Efficacy of Programmed Intermittent Bolus Epidural Analgesia in Thoracic Surgery

A Randomized Controlled Trial

M. Higashi; K. Shigematsu; E. Nakamori; S. Sakurai; K. Yamaura


BMC Anesthesiol. 2019;19(107) 

In This Article


This randomized prospective study was approved by the Institutional Review Board (IRB No. 15–9-06) of Fukuoka University Hospital, Fukuoka, Japan, and was registered in the clinical trials database UMIN (ID 000019904) on 24 November 2015. Written informed consent was obtained from all patients.


Patients undergoing open lung lobectomy or partial lobectomy at the Fukuoka University Hospital, Fukuoka, Japan between March 2016 and March 2017 were recruited. The exclusion criteria were age < 20 years and contraindication for epidural anesthesia. Patients were randomly divided into a CEI or PIB group by computer generated randomization using Excel 2013 (Microsoft Inc., Redmond, WA) by KY (Figure 1).

Figure 1.

Flow chart of this study

The study was discontinued when epidural analgesia was ineffective, when the patient's hypotension continued even after the administration of vasopressor, or when motor paralysis appeared owing to epidural analgesia.


Under standard monitoring, thoracic epidural anesthesia was performed at Th4–6 in the lateral position. An 18G epidural Tuohy needle (Uniever®, Unisis Corp., Saitama, Japan) was used, and the epidural space was identified using the loss-of-resistance technique. A 20G epidural catheter (Uniever®, Unisis Corp.) was inserted 5 cm to the head side. Following a 3-mL test dose of 1% mepivacaine, the epidural catheter was fixed.

General anesthesia was induced with intravenous fentanyl (2 μg/kg), propofol (1 mg/kg), and rocuronium (0.9 mg/kg) and was maintained with sevoflurane (1.5–2%) and remifentanil (0.1–0.2 μg/kg/min). Fentanyl was used intravenously up to 5 μg/kg. A local anesthetic via the epidural catheter was not used during the operation.

After surgery, all patients were extubated in the operating room, observed in the post-anesthesia care unit for 30 min to 1 h, and then transferred to the ward.


At the end of surgery, a 5-mL initial dose of local anesthetic (ropivacaine 2 mg and fentanyl 2 μg in 1 ml) was administered via the epidural catheter after closure of thoracotomy in both groups.

The study protocol is shown in Figure 2. In both the PIB and CEI groups, a pump (CADD-Solis ambulatory infusion pump, Smith Medical, St Paul, MN, USA) was used. In the CEI group, patients received continuous infusion of the local anesthetic at a rate of 5.1 mL/90 min (3.4 mL/h). In the PIB group, the pump delivered the local anesthetic at a dose of 5.1 mL every 90 min. The PCA system was programmed to deliver a 3-mL bolus of the local anesthetic with a lockout interval of 15 min in both groups.

Figure 2.

Study protocol. PIB: programmed intermittent bolus, PCA: patient controlled analgesia, CEI:continuous epidural infusion

The primary endpoints were the frequency of PCA and total dose of local anesthetic during 36 h of postoperative period. The secondary endpoints were pain intensity, frequency of rescue analgesics, including nonsteroidal anti-inflammatory drugs and acetaminophen, adverse events, hypotension, and postoperative nausea and vomiting (PONV). Hypotension was defined as systolic blood pressure (SBP) 20% less than the baseline value or less than 90 mmHg. The onset of adverse reactions and use of rescue analgesics postoperatively were examined. The pain intensity was assessed using a visual analog scale (VAS) during rest, deep breathing, cough, and movement.

Statistical Analysis

Continuous variables are expressed as mean ± standard deviation or median [lower–upper quartiles]. From the results of preliminary study, total dose of local anesthetics in CEI was 24 ml more than in PIB, and SD of CEI was 32. Based on these results, we estimated that the following: SD = 32, Δ = 0.78, α = 0.05, and beta = 0.2. The required number of cases was estimated to be 21 for each group. We considered a 10–20% dropout rate; therefore, 50 patients were enrolled. Differences between groups were examined for statistical significance by using student's t-test after logarithmic transformation. Student's t-test, the chi-square test, and the Mann–Whitney U test were used for statistical analyses. A P-value < 0.05 was considered statistically significant.