Impact of opioid addition on procedural conditions during colonoscopy: a randomized trial comparing propofol-based sedation protocols
Article information
Abstract
Background/Aims
Propofol is the most effective sedative for colonoscopy; however, opioids do have several adverse effects that need to be discussed. The objective of this study was to compare the ease of colonoscopy during propofol-based sedation with and without fentanyl, while closely monitoring ventilatory data and the safety of the procedure.
Methods
This prospective single-center trial randomized 50 patients who underwent minor colonoscopies. The propofol group received sedation exclusively via propofol, whereas the propofol+fentanyl group was premedicated with 1 µg/kg fentanyl. Patients were monitored using a bioimpedance ventilatory monitor, and both the patients and endoscopists were questioned regarding their level of satisfaction.
Results
The endoscopists reported a higher level of ease with the colonoscopy procedure (mean on a 5-point scale, 1.2 vs. 1.72; p=0.028) and the overall patient satisfaction score was higher (1.15 vs. 1.28, p=0.026) in the propofol+fentanyl group. No significant differences were observed in the ventilatory parameters between the groups. No major adverse events were reported in any of the groups.
Conclusions
The inclusion of fentanyl leads to enhanced levels of satisfaction for both the endoscopist and the patient, without any impact on ventilation and safety. The use of bioimpedance monitoring of ventilation during colonoscopy is a suitable approach that may enhance the safety of procedural sedation.
INTRODUCTION
New monitoring techniques are being developed to enhance patient safety during routine diagnostic and therapeutic procedures that involve deep sedation. Several studies have indicated that noninvasive bioimpedance monitoring of ventilation is a precise method1,2 that may help reduce the risk of potential adverse events during sedation.3,4 This approach may prove more beneficial than end-tidal carbon dioxide (EtCO2) monitoring.5
There are some variations in the protocols for sedation for colonoscopy, as well as in personnel settings.6 For an extended period, sedation with benzodiazepines and opioids has been regarded as the gold standard7 and propofol has been administered exclusively by anesthesiologists.6 This is true for many countries. Nevertheless, evidence suggests that propofol may be a more suitable option than midazolam for colonoscopy sedation, as it has been demonstrated to facilitate a faster recovery time,8-11 enhance patient comfort9-11 and, in instances where deeper sedation is required for patients with inflammatory bowel disease, it may improve the ease of colonoscopy.12 Another meta-analysis showed that propofol improves the satisfaction scores of endoscopists, but might slightly increase the risk of hypotension.13
Considering these findings, recent recommendations and studies have suggested the administration of propofol by trained endoscopists or certified nurses as a suitable approach.14,15 Although propofol sedation administered by non-anesthesiologists is associated with a higher risk of cardiopulmonary adverse events16 or awareness,17 the methods are improving and non-anesthesiologist providers of sedation are gaining more experience.18 Consequently, a recent meta-analysis concluded that non-anesthesiologists can safely provide propofol sedation without substantially compromising the quality of the procedure.17 Intermittent bolus administration of propofol was found to provide similar anesthesia with a risk profile comparable to that of continuous infusion.19,20
The main factor limiting the widespread use of propofol in endoscopic clinical practice is the lack of an antidote.21 Furthermore, the pharmacokinetics need to be taken into consideration; after a single dose, the offset of the effect is mainly due to redistribution, while in cases of prolonged administration, the transfer to and from slower compartments, such as lipid tissue, can impact the time to the recovery of consciousness.22 However, the pharmacokinetic profile of propofol is still considered more favorable than that of other hypnotics.22,23 Recently evaluated alternative sedation protocols include the use of dexmedetomidine,24 dexmedetomidine with ketamine,25 remimazolam,26 lidocaine,27 or ciprofol.28
A lot of recent discourse has centered on the perioperative use of opioids, underscoring the rising prevalence of these medications and the potential adverse effects associated with them.29 Consequently, interest in the development of opioid-sparing or opioid-free approaches to perioperative care has risen.30-32 In contrast, it is crucial to acknowledge that adequate perioperative analgesia is a fundamental requirement and that the advantages of opioid-sparing approaches over conventional treatments remain inconclusive.31
According to recent guidelines, monitoring a patient's oxygen saturation is recommended and ventilation monitoring should be considered.33 Desaturation has been linked to adverse effects including an increased risk of postoperative delirium.34 Administration of supplementary oxygen can result in undetected periods of hypoventilation and carbon dioxide retention when using pulse oximetry.35
The objective of this prospective randomized trial was to compare two sedation protocols during colonoscopy using a noninvasive respiratory volume monitor to assess the safety profile of the patient when an opioid is administered.
METHODS
Patient selection and randomization
From December 2023 to March 2024, 50 consecutive consenting patients preparing for a planned minor colonoscopy procedure who met the inclusion and exclusion criteria (Table 1) were enrolled in this prospective randomized study. The study was registered at ClinicalTrials. gov (NCT06060626; September 25, 2023). After enrollment, patients were randomly assigned to one of two study groups using a computer-generated randomization scheme. Patients in the propofol+fentanyl group underwent sedation via a combination of fentanyl and propofol, whereas those in the propofol group underwent sedation solely via propofol.
Inclusion and exclusion criteria of the study
Study protocol and data collection
In both groups, standard monitoring including electrocardiography, noninvasive blood pressure measurement, and pulse oximetry was coupled with a noninvasive bioimpedance ventilation monitor (ExSpiron 2Xi; Respiratory Motion Inc.). Following the establishment of an intravenous access and the measurement of baseline vital function parameters, patients randomized to the propofol+fentanyl group received a bolus of 1 µg/kg of intravenous fentanyl, rounded to the nearest whole milliliter (50 µg). Consequently, patients in both groups were sedated with titrated intermittent injections of propofol to achieve the desired level of sedation at RASS-2. The investigator administering fentanyl was present with the randomization sheet during endoscopy to ensure the possibility of unblinding in case of an adverse event. The second investigator administering the propofol and the endoscopist were blinded to the randomization.
Throughout the procedure, vital parameters were collected at five-minute intervals, and respiratory data from the ExSpiron monitor were exported at a sampling frequency of 0.2 Hz. The monitor presents the measured tidal volume and respiratory rate and calculates the actual minute ventilation. The built-in software also calculates a “predicted minute ventilation,” which is a theoretical resting minute ventilation given the patients sex, age, height, and weight. Consequently, the actual minute ventilation is divided by the predicted minute ventilation, obtaining a percentage value representing the ratio of the actual minute ventilation to the normal value. Furthermore, ventilatory data were collected at two-minute intervals and incorporated into the study protocol as a backup measure.
If the patient's oxygen saturation level declined to below 92%, oxygen supplementation was initiated at a rate of 2 L/min. If the flow rate was insufficient, it was increased in increments of 2 L/min. If ventilation measured by ExSpiron dropped below 25% of predicted minute ventilation, the patient was first stimulated to breathe through verbal contact. If insufficient, a jaw thrust was performed to relieve upper airway tract obstruction and stimulate breathing. At the same time, propofol administration was discontinued until the level reached a value higher than 25% again. The protocol included provision for bag-valve-mask ventilation in the event of further degradation and desaturation, although this was not required during the study period.
Endoscopist satisfaction survey
Once the procedure was complete, the endoscopist was invited to provide feedback on the level of satisfaction with sedation using a 5-point Likert scale. The scale was presented to the endoscopist as follows: very satisfied–satisfied–neutral–unsatisfied–very unsatisfied.
The “very satisfied” mark meant that the endoscopy was done smoothly without any interruption due to insufficient sedation. The “very unsatisfied” mark was never used by the endoscopists, but it would have meant that the surgery could not have been performed due to insufficient sedation. The protocol was unblinded to the endoscopist’s response and the eventual dosage of fentanyl was noted in the anesthesia protocol.
Patient satisfaction survey
After complete recovery from sedation, the patients were asked to fill out a questionnaire on the following three questions with 5-point Likert scaling: (1) How painful was this endoscopy for you?: not painful–minimally painful–a little painful–moderately painful–very painful. (2) How pleasant was the course of the endoscopy for you?: pleasant–neutral–slightly unpleasant–very unpleasant–extremely unpleasant, I do not want to repeat such a procedure. (3) Do you agree that the procedure left you with a feeling of privacy breach and your feelings were not respected?: I do not agree at all–I do not agree–I do not know–I agree–I agree completely. As recommended, when using similar sets of Likert scores, we averaged the three marks to obtain a composite Likert score36 to summarize the opinions of the patients.
Power analysis
The sample size for this study was determined using a power analysis calculation.37 Based on our clinical experience, we hypothesized that the use of fentanyl would significantly enhance endoscopic conditions. It was predicted that the endoscopist satisfaction score would be 2 on average in the propofol+fentanyl group and 2.5 on average in the propofol group, with a standard deviation of 0.5. A sample size of 32 patients was predicted for an enrollment ratio of 1, alpha value of 0.05, and power of 80%.
Statistical analysis
All statistical analyses were performed using R ver. 4.1 (R Foundation for Statistical Computing). Normality of the distribution was assessed using the Shapiro-Wilk test. The Kruskal-Wallis H test was used for non-normally distributed quantitative data, and the measured values were presented as medians (inclusive interquartile range). Among these data, the arithmetic mean was calculated using Likert scales and the number of interventions performed, as the clarity of the comparison of medians with interquartile ranges would be severely affected by the minimal variability of values. The use of mean values is mentioned in the text. As the age, height, and weight of patients are generally considered normally distributed variables, they were presented as arithmetic means with standard deviations. Categorical variables are presented as a percentage from the study group (%) and were tested by the χ2 test. Statistical significance was defined as p<0.05.
Ethical statements
The study was approved by the Ethics Committee of the Military University Hospital Prague, REC reference: 108/18-75/2023. Informed consent was obtained from all the participants. All methods were performed in accordance with the relevant guidelines and regulations.
RESULTS
All 50 enrolled patients were included in the data analysis. No major adverse events were reported. In all cases, the surgical procedure was completed without violating the blinding protocol. There were no notable differences in the demographic characteristics between the two study groups (Table 2).
Demographic characteristics of the studied groups
Endoscopist and patient satisfaction
Endoscopist satisfaction was more favorable in the propofol+fentanyl group than in the propofol group, with mean grades 1.2 and 1.72 respectively (p=0.03). A histogram of endoscopist satisfaction is shown in Figure 1.
Endoscopist satisfaction Likert scoring comparing the propofol+fentanyl group and the propofol group, dashed lines presenting respective means. Score 1 indicates highest satisfaction.
In all categories of patient satisfaction separately, the mean grades followed a statistically insignificant trend of preference of the fentanyl sedation–pain 1.12 vs. 1.16; pleasant 1.12 vs. 1.36; privacy 1.2 vs. 1.32. This resulted in a statistically significant difference in the pooled patient satisfaction marks, as expressed by the composite Likert score–mean, 1.15 vs 1.28 (p=0.03) (Fig. 2).
Patient composite Likert scoring for satisfaction comparing the propofol+fentanyl group and the propofol group, dashed line presenting respective means. The composite score is calculated as an arithmetic mean of the individual satisfaction marks, thus leading to unround results. For better visualization, the results are rounded and presented as a histogram. Score 1 indicates highest satisfaction.
Pulse oximetry and oxygen supplementation
In the propofol group, 5 patients did not need oxygen therapy, 17 achieved normal saturation with a maximal flow of 2 L/min, and 3 required a maximum flow of 4 L/min. In the propofol+fentanyl group, no oxygen supplementation was needed in 6 patients, 14 had a maximal flow of 2 L/min, and 5 achieved a maximum flow of 4 L/min. Refractory desaturation was not observed. No statistically significant differences were observed between the results.
Ventilation parameters
Ventilatory parameters in the two study groups are presented in Table 3. The drop in minute ventilation required a mean of 0.8 interventions (verbal stimulus or jaw thrust) in the propofol+fentanyl group, while the propofol group needed 1.2 interventions per patient (p=0.2).
Ventilatory parameters in the two study groups
Propofol doses
The propofol dose was slightly higher in the propofol group, but the difference was not statistically significant, 0.140 (0.110–0.205) vs. 0.108 (0.092–0.141) mg/kg/min (p=0.43).
DISCUSSION
The objective of this study was to investigate the differences between two sedation protocols for predicting difficult colonoscopy using a noninvasive ventilation impedance monitor. Both sedation protocols, which included ventilatory monitoring, resulted in a satisfactory level of sedation and a high level of safety with no severe adverse events. The addition of fentanyl resulted in enhanced patient satisfaction and a more straightforward colonoscopy experience, as reported by endoscopists.
It was anticipated that all the patients would present with a potentially challenging colonoscopy, as identified by the endoscopist. This was predominantly attributable to a history of uncomfortable colonoscopy experiences or known intra-abdominal adhesions. Despite the introduction of multiple scoring systems for predicting the difficulty of unsedated colonoscopy based on patient sex, body mass index, age, or bowel preparation,38-40 these are not typically employed to predict the necessity of deeper sedation provided by propofol.
The findings of other studies are consistent with the selection of endoscopists involved in our investigation, emphasizing the challenges encountered in patients with a history of abdominal surgery (particularly hysterectomy) or constipation.41,42 Despite the existence of scoring systems, none of them were deemed adequate for conducting the study. Consequently, the prediction of painful reactions to colonoscopy was primarily based on the department’s standard practices, which may have affected the homogeneity of the study group. Furthermore, a history of painful colonoscopy was the primary criterion for inclusion as perceived by endoscopists. This criterion is inherently heterogeneous, encompassing not only true algic reactions resulting from intra-abdominal adhesions, but also cases of anxiety, which can initially be challenging to differentiate.
An exploration of the feasibility of bioimpedance ventilatory monitoring in clinical practice revealed no problems in performing sedation in this setting. During the study, patients always responded to stimulation and diminution of sedation doses in case of hypoventilation, according to the study protocol, and no major respiratory events were recorded in either group. Thus, we believe that this monitoring might be a useful guide for performing sedation in clinical practice.
The introduction of fentanyl resulted in a notable improvement in patient satisfaction coupled with enhanced endoscopic conditions. This finding is consistent with those of previous research, indicating that the addition of fentanyl to midazolam-based sedation improves sedation quality as perceived by endoscopists.43 Nevertheless, inadequate sedation was not a justification for termination of the colonoscopy procedure in the propofol cohort.
Only one colonoscopy was prematurely abandoned during the procedure because of severe diverticulosis. Despite incomplete colonoscopy, the complexity of the procedure was high, necessitating a deeper plane of sedation; thus, the patient was still included in the final analysis. For the other 7 procedures (3 and 4 in the propofol+fentanyl and propofol group, respectively), the endoscopist decided to perform only a partial colonoscopy for a more complex pathology (for example, a circular lesion or a post-surgical stricture to dilate). Because the similarity with classical colonoscopy was large and the procedures were equally distributed, patients were included in the analyses.
The improved perception of sedation with fentanyl by the endoscopist and patient was not associated with any significant differences in the ventilation profile or risk of adverse events. The clinically insignificant trends noted the well-documented effect of decreased respiratory rate related to fentanyl44-46 which might explain the non-significantly higher minute ventilation in the propofol group, which was also connected to hyperventilation due to diminished comfort.
Considering the recommendations for opioid usage reduction,29-32 these results may contribute to the clinician’s decision-making process. While enhancing patient satisfaction and streamlining the colonoscopy process, it is feasible to perform most colonoscopies without fentanyl. This approach can be employed to mitigate the potential risks associated with opioid use in specific patient populations, including those with obstructive sleep apnea, morbid obesity, known postoperative nausea, or opioid addiction.47,48
In conclusion, the addition of fentanyl to propofol-based sedation for expected difficult colonoscopy increases the ease of the procedure and patient satisfaction without affecting ventilation parameters or safety profile. Bioimpedance ventilation monitoring is a suitable and simple tool for monitoring deep plane sedation during colonoscopy, thus allowing the safe administration of a combination of propofol and opioids.
Notes
Conflicts of Interest
Stepan Suchanek is currently an editorial board member of Clinical Endoscopy; however, he had not been involved in the peer reviewer selection, evaluation, or decision process of this article. David Novotny, Jan Palenik, Tomas Tyll, Nadija Brodyuk, and Michal Sotak have no potential conflicts of interest.
Funding
This study was supported by the Czech Technical University research grant SGS23/198/OHK4/3T/17.
Acknowledgments
The authors would like to thank all the anesthesia and endoscopy nurses and technicians of the hospital for allowing a smooth course of the study. The authors would also like to thank Tereza Novotna Jaromerska for her help with statistical evaluations. We also thank the IMEDEX Technology Scout company for lending the ExSpiron monitor to the department and providing technical support during the study.
Author Contributions
Conceptualization: TT, MS; Data curation: DN; Formal analysis: DN; Investigation: DN, JP, NB, SS, MS; Methodology: DN, TT, MS; Project administration: MS; Supervision: TT, SS, MS; Validation: TT; Visualization: DN; Writing–original draft: DN; Writing–review & editing: all authors.