Conscious sedation is an accepted method of anxiety control in dentistry. Its use in the UK is likely to increase following the recent guidelines issued by the General Dental Council and the Department of Health which encourage replacement of general anaesthesia with utilization of sedative techniques in anxious patients [1,2]. Midazolam is currently the most widely used intravenous (i.v.) sedative drug as it has a rapid onset, a short elimination half-life and does not have any significant pharmacologically active metabolites . However, benzodiazepines exhibit wide interindividual variability in both pharmacokinetic and pharmacodynamic responses . In addition, patients vary with regard to normal baseline anxiety levels (trait anxiety) and in their response to anxiogenic stimuli such as dental treatment (state anxiety) . As a result of this variation in response, the usual method of administering these drugs is by injecting small increments until the patient appears to be adequately sedated.
Patient-controlled analgesia is already an established technique for the relief of postoperative pain. It provides effective analgesia by matching the opioid dose to individual patient requirements [5,6]. We have not been able to find any previous study that evaluates the feasibility of patient-controlled sedation with midazolam in patients undergoing restorative dentistry. The objective of this study was to determine whether patient-controlled sedation was a suitable alternative to operator-controlled sedation in restorative dentistry.
Institutional Ethics Committee approval was obtained and all patients gave informed written consent to participate in the study. Forty ASA I-II patients requiring i.v. sedation for restorative dentistry were recruited to the trial. Patients were randomly allocated to receive i.v. midazolam, given by the operator (operator-controlled sedation) or else to administer it themselves. Self-administration (patient-controlled sedation) involved use of a Graseby 3000® syringe pump (Graseby Medical Limited, Herts, UK), on the first visit and the alternate method on the second visit. As both methods of sedation were quite different it was not possible for the study to be double-blinded. However, the operator administering the sedation had no prior knowledge of the dose of midazolam used on the first visit. Each patient received a full explanation on the use of the syringe pump. The treatment was carried out by two of the authors (K. A. Bavisha and A. R. Leon) in the Restorative Dentistry Hospital Unit of The Royal London Hospital. The type of local anaesthesia and the operator was the same and the dental procedure was similar on both visits. Blood pressure (BP), heart rate (HR) and arterial oxygen saturation were recorded using a Dinamap® Compact TS (Johnson & Johnson Medical Inc., Tampa, FL) integrated monitor after the patient had (a) settled into the dental chair; (b) after the first dose of midazolam and (c) at the time of sedation. These values were also recorded while (d) the local anaesthetic was administered; (e) during placement of the rubber dam in the mouth; (f) during drilling and (g) restoration of the tooth. The values were also recorded at (h) the end of the dental procedure and again (i) when the patient was judged fit for discharge.
With operator-controlled sedation the midazolam was given in 1 mg increments with a 1 min interval until the operator judged that the patient was suitably sedated for treatment to start. With patient-controlled sedation a Graseby syringe pump was used to deliver midazolam 1 mg on every press of the button, but with a lockout period of 1 min. With both methods the dose of midazolam and time required to achieve adequate sedation were recorded. Further increments, total dose used, sedation scores, the duration of the dental procedure and time to fitness for discharge were also noted. Both the dentist and an anaesthetist observer graded the level of sedation as: 1, patient awake and co-operative; 2, patient visibly drowsy; 3, eyes closed, responsive to verbal instruction; 4, eyes closed, responsive to physical stimulation; 5, eyes closed, unresponsive to physical stimulation. At the end of treatment the level of consciousness and ability to walk in a straight line without swaying were assessed at intervals. Patients were judged fit for discharge when they appeared and felt themselves to be fully alert and were able to walk unaided. On each occasion, at this stage the patient was asked to complete a questionnaire relating to their experience during treatment under sedation. They were then discharged into the care of a responsible adult and provided with standard written post sedation instructions regarding their activity over the following 24 h.
Statistical analyses were with analysis of variance, paired t-test, unpaired t-test and McNemar's test where appropriate. The results are given as mean ± SD.
Of the 40 patients recruited 35 completed the trial - the other five patients did not return to the hospital to complete treatment. The average age of this study population was 33.3 (±7.9 SD) yr ranging from 20 to 48 yr. The majority of patients were females (n = 21). The mean dose of midazolam required to induce adequate sedation, the total dose administered, the mean sedation score at adequate sedation and the mean sedation score during treatment are shown in Table 1. The mean initial dose to induce sedation was 3.6 mg (±1.9 SD) in the patient-controlled sedation group and 3.8 mg (±1.5 SD) in the operator-controlled sedation group, the mean time to the onset of adequate sedation for the patient-controlled sedation group was 6.4 (±2.7 SD) min and for the operator-controlled sedation group it was 6.2 (±3.6 SD) min, P > 0.05. The mean total dose of midazolam administered was 7.9 (±4.2 SD) mg in the patient-controlled sedation group and 4.2 mg (±1.8 SD) in the operator-controlled sedation group, P < 0.05. The mean sedation score at the start of treatment was 1.7 (±0.9 SD) in the patient-controlled sedation group and 2.0 (±0.9 SD) in the operator-controlled sedation group and the mean sedation score during treatment was 2.5 (±1.1 SD) in the patient-controlled sedation group and 2.1 (±1.0 SD) in the operator-controlled sedation group. The maximum sedation score throughout treatment for both methods was four and this only occurred in six patients; five occurred in the patient-controlled sedation group and one in the operator-controlled sedation group. No patient had a sedation score of 5. There was no significant difference between the sedation scores between the two techniques.
Induction, operating and recovery times are summarized in Table 1. The mean time to complete dental treatment was 39.3 min (±9.3 SD) for the patient-controlled sedation group and 32.1 min (±12.0 SD) for the operator-controlled sedation group (P > 0.05). The mean time from the end of treatment to fitness for discharge was 15.37 min (±11.9 SD) for the patient-controlled sedation group and 8.54 min (±9.5 SD) for the operator-controlled sedation group (P < 0.05). The mean time from adequate sedation to fitness for discharge was 53.2 min (±16.3 SD) for the patient-controlled sedation group and 47.7 min (±12.0 SD) for the operator-controlled sedation group (P < 0.05).
Figure 1 shows the pulse oximeter readings during patient-controlled sedation and operator-controlled sedation recorded at specific stages of the procedure. There were no significant differences in SPO2 readings between the two groups at any point during the procedure (P > 0.05). No supplemental oxygen was administered to any patients during the procedures. Figure 2 shows the systolic and diastolic BP and the HR at various stages of treatment. There was no significant difference in these variables between the two groups (P > 0.05). The results of the patient questionnaire are shown in Table 2. There was no difference in recall of the injection of local anaesthetic between the two methods but significantly fewer patients receiving patient-controlled sedation remembered the drilling, (P ≤ 0.001). Significantly more patients receiving patient-controlled sedation felt relaxed throughout the treatment (P ≤ 0.038).
A number of studies have compared patient-controlled sedation with operator-controlled sedation using midazolam [7,8] but there have been no studies comparing the two methods in restorative dentistry using the crossover design. Our results were similar in terms of physiological variables to those found by previous investigators during third molar extraction . Arterial oxygen saturation remained above 90% in all patients and there was no significant difference in saturation between the two methods. No supplemental oxygen was given during the sedation period and oxygen administration was not required at any stage. Following the onset of sedation there was an insignificant fall in BP and no difference in cardiovascular stability between the two methods. There was no difference in the initial dose of midazolam required to induce sedation and it was notable that the mean time to the onset of sedation was similar for both methods suggesting that patient-controlled sedation did not require more time. This is in contrast to the finding of Rodrigo and Tong  who found an onset of patient-controlled sedation of 9 min compared with 5 min with operator-controlled sedation. However, we used a total dose of midazolam that was significantly higher in the patient-controlled sedation group than in the operator-controlled sedation group. A previous study comparing patient-controlled sedation with operator-controlled sedation  found no difference in total midazolam dose administered during third molar surgery, but a more recent similar study  from the same centre found that higher doses of midazolam had to be administered to the patient-controlled sedation group. The same authors postulated that many operator-sedationists aim to adhere to the guidelines of the Report of an Expert Working Party (Poswillo Report)  which recommended that midazolam should be administered as a titrated dose to a defined end-point. Another reason for limiting the use of incremental doses of midazolam is that some dental units do not have purpose built recovery areas and a prolonged recovery period reduces the number of patients who can be treated. In our study the time to discharge in the patient-controlled group was almost double that in the operator-controlled group.
The results of Rodrigo and Chow 1996  suggest that a single dose technique may not be adequate for the needs of all patients. We found that more patients in the patient-controlled sedation group felt relaxed throughout treatment, but they gave themselves more midazolam. This could suggest that some patients might have felt inadequately sedated in the operator-controlled sedation group. It is more likely that nervous patients, who would prefer to be unconscious during treatment, would administer greater doses of midazolam in order to achieve deeper levels of sedation. An additional possible reason for the higher incidence of patients feeling relaxed could be the fact that patient-controlled sedation allows a degree of behavioural, cognitive and decisional control over events which may manifest itself in the lessening of anxiety .
A major concern with patient-controlled sedation has always been the risk that patients may render themselves unresponsive. In this study the mean maximum sedation scores were 2.5 for the patient-controlled sedation group and 2.1 for the operator-controlled sedation group and showed no statistical difference between the two methods. In the six cases where the sedation score of four was found (five in the patient-controlled sedation group and one in the operator-controlled sedation group), dental treatment ceased until the patients responded to verbal command. In no patient was this longer than 2 min and in no case was any active intervention required.
We have shown in this study, using a crossover design, that patient-controlled sedation is a suitable alternative to operator-controlled sedation during restorative dentistry. The initial dose required to induce sedation was similar for both methods, but more midazolam was administered by the patient-controlled sedation group during the procedure and more of the patients in the patient-controlled sedation group felt relaxed throughout treatment. The sedation scores were similar with the two methods, which implied that these patients did not indiscriminately continue administering the drug to produce loss of consciousness.
1. General Dental Council. Maintaining Standards, Guidance to Dentists on Professional and Personal Conduct.
London, UK: General Dental Council, Revised November 2001.
2. A Conscious Decision: Report of an Expert Group.
London, UK: Department of Health, July 2000.
3. Dundee JW, Halliday NJ, Harper KW, Brogden RN. Midazolam: a review of its pharmacological properties and therapeutic use. Drugs
4. Hillestad L, Hansen T, Melsom H, Drivenes A. Diazepam metabolism in normal man. 1. Serum concentrations and clinical effects after intravenous, intramuscular, and oral administration. Clin Pharmacol Ther
5. Corman HH, Hornick EJ, Kritchman M, Terestman N. Emotional reactions of surgical patients to hospitalization, anesthesia and surgery. Am J Surg
6. Sechzer PH. Objective measurement of pain. Anesthesiology
7. Osborne GA, Rudkin GE, Curtis NJ, Vickers D, Craker AJ. Intra-operative patient-controlled sedation. Comparison of patient-controlled propofol with anaesthetist-administered midazolam and fentanyl. Anaesthesia
8. Zacharias M, Hunter KM, Luyk NH. Patient-controlled sedation using midazolam. Br J Oral Maxillofac Surg
9. Rodrigo MR, Tong CK. A comparison of patient and anaesthetist controlled midazolam sedation for dental surgery. Anaesthesia
10. Rodrigo C, Chow KC. Patient-controlled sedation: a comparison of sedation prior to and until the end of minor oral surgery. Aust Dent J
11. Poswillo D. General Anaesthesia, Sedation and Resuscitation in Dentistry. Report of an Expert Working Party
1990. Prepared for the Standing Dental Advisory Committee. London, UK: Department of Health.
12. Dell R. A review of patient-controlled sedation. Eur J Anaesthesiol