Sole Use of Dexmedetomidine Has Limited Utility for Conscious Sedation during Outpatient Colonoscopy

This prospective, randomized, single-blind study was conducted with a population of patients undergoing ambulatory elective colonoscopy. The study was approved by the Ethical Committee of the Silesian University School of Medicine (Katowice, Poland), and written, informed consent was obtained from all of the participants. We intended to enroll 90 patients, but the study was interrupted before the planned number of patients had been recruited because of adverse events (described in the Results section).

A total of 64 adult patients (aged 18–60 yr; American Society of Anesthesiologists physical status I and II) were studied. The exclusion criteria were as follows: pregnancy, history of colon surgery, psychiatric or emotional disorder, chronic use of or addiction to opiates or sedatives, previous adverse reactions to any medication used in this study, and pain unrelated to the gastrointestinal tract.

A computer-generated randomization list was used to assign patients to one of three study groups. In the first group (group D, n = 19) patients received conscious sedation with an intravenous infusion of 1 μg/kg dexmedetomidine (Precedex®; Abbott Laboratories, North Chicago, IL) administered via  a syringe infusion pump (Terufusion TE-312®; Terumo, Tokyo, Japan) over a 15-min period before colonoscopy, followed by an infusion of 0.2 μg · kg⁻¹· h⁻¹. Drug infusion was discontinued if one of the following adverse events was observed: apnea lasting longer than 30 s, hemoglobin oxygen saturation (Spo ²) lower than 90%, decrease of heart rate (HR) below 50 beasts/min, mean arterial pressure (MAP) below 70% of its initial value, and sedation that made verbal contact with the patient impossible.

In the second study group (group P, n = 21), patients received 1 mg/kg intravenous meperidine (Dolargan®; Chinoin, Budapest, Hungary) and 0.05 mg/kg intravenous midazolam (Sopodorm®; Solco-Basel, Basel, Switzerland) before the colonoscopy.

In the third study group (group F, n = 24), the patients were not sedated before the procedure but were given 0.1–0.2 mg intravenous fentanyl (Fentanyl®; Polfa-Warsaw, Warsaw, Poland) in response to pain. Oxygen was delivered by Venturi facemask (fraction of inspired oxygen = 0.35) to all patients throughout the procedure. A jaw thrust maneuver was applied to the patients if necessary. All colonoscopies were performed by the same experienced endoscopist with a video colonoscope (EC 3840 FK2; Pentax Corp., Tokyo, Japan). Sedation and monitoring were performed by the same anesthesiologist in all cases.

The following parameters were measured continuously and recorded every minute: HR, MAP, Spo ², and respiratory rate (DASH 3000® monitor; GE Marquette Medical Systems, Milwaukee, WI). The recorded data were analyzed and averaged over the following time intervals: baseline—at least 3 min before the first drug administration or starting colonoscopy; colonoscopy—during colonoscopy; and postprocedure—from the end of the colonoscopy to the time at which a score of 10 points on the Modified Post Anesthesia Discharge Scoring System (MPADSS) scale was reached (table 1).11 

The quality of analgesia was assessed using the 11-point (0–10) Numerical Pain Rating Scale (NRS),12in which 0 represents no pain at all and 10 represents the worst pain imaginable. If the patient reported pain exceeding 6 on the NRS, intravenous fentanyl in doses of 0.1 mg was administered.

The evaluation of the quality of sedation was based on a five-point Observer’s Assessment of Alertness/Sedation Scale.13The level of recovery from anesthesia and the return of psychomotor fitness were studied using the MPADSS scale (table 1). The duration of colonoscopy and the time from the end of the procedure to the moment of scoring 9 and 10 on the MPADSS scale were measured. The discharge criteria required that the patients be awake and alert with stable vital signs, be able to ambulate without assistance, and be free of side effects (i.e. , a score of 10 on the MPADSS scale).

Administration of any medication apart from the study protocol and occurrences of complications and side effects (strong vertigo, nausea, vomiting) were recorded. In case of serious adverse events, the study protocol provided for hospital admission to an observation ward for minimum of 12 h. Patients were also asked about their willingness to undergo a repeat procedure with the same sedation regimen in the future if required.

Results were expressed as numbers of occurrences, percentages, and mean ± SD for continuous variables. In some cases, the data were presented as percent changes for clarity. Categorical data were analyzed using the chi-square test with a Yates correction or Fisher exact test, where appropriate. NRS scores were not normally distributed and were compared between groups using a nonparametric Kruskal-Wallis test. The differences in continuous parameters such as patient characteristics, preoperative data, time intervals, and amounts of supplemental fentanyl were analyzed using one-way analysis of variance. Repeated-measures analysis of variance was used to test for the difference between groups in HR, MAP, respiratory rate, and Spo ²over time. Correction with post hoc  tests (Bonferroni method) was used because repeated measurements of a single variable were tested over time. A P  value of less than 0.05 was considered significant. Statistical analyses were performed using Statistica for Windows version 6.0 software (StatSoft, Tulsa, OK).

The study groups were comparable regarding demographic characteristics, indications for colonoscopy, initial vital signs, and American Society of Anesthesiologists physical status (table 2). In all patients, endoscopy was completed. Because of adverse events in group D, enrollment was halted after 64 subjects. In all groups, negligible Spo ²and respiratory rate variations were observed (fig. 1). A jaw thrust maneuver had to be applied in 6 of the cases (29%) in group P. This was not required in groups D and F.

In group D, significantly larger HR and MAP decreases were observed during and after colonoscopy as compared with groups F and P (fig. 1). In group D, the postcolonoscopy MAP values were on average 26% lower than baseline (vs.  14% in group P and 3% in group F). Blood pressure decreased to approximately 50% of the initial value in four cases in group D. One of these patients was taking perindopril because of hypertension. In contrast, the maximum change in MAP did not exceeded 35% in group P and 30% in group F. Similarly, in group D, the postcolonoscopy mean HR was 17% lower as compared with preprocedure value readings (vs.  9% in group P and 7% in group F). In two cases from group D (one with mild ischemic heart disease treated with pentaerythritol and trimetazidine), HR decreased from 77 and 86 beats/min to 40 beats/min. In the other study groups, the minimum HR was not lower than 50 beast/min. In four cases in group D, bradycardia was treated with atropine.

The lowest Observer’s Assessment of Alertness/Sedation Scale score during the study was 3 points in all groups. The NRS score was not different among the study groups (P = 0.951; fig. 2). The average use of supplemental fentanyl in cases of inadequate analgesia was 0.04 ± 0.05 mg in group P, which was significantly lower than that in group D (0.1 ± 0.1 mg) and group F (0.18 ± 0.1 mg) (P = 0.0005). Fentanyl was administered in 9 cases in both groups D and P and to 19 patients in group F. Vertigo and nausea, followed by vomiting treated with ondansetron, occurred in 5 cases in group D. No vertigo or nausea was observed in groups P and F.

The average duration of colonoscopy in group D (13 ± 8.6 min) was significantly longer than in groups P (9 ± 4.8 min) and F (8 ± 4.7 min) (P = 0.0038). In group D, the time required to reach home discharge readiness (10 points on the MPADSS scale) was significantly longer (85 ± 74 min) compared with the other groups (39 ± 21 min for group P and 32 ± 13 min for group F; P = 0.007; fig. 3). Group D patients were delayed in achieving MPADSS scores of both 9 and 10 (i.e. , ready for discharge to home) compared with the other groups (fig. 3). Discharge was delayed in group D patients because of the following factors: prolonged drowsiness (six cases), dizziness (two cases), nausea/vomiting (three cases), low blood pressure (two cases).

According to the study protocol, three patients from group D were admitted to the hospital for 12-h observation because of safety concerns. One patient, with no previous history of cardiac disease or symptomatic arrhythmia, experienced ventricular bigeminy with symptoms of chest pressure, blood pressure of 80/50 mmHg, and near-syncope. Two other patients (one with an American Society of Anesthesiologists physical status of I and one with an American Society of Anesthesiologists physical status of II due to diabetes) experienced profound hypotension accompanied by sustained nausea and vomiting that was resistant to ondansetron therapy. Oxygen via  facemask, intravenous fluid therapy, and ephedrine was implemented.

One patient in group F (4.5%) and four patients in group D (22.2%) stated that they would decline the same treatment during future colonoscopies, whereas in group P, all patients stated they would opt for the same pain treatment procedure.

Methods used to assure patient comfort during painful and embarrassing diagnostic procedures, such as colonoscopy, should focus on effective pain relief, safety, lack of recall of unpleasant aspects of the procedure, and, finally, on rapid patient recovery and readiness for home discharge.

The results of this study indicate that all techniques studied provided adequate pain relief in patients during colonoscopy. The level of analgesia was similar in all study groups, and the NRS score was less than 4. This result compares favorably with the analgesic effects of other commonly used methods of analgesia during colonoscopy.14,15Unfortunately, in 47% of cases, adequate pain relief in patients receiving dexmedetomidine could be achieved only with supplemental fentanyl. Reports suggest that dexmedetomidine can produce general anesthesia when administered in high doses that are often associated with bradycardia, hypotension, and prolonged recovery.16 

The impetus for exploring the use of dexmedetomidine during ambulatory colonoscopy was the possibility that adequate sedation could be provided with minimal respiratory depression.9,17This study confirmed the lack of clinically significant respiratory effects, including apnea, hypoxemia, or airway obstruction.

It has been suggested that dexmedetomidine may have favorable hemodynamic properties in patients with moderate hypertension and with coronary heart disease.3,4We found that a dose sufficient to produce sedation caused a statistically significant decrease in arterial blood pressure and bradycardia during colonoscopy. These reactions were more profound than those observed in the other study groups. Vasovagal reactions have been reported in 0.8% of unsedated patients undergoing colonoscopy as a consequence of stretching of the colon and mesenteric attachments from looping of the instrument shaft.18It is possible that dexmedetomidine may have exacerbated these vagal effects.19In our study, pronounced bradycardia and hypotension was relatively common in patients receiving dexmedetomidine. Hemodynamic instability was also observed after the procedure. Similar hemodynamic effects can be observed in patients sedated with dexmedetomidine in the intensive care unit and may represent a significant limitation in the clinical use of this agent.20,21Pronounced hypotension and bradycardia of sufficient severity to necessitate intensive medical interventions and prolonged observation occurred in 3 of 19 patients receiving dexmedetomidine and was the reason for prematurely halting study enrollment.

Another important and frequently discussed factor influencing the general evaluation of ambulatory practice is rapid home discharge. In this study, the times to discharge readiness were significantly longer when dexmedetomidine was used, in some patients requiring several hours. This observation reflects the pharmacokinetic properties of dexmedetomidine, which has an elimination half time of approximately 2 h.10The most frequent reasons for delay in reaching the target MPADSS score in this group included variations in arterial pressure and HR exceeding 20–40% of baseline, prolonged sleepiness, weakness, and nausea. The increased frequency of the adverse events in patients receiving dexmedetomidine can only be partially explained by the use of supplemental opiate because fentanyl was administered in cases of inadequate analgesia in all three study groups.

In the current investigation, the duration of the procedure was significantly longer in patients receiving dexmedetomidine therapy as compared with the other study groups. The adequacy of sedation affects both the technical aspects and the duration of colonoscopy.22Dexmedetomidine may also inhibit intestinal transit,23which could cause technical difficulties with colonoscopy. The study protocol was not designed to assess the satisfaction of the endoscopist.

In conclusion, in patients undergoing colonoscopy, dexmedetomidine provides a relatively satisfactory level of analgesia and sedation without clinically notable adverse respiratory effects. However, compared with commonly used sedation regimens, dexmedetomidine is associated with the frequent requirement for supplemental fentanyl, sometimes profound hypotension and bradycardia, and prolonged recovery time. These side effects may limit its usefulness for this indication.