Study design
This is a retrospective, cohort study of existing medical data that were prospectively entered into an institutional perioperative analgesia and nerve block database. The hospital’s institutional review board approved the data review (No. S-K 1574), and patient informed consent was waived.
Participants
We searched medical records of patients receiving open hepatectomy for hepatic tumor with a J-shaped subcostal incision from Jan 2019 to Dec 2019 in the database. Patients receiving intermittent bolus paravertebral blocks (0.5% ropivacaine 25ml before surgery plus 0.125ml•kg− 1 0.2% ropivacaine bolus per hour after surgery) and self-controlled intravenous bolus morphine pumps for postoperative 48 hours were recruited into the paravertebral block (PVB) group. Patients receiving only self-controlled intravenous bolus morphine pumps for postoperative 48 hours, with matched age, sex and body mass index were selected as control group.
Patients comorbid with other pain diseases, received analgesic techniques apart from the studied analgesia program, or failed to complete the analgesia program or follow-up due to reasons that were not related to perioperative analgesia, for instance, postoperative hemorrhage, surgical site infection or self-withdraw, were excluded from the study.
Baseline data collection
The hospital’s perioperative analgesia team consisted of a group of experienced anesthesiologists, they provided all kinds of analgsic services to the surgical patients, including preparing analgesia pump, performing nerve blocks, postoperative follow-up and data recording.
Baseline data included age, sex, body mass index, American Society of Anesthesiologist (ASA) classification and preoperative surgical site pain score evaluated by a 11 points-numerical rating scale (NRS). The NRS pain score was evaluated and recorded by the anesthesiologists during preoperative visit, with 0-point indicating no pain and 10-point indicating the maximum degree of insufferable pain. Patients were also educated with a pain control goal of NRS < 4. Laboratory examination data that were collected included preoperative liver, renal and coagulation function and complete blood count.
Paravertebral block
In the PVB group, all paravertebral blocks were performed by the perioperative analgesia team members before anesthesia. The paravertebral block was performed using an in-plane approach(15) to insert a 21-gauge 10 cm needle (PlexoLongNanoline; Pajunk Inc, Geisingen, Germany) into the T8 paravertebral space between the internal intercostal membrane and the pleura under the guidance of ultrasonography (CX-50, Philips Inc., USA). After a negative aspiration test, 25 ml of 0.5% ropivacaine was slowly injected into the paravertebral space. Then, a catheter (PlexoLongNanoline; Pajunk Inc, Geisingen, Germany) was inserted through the needle into the paravertebral space with a depth of 1–2 cm, tunneled subcutaneously and secured with Bio-gel to the patients’ back. The catheter was connected to a programmable, portable, electronic infusion pump (Apon ambulatory infusion pump ZZB-I, Jiangsu Apon Medical Technology Co., Ltd.) to deliver a bolus of 0.125 ml•kg− 1 0.2% ropivacaine per hour, commencing immediately after surgery. In the control group, no nerve blocks were performed before or after the surgery.
Intraoperative data collection
Both PVB and control groups received general anesthesia with endotracheal intubation. All patients recorded in the studied database adopted similar general anesthesia plan. Induction was performed with intravenous fentanyl, propofol and rocuronium. Anesthesia was maintained with sevoflurane and a 50%O2-50%N2O mixture to maintain a BIS index within 40–60. All open hepatectomies were performed by the same group of surgeons. The open hepatectomy with a J-shaped incision consisted of a right subcostal incision with a medial cranial extension to the xiphoid process and a variable right lateral extension with transection of the oblique abdominal musculature.(16)
Intraoperative data that were collected included the baseline and mean value of heart rate and blood pressure, mean sevoflurane concentration, intraoperative medication and fluid volume. All vital signs and sevoflurance concentration were collected by the monitor and ventilator in real-time and transmited to the database via electronic cables at 5 minutes interval during the surgery. All medications and fluids were recorded by the anesthesiologists during the surgery.
Postoperative data collection
Both PVB and control groups’ patients received self-controlled intravenous bolus morphine pump (Gemstar, Hospiria Inc., USA) after extubation. Intravenous morphine was given with no background infusion, patient-controlled bolus of 1–2 mg, 5-minute lockout interval and an upper limit of 8mg per hour. If the patients still complaint about pain with the upper limit dose, tramadol was given as rescue analgesia.
Postoperative paravertebral and intravenous analgesia were provided for 48 hours. The perioperative analgesia team members visited the patients and assessed the outcomes in the ward at postoperative 2, 4, 12, 24 and 48 hours and follow-ups were done by telephones at postoperative three months.
Primary outcome was postoperative 48 hours cumulative morphine consumption collected from electronic pump record. Secondary outcomes included: (1) cumulative morphine consumption at postoperative 2, 4, 12 and 24 hours and total number of rescue analgesia; (2) pain NRS score at rest and on movement at postoperative 0, 2, 4, 12, 24 and 48 hours; (3) opioid adverse effects including nausea, vomiting, pruritus, respiratory depression, bowel movement recorded as gas time and urinary retention recorded as Foley catheter removal time; (4) recovery data that were evaluated at postoperative 48 hours included drowsiness, thirsty, cold feeling, cognitive decline and shiver. These indices were evaluated using a 0–3 points Likert scale with 0 point defined as none, 1-point defined as mild, 2-point defined as moderate and 3-point defined as severe. Emergence, analgesia and overall satisfactions were evaluated using a 1–5 points Likert scale with 1-point defined as very unsatisfied, 2-point defined as unsatisfied, 3-point defined as no comments, 4-point defined as satisfied and 5-point defined as very satisfied; (6) length of hospital stay; (7) recovery data that were evaluated at postoperative three months included incidences of hypoesthesia, numbness and pain, pain NRS scores at rest and on movement, pain characteristics including throbbing, aching, pricking and stabbing pain, and the incidence of sleep disorder.
Statistical analysis
Study sample size was calculated based on a pilot study of ten patients receiving the studied analgesia program, that were randomly selected from the database via computer. The mean postoperative 48 hours cumulative morphine consumptions were (16.8 ± 13.5) mg and (32.7 ± 19.2) mg in the PVB and control groups, respectively. Thirty-five patients were required in each group to achieve an α level of 0.01 and β level of 0.9.
Statistical analysis was performed using SPSS for Mac version 23.0 (IBM Corp., Armonk, NY, USA). Normality was tested using the Q-Q plots. Normally distributed variables were expressed as mean ± SD, non-normally distributed variables were expressed as median (quartile), and categorical variables were expressed as frequency (percentage). Normally distributed continuous data were analyzed using the students t test and non-normally distributed continuous data were analyzed using the Mann-Whitney U tests. Categorical data were compared using the Chi-square test when the expected cell counts > 5, otherwise the Fisher’s exact test was used. All tests were two-tailed, and a P value less than 0.05 was considered statistically significant. The Bonferroni correction was performed on the raw P value where applicable. For the multiple comparison of cumulative morphine consumption and pain score at different time points, the P values for the 0.05 level of significance were adjusted to 0.01 and 0.008 based on the number of observed time points.