INTRODUCTION

The use of sedatives in gastrointestinal endoscopies is known to reduce patient discomfort and anxiety as well as increasing tolerance and acceptance. This in turn increases the satisfaction of the endoscopist due to the reduction in patient movement, which allows for a better visualization of the mucosae and also decreases therapeutic risk ¹. In the absence of a specific sedation protocol, numerous studies have been conducted with currently available sedatives and relevant guidelines have been published ^1,2.

Benzodiazepines, either alone or combined with opioids, have been used for decades and constitute the traditional regimen for sedation in the clinical practice ^3,4. More recently, propofol has gained popularity due to its pharmacokinetic characteristics. Despite the extensive use of propofol, caution should be taken when administrated by non-anesthetists as there is no antagonist and there is also a potential to induce general anesthesia. Although, its safety in the hands of trained endoscopists has been demonstrated in numerous studies ^{5,6,7,8,9,10,11,12}.

In our experience, performing diagnostic gastroscopies with low doses of midazolam in combination with propofol can provide advantages, including both the reduction of the required dose of propofol without increasing adverse effects as well as maintaining high levels of patient satisfaction. From this perspective, this study was performed to compare two strategies for intravenous (IV) sedation in diagnostic upper gastrointestinal endoscopy (UGE).

PATIENTS AND METHODS

A prospective, randomized and double-blind study was performed in which the use of propofol alone was compared with the use of midazolam plus propofol in a diagnostic upper GI endoscopy (UGE) diagnoses. The main objectives were to evaluate safety and efficiency, while the secondary objectives included the evaluation of the quality of the endoscopy and patient satisfaction as follows:

• Safety (frequency of the following complications: hypoxemia, bradycardia and hypotension). Systolic blood pressure below 80 or a 30% decrease from the baseline value was considered as hypotension. A decrease greater than or equal to 20% of the normal heart rate was considered as bradycardia. Hypoxia was defined as oxygen saturation below 90%.

• Efficiency (the time lapse between the initial administration of any type of sedation IV and patient discharge from the hospital). The following parameters were measured and recorded during the procedure in order to evaluate the impact of sedation on efficiency in our Endoscopy Unit:

• Quality of the endoscopy (the evaluation and score of the exploration by two endoscopists according to the appropriate visualization). Two independent physicians evaluated the exploration, the endoscopist who carried out the endoscopy and a second gastroenterologist who was present during the procedure but not involved in the endoscopic process. A non-validated scale was used as a reference as described by Meining et al. (13) in 2007. The following steps (or parameters) were identified during the performance of UGE: P1: passage of the endoscope through the throat; P2: visualization of the esophagus; P3: assessment of the proximal cardiac folds; P4: passage through the stomach towards the pylorus along the greater curvature; P5: passage through the pyloric sphincter; P6: a complete evaluation of the duodenal bulb; P7: introduction of the scope and visualization of the descending duodenum; P8: complete evaluation of duodenal folds; P9: a complete visualization of the antrum; P10: visualization of the angular incisures; P11: execution of the retroflexion maneuver; P12: retroflexion visualization of the fundus and cardia; P13: visualization of the body and lesser curvature and P14: removal through the esophagus.

• Each step of the endoscopy was evaluated subjectively by the endoscopists on a quality scale ranging from 1 to 6 (6 excellent; 5 good; 4 moderate; 3 sufficient; 2 poor and 1 unacceptable) with the purpose of discovering if the quality of the endoscopy was significantly influenced by the sedative used. The overall scores for the complete examination of the three separate organs (esophagus, stomach and duodenum) were also included in the evaluation as follows: P15: overall general assessment of the upper gastrointestinal tract; P16: general assessment of the esophagus; P17: general assessment of the stomach and P18: general assessment of the duodenum.

• Patient satisfaction (the opinion of the patient was expressed in a questionnaire that was completed before discharge). In order to measure patient comfort after the examination, participants were given a survey to be completed before discharge with regard to the quality of sedation and the general experience during the procedure. A visual analogical scale was used that ranged from zero (very unsatisfactory) to 10 (completely satisfactory). This questionnaire also evaluated the level of amnesia and when appropriate, the patients' willingness to undergo the same procedure in the future (Fig. 1).

Fig. 1 Survey filled out by patients before being discharged. 

RESULTS

Patients and medication used

Eighty-three patients were included in the study and randomized. Of these, 42 received the placebo plus propofol (group A) and 41 received midazolam plus propofol (group B). There were no significant differences between the two groups in terms of age, weight, sex and ASA (Table 1). The most common symptoms that led to an UGE exploration were epigastric pain and dyspepsia and there were no significant differences between the two groups. The propofol dose was significantly lower (p < 0.01) in group B patients (midazolam plus propofol) in comparison to group A patients (placebo plus propofol). The average dose of propofol was 115.90 mg (SD 40.57) in group A and 83 mg (SD 40.34) in group B (Table 1).

Table 1 Patient characteristics. Safety profile 

The characteristics of the patients were similar in both groups and the dose of propofol was significantly higher in group A. There were no differences in the safety profile. ns: not significant; SD: standard deviation.

Analysis of time in relation to the procedure

The average induction time (time from commencing sedation to the initiation of the exploration) in the placebo plus propofol group (group A) was 4.07 minutes (SD 1.65) and 3.00 minutes (SD 1.30) in the midazolam plus propofol group (group B), which was statistically significant (p = 0.001). However, the times of the duration of the procedure, recovery and overall duration of the UGE were not significantly different (Fig. 2).

Fig. 2 Average time expressed in minutes of Induction time: defined as the time elapsed between initiating sedation and beginning the exploration. Sedation time: the time between the administrations of the first sedation dose to the final dose. Duration of the procedure: defined as the time between the passage of the endoscope through the throat until its complete removal. Total time of the UGE from the initiation of sedation to the complete removal of the endoscope. Recovery time: the time between the end of the endoscopy procedure and patient discharge. Overall duration of the procedure: defined as the time elapsed from the administration of the first sedation dose until recovery and patient discharge (m: minutes; ns: not significant). 

Post-procedure evaluation: quality of the exploration and patient satisfaction

The quality of the exploration according to the score of the endoscopic evaluation evaluated by physicians was similar and there were no significant differences between the two groups, regardless of the sedative administered (Table 2). The acceptance rate of the patients was similar in both groups. The average satisfaction score was 9.85 and 9.92 (SD 0.52 and 0.35) out of 10 in groups A and B, respectively. In group A, 3 patients (7.3%) had some memory of the endoscopy; one remembered the entry of the endoscope and two remembered the removal. None of the patients in group B remembered the endoscopy. Although it was not an objective of the study a priori, it is worth noting that both groups had similar amnesic properties. Forty patients (97.6%) in group A and 39 patients (100%) in group B expressed their willingness to repeat the procedure in the future if it were necessary, indicating their satisfaction with the administered sedation regimen.

Table 2 The median scores awarded by each physician in each step of the endoscopy 

P1: passage of the endoscope through the throat; P2: visualization of the esophagus; P3: assessment of the proximal cardiac folds; P4: passage through the stomach towards the pylorus along the greater curvature; P5: passage through the pyloric sphincter; P6: a complete evaluation of the duodenal bulb; P7: introduction of the scope and visualization of the descending duodenum; P8: complete evaluation of duodenal folds; P9: a complete visualization of the antrum; P10: visualization of the angular incisures; P11: execution of the retroflexion maneuver; P12: retroflexion visualization of the fundus and cardia; P13: visualization of the body and lesser curvature and P14: removal through the esophagus. P15: overall general assessment of the upper gastrointestinal tract; P16: general assessment of the esophagus; P17: general assessment of the stomach and P18: general assessment of the duodenum. Score range: 1-6 (1: unacceptable-6: excellent). ns: not significant.

DISCUSSION

In recent years, many studies and meta-analyses have compared traditional sedation with sedation based on the administration of propofol ^18,19,20. Some disadvantages of propofol such as over-sedation, negative cardio-respiratory effects or pain in the injected area, especially if the vein used is small ^21,22,23,24, have led to the study of propofol in association with small doses of opioids or benzodiazepines. These evoke a synergetic action, without increasing the side effects and reduce the total required dose of propofol ^25,26. This protocol, which aims to achieve a moderate-profound sedation level, has been described in the recent medical literature as "balanced propofol sedation" (BPS) ^27,28,29 and has been compared to propofol administration as the only sedative ^{29,30,31,32,33}.

The combination of midazolam and propofol in our study diminished the required dose of propofol significantly. Group A were administered 115 mg, compared to an average of 83 mg in group B. However, the average dose of propofol in group A (115 mg) was lower than that reflected by other studies in which propofol was the only sedative agent ^13,29,34,35. Although, it must be noted that their patients achieved a deep sedation level. With regard to midazolam, there was a tendency to use lower doses in most other studies than that used in this study in combination with propofol ^12,29,30. The dose of midazolam in the present study was established based on the current overall experience of the professionals within our unit. The same is true with regard to the chosen dose of propofol.

When both sedation regimens were compared, we found that the adverse effects were scarce and hardly relevant and comparable between the groups. Depressed respiration and adverse cardiovascular effects such as hypotension or bradycardia were principally associated with propofol. Bradycardia constituted the most frequent adverse effect in our series, particularly in the propofol group and occurred in 7 (17.1%) and 3 (7.5%) patients in groups A and B, respectively. This data proves the safety of both regimens previously mentioned in other reports which compared the administration of propofol alone with BPS ^{12,29,30,31,32,33}.

Our data shows that the BPS protocol that combines low doses of midazolam with propofol (group B), achieved a quicker sedation than propofol administered as monotherapy (group A). Although the induction time difference was statistically significant, it was an isolated element as there were no other statistically significant differences in the duration of the procedure, recovery time or the overall duration of the UGE. The induction time with midazolam plus propofol (BPS) in group B was three minutes and was similar ^27,28 or slightly higher ^29,30,33 than that reported by other studies which also used BPS. Cohen et al. ³⁶ found that the addition of narcotic and benzodiazepine to propofol did not appreciably alter the induction time compared with other studies which used only propofol. In group A (placebo plus propofol), the sedation induction time (4.07 minutes) was very similar to that described in other studies ^30,33,18. However, in general, data collected from other studies that used propofol as a monotherapy report lower induction times (2-3 minutes) ^29,34. Currently, the tendency is to begin with an initial bolus of 0.5-1 mg/kg and therefore it is likely that the induction time would have been less in group A. In this sense, it could be argued that after administering midazolam, we should have waited 2-3 minutes for it to take effect, which would result in a prolonged induction time in group B. However, if we had waited 2-3 minutes after administering the placebo in one group and midazolam in the other, the blinded nature of the study would have been jeopardized, as the endoscopist would have been aware of the greater sedation level of patients in group B before the second bolus was administered. In any case, the time elapsed between the initiation of sedation to the introduction of the endoscope (4 or 3 minutes in each group A and B, respectively) was sufficient for midazolam to begin to take effect.

Although the recovery time for the patient was shorter using propofol alone, the differences were not significant. Other studies have indeed described a synergic reaction when combining propofol with midazolam. Thus prolonging the patients´ recovery time significantly (30-33). In this sense, the study carried out by Levitzky et al. demonstrates the substantial advantage of BPS over traditional sedatives ^27).

The quality of an endoscopic evaluation is essential in order to provide a precise diagnosis. Along these lines, we concluded that there were no differences between the two regimens with regard to achieving an adequate endoscopic visualization associated with the tolerance of insufflations. Meining et al. ¹³ used an identical scale to compare the administration of propofol in one arm and midazolam in the other. This study concluded that sedation with propofol greatly improved the quality of the endoscopy, while midazolam caused more belching, retching and a lower tolerance of inflation of the stomach.

Patient satisfaction constitutes an important advantage to the medication. There is no validated survey with regard to post-endoscopy patient satisfaction. The mGHAA-9 has been used to evaluate subjective aspects of endoscopy centers and hospital systems, although it is considered insufficient to evaluate patient satisfaction with the sedation ^37,38. Patient satisfaction was similar in both groups of our study. Therefore, we may conclude that the combined use of propofol plus midazolam does not reduce patient satisfaction. Some studies have demonstrated a higher level of satisfaction in regimens that include propofol sedation ^27,30,34. Amnesia is a highly desired effect for endoscopists. Both midazolam and propofol have amnesic effects, although the latter to a lesser degree ³⁶. Memories of the procedure was slightly lower in the BPS group, which was not statistically significant. However, this reflects the results obtained by Molina-Infante et al. ³⁰, which compared monotherapy with propofol to propofol plus midazolam.

Our study does have some limitations. Only ASA I and II patients under 80 years of age were included in the study. Other studies, have included higher risk and older patients and obtained similar results in terms of safety ^6,12,30,31. Furthermore, this trial was designed to maintain a moderate sedation level and the exact level of sedation was not measured using the observer's assessment of alertness/sedation scale [OASS] ¹⁴. Therefore, we cannot discard the possibility of deep sedation at some point during the procedure. This is particularly true for group A, as BPS allows a more moderate sedation than propofol as a monotherapy ^18,29.

Furthermore, the use of patient satisfaction surveys immediately after the procedure has been proven to overestimate the perceived satisfaction. Even though it is the simplest and perhaps the cheapest method which ensures patient response, satisfaction may decrease when surveys are filled out days after the endoscopic procedure. In this sense, the study of Lin et al. ³⁹ compared both scenarios and concluded that despite the potential weakness of pre-discharge surveys, the magnitude of the differences between them was relatively small. In fact, there were significant values in only four of eleven questions of the survey. The survey used in our study was anonymous, as a non-anonymous survey might lead patients to evaluate the sedation more positively. This may help to diminish the differences found in the aforementioned study. In addition, the time at which the satisfaction survey was performed to compare the two sedation protocols was the same, before discharge from the hospital. Therefore, even though the survey may be biased to be more positive, there should be no bias when comparing both groups. Another issue might be the possible state of euphoria caused by the benzodiazepines, which may cause a higher score in the level of satisfaction in group B.

In conclusion, despite these limitations, the midazolam combined with propofol sedation regimen used in our study provided the same benefits of monotherapy with propofol without increasing adverse effects. This protocol offers a rapid sedation and recovery with a similar quality of the exploration and patient satisfaction, enhancing the existing body of evidence, which demonstrates the safety of endoscopic sedation supervised by endoscopists.