Chinese Medical Journal 2007;120(22):2036-2038
Unconscious sedation/analgesia with propofol versus conscious sedation with fentanyl/midazolam for catheter ablation of atrial fibrillation: a prospective, randomized study

Correspondence to:MA Chang-sheng,Department of Cardiology, Beijing Anzhen Hospital, Capital Medical University, Beijing 100029, China (Tel: 86-10-64456412. Fax:86-10- 64456078.
atrial fibrillation; ablation; sedation

Catheter ablation of atrial fibrillation (AF) has been increased dramatically recently.1 However, it is an unpleasant procedure with intolerable pain without sedation. Propofol and fentanyl/midazolam have been widely used in painful clinical examination and cardiovascular procedures with established safety and efficacy.2,3 Propofol, alfentanyl and midazolam were administrated for catheter ablation in some electrophysiological labs for a less painful procedure.4 However, there is few published work on the sedation regimen for catheter ablation of AF.

Catheter ablation of AF guided by 3D electroanatomical nonfluoroscopic mapping systems (CARTO, Biosense Webster, USA) has been widely used.1 If the reference patch on the back of the patient is moved, the left atrial geometry of the 3D mapping system will lose the precision in guiding the catheter. We carried out a prospective, randomized study to test the hypothesis that conscious sedation with fentanyl/midazolam for catheter ablation of AF had advantages over unconscious sedation/analgesia with propofol according to adverse influence on the ablative procedure.


After obtaining written informed consent, 120 consecutive patients (American Society of Anesthesia class I-II) presenting to our center for the first-time catheter ablation of AF between January 2006 and May 2006 were enrolled prospectively. The left atrial geometry was reconstructed using CARTO system with a 3.5 mm tip ablation catheter (Navi-Star, ThermcoolTM, Biosense-Webster, USA). The ablative protocol consisted in achieving two continuous circular lesions around ipsilateral pulmonary veins has been described previously in details.1 Recurrence was assessed by 12-lead electrocardiogram (ECG) and 24-hour Holter recordings before discharge and at 1, 3, 6 and 12 months. Any episode of confirmed atrial tachyarrhythmia lasting for at least 30 seconds beyond one month after catheter ablation was considered as an arrhythmia recurrence.

The patients were fixed on the operation bed by two restraint straps around the elbows and laps in a fasting state. All patients were assigned to receive propofol or fentanyl/midazolam by random lotting. Propofol or fentanyl/midazolam was administrated by an electrophysiologist who had been trained in administration of anesthetic drugs under the direction of an anesthesiologist before the left atrial geometry was reconstructed in the CARTO system. Oxygen was delivered by Venturi face mask (FiO2=0.37) to all patients throughout the procedure. The jaw thrust maneuver was applied to all patients who snored during the sedation. Oropharynx airway was applied if snore could not be terminated by the jaw thrust in the propofol group. When respiratory distress or hypoxia occurred, all sedation medicines were withdrawn, and a ventilation face mask with a manual resuscitation bag or endotracheal intubation and machine ventilation support if neccessary was applied. The values of physiologic parameters such as blood pressure, pulse oxygen saturation, heart rate and respiratory rate were documented every ten minutes. Hypotension was defined as a decrease in systolic and/or diastolic blood pressure by greater than 20 mmHg, or systolic blood pressure was less than 90 mmHg. Bradycardia was defined as a heart rate less than 50 beats per minute. Hypoxia was defined as pulse oxygen saturation less than 90%. Respiratory distress was defined as cessation of airflow at the nose and mouth for at least 30 seconds and pulse oxygen saturation was less than 80%.

In the propofol group, an initial slow intravenous administration of 1 mg/kg propofol (Diprivan, Astra- Zeneca, Caponage, Italy) was followed by continuous infusion with a syringe pump at a starting maintenance dose of 4 mg∙kg-1∙h-1 and was titrated to achieve adequate motionless and painless. In the fentanyl/midazolam group, intravenous administration of 0.8 µg/kg fentanyl (Yichang Renfu, China) was followed by 1 µg∙kg-1∙h-1 fentayl continuous infusion with a syringe pump and 0.03 mg/kg bolus of midazolam (Jiangsu Enhua, China) was given slowly over 1 minute as an initial dosage. If the patients felt pain, the maintenance dose of fentanyl was increased to less than 2 µg∙kg-1∙h-1. Another 1 mg midazolam was added if the pain was not relieved by the adjustment of fentanyl.

The Ramsay assessment scale5 was used to assess the grade of sedation every thirty minutes. Intensity of pain was assessed by the patients with the 11-point (0-10) Numerical Pain Rating Scale (NRS).6 The recovery time was defined as the time elapsed between the last dose of medication was given and the patient could perform mandatory action. The frequency of the reference patch movement according to unsatisfactory sedation and the delay time consumed in unsatisfactory sedation were accumulated. A nurse, who was responsible for documentation, used a stopwatch to record start time, recovery time, delay time, and finish time for each individual. The operator satisfaction survey was carried out after the ablative procedure. The operator satisfaction score was defined as: I, the procedure was completed without any adverse influence of the patient’s sedation; II, the procedure had to stop no more than three times for the disadvantage of sedation, or the patients’ breath had minor effect on the procedure; III, the procedure had to stop more than three times or the patients’ breath had so severe influence on the procedure that it was difficult for the catheter to reach the target point; IV, the geometry was moved for the patient’s restless and the geometry had to be reconstructed, or the sedation had to be abandoned for anaesthetic adverse effects. The procedure time, radiofrequency time, success rate of pulmonary veins isolation, success rate and incidence of complications of the catheter ablation of AF were compared between the two groups.

Statistical analysis
All analyses were performed using SPSS software version 10.0. All continuous data are presented as mean ± standard deviation. Unpaired independent samples t test (normal distribution) or Mann-Whitney U test (non-normal distribution) was used for comparison of continuous variables. Chi-square test and Fisher exact test were used for categorical variables. A P value <0.05 was considered statistically significant.


Sixty of the 120 patients were randomly assigned to the propofol group. The gender, age, weight, AF duration, AF type, proportion of hypertension and structural heart disease, serum level of creatinine, serum level of alanine amiotransferase and serum level of albumin did not differ significantly between the two groups. The mean dose of propofol was (8.0±2.2) mg∙kg-1∙h-1, the mean dose of fentanyl was (1.4±0.3) µg∙kg-1∙h-1, and the mean dose of midazolam was (2.7±1.0) mg. There were no significant differences in radiofrequency time, procedure time and sedation time between the two groups. The total sedation time was (134±32) minutes in the propofol group and (127±24) minutes in the fentanyl/midazolam group (P=0.189). The recovery time from sedation was (6.2±2.8) minutes in the propofol group, significantly longer than that of (0.8±1.8) minutes in the fentanyl/midazolam group (P<0.001).

A total number of 528 Ramsay score determinations were obtained. Most of the patients in the propofol group underwent unconscious sedation with the Ramsay score IV to V, while most of the patients in the fentanyl/midazolam group underwent conscious sedation with the Ramsay score II to III. Numerical Pain Rating Scale in the propofol group was significantly lower than that in the fentanyl/midazolam group (P<0.001). Forty-eight of the 60 (80%) patients in the propofol group were satisfied with the painless procedure, and 56 of the 60 (93.3%) patients in the fentanyl/midazolam group had tolerable pain.

The incidences of hypotension, hypoxia, cough and nausea were 28.3%, 21.7%, 15.0%, 0, in the propofol group and were 10.0%, 6.7%, 0, 11.7% in the fentanyl/midazolam group, respectively. There were significant differences in the incidences of the aforementioned complications between the two groups. Hypoxia usually happened when patients had snore, three of the four patients with snore were retrieved by jaw thrusts in the fentanyl/midazolam group and nine of the 13 patients were retrieved by oropharynx airways in the propofol group. Two cases of respiratory distress occurred in the propofol group and one occurred in the fentanyl/midazolam group. Two of the three were retrieved by face mask ventilation, the other one in the propofol group suffered from cardiac arrest, and was saved with endotracheal intubation and direct current defibrillation (360 J).

There was no significant difference in the frequency of the reference patch movement caused by sedation between the two groups. The time delayed on account of sedation problem was longer in the propofol group ((4.4±1.8) minutes vs (3.2±1.8) minutes, P=0.005). The operator was more satisfied with the fentanyl/midazolam sedation than the propofol according to the satisfaction score (P=0.002).

There was no significant difference in the success rate of pulmonary veins isolation between the propofol group and the fentanyl/midazolam group (98.7% vs 98.3%, P=1.000). After (7.3±3.2) months follow-up, the success rate of the catheter ablation of AF was 78.3% in the propofol group and 76.7% in the fentanyl/midazolam group (P=0.827). The incidences of complications, mostly groin hematoma were 13.3% and 8.3% respectively in the two groups without significant difference.


The main findings of this prospective, randomized study are as follows. The incidence of transient anaesthetic complications is higher in the unconscious sedation/analgesia group. Conscious sedation with fentanyl/midazolam is more suitable for catheter ablation of AF with fewer anaesthetic complications and less adverse influence on the ablative procedure. However there were no significant differences in the success rate and incidence of complications of the catheter ablation of AF between the two groups.

As the data shown in Gasparovic and cowokers’ study,7 propofol had more effects on the hemodynamic and respiratory parameters. Continuous and large dose infusion of propofol in the study could evoke hypotension and hypoxia. Nausea and vomiting were usual adverse effects after anesthetic.8,9 The incidence of nausea was lower in the propofol group, because propofol had clear advantage in reducing nausea as shown in Castano and cowokers’ study.10 Cough was another adverse effect of propofol, which often occurred in the early phase of propofol medication. Cough could evoke the movement of the reference patch, which made it necessary for the magnetic field to be revised. Cough also added the risk of cardiac perforation when the catheter was placed in left atrium. However, there were no effective methods to resolve this problem yet. Sedation with propofol had to be abandoned in one patient because of the intensive refractory cough.

To keep a steady procedure is a key metric in the sedation evaluation. It took more time wasting in the procedure on account of the sedation problems in the propofol group than in the fentanyl/midazolam group. Most of the delay in the propofol group was due to the management of the sedation complications and the rectification of the reference patch movement. However, it was wasted in the pause of ablation and medication titration because of patient’s pain in the fentanyl/midazolam group. The assessment of the operator was better in the fentanyl/midazolam group. This is not only because the delay is longer on account of sedation in the propofol group, but also the breath of the patients in the unconscious propofol sedation group, especially when snoring was deep and unsteady, which made it difficult for the catheter to reach the target point steadily.

The sedation protocol was not meant to gain the same level of sedation between the two groups. The investigation presented in this paper has compared not only the exact sedation effects of the two medication regimens, but also their influence on the ablation procedure. However, the score of the operator’s assessment was defined by our own experience due to the lack of established scoring system. Because the distinct medication methodologies were distinguished, it was hard to use double-blind method to evaluate the two sedation regimens. We found that the success rate of the isolation of the pulmonary veins were comparable between the two groups. Further study to compare the influence of the two sedation regimens on the electrophysiological characteristic of AF should be carried out in the future.


1. Ma CS, Dong JZ, Liu XP, Long DY, Fang DP, Hu FL, et al. Pulmonary vein tachycardia after pulmonary vein isolation in patients with atrial fibrillation. Chin Med J 2006; 119: 551-556.[PubMed]

2. Mitchell AR, Spurrell PA, Gerritse BE, Sulke N. Improving the acceptability of the atrial defibrillator for the treatment of persistent atrial fibrillation: the atrial defibrillator sedation assessment study (ADSAS). Int J Cardiol 2004; 96: 141-145. [PubMed]

3. Wang Y, Yue Y, Sun YH, Wu AS. Can bispectral index or auditory evoked potential index predict implicit memory during propofol-induced sedation? Chin Med J 2006; 119: 894-898. [PubMed]

4. Sharpe MD, Dobkowski WB, Murkin JM, Klein G, Guiraudon G, Yee R. Alfentanil-midazolam anaesthesia has no electrophysiological effects upon the normal conduction system or accessory pathways in patients with Wolff-Parkinson-White syndrome. Can J Anaesth 1992; 39: 816-821.[PubMed]

5. Ramsay MA, Savege TM, Simpson BR, Goodwin R. Controlled sedation with alphaxalone-alphadolone. BMJ 1974; 2: 656-659.[PubMed]

6. Chapman CR, Casey KL, Dubner R, Foley KM, Gracely RH, Reading AE. Pain measurement: an overview. Pain 1985; 22: 1-31. [PubMed]

7. Gasparovic S, Rustemovic N, Opacic M, Premuzic M, Korusic A, Bozikov J, et al. Clinical analysis of propofol deep sedation for 1104 patients undergoing gastrointestinal endoscopic procedures: a three year prospective study. World J Gastroenterol 2006; 12: 327-330. [PubMed]

8. Raeder J, Dahl V, Bjoernestad E, Edlund G, Modin S, Naucler E, et al. Does esomeprazole prevent post-operative nausea and vomiting? Acta Anaesthesiol Scand 2007; 51: 217-225. [PubMed]

9. Nitahara K, Sugi Y, Shono S, Hamada T, Higa K. Risk factors for nausea and vomiting following vitrectomy in adults. Eur J Anaesthesiol 2007; 24: 166-170. [PubMed]

10. Castano I, Carceles MD, Alonso B, Sanchez-Rincon I, Alfaro JM, Lopez F. Infusion of propofol as anesthetic and antiemetic drug. Rev Esp Anestesiol Reanim 1995; 42: 257-260. [PubMed]

  1. National Natural Science Foundation of China,No. 30500202 and No. 30570724;