Risk factors of different types of acute pain after laparoscope-assisted vaginal hysterectomy

2.1 Study design and participant recruitment

This prospective, observational cohort study was approved by the Ethics Committee (2019-030) in Clinical Research of the First Affiliated Hospital of Wenzhou Medical University and was registered at China Clinical Trial Registry: ChiCTR-ROC-17013036. Patients scheduled for LAVH at the Gynecology Department at First Affiliated Hospital of Wenzhou Medical University between January, 2017 and October, 2018 were enrolled in the study on the day before surgery. Written informed consent was obtained from every participant by preoperative interview on the same day; subsequently, data of patient characteristics were collected. Data of futher information about the surgery, anesthesia and pain were collected by researcher on the day of operation and 24 h, 48 h after operation. Data of all patients who completed the entire study were submitted to analysis.

From January, 2017 to October, 2018, indices of potential acute postoperative pain were recorded in a single database. Patients scheduled for LAVH under general anesthesia were recruited; at the time of registration, eligible patients were $>$ 18 years-old. Those with serious psychological disorder and other significant cardiovascular, respiratory, hepatic and kidney diseases were excluded; other exclusion criteria were limitations of self-expression or language problems, pregnant mothers, and nursing mothers.

2.2 Types of outcomes

For development of risk assessment models, the primary outcome was incidence and intensity of acute postoperative pain, including somatic pain, visceral pain, shoulder pain, and LBP. Intra-abdominal pain (visceral pain) was defined as pain within the abdomen, which may be deep, dull, and more difficult to localize, but that may be predominantly located in the hypogastrium. Somatic pain was a superficial pain localized in the wounds that were sensitive to touch. Shoulder pain was defined as pain sensation in the shoulder(s) [9, 17]. Low back pain (LBP) was defined as pain located between the 12th rib and the inferior gluteal folds, with or without leg pain according to a previous study [18].

Pain intensity was assessed using NRS from 0 to 10, with zero representing no pain and 10 representing worst imaginable pain [19], and categorized into two groups: no pain or mild pain, 0 to 3; moderate to severe pain, 4 to 10 [8, 20].

The secondary outcomes were (1) postoperative complications, such as respiratory depression, bleeding, and fever and (2) length of hospital stay, exhaust time, and time to get out of bed.

2.3 Data collection

Data collection and processing were completed by three anesthesiologists and four college students; preoperative data were collected and checked by two college students and entered by two other college students all who completed 1 week’s training for their respective tasks. Two anesthesiologists were responsible for intraoperative anesthesia and administration of analgesia, and two university students, for intraoperative data collection. Follow-up was conducted by another anesthesiologist who was blinded to preoperative and intraoperative data, which included pain assessment, and management of postoperative pain and other complications.

Variables as potential risk factors for acute postoperative pain were identified based on available literature. Preoperative variables were age, BMI, education, occupation, exercise habit, individual disease history, obstetrical history, previous surgical history, history of preoperative chronic pain, type of medical insurance, main indication for surgery, and anxiety level. Data of patient’s characteristics were obtained by interview on the day before surgery.

Patient’s age, height and weight, occupation, vital signs, obstetrical history, individual disease history, and surgical history were obtained by review of the medical record, and the data were verified for accuracy by staff who conducted preoperative interview. BMI was categorized dichotomously with cutoff at 24 [21, 22]. Subjects were classified by dichotomous variable according to previous major surgery with or without abdominal surgery.

In addition, history of pre-existing pain was assessed, which included aspects of its location, duration, and intensity. Patients who experienced pain for a period of $>$ 3 months at $\ge$ 15 days per month with intensity of $\ge$ 2 on an 11-point NRS by preoperative interview were considered to have pre-existing chronic pain [23]. In the questionnaire, preoperative pain was classified according to the location of pain such as the head, cervical vertebra, waist, back, chest, abdomen, and limbs [24].

2.4 Anesthesia, surgical procedures and analgesia

In all patients, similar regimens of general anesthesia and prophylactic analgesia were administered without any premedication on the day of surgery. Patients were monitored with standard procedure by bispectral index (BIS) for depth of anesthesia. General anesthesia was induced with a combination of sulfentanyl (0.4 $\mu$g.kg${}^{-1}$) and propofol (2-3 mg.kg${}^{-1}$). Next, rocuronium (6-8 mg.kg${}^{-1}$) was administered to allow tracheal intubation, and further boluses of rocuronium were administered to maintain the train-of-four at 0-10%. Anesthesia was maintained with oxygen in air (1 : 2), propofol, remifentanil, and sevoflurane. Morphine (0.1 mg.kg${}^{-1}$) and palonosetron (0.25 mg) were administered 30 minutes before the end of the operation when administration of the inhalation anesthesia was stopped. During closure of the skin, remifentanil infusion was discontinued [25], and after closure, 10 mL of 0.75% ropivacaine local anesthesia was infiltrated at the incision site.

All surgical procedures were performed by or under supervision of experienced surgeons according to standard protocol with laparoscopic ports (1 $\times$ 10 mm and 2-3 $\times$ 5 mm). The uterus was removed vaginally by volume-reducing techniques. During surgery, intra-abdominal pressure was maintained at 12-13 mmHg, and at the end of procedure, CO2 was evacuated by manual compression of the abdomen through the open laparoscopic ports. Intraoperative procedures and diagnosis, duration of operation, bleeding volume, and intraoperative complications were noted.

At the post anesthesia care unit (PACU), patients were titrated with morphine by the anesthesiologist as follows: Level of subjective pain intensity was obtained at 5 minutes’ interval, and repeat 1-2 mg bolus dose of morphine was administered until NRS score of $<$ 4/10 was achieved, indicating adequate level of analgesia based on responses of the patients to questions by the anesthesiologist [26].

One hour later, the patients were moved to the ward; if recurrence of NRS score $\ge$ 4/10 was observed, 50-mg suppository of indomethacin was administered though the rectum, and if the score of $\ge$ 4/10 persisted, IM pethidine injection was administered to relieve pain [27]. The total amount of analgesia during postoperative period (including that at PACU) and any adverse events were recorded in each patient.

Pain was assessed by the study team retrospectively on postoperative day 1 and 7 at 17:00 [28], using 11-point NRS for different types of acute postoperative pain according to the method in a previous study for average pain at rest and at movement, and worst pain [12]. For assessment of the predictors of and risk factors for intense pain scores after surgery, we considered that highest pain intensity is generally attained on postoperative day 1, and highest pain scores are obtained at interview itemizing maximum pain after surgery [15]. Based on these facts, we selected maximum pain during 7 days as the measure of severe postoperative pain in the prediction analysis [23].

2.5 Statistical analysis

In our hospital, every year, approximately 1500 patients undergo LAVH surgery. Based on mean value of maximum acute postoperative pain after LAVH of approximately 4.5 (NRS 0-10), standard deviation of 2, and permissible error $\le$ 0.5, in total, 207 patients were required, with expected loss to follow-up of 10%.

Statistical analysis was performed with Stata 15. Baseline data are expressed as mean (SD) for continuous variables, or number of patients (%) for binary or categorical data. Normality was determined using Kolmogorov-Smirnov tests. Box plot or scatter plot was used to detect outliers. Collinearity was examined by variance inflation factor (VIF); according to NRS score, acute postoperative pain was categorized into mild pain and moderate to severe pain [29]. Categorical data were compared using Fisher’s exact test. Continuous variables were compared using Student’s t-test (parametric) and Mann-Whitney U-test (nonparametric). Assumption of linearity was examined using empirical logit plots to preserve degrees of freedom. All variables with P-value $<$0.20 in univariate analysis were entered into multivariable logistic regression models; those with P $\le$ 0.05 by multivariable logistic regression were considered significant, and the odds ratio (OR), 95% CIs, and P-value were reported. Discrimination of the model was quantified by the area under the receiver operating characteristic (ROC) curve [30]. All statistical tests were two-sided, and P $<$ 0.05 was considered to be statistically significant.

3.1 Study sample

A total of 407 patients were screened, among who, 120 patients were missing complete records or without records: Forty-two (35%) patients were excluded because of change in the mode of surgery, 38 patients did not conform to the inclusion criteria, and three patients (2.5%) were lost to follow-up. Finally, 287 remaining patients with complete baseline data were enrolled. Flowchart of exclusion and loss to follow-up is shown in Fig. 1.

Fig. 1.Flowchart of the patients.

The mean [minimum-maximum] age at the time of inclusion was 49.20 [33-70] years. In the cohort of 287 patients, history of hypertension was noted in 61 patients (21.3%), and diabetes, in only 17 patients (5.9%); history of abdominal surgery, such as cesarean section, hysteromyoma decollement, cholecystectomy or appendectomy was noted in 74 (25.8%) patients [31, 32]. Majority of the patients were without much anxiety on preoperative day 1, and among those, self-report of mild anxiety was obtained in 74.2% patients. In 253 (88.2%) patients, birth was through vaginal route, and in 34 patients (11.8%), it was by cesarean section (Table 1).

3.2 Preoperative pain

In the study cohort, preoperative chronic pain was observed in 84 patients: headache in 23 (8.0%) patients, cervical pain in 13 (4.5%) female patients, low back pain in 46 (16.0%) patients, shoulder pain in four (1.4%) patients, stomachache in one patient, and pain of the limbs in five (1.7%) female patients; in addition, dysmenorrhea was observed in 100 (34.8%) female patients. Nevertheless, no chronic pain was reported by 129 patients before surgery.

Different types of moderate to severe pain ($\ge$ 1 type) were observed in 217 patients: one type in 116 patients, two types in 81 patients, three types in 17 patients, four types in 3 patients. Moreover, no pain or mild pain was reported in 70 patients, which suggests achievement of insufficient analgesia in our cohort.

3.3 Postoperative pain

Among 254 patients with postoperative visceral pain, the pain was located at the middle of the pelvic area in majority of the patients and with moderate to severe degree in 125 women. Within 2 days after surgery, low back pain was reported in 220 patients, of which 108 patients were cases of moderate to severe low back pain, and 11 patients were worst cases of severe low back pain such as spine fracture. After delivery of local anesthesia to the area of the laparoscopic ports, severe somatic pain was revealed in only six patients. Moreover, in our study, shoulder pain was observed in 44 (15.3%) patients (Table 2).

In the total cohort, on postoperative day 1, the mean NRS score of maximum postoperative visceral pain was 3.71, and that of maximum acute postoperative low back pain was 3.17, and no significant difference was observed between the patients with visceral pain and those with low back pain; in addition, the mean NRS score of maximum acute postoperative somatic pain was 1.72. With regard to the pain location/type, visceral pain and low back pain was observed in majority of the patients, followed by somatic pain and shoulder pain.

3.4 Postoperative analgesia

At PACU, analgesics were administered in 35 (12.3%) patients, and maximum pain was 5 on NRS (0-10); particularly, on postoperative day 1, analgesics were administered in 93 (32.4%) patients, and among those, repeat administration of analgesic agent was required in 34 (36.56%) patients. After administration of pethidine, there was no recurrence of serious pain in almost all patients.

3.5 Risk factors of acute postoperative visceral pain

Among 24 potential risk factors included in univariate analysis, six factors were associated with acute postoperative visceral pain with statistical significance (P $<$ 0.2): medical insurance type, hypertension, diabetes, BMI, main indication for surgery, and postoperative nausea and vomiting (Table 3). All those combined were submitted to multivariate analysis; in the results, BMI, main indication for surgery, and postoperative nausea and vomiting were associated with acute postoperative visceral pain (P $<$ 0.05) (Table 6).

3.6 Risk factors of acute postoperative low back pain

Among the risk factors of low back pain included in univariate analysis, 11 factors were associated with acute postoperative low back pain with statistical significance (P $<$ 0.2): age, medical insurance type, history of abdominal surgery, mode of delivery, BMI, main indication for surgery, preoperative chronic pain, anxiety, the number of trocar, bleeding volume, and the time of indwelling catheter (Table 4). All those combined were submitted to multivariate analysis; in the results, BMI, preoperative chronic pain, anxiety, and the number of trocar were associated with acute postoperative low back pain (P $<$ 0.05) (Table 6).

3.7 Risk factors of acute postoperative somatic pain

In the univariate analysis, 10 factors were associated with acute postoperative somatic pain with statistical significance (P $<$ 0.2): age, occupation, hypertension, diabetes, history of abdominal surgery, mode of delivery, level of anxiety, pelvic adhesions, the number of trocar, and time of pelvic drainage (Table 5). All those combined were submitted to multivariate analysis; in the results, only occupation, pelvic adhesions, and history of abdominal surgery were associated with acute postoperative somatic pain (P $<$ 0.05) (Table 6).

Somatic pain was a significant factor of the length of hospital stay (P = 0.04), and visceral pain and somatic pain were significant factors of the increase of the time to get out of bed after surgery (each, P $<$ 0.01). Nevertheless, none of the pain types was an influencing factor of the exhaust time (P $>$ 0.05).