Gastrointestinal endoscopy procedures, including esophagogastroduodenoscopy, colonoscopy, endoscopic ultrasound (EUS), and endoscopic retrograde cholangiopancreatography (ERCP), play a key role in the diagnosis and treatment of various gastrointestinal conditions. These techniques allow healthcare professionals to directly visualize and evaluate the health of the digestive system, detect abnormalities, and collect tissue samples for further analysis. They facilitate the early detection and diagnosis of conditions such as peptic ulcers, gastrointestinal bleeding, inflammatory bowel disease, polyps, and tumors.¹

According to the American Cancer Society, in the United States, 348,840 new cases of digestive system cancers were diagnosed in 2023. The same report noted 172,010 deaths due to cancers of the digestive system that year.² Early detection is crucial for the prevention of new cases and deaths. With the increase in screening and surveillance worldwide, the number of gastrointestinal endoscopies performed is similarly rising.

Sedation is commonly prescribed for gastrointestinal endoscopy to alleviate patient discomfort, thereby increasing patient compliance and contributing to the success of the procedure.^3,4 Various methods are used to administer sedation for gastrointestinal endoscopy, including manual intermittent bolus, manual infusion, and the most recently developed approach: target-controlled infusion (TCI).^5–7 Several studies have compared these techniques. The present review focuses on the use of TCI as a new modality in facilitating gastrointestinal endoscopy.

Written informed consent was not required because no patient data were included in the manuscript.

In modern medicine, the appropriate administration of drugs is crucial for optimizing therapeutic outcomes and minimizing side effects. Traditionally, drug administration was performed manually; however, this approach is often inaccurate, as it introduces variability in drug concentrations and the potential for dosing errors. TCI is a computer-controlled method of intravenous drug administration that utilizes real-time pharmacokinetic models to achieve and maintain predetermined target drug concentrations in the bloodstream.⁸ The TCI system determines infusion rates using patient-specific parameters such as body weight, age, and pharmacokinetic variables, ensuring precise drug delivery.⁹

The basic mechanism of TCI is grounded in the pharmacokinetic principles of drug volume and clearance. When a drug is administered intravenously, it is distributed throughout the body’s volume, which includes plasma and other tissues, such as muscle and fat. The drug is then eliminated from the plasma and tissues through the process of clearance. The pharmacokinetics of intravenous anesthetic drugs can be described by a three-compartment model of drug distribution. These three areas are the central compartment (V1, representing plasma), the rapidly equilibrating compartment (V2, referring to well-perfused tissues such as muscle), and the slowly equilibrating compartment (V3, denoting fat tissue).⁹

Each compartment has a distinct clearance pathway. Clearance from the central compartment through elimination or metabolism is termed CL1. Clearance between the central and second compartments is described as CL2, and clearance between the central and third compartments is denoted as CL3.⁹ Notably, a discrepancy has been observed between the time when a drug reaches its target plasma concentration and when it achieves its clinical effect. The delay in concentration equilibrium between the plasma and the active sites within the central nervous system or brain, also known as the effect site, results in a postponed clinical effect. This delay can be quantified with the constant keo. The relationships involved in the three-compartment model of an intravenous drug and its effect site are illustrated in Fig. 1.^9,10

Figure 1. The three-compartment model of intravenous pharmacokinetic theory delineates the relationships between volume and clearance among compartments and their connection to the effect site. CL1, clearance from the central compartment through elimination or metabolism; CL2, clearance between the central and second compartments; CL3, clearance between the central and third compartments.

After an intravenous bolus, plasma drug concentrations typically exhibit an exponential decline. Maintaining adequate intravenous anesthesia requires that drug concentrations in the brain remain balanced with plasma levels. TCI addresses this by solving complex equations that describe the distribution of agents between compartments, enabling rapid adjustment of target levels to achieve the desired clinical effect. Various validated models are available for use, with the Marsh and Schnider models being the most common. These pharmacological models were derived using differing methodologies, include distinct parameters, and can produce significantly different results when applied to determine the drug infusion rate in a TCI system.^8,10

For mathematical calculations, the Schnider model incorporates sex, body weight, age, and height, leading to the use of lean body mass.¹¹ In contrast, the Marsh model includes only body weight.¹² When applying TCI with either model, the desired target drug concentration can be programmed for either plasma or effect-site targeting. The programming of the early systems was limited to plasma targeting. However, analysis soon revealed the time delay between the achieved plasma concentration and the actual clinical effect. This observation led to the incorporation of the additional factor keo into the original three-compartment model to enable the prediction of effect-site concentration.⁸

Drug selection, dose, and target concentration

The combination of midazolam and fentanyl has traditionally been used for sedation during gastrointestinal endoscopy.⁴ However, propofol is increasingly preferred due to its rapid onset and quick recovery times, along with a safety profile and satisfaction levels that are comparable to, if not better than, those of midazolam.^4,13 Additionally, the analgesic and anti-emetic effects of propofol are advantageous for patients undergoing same-day procedures.¹⁴ In the studies of TCI for gastrointestinal endoscopy discussed in this article, propofol was administered either alone or in combination with midazolam and/or fentanyl.

Fanti et al¹⁵ investigated the application of TCI for ERCP, employing a plasma targeting approach with an initial concentration of 4 µg/mL. The desired level of sedation was determined to have been reached when the patient did not respond to nasal cannula placement. The target concentration of TCI was adjusted in increments of 0.5 µg/mL, within a range of 2–5 µg/mL, to maintain an appropriate sedation level throughout the procedure. If the TCI concentration reached 5 µg/mL and the patient still exhibited signs of pain, fentanyl was administered in doses ranging from 50 to 100 µg. When using this protocol for TCI propofol dosing, 98% of endoscopists and nurse assistants rated the sedation as satisfactory.¹⁵

In a subsequent study, Fanti et al¹⁶ reported the use of TCI propofol for EUS. Here, they similarly employed plasma targeting with an initial concentration of 4 µg/mL until the desired level of sedation was achieved, as measured by an Observer’s Assessment of Alertness/Sedation (OAA/S) scale score of 2. The study compared a combination of midazolam and TCI propofol to TCI propofol alone (with placebo). Midazolam was administered at a dose of 0.02 mg/kg prior to the initiation of TCI propofol. The findings indicated no significant difference in the total amount of propofol used or the incidence of complications between the two groups.¹⁶

Hsu et al¹⁷ compared the use of TCI propofol alone with TCI propofol combined with fentanyl and midazolam in patients undergoing gastrointestinal endoscopy, which included both colonoscopy and upper gastrointestinal endoscopy. For both groups, the TCI was set to effect-site targeting with an initial target concentration of 4.0–5.0 µg/mL for upper gastrointestinal endoscopy and 2.0–3.0 µg/mL for colonoscopy. In the combination group, participants received 1–2 mg of midazolam and 25–50 µg of fentanyl before the initiation of TCI propofol. In the TCI propofol-only group, participants could be administered 25–50 µg of fentanyl during the procedure as a rescue medication. The study revealed significant differences; specifically, the combination group displayed lower propofol consumption, a reduced incidence of hypotension, and shorter recovery and discharge times. However, overall post-procedural satisfaction was comparable between groups.¹⁷

Cuiabano et al⁶ administered 1 µg/kg of fentanyl prior to initiating TCI with the Schnider pharmacokinetic model. They set an initial effect-site target of 2 µg/mL, which could be incrementally titrated by 0.5 µg/mL until the desired level of sedation was achieved. This desired level was characterized by the patient being unresponsive to prodding or shaking, equivalent to a score of 1 on the OAA/S scale.⁶

In a comparison of propofol delivery with TCI versus manual infusion, Vučićević et al⁷ administered midazolam and fentanyl prior to administering propofol. Midazolam dosing was based on body weight: individuals weighing up to 70 kg received 2 mg, while those over 70 kg received 3 mg. Similarly, fentanyl dosing was weight-dependent: patients weighing 50–60 kg received 50 µg, those weighing 60–80 kg received 75 µg, and those whose weights exceeded 80 kg received 100 µg. For the group receiving TCI, the protocol was then initiated by setting the pump to effect-site targeting, with an initial target concentration of 2.5 µg/mL. This concentration could be titrated up or down in increments of 0.5–1 µg/mL until the patient no longer responded to verbal commands, corresponding to a modified OAA/S (MOAA/S) scale score of 2.⁷

Hsu et al¹⁸ compared the use of low versus high effect-site target concentrations of TCI propofol during endoscopy. Prior to initiating TCI, they administered both midazolam (0.04 mg/kg) and fentanyl (0.5 µg/kg). Participants in the low-concentration group received TCI with an effect-site target concentration of 1.5–2.5 µg/mL, whereas those in the high-target group received 3.0–4.0 µg/mL. The results indicated that post-procedure satisfaction did not differ significantly between groups. However, the incidence of hypotension was significantly higher in the group with the high target concentration.¹⁸

Studies utilizing propofol TCI for gastrointestinal endoscopy have consistently revealed that this administration method, whether used alone or in combination with midazolam and/or fentanyl, provides satisfactory sedation. The addition of midazolam and fentanyl may be advantageous in reducing the total propofol dose required, decreasing the risk of hypotension, and shortening recovery and discharge times.¹⁷ Effect-site targeting was more commonly employed than plasma targeting, with concentrations ranging from 1.5 to 5.0 µg/mL. A higher concentration is considered appropriate for upper gastrointestinal endoscopy, which involves oral probe insertion and tends to be more irritating than colonoscopy, where the probe is inserted rectally. Nearly all studies titrated the propofol concentration by 0.5 to 1.0 µg/mL. Propofol TCI has been demonstrated to effectively induce moderate to deep sedation during gastrointestinal endoscopic procedures, achieving OAA/S or MOAA/S scale scores of 1 to 2.

TCI versus other techniques

Several studies have compared methods of providing sedation during gastrointestinal endoscopy. The conventional and most widely used technique is manual intermittent bolus administration. This approach requires the anesthesia provider to inject an initial dose of a sedative (with or without analgesics such as opioids) during induction, and then maintain the level of sedation by manually and intermittently administering additional doses throughout the procedure. As the practice of continuous infusion for total intravenous anesthesia has become more routine, it has also been adopted in endoscopy, particularly for longer procedures such as ERCP. Recently, TCI has been employed in several studies for gastrointestinal endoscopy, and comparisons have been made among these three techniques.

In a randomized controlled trial involving 100 patients undergoing ERCP and EUS, participants were divided into two groups: those receiving a bolus and those receiving medication via a perfusor pump.¹⁹ The study concluded that the techniques were comparable regarding the total dose of propofol administered, the incidence of hypoxemia and hypotension, patient cooperation, and the quality of recovery. The only significant difference observed was in recovery time, which was shorter in the bolus group than in the perfusor group.¹⁹

Intermittent bolus injection and TCI have also been compared in the context of propofol delivery for colonoscopy.⁶ This trial, involving 50 participants, investigated the safety and efficacy of these administration techniques. Researchers measured and compared the number of propofol dose adjustments, the need for safety maneuvers (such as ensuring airway patency and providing ventilation assistance), instances of patient agitation, recovery time, and the total dose of propofol administered. The findings revealed that TCI reduced the need for dose adjustments and the rate of patient agitation. In contrast, the manual intermittent bolus approach was associated with a faster recovery time and a lower total propofol dose.⁶

Another study compared the use of TCI with manual infusion during colonoscopy, aiming to evaluate patient safety and endoscopist comfort.⁷ The researchers found that the mean arterial pressure in the manual infusion group was significantly lower than that in the TCI group at both the 10th minute and the end of the procedure, although it did not drop below 86.5 mmHg. Oxygen saturation was also significantly lower in the manual infusion group at minutes 5 and 15, but it remained at or above 97%. In contrast, the respiratory rate was significantly lower in the TCI group at minute 5 and at the conclusion of the colonoscopy, although it did not fall below 12 breaths per minute. Additionally, patients in the TCI group reached a deeper level of sedation at minute 5 without an elevated risk of desaturation. Endoscopist comfort showed no significant difference between the two groups.⁷

Based on the available studies, the use of propofol TCI is comparable to manual intermittent bolus administration and manual infusion in terms of satisfaction and overall hemodynamic stability during endoscopic procedures.

Few guidelines for sedation and anesthesia in gastrointestinal endoscopy address the use of TCI for administering propofol during endoscopic procedures. The Standards of Practice Committee of the American Society for Gastrointestinal Endoscopy has published the “Guidelines for Sedation and Anesthesia in GI Endoscopy,” which lists propofol as a preferred agent for endoscopy. However, the guidelines do not detail the methods of administration.²⁰ The publication “Non-anesthesiologist Administration of Propofol for Gastrointestinal Endoscopy: European Society of Gastrointestinal Endoscopy, European Society of Gastroenterology and Endoscopy Nurses and Associates Guideline – Updated June 2015” briefly mentions the use of propofol TCI. This resource suggests administering propofol either through manual intermittent bolus infusion or via continuous infusion, including TCI, but it offers no specific recommendations on settings or dosages.²¹

Future guidelines for sedation and anesthesia in gastrointestinal endoscopy should be refined with detailed descriptions of the various techniques available for sedative administration, including TCI as an alternative to conventional manual bolus infusion. These guidelines should also describe specific sedation targets, approximate dose titration parameters, and the timing for adjusting the dose or target concentration, if necessary. Further studies of TCI in gastrointestinal endoscopy should aim to determine optimal target concentrations for various types of procedures. Procedures involving oral probe insertion are expected to require a higher target concentration for induction, which could then be gradually reduced during maintenance. Standardized guidelines for these adjustments would be beneficial, as excessively high target concentrations can result in hemodynamic instability and respiratory depression.

TCI is a validated method for intravenous sedation. Propofol has become the preferred agent for ambulatory anesthesia due to its rapid onset, quick recovery profile, and anti-emetic and analgesic effects. Accordingly, propofol TCI is suitable for gastrointestinal endoscopy, given that most patients are outpatients receiving same-day treatment. The additional administration of midazolam and fentanyl appears to reduce the total dose of propofol required, decrease the risk of hypotension, and facilitate shorter recovery and discharge times. The target propofol concentration range is typically between 1.5–5.0 µg/mL, with effect-site targeting predominating. Relative to conventional and widely practiced techniques, such as manual intermittent bolus administration and manual infusion, propofol TCI demonstrates comparable safety and efficacy. Further research is needed to develop standardized guidelines for propofol TCI sedation in gastrointestinal endoscopy.

None.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

No potential conflict of interest relevant to this article was reported.