This study investigated the effects of pre-procedural anxiety (assessed using the Beck Anxiety Inventory) on sedative requirements in 135 patients undergoing sedation for colonoscopy. Deep sedation was defined as loss of consciousness and no response to colonoscopy, and was achieved by target-controlled infusion of propofol. Patients' characteristics, baseline haemodynamic profiles, Beck Anxiety Inventory scores, effect-site propofol concentration at loss of consciousness and characteristics of recovery were recorded. No correlations were found between Beck Anxiety Inventory scores and effect-site propofol concentration at loss of consciousness or baseline haemodynamic profiles. There was no statistical difference in the characteristics of recovery among patients with different levels of anxiety. In conclusion, in patients receiving deep sedation for colonoscopies, the level of pre-procedural anxiety did not relate to the sedative requirement or post-procedural recovery characteristics.
Colonoscopy is often viewed as an invasive procedure with the potential for embarrassment, discomfort and worry related to potential findings . These concerns can result in anxiety that unfavourably decreases patient cooperation and satisfaction with the procedure . The use of sedation for colonoscopy not only enhances patient tolerance and satisfaction but also increases the willingness of patients to return for repeated examination . Deep sedation for colonoscopy, in which there is no response to colonoscopy and verbal contact is lost, may be routinely practised in some countries and centres  as opposed to conscious sedation where the patient may respond verbally or with movement to the stimulus. This level of sedation may allow the colonoscopist to dedicate more time and attention to examining the colon during colonoscopy. It has been found that with the use of deep rather than conscious sedation, more advanced lesions may be found  and more patients are likely to have complete colonoscopies .
Previously, several studies reported a correlation between level of anxiety and sedative or anaesthetic requirement when conscious sedation or anaesthesia is being used [6-8]. This seems reasonable and by extension, pre-procedure anxiety should also affect the sedative requirements for patients who are to receive deep sedation. If so, this information might help anaesthesia providers to determine the effective dosages to ensure rapid and comfortable sedation. Our hospital routinely administers propofol through target-controlled infusion to achieve deep sedation for individuals who are to receive colonoscopies. Therefore, in this study, we evaluated the effects of pre-procedure anxiety on sedative requirements for propofol in patients receiving deep sedation for colonoscopies. We also investigated the potential effects of anxiety on their recovery.
The protocol for this study was approved by the research ethics committee of Chang Gung Memorial Hospital. For this prospective study carried out in a single institution from December 2010 to December 2011, we recruited patients between 18 and 80 years old receiving planned colonoscopies. Patients with a history of colonic resection or any other intra-abdominal surgery, a history of a psychiatric disease or psychotropic drug prescription, a predicted difficult airway or allergy to propofol were not studied. Patients with insufficient bowel preparation (large amounts of solid faecal material found or < 90% of mucosa seen as graded by the endoscopists)  and those declining deep sedation were also not studied. Written informed consent for the study was obtained from all patients.
All patients were instructed to take a standard colon preparation agent the day before the examination. We recruited patients on the day the elective colonoscopy was to be performed. We collected demographic data and ascertained prior experience of surgery or colonoscopy. We also administered the Beck Anxiety Inventory  before each patient was brought to the operating room. This is a 21-item self-report questionnaire in which the respondent is asked to rate how much each symptom has bothered him/her over the week leading up to the procedures. The symptoms are rated on a four-point scale, ranging from ‘not at all’ (0) to ‘severely’ (3). The values for each item are summed yielding a total score ranging between 0 and 63 points. A Beck score of 0–7 represents a minimal level of anxiety, 8–15 mild, 16–25 moderate and 26–63 severe anxiety. No sedative premedication was employed. Data were collected by an anaesthetist (SEJ) not involved in sedation procedures.
When the patient arrived in the operating room, a 22-gauge iv catheter was inserted in the right forearm and 0.9% saline infusion was used to keep the iv line open. All patients were monitored following a standard protocol established by our hospital, which included electrocardiogram, continuous pulse oximetry and non-invasive blood pressure measured at 5-min intervals. After baseline haemodynamic profiles were obtained, the patient was placed in the left lateral position. Supplemental oxygen (6 l.min−1) was administered through a face-mask. Patients received scopolamine butylbromide as an antispasmodic premedication and supplemental alfentanil 10 μg.kg−1 as an analgesic premedication. Before infusion of propofol, 40 mg lidocaine was administrated iv to reduce pain on injection of propofol. Propofol was administered iv using the Base Primea™ system (Fresenius, France), which targeted effect-site concentrations (an estimate of the drug concentration at its site of action) of propofol using Schneider's pharmacokinetic model . The Base Primea system displays predicted effect-site concentration.
The propofol infusion was started with an initial effect-site concentration of 2 μg.ml−1, and increased by 0.5 μg.ml−1 every 30 s until patients lost the eyelash reflex and exhibited no response to a verbal command. This clinical endpoint was defined as loss of consciousness. The effect-site concentration of propofol and induction time required for loss of consciousness (LOC) was recorded. This initial effect-site concentration was chosen based on a previous study that found the EC50 (half maximal effective concentration) for effect-site propofol concentration at LOC to be 2.7 μg.ml−1 in the Chinese populations . After LOC was achieved, colonoscopy was commenced. The goal of this study was to achieve deep sedation, defined as a level of sedation whereby patients lost consciousness and were unable to respond to the stimulus from colonoscopy. If the desired sedation level was not achieved during colonoscopy (indicated by such signs a purposeful muscular movement or eye opening), the target effect-site concentration was increased in steps of 0.5 μg.ml−1. If no presence of purposeful muscular movement or cardiopulmonary depression was found, it was decreased by 0.5 μg.ml−1. To reduce inter-individual variability in the assessment of level of sedation, a single anaesthetist (KCC), who was blinded to the results of the Beck score, performed the anaesthesia procedures to achieve deep sedation in all patients. If there were any signs of airway obstruction or respiratory depression, a simple jaw thrust or chin lift was performed. Positive pressure ventilation was performed as required in the event of hypoxaemia (SpO2 < 90%). Ephedrine (8 mg) was administered iv if systolic blood pressure fell by > 20% baseline value or was < 90 mmHg.
All colonoscopies were performed by one of three endoscopists, each of whom had performed more than 200 colonoscopies before participating in the study. A standard colonoscope was used for all examinations. Each individual endoscopist documented the caecal intubation based on the view of the ileocaecal valve and appendiceal orifice. The procedural time, spasm score (1 = no spasm encountered; 5 = marked, long waiting and very difficult to examine)  and difficulty score for colonoscopy (1 = easy; 5 = very difficult) which were assessed by the endoscopist immediately after the procedure. The endoscopist was also blinded to the results of the Beck score.
At the end of the procedure, the infusion pump was discontinued and the total dosage of propofol required was recorded. The time needed to regain consciousness was assessed by requesting patients to open their eyes after discontinuing the infusion pump. After the patient had fully recovered in the recovery room, the investigator who was blinded to the results of the Beck score recorded the willingness of the patient to repeat colonoscopy (yes/no), the patient satisfaction score with the sedative technique on a scale of 1–5 (the higher the score, the greater the level of satisfaction) and physical complaints (dizziness, abdominal fullness and abdominal pain) after the procedure. Patients were discharged from the post-anaesthesia care unit when they met the discharge criteria.
The main association that we examined was between the effect-site propofol concentration at LOC and pre-procedure anxiety. Based on our preliminary data, we presumed a correlation coefficient of 0.25. We needed at least 120 patients to set a significance level of 0.05 (two-sided) and achieve a power of 0.80. To compensate for possible dropouts, we enrolled 144 patients. The association between pre-procedure anxiety and effect-site propofol concentration at LOC was assessed by Spearman correlation coefficient. Categorical variables were analysed using the chi-squared test. Non-continuous data were analysed using the Kruskal–Wallis test. Statistical significance was set at p < 0.05. Data were analysed using SPSS version 13.0 (SPSS Inc., Chicago, IL, USA).
A total of 144 patients successfully completed the Beck Anxiety Inventory during the study period. Nine had insufficient bowel preparation, and were not studied, leaving 135 participants whose data were included into our statistical analysis.
The patients' characteristics, baseline haemodynamic profiles, prior surgical experience and prior colonoscopy experience are shown in Table 1. No correlation between Beck scores and baseline haemodynamic profiles was found (systolic blood pressure: ρ = 0.063, p = 0.466; diastolic blood pressure: ρ = 0.023, p = 0.79; heart rate: ρ = 0.156, p = 0.071). Table 2 summarises the results of sedation and level of pre-procedural anxiety. No correlation was found between Beck scores and the effect-site propofol concentration at LOC (Fig. 1). The probability of LOC and effect-site propofol concentration is shown in Fig. 2. Additional analysis was performed to examine the associations between (a) the propofol dose and anxiety, and (b) induction time and anxiety. However, no correlations were found for these comparisons (ρ = −0.04, p = 0.647 and ρ = −0.046, p = 0.595, respectively). The correlation between time required for regaining of consciousness and Beck scores was significant (Fig. 3).
Table 1. Patients' characteristics, baseline haemodynamic profiles and prior experience of surgery or colonoscopy for 135 patients. Values are as median (IQR [range]), number or number (proportion)
Table 2. Sedation results and level of anxiety. Values are median (IQR [range]) or number (proportion)
LOC; loss of consciousness; Beck score; Beck Anxiety Inventory score
Effect-site propofol concentration at LOC; μg.ml−1
3.6 (3.4–3.8 [2.2–5.1])
Induction time for LOC; s
102 (89–127 [40–262])
Total propofol dosage; mg.kg−1
3.0 (2.4–3.9 [1.0–6.5])
Propofol dosage rate; mg.kg.min−1
0.3 (0.2–0.3 [0.1–1.1])
Time required for regaining of consciousness; min
6 (4–10 [0–32])
Minimal (Beck scores 0–7)
Mild (Beck scores 9–15)
Moderate (Beck scores 16–25)
Severe (Beck scores > 26)
All colonoscopies were performed without complications. The examination time for colonoscopy was 11.6 (7.8–16.1 [3.5–35.1]) min. The caecal intubation rate was 96.3% (130/135). The difficulty score for colonoscopy and spasm score assessed by endoscopists were 2 (1–3 [0–5]) and 2 (1–2 [1-5]), respectively. The recovery characteristics, which included post-procedure physical complaints, willingness to repeat the procedure and patient satisfaction score with sedative technique, were assessed. The cohort of patients was next categorised into three groups based on their anxiety scores to assess the potential effect of anxiety on recovery characteristics: (1) minimal anxiety (Beck score 0–7) (n = 87); (2) mild anxiety (Beck score 8–15) (n = 35); and (3) moderate-to-high anxiety (Beck score > 15) (n = 13). There were no significant differences in the frequency of post-procedure physical complaints, patient satisfaction score or willingness to repeat the colonoscopy between these groups (Table 3). During the study period, no patient required assisted ventilation.
Table 3. Recovery characteristics after colonoscopy. Values are number (proportion) or median (IQR [range])
Minimal anxiety (n = 87)
Mild anxiety (n = 35)
Moderate-to-severe anxiety (n = 13)
Minimal anxiety = Beck scores of 0–7; Mild anxiety = Beck scores of 8–15; and Moderate-to-severe anxiety = Beck scores > 15.
The satisfaction score was assessed on a scale of 1–5. The higher the number, the greater the level of satisfaction.
We found no association between pre-procedural anxiety and sedative requirements for deep sedation in patients receiving colonoscopies. Also, the post-procedural recovery characteristics were similar in our patients with different level of anxiety. Patients were satisfied with the sedative technique and a > 96% caecal intubation rate was achieved.
Recently, Kil et al. found that propofol requirements correlated with both situational (i.e. state) and baseline (i.e. trait) anxieties during light-to-moderate levels of sedation (bispectral index (BIS): 75–85), but correlated with baseline anxiety only during deeper sedation (BIS: 65) . However, during general anaesthesia, no correlation between pre-operative anxiety and anaesthetic requirement (MAC-hour of sevoflurane) was found. Their findings may suggest that the effects of anxiety on sedative requirements may be less when the level of sedation is deepened. A lighter level of sedation (conscious sedation) may allow the effects of anxiety to be manifested during uncomfortable procedures [7, 8].
In this study, LOC was used as a clinical indicator of deep sedation and at this level of sedation, a BIS < 60 may be observed [12, 15, 16]. The suppression of consciousness achieved in our participants may, therefore, be comparable to that required for general anaesthesia , and some may choose to refer to ‘deep sedation’ as ‘general anaesthesia’. We argue that an additional requirement for the latter is loss of movement to a standard surgical skin incision; for interventions like colonoscopy where that degree of stimulation does not occur, then the term ‘deep sedation’ is appropriate. For example, Maranets et al.  found that trait anxiety was correlated with the amount of propofol administered during general anaesthesia when propofol, alfentanil and N2O were used concurrently to maintain a BIS value of 40–60. Their results indicate possible differences between the states of deep sedation and anaesthesia, or might be related to their small sample size (n = 57), or to their use of iv hypnotic and N2O to maintain the depth of anaesthesia. We used only a single iv agent to achieve the sedation.
Colonoscopy is an invasive procedure with the potential for discomfort. Even in colonoscopy patients who have received procedures to achieve deep sedation, physical complaints such as abdominal fullness, pain or dizziness may occur afterwards. These physical complaints related to either colonoscopy or sedation may reduce a patient's willingness to repeat the examination. Previous research studying patients receiving lumbar surgery has suggested that pre-operative anxiety independently predicted more postoperative physical complaints . Our study found similar recovery characteristics regardless of the level of pre-procedural anxiety. The inconsistency between these two studies might be attributed to the fact that patients receiving colonoscopy may experience less psychological and physical stress than those undergoing surgical procedures.
The effect-site propofol concentration associated with a 50% probability of LOC in this study was 3.6 μg.ml−1, a result consistent with the results of Iannuzzi et al., who reported that the EC50 for effect-site propofol concentration at LOC was 4.14 μg.ml−1 in unpremedicated patients . It is well known that there is a large inter-individual variability in the effect-site propofol concentration for obtaining a clinical endpoint. For example, Iwakiri et al. reported that there is an approximately sixfold inter-individual difference between the minimum (0.7 μg.ml−1) and maximum effect-site concentration (4.8 μg.ml−1) at LOC . Previously, Hong et al. reported that the effect-site propofol concentration required for conscious sedation correlated with the level of anxiety , suggesting that this psychological variable may be a confounding factor for effect-site concentration during conscious sedation. However, this correlation was not found in our study of patients under deep sedation.
Although the Beck Anxiety Inventory is a well-accepted self-reported measure of anxiety in adults and adolescents for use in both clinical and research settings, its use in our study may make it difficult to compare results with those of previous studies [6-8, 14] in which the Spielberger State-Trait Anxiety Inventory was used . We chose the Beck Anxiety Inventory because first, it was designed to minimise overlap between depression and anxiety scales  and second, pre-procedural anxiety can increase the levels of stress hormones and catecholamines [21, 22], which in turn may increase the symptoms of physiological arousal. These are reflected in the ‘somatic subscale measures’ of the Beck Anxiety Inventory.
Although deep sedation provided by the operating surgeon during colonoscopy may be more cost-effective, safety issues require the presence of an anaesthetist in our institute. Colonoscopists can therefore dedicate attention to examining the colon; we found a caecal intubation rate of 96.3%, comparable with one previous report .
Some electroencephalographic (EEG) monitors such as BIS monitors have been used as an adjunct to assess sedative requirements in several studies [6, 7, 14]. However, there is still wide inter-individual variability and time lags for EEG monitors in the prediction of LOC [15, 24], which may limit their use in pharmacodynamic studies. In addition, it has been found that, when deep sedation is intended, the use of a BIS monitor as an adjunct did not influence the dose of propofol administered . These findings suggest that the use of LOC as a clinical indicator to assess the level of sedation may not have biased our results.
This study has some limitations. First, we did not collect data on the frequency of adverse events during deep sedation for colonoscopy. Previous studies have reported that these adverse sedation-related events are uncommon during colonoscopy and most of them are transient [23, 26]. Secondly, LOC was used as a clinical indicator to assess the level of sedation. As sedation comprises a continuum of states, the potential for bias in the anaesthetist's control of level of sedation may exist and introduces a potential bias into some of the outcome. Thirdly, EEG monitors were not used in this study. Their use might have made our results more comparable with previous studies. Finally, we did not record the propofol concentrations on recovery and this might have enabled us to examine any hysteresis in the plot in Fig. 2.
In conclusion, in patients receiving deep sedation for colonoscopies, the level of pre-procedural anxiety did not relate to the sedative requirement or post-procedural recovery characteristics. Additional evaluations may be needed to confirm the effect of pre-procedural anxiety on sedative requirement under conscious sedation for colonoscopy.
We thank Yu-Ru Kou PhD for his advice regarding the statistical analysis of our data.
No external funding or competing interests declared.