15 (0–92) mm, respectively (P0.05). Some patients did not administer the sedative when available while others in the saline group would have benefited from anxiolytic drugs. While both drugs prevented an increase in heart rate, only midazolam prevented an increase in arterial pressure during surgery. (Br. J. Anaesth. 1996;77:370–374) Key words Hypnotics benzodiazepine, midazolam. Anaesthetic techniques, regional, peribulbar. Sedation, patient-controlled. Anaesthetics i.v., propofol. Surgery, ophthalmological. Age factors.

Cataract is predominantly a disease of old age1 2. The number of elderly patients in the UK, together with earlier referral of patients for treatment, is increasing the demand for cataract surgery3 4, using day-care facilities5 6. Despite the availability of new anaesthetic drugs which allow rapid recovery, for example propofol, desflurane and sevoflurane, the use of local anaesthesia is increasing as it provides not only satisfactory operating conditions, but also advantages in terms of health economics5. A major disadvantage of local anaesthesia alone is anxiety, so that patients may be restless when immobility is required7. This may be overcome with anxiolytic drugs causing only minimal sedation. Unfortunately, a patient’s response to a predetermined dose of a sedative drug may be unpredictable, particularly in the elderly where only relatively small doses are required8–10. An alternative would be patient-controlled sedation (PCS) which has been used successfully during surgery performed under local anaesthesia11–14. In early studies, the technique included an initial predetermined dose of the sedative drug before the patient took over which, by allowing the possibility of initial

over sedation, may be a disadvantage. Recently, a modified patient-controlled analgesia pump (Graseby 3300, Watford, UK) without a lock-out period has been used to allow small doses of either propofol or midazolam to both initiate and maintain sedation15. The only restriction of access to the drug is the duration of injection of each dose, which was 5.4 s in these studies15–17. Patients use a much wider range of doses to achieve similar levels of psychomotor depression, with maintenance of verbal communication, than would be expected during physician administration. We have evaluated this new method of total patient-controlled sedation in an elderly population undergoing cataract surgery during local anaesthesia.

Patients and methods After obtaining approval from the Ethics and Research Committee (No. L94/1) and written informed consent, we enrolled 77 patients in the study. Exclusion criteria included age greater than 85 yr, clinical evidence of heart failure, severe pulmonary disease, sedative medication within the previous month, difficulty with language or communication, and poor vision in the nonoperated eye. Patients were allocated randomly to one of three groups by opening a sealed numbered enveloped containing the study group (120, 4-cm square labels, 40 each labelled midazolam, propofol, and placebo, were folded into four and mixed thoroughly in a box approximately 30 cm square by stirring and shaking, and were then taken out, placed in an envelope and numbered in the selected order). Three senior registrar anaesthetists and four ophthalmic surgeons of senior registrar or consultant status participated in the study. This was a single-blinded study with patients in the saline group informed that they were not receiving any sedation while the research nurse and the surgeons were blinded. We felt that it was inappropriate and unethical to allow patients to try to sedate themselves with saline.

C. K. PAC-SOO*, MB, CHB, FRCA, C. CARR, MB, CHB, FRCA, Moorfields Eye Hospital, City Road, London EC1V 2PD. S. DEACOCK, MB, CHB, FRCA, G. LOCKWOOD, MB, CHB, BSC, FRCA, J. G. WHITWAM, MB, CHB, FRCP, FRCA, PHD, Anaesthetic Department, Royal Postgraduate Medical School and Hammersmith Hospital, Du Cane Road, London W12 0NN. Accepted for publication: April 24, 1996. *Present address for correspondence: Anaesthetic Department, Royal Postgraduate Medical School and Hammersmith Hospital, Du Cane Road, London W12 0NN.

Sedation and cataract surgery SEDATION EQUIPMENT

The use of the sedation pump (modified Graseby 3300, Watford, UK, no lock-out, infusion rate 200 ml h1) was explained to the patients. They were told that they may not feel the need to administer any drug. If they started to feel less relaxed or anxious they should press the button as often as required until they were no longer anxious and “felt good”, and while they could press the button again whenever they felt they needed more drug, they should not attempt to become unconscious. The pump contained 30 ml of one of three solutions: 0.9% saline, with or without midazolam or propofol 10 mg (1%). It was set to administer 0.33-ml aliquots of solution, that is midazolam 0.1 mg or propofol 3.3 mg. A 21-gauge cannula was inserted into a vein in the forearm on the contralateral side to the arterial pressure cuff and connected to the patient-controlled sedation pump via a primed infusion cannula (standard intravascular monitoring catheter length 150 cm, internal diameter 1 mm). Patients who were allocated to the saline group were then informed that the pump did not contain any sedative drug and that they were the control group. The number of demands, and successfully administered boluses of drugs, were recorded. PATIENT ASSESSMENT

Sedation The level of sedation was graded before performing peribulbar block by an independent “blinded” research nurse using the following scale: (1) awake, anxious; (2) awake, not anxious; (3) sedated, speech slurred; and (4) eyes closed with appropriate response to verbal command15. Awake patients were asked by the nurse if they felt anxious. Intraoperative assessment was made by the operating surgeon. Anxiety Two hours after surgery, patients were reviewed by a different anaesthetist. The level of anxiety experienced in the intraoperative period by the patients was assessed retrospectively using a linear visual analogue scale (VAS) from 0 to 100 mm, where 0feeling normal and not anxious and 100extreme anxiety. Psychomotor test The use of the psychomotor test equipment was explained during an initial period of familiarization. Two ascending and descending readings were obtained for critical flicker fusion, a simple response time was repeated six times and a computerized digit span memory test was also applied. The psychomotor tests were performed by a nurse trained in the use of a portable notebook computer (Compaq Contura, Compaq Computer Corporation, Singapore). Baseline tests were performed before the start of each study and repeated 10, 30, 60 and 120 min after the end of surgery. It was not possible to apply the tests satisfactorily when the treatment pathway had started, because of the speed of events, changing conditions and limited access to the patients. MONITORING

The electrocardiogram (Cardiorecorder 364 AM, UK), heart rate and arterial oxygen saturation (SpO2 )

371 (Novametric 515 A, USA) were monitored continuously together with intermittent measurements of non-invasive arterial pressure (Dinamap Crition A 45, UK). Baseline readings were recorded before sedation and repeated at 5-min intervals thereafter, including the phase I recovery area after surgery. PERIBULBAR BLOCK

All patients, whether receiving saline, sedation or those not taking sedation, indicated that they were ready for the block when assessed within 5–10 min after connection of the sedation apparatus, which in the saline group was there to blind the observers. After the research nurse had assessed the level of sedation, which took approximately 2–4 min, a peribulbar local anaesthetic block was performed on the operated eye using 8 ml of a mixture of equal volumes of 0.5% bupivacaine and 2% lignocaine to which 150 iu of hyaluronidase had been added. A Honan balloon was applied for 20 min to facilitate spread of the local anaesthetic into the muscle cone and subsequently to reduce intraocular pressure. Additional local anaesthetic was administered when required to ensure satisfactory block. PROCEDURE

After transfer to the operating theatre, the surgical field was prepared. Oxygen was delivered at 6 litre min1 by a cannula placed close to the nostrils under the drapes. The PCS apparatus was positioned near the patient's feet under the surgical drapes but easily accessible to the anaesthetist. Before surgery, the surgeon, who was blinded to the sedative technique, assessed the degree of ocular movement using a recognized scale18. In the intraoperative period, patients maintained their own level of anxiolysis and the anaesthetist was present throughout. The surgeon ascertained that patients were not over sedated by maintaining verbal contact. At the end of the operation, the surgeon recorded the quality of the local anaesthesia and also the patient’s cooperation during surgery. Assessment of cooperation was left to the discretion of the surgeon and was considered good if the patient responded appropriately to his request and did not move unexpectedly during the procedure. Sedation was then discontinued and the duration of use of the sedation pump and surgery were recorded. Patients were then transferred to the recovery unit (phase I) where they remained until completion of the study. After step down (phase II), patients were reviewed by a second anaesthetist before discharge. STATISTICAL ANALYSIS

Analysis of variance (ANOVA) was applied to normally distributed data. The difference between the highest systolic and diastolic pressures and heart rat and baseline pressure in each patient were compared between groups using one-factor ANOVA. Within each group, comparison with baseline was made using Student’s paired t tests with Bonferroni's correction. Sedation scores were compared using chi-square analysis and the median anxiety scores were compared using the Mann–Whitney U test. The results of the psychometric tests were analysed using

372

British Journal of Anaesthesia

Friedman’s two-way analysis of variance for repeated measurements. P0.05 was considered statistically significant.

Table 2 Duration of surgery and patient-controlled sedation (PCS) (mean (SD), total amount of drug used (median (range)), number of patients who could not recall operative events and quality of anaesthesia and co operation of the patients in the three groups, no significant differences between groups

Results Two patients were removed from the study; one became over sedated with midazolam and its effects were reversed with flumazenil 0.5 mg i.v. before the local anaesthetic block was performed; the other in the non-sedated group, suffered a posterior capsular tear with vitreous loss. There were no significant differences in patient characteristics between the groups (table 1) and no differences in duration of surgery, quality of local anaesthesia (table 2) or ocular muscle paralysis. In one patient in the midazolam group and two in the propofol group, local anaesthetic action wore off during surgery and was supplemented with 1% topical amethocaine. Cooperation of the patients, assessed by the surgeon, was good, except for one non-sedated patient, three patients in the midazolam group and one in the propofol group (table 2). The median total doses of midazolam and propofol were 2.54 mg and 87.4 mg, respectively (table 2). Three patients in the propofol group did not use the pump but their anxiety scores were 0–6 mm on the VAS. Three patients in the midazolam group fell asleep during surgery but were awakened immediately by verbal communication, without distress or restlessness, and with appropriate responses to instructions. Sedation scores before performing peribulbar block were grouped as inadequate (group 1) and adequate (groups 2–4). There were significantly more patients in the saline group who were inadequately sedated compared with those in the midazolam and propofol groups (table 3). There was no significant difference in the degree of sedation between the midazolam and propofol groups and only two patients were considered retrospectively to have used too little sedation. Overall, the non-sedated group was significantly more anxious during surgery. They had a median VAS of 15 mm compared with 5 mm in the midazolam and propofol groups (table 3). However, seven (27%) patients in the non-sedated group when assessed after surgery had a VAS anxiety score of 0 mm. The results of the psychomotor tests, which were performed before and after the procedure, were not significantly different either between the three groups or within each group (table 4). Three patients in the midazolam group had complete amnesia for intraoperative events. Only one patient (in the midazolam group) developed an arterial saturation of less than 90% for more Table 1 Patient characteristics for the three groups (mean (SD or range) or number)

Age (yr) Weight (kg) Sex (M:F) ASA (I/II/III)

Non-sedated group (n26)

Midazolam group (n24)

Propofol group (n25)

70 (52–85) 71 (15.3) 13:13 13/13/0

70.3 (44–85) 70.1 (13.4) 12:12 14/10/0

67 (45–82) 70 (14.8) 11:14 9/16/0

Duration of surgery (min) Duration of PCS use (min) Amount of drug used before block (mg) Total amount of drug used (mg) Complete amnesia during surgery Yes No Quality of local anaesthesia Good Poor Cooperation of patient during surgery Good Bad

Non-sedated Midazolam group (n26) group(n24)

Propofol group (n25)

43.9 (23.5)

40.1 (14.4)

42.6 (15.2)

0

93.3 (33.4)

89.6 (20.1)

0

1.62 (0.1–4.8)

40.1 (0–145)

0

2.54 (0.1–6.0)

87.4 (0–195)

0 26

3 21

0 25

25 1

23 1

23 2

25 1

21 3

24 1

Table 3 Assessment of sedation and anxiety. Degree of sedation before peribulbar block and anxiety experienced during surgery (retrospective) using a visual analogue scale ((VAS) of 0–100 mm). Sedation state was analysed using the chi-square test and anxiety score using the Mann-Whitney U test. *P  0.05 Propofol Midazolam Non-sedated group (n  26) group (n  24) group (n  25) Sedation score 1 2 3 4 Inadequate (1) Adequate (2+3+4) Anxiety score (mm) (median (range))

9 17 0 0 9*

2 15 6 1 2

0 19 5 1 0

17*

22

25

15 (0–92)*

5 (0–38)

5 (0–25)

than 1 min. This occurred 20 min after initiation of sedation while breathing air. Saturation improved immediately when the patient was asked to inhale a few deep breaths and remained greater than 92% throughout the remainder of the study. There were no significant differences in mean control systolic and diastolic arterial pressures, which were 148.8 (SD 17.3) mm Hg and 86.2 (21.5) mm Hg in the placebo group, 157.5 (13.9) mm Hg and 85.7 (11.4) mm Hg in the midazolam group, and

Sedation and cataract surgery

373

Table 4 Psychometric studies. Effect of midazolam, propofol and placebo on psychometric tests before sedation and after surgery (mean (SD)). SRT  Simple response time, CFF  Critical flicker fusion, DS  Digital span memory test (maximum number of digits scored correctly). Data were compared using two-way analysis of variance. No significant differences between groups Time after surgery(min) Group SRT (ms) Placebo Midazolam Propofol CFF (Hz) Placebo Midazolam Propofol DS (No.) Placebo Midazolam Propofol

n

Before sedation

10

30

60

120

26 24 25

331(104) 335 (74) 375 (158)

368(183) 477(456) 535 (617)

363 (157) 447 (472) 346 (112)

353 (140) 344 (123) 352 (114)

367 (185) 350 (133) 334 (87)

26 24 25

34.0 (3.0) 34.0 (2.3) 34.0 (2.9)

34.1 (3.0) 34.3 (3.8) 34.0 (2.1)

34.0 (3.4) 33.9 (4.3) 34.6 (2.7)

34.4 (3.1) 35.3 (3.8) 34.0 (2.4)

33.8 (3.4) 33.9 (2.2) 35.0 (2.9)

26 24 25

5.0 (1.3) 5.5 (1.1) 5.8 (1.0)

5.0 (1.1) 4.5 (1.1) 5.2 (1.1)

5.1 (1.4) 5.2 (1.0) 5.2 (1.1)

5.4 (1.4) 5.0 (1.0) 5.6 (1.0)

5.3 (1.2) 5.0 (1.3) 5.6 (1.3)

150.1 (18.8) mm Hg and 86 (12.1) mm Hg in the propofol group. In the saline and propofol groups there were significant peak increases of 14.6 and 9.4 mm Hg in mean systolic pressure (P0.05 in each). In contrast, in the midazolam group, mean arterial pressure changed by only 2.4 mm Hg (ns). In all groups arterial pressure returned to near control values by the end of the procedure. There were no significant changes in mean diastolic pressure. One non-sedated patient was restless and his systolic and diastolic arterial pressures increased to 200 and 120 mm Hg, respectively, and were restored to normal control levels with labetalol (total dose 30 mg i.v.). The patient was not known to suffer from hypertension but was extremely anxious during surgery, with a retrospective anxiety score of 92 mm on the VAS. Mean control heart rates were 73.2 (10.8), 72.4 (11.4) and 74.4 (11.1) beat min1 in the saline, midazolam and propofol groups, respectively. There were no significant changes in heart rate throughout the study in the midazolam and propofol groups, and although the unsedated group demonstrated a significant peak increase in mean heart rate during surgery, it was only 4.2 beat min1 and therefore not clinically important. Heart rate exceeded 90 beat min1 in approximately 33% of patients.

Discussion We have demonstrated that elderly patients can use patient-controlled sedation during cataract surgery to induce and maintain anxiolysis and conscious sedation with minimal drowsiness (table 3). The term sedation used in the study dose not refer to the degree of drowsiness but rather describes a state of anxiolysis. When patients feel very relaxed, they can become very drowsy but verbal communication is retained. This method of sedation is as safe as other methods and may have advantages over other techniques where an anaesthetist or other practitioner induces and maintains sedation7–9. Both midazolam and propofol were effective in obtunding an increase in heart rate during surgery, which occurred in the non-sedated group, but only midazolam prevented an increase in arterial pressure, which confirms a previous report in patients undergoing cataract surgery7.

Clearly sedation is not required in all patients undergoing cataract surgery with local anaesthesia but although the 26 patients in the control group managed without sedation, some were very anxious, as demonstrated retrospectively by the VAS anxiety scores. All patients in the midazolam group received some drug even though their median anxiety score (VAS) was only 5 mm (range 0–38 mm) (table 3). In contrast, three patients in the propofol group with VAS anxiety scores less than 6 mm did not receive any drug and seven (27%) patients in the nonsedated group had an anxiety score of 0 mm during surgery. Therefore, 10 of 75 patients (13%) did not require any sedation and for several other patients, drug consumption was minimal (table 2). Thus as the need for sedation is unpredictable and clearly there is no requirement for routine use, a case may be made for demand sedation in all patients. In these situations security and confidence are provided by patients’ knowledge of the immediate availability of sedation should they become acutely anxious15. Moreover, intervention with sedative drugs administered i.v. by anaesthetists or surgeons can be disastrous when surgery has started. This has been demonstrated in a study where administration of infusions of either midazolam or propofol by anaesthetists during cataract surgery, to achieve a predicted level of sedation with cooperative dysarthria, caused loss of airway patency in some patients19. In one patient in the midazolam group, verbal communication was lost before peribulbar block was performed because of “stacking” of doses to a total of more than 5 mg. This is predictable in some elderly patients with slow circulation times. It is caused by the slow onset time of midazolam20 and is reversed easily with flumazenil. A similar problem during cataract surgery has been reported after administration of midazolam 20 ␮g kg1 i.m. where 33% of patients fell asleep8. The absence of a formal lock-out time implies that the time between successful demands was only 5.4 s, that is the duration of injection of each bolus. In previous studies using the same demand sedation system in healthier younger patients, none become over sedated15. In this study, no patient receiving propofol, which has a shorter onset time14 21, became over

374 sedated. Thus although midazolam is very effective at relieving anxiety and suppressing a cardiovascular response to surgery, in elderly patients small doses and a lock-out time during patient-controlled sedation are to be recommended. In terms of respiratory depression, our sedation technique compared favourably with other studies. For example, apnoea was reported in five (16.1%) patients who received fentanyl 0.5 ␮g kg1 and midazolam 0.0015 mg kg1 i.v.9. Both midazolam and propofol depress hypoxic ventilatory drive in humans22 23 and although our method of sedation appears to be better than others9 10, we recommend routine use of supplementary oxygen whenever possible, as the alternative of deep breathing may cause coughing. Three (12.5%) patients in the midazolam group were completely amnesic for all surgical events. This is less than in a previous study where premedication with temazepam 10 mg orally combined with midazolam in a mean total dose 1.41 mg for cataract surgery resulted in an incidence of amnesia of 29.4%19. There were no significant differences in the results of the psychomotor tests before and after surgery which, on average, lasted 40 min (table 2). Because of clinical constraints, the group of psychomotor tests could not be reapplied in a valid way until the postoperative period. Patients administered only small amounts of either midazolam (median 2.54 mg) or propofol (mean 87.4 mg) (table 2), approximately 50% of the dose before peribulbar block and most of the remainder at the beginning of surgery. As the psychomotor tests were performed after the end of the procedure, that is on average more than 50 min after the last dose of either drug, it is not surprising that potential small differences between the drugs were not detected. Also, in another study where drugs were administered in predetermined doses by an anaesthetist, recovery measured by the response time was similar for both midazolam and propofol24. Midazolam is the recommended anxiolytic drug for use by nonanaesthetically trained practitioners25 and has the advantage of the possibility of immediate reversibility of effect by flumazenil. Nevertheless, patients should be supervised continuously by a competent person, as has been recommended recently for propofol26.

Acknowledgements The study was funded by the Locally Organized Research Scheme, conducted at Moorfields Eye Hospital, City Road, London EC1V 2PD.

References 1. De Natale R, Flammer J. Lens opacity: a population study. International Ophthalmology 1992; 16: 1–5. 2. Klein BE, Klein R, Linton KL. Prevalence of age-related lens opacities in a population. The Beaver Dam eye study. Ophthalmology 1992; 99: 546–552. 3. Ferguson JA, Goldacre MJ, Henderson J, Bron AJ. Ophthalmology in the Oxford region: Analysis of time trends from

British Journal of Anaesthesia linked statistics. Eye 1991; 5: 379–384. 4. Batterbury M, Khaw PT, Hands R, Elkington AR. The cataract explosion: The changing pattern of diagnoses of patients attending an ophthalmic outpatient department. Eye 1991; 5: 369–372. 5. Strong NP, Wigmore W, Smithson S, Rhodes S, Woodruff G, Rosenthal AR. Daycase cataract surgery. British Journal of Ophthalmology 1991; 75: 731–733. 6. Watts WT, Pearce JL. Day-case cataract surgery. British Journal of Ophthalmology 1988; 72: 897–899. 7. Gobeaux D, Sardnal F. Midazolam and flumazenil in ophthalmology. Acta Anaesthesiology Scandinavia 1990; 34 (Suppl. 92): 35–38. 8. Virkkila MEJ, Ali-Melkkila TM, Kanto JH. Premedication for outpatient cataract surgery: a comparative study of intramuscular alfentanil, midazolam and placebo. Acta Anaesthesiologica Scandinavica 1992; 36: 559–563. 9. Chung F, Cheng DC, Seyone C, Dyck BJ. A randomised comparison of midazolam and diazepam injectable emulsion in cataract surgery. Canadian Journal of Anaesthesia 1990; 37: 528–33. 10. Forrest JB, Lam L, Woo J, Rifkind A. Oxygen desaturation in elderly patients during cataract surgery. Canadian Journal of Anaesthesia 1990; 37: S50. 11. Osborne GA, Rudkin GE, Curtis NJ, Vickers D, Cracker AJ. Intra-operative patientcontrolled sedation. Comparison of patient-controlled sedation with anaesthetist administered midazolam and fentanyl. Anaesthesia 1991; 46: 553–556. 12. Park WK, Watkins PA. Patient-controlled sedation during extradural anesthesia. Anesthesia and Analgesia 1991; 72: 304– 307. 13. Rudkin GE, Osborne GA, Curtis NJ. Intra-operative patientcontrolled sedation. Anaesthesia 1991; 46: 90–92. 14. Rudkin GE, Osborne GA, Finn BP, Jarvis DA, Vickers D. Intra-operative patientcontrolled sedation. Comparison of patient-controlled propofol with patient-controlled midazolam. Anaesthesia 1992; 47: 376–381. 15. Cook LB, Lockwood GG, Moore CM, Whitwam JG. True patient-controlled sedation. Anaesthesia 1993; 48: 1039–1044. 16. Whitwam JG, Wang C. Sedation and sedoanalgesia. In: Whitwam JG, ed. Day-case Anaesthesia. Oxford: Blackwell Scientific, 1994; 262–274. 17. Cook LB, Whitwam JG. Patient controlled sedation. In: Whitwam J, ed. Day-case anaesthesia. Oxford: Blackwell Scientific, 1994; 275–288. 18. Mein J, Harcourt B. Ocular movements. In: Mein J, Hartcourt B, eds. Diagnosis and Management of Ocular Motility Disorders. Oxford: Blackwell Scientific, 1986; 64–86. 19. Salmon JF, Mets B, James MF, Murray AD. Intravenous sedation for ocular surgery under local anaesthesia. British Journal of Ophthalmology 1992; 76: 598–601. 20. Reves JR, Fragen RJ, Vinik HR, Greenblatt DJ. Midazolam pharmacology and uses. Anesthesiology 1985; 62: 310–324. 21. Sebel PS, Lowdon JD. Propofol: A new intravenous anesthetic. Anesthesiology 1989; 71: 260–277. 22. Alexander CM, Gross JB. Sedative doses of midazolam depress hypoxic ventilatory responses in humans. Anesthesia and Analgesia 1988; 67: 377–382. 23. Blouin RT, Seifert HA, Babenco HD, Conard PF, Gross JB. Propofol depresses the hypoxic ventilatory response during conscious sedation and isohypercapnia. Anesthesiology 1993; 79: 1177–1182. 24. Polster MR, Gray PA, O'Sullivan G, McCarthy RA, Park GR. Comparison of sedative and amnesic effects of midazolam and propofol. British Journal of Anaesthesia 1993; 70: 612–616. 25. Royal College of Surgeons of England. Report of working party. Guidelines for sedation by non-anaesthetist. London: Royal College of Surgeons of England, 1993. 26. Thorpe SJ, Balkvishnan VR, Cook LB. Observations on patient-controlled administration of propofol. British Journal of Anaesthesia 1996; 76: 579P.