Cocktail sedation containing propofol versus conventional sedation for ERCP: a prospective, randomized controlled study

In order to facilitate a high success rate and avoids patients’ discomfort from a long procedure, endoscopic retrograde cholangiopancreatography (ERCP) requires a good moderate to deep level of sedation [1, 2]. Propofol has been increasingly used for sedation in many gastrointestinal procedures including ERCP because it has a shorter both half-life and results in a shorter recovery time than conventional sedation (opioid and/or benzodiazepine)[3–6]. In addition, it can be safely administered by non-anesthetist e.g. well-trained nurses with or without endoscopist’s advice [7–9]. Because of its narrow therapeutic window, the level of conscious sedation can easily go deeper from moderately deep sedation to near general anesthesia. Therefore, propofol as a sole agent can cause oversedation and apnea [4, 10, 11]. A combination of benzodiazepine, and propofol has been shown to reduce the dose requirement of propofol while maintaining comparable efficacy [12, 13]. This in turn may reduce the risk of oversedation by propofol [14, 15]. Moreover, at a moderate sedation level, propofol does not have analgesic property, patients undergoing ERCP under propofol sedation alone could be suffered from pain. A cocktail regimen containing meperidine/midazolam and propofol may add additive analgesic effect during ERCP. Furthermore, administering propofol by infusion pump provides constant plasma level of propofol and may reduce adverse effect of overdosing by a bolus dose. However, data for cocktail sedation in ERCP is limited to only a bolus propofol dosing [16, 17].

This study was aimed to compare the differences in terms of recovery time, patients’ satisfaction, and sedation-related adverse events between conventional sedation (meperidine and midazolam), and cocktail regimen (meperidine, midazolam and continuously infused propofol) in patients undergoing ERCP.

Over the last decade, we have seen an increase in propofol use for sedation in many GI endoscopies [2, 3]. The most important advantage of this agent is its fast recovery profile. In addition, non-anesthetists are able to give this agent. However, its narrow therapeutic window with a possible risk of apnea is the main concern. Thus, adequate experience of those who can administer this agent and early detection of overdosing skill are mandatory for patients’ safety. Propofol sedation at a moderate level is safe and well accepted for many diagnostic GI procedures [1–3]. However, the risk will increase if the level of conscious sedation is designed at moderate to deep sedation. Practically, this level of sedation is preferred for ERCP sedation. Our study confirmed the benefit of cocktail sedation containing propofol infusion over the conventional sedation in many categories including recovery profiles, patients’ tolerance, and satisfaction. Another hypothetical benefit in the present study was decreasing the rate of desaturation by reducing propofol dosage. We hypothesized that by giving continuous perfusion of propofol, the constant level of propofol would not exceed the therapeutic window and this in turn might result in an acceptable desaturation rate that comparable to the conventional sedation. Although, the dosage of proposal used in this study was lower than that of propofol infusion alone in the same population reported by us previously (6.2 mg/kg/hr vs. 9.4 mg.kg.hr) [6]. Unfortunately, the present study failed to demonstrate the lower incidence of desaturation of this regimen over the conventional sedation. We observed a higher percentage of desaturation in the cocktail group than the conventional group (SpO2 <90%; 58.3% vs. 31.4%, p <0.01 and SpO2 < 85%; 8.7% vs. 1.0%, p < 0.019). However, these desaturation episodes were very short and did not interfere the continuity of our ERCP procedures. It may be implicated that reducing the dose of propofol was not equivalent to decreasing the risk of oversedation. We speculate that there are some pitfalls in our study that require further investigation. First, prophylactic oxygen supplementation was not routinely given to any of our patients. During the study period, all desaturated patients recovered immediately after oxygen cannula was given and only 2 L/min was required. There was no need for assisted ventilation in any patients. Other previous studies that gave a routine oxygen supplementation showed a lower rate of desaturation in the propofol group (5-24%) [12, 13, 17]. A prophylactic oxygen supplementation may apply to prevent this adverse event [19], however its discriminate use may expose patients to delays in the diagnosis of disventilation by maintaining apneic oxygenation during hypoventilation [20]. Second, meperidine as the choice of narcotic agent in this study may result in a higher rate of desaturation than other short acting and milder narcotic agents such as pentazocine and fentanyl. Pentazocine is a synthetic narcotic that designed to have a lower rate of respiratory suppression [21]. Ong, et al. used a combination of pentazocine and propofol for sedation in ERCP patients [17]. They demonstrated a lower rate of desaturation (<5%) than ours. In addition, fentanyl is another narcotic with shorter half-life than meperidine. When compared with conventional sedation, a combination of propofol and fentanyl resulted in a faster recovery with similar safety even when these combinations were given in cirrhotic patients who underwent upper GI endoscopy [22]. Lastly, hypercapnia usually develops before desaturation occurs [23]. If we have used capnography in our study, we might have detected earlier cases of propofol overdose and desaturation rate might have been lower. Qadeer, et al. demonstrated that capnography is helpful for a better detection of apnic episode when compared with visual respiratory observation as confirmed by a lower hypoxic rate in the patients who underwent ERCP/EUS with capnography on board [24].

Cardiovascular adverse event related to propofol based sedation is another concern in non-anesthetist administered technique. Although, we observed a higher rate of dropping in systolic blood pressure to less than 90 mmHg in the cocktail group than the conventional group (14% vs. 5%), this rate of hypotension was not statistically significant and still comparable to the previous studies (13%) [5, 6]. Fortunately, these hypotensive episodes did not lead to any significant clinical cardiovascular compromise. We speculate that, histamine release causing hypotension can occur after a rapid administration of meperidine [21]. Propofol itself can also cause hypotesion by relaxing vascular smooth muscle [25]. Thus, the combined hypotensive effect may be potent especially during the initial phase of sedation. Alternatively, cocktail sedation containing fentanyl may cause a lower rate of hypotensive effect since fentanyl does not cause an over-release of histamine. However, further comparison study is required to confirm this hypothesis.

One of the limitations is that the level of sedation on each patient was controlled by a sedation nurse who was unblinded to sedative agents. Perhaps, there might be a bias in giving sedative agents at the doses that she preferred to be safe. However, this was more practical to real life practice since the level of sedation must be adjusted to the interest of patient safety as the first priority.

Although, we observed that the recovery time in the present conventional group (12.89 ± 12.87 minutes) seemed to be shorter than previous studies including our previous study (29–70 minutes) [4–6, 10, 25], the present study showed that the recovery time in the cocktail group was still significantly faster than the conventional group (9.67 ± 4.86 minutes vs. 12.89 ± 12.87 minutes, p = 0.045).We speculate that the overall shorter in procedural time was because of only non-complex ERCP such as common bile duct stone and distal common bile duct obstruction were recruited in the study. The low level of complexity in this seires was confirmed by a very low number of failed ERCP in both groups (0-2%).

Our patients in the cocktail group reported a significant higher satisfaction score than the patients in the conventional group (93 vs. 88, p < 0.001). The result was in line with a study by Lee, et al. that showed a non-significantly better satisfaction by patients when cocktail regimen was used for ERCP [16]. Moreover, their study demonstrated a significantly better satisfaction by both endoscopists /nurses when propofol was used for sedation. During our study, we did not encounter any undesirable movement that would interfere the procedure outcome or delay the procedure time of any of the patients from both groups. However, the limitation in our study is that our protocol did not have a formal intra-procedural assessment by sedation team. Therefore, a complete evaluation of patients’ cooperation may not be assessed from this study.

In conclusion, the present study shows the superiority of cocktail sedation containing propofol infusion, midazolam, and meperidine over conventional sedation in terms of recovery profiles, patients’ tolerance, and satisfaction. Although there was no serious adverse event in the cocktail group, the results of the study actually showed the cocktail sedation increased the rate of desaturation, which implicated that reducing the dose of propofol in this cocktail protocol was not equivalent to decreasing the risk of oversedation. Adjustment in the protocol that may overcome this problem is required. We propose on using fentanyl or pentazocine instead of meperidine in the regime. Lastly, a prophylactic oxygen supplementation and monitoring with capnography may be of help.