Patients
This study was conducted according to the STROBE recommendations [17, 18]. The study protocol was approved by the Institutional Ethics Committee of Xinqiao Hospital, Third Military Medical University, Chongqing, China. Prior to the enrollment of patients, written informed consent was obtained from all the patients, and the study was registered on Clinicaltrial.gov (ID: NCT03009955).
Patients were included according to established inclusion and exclusion criteria. From January to May 2017, 168 Chinese patients, aged 20 to 40 years scheduled to undergo elective cesarean section with a transverse incision were recruited for this study (Figure 1). Patients who had a gestational age of 37 to 40 weeks and singleton pregnancy, voluntarily receiving intravenous patient-controlled intravenous analgesia (PCIA) treatment, and classified as having ASA physical status scale I-II were eligible for participation. The reasons for an elective cesarean section in a primipara included the patient's own choice, preoperative complications including malpresentation (breech and transverse positions and compound presentation), placenta previa, uterine inertia, gestational diabetes, chronic or gestational hypertension, and preeclampsia. For the multiparas, the indication for a cesarean section was a previously scarred uterus. Only those who were undergoing their first repeat cesarean deliveries were included. Exclusion criteria included a history of chronic pain disorder, recent or chronic opioid use, substance abuse, heavy smoking (>30 pack-years) [19] or alcohol dependence, absolute or relative contraindication to subarachnoid space block anesthesia, history of prior pelvic or abdominal surgery, or severe pregnancy complications, such as heart disease, brain disease, liver disease and kidney disease, that were life-threatening and required emergency treatment prior to the cesarean section.
Anesthesia and Analgesia Management
Cardiac rhythm via electrocardiography, mean arterial pressure, and pulse oxygen saturation were monitored after the patients entered the operating room. Standardized anesthesia was administered by an experienced anesthetist, and the operations were conducted by a single surgical team using the same standardized technique. Spinal anesthesia, via a subarachnoid space block at the L3–4 interspace, was administered using 0.66% ropivacaine (20 mg).
After the fetal section and once daily after the surgery, oxytocin (20 IU in 500 mL of saline) was routinely administered while the patient was admitted to the obstetrics ward. PCIA was started immediately after surgery with a mixture of hydromorphone (0.2 mg/kg), flurbiprofen (4 mg/kg), and 0.9% normal saline at a dose volume of 200 mL, using a controlled infusion pump. The pump was programmed to a loading dose of 2 mL, background infusion rate of 2.0 mL/h, and PCIA dose of 1 mL, with a lockout period of 15 min. For the prevention of the postoperative nausea and vomiting, 3 mg of droperidol was administered at the outset of PCIA. Patients were monitored for 6 hours in the postanesthesia care unit of the obstetrics ward after surgery. When pain was treated inadequately, the patients were administered additional pain treatment with tramadol 50 mg in a timely manner.
Outcome Measurements
Standardized training for follow-up assessment was performed for all included investigators. A pain visual analogue scale (VAS score; 0–100, where 0 is defined as no pain and 100 as maximum pain) was used to evaluate postoperative pain at 4, 8, 12, 24, and 48 hours. The primary outcome was the incidence of inadequate analgesia (defined as a pain VAS score ≥40) [20] during the postoperative 48 hours. Pain caused by abdominal incision at rest and during mobilization (during coughing) was assessed using the VAS. Visceral pain was also assessed using the VAS. For visceral pain, the subjects were asked to report the pain induced by uterine contractions and were informed that the visceral pain could be enhanced when oxytocin was given. The duration of pain according to the patient’s self-reported time and PCIA consumption for 48 hours after surgery were recorded.
Assessment with the hospital anxiety and depression scale (HADS) before the operation was performed in all patients. The HADS includes 14 assessments, including the symptoms of anxiety and depression (seven items scored 0 to 3 in each subscale, yielding a range of 0–21) with subscale scores of 8 indicating possible anxiety or depression [21, 22]. The intraoperative amount of blood loss, neonatal Apgar score, weight and height of the newborn, and surgery time were recorded. The Ramsay sedation score, respiratory rate, pulse oxygen saturation, systolic pressure, diastolic pressure, and heart rate were recorded before surgery and during the postoperative 48 hours. Early walking time (determined by the time point when patients could ambulate) was also recorded. The sleep quality (rated as good or poor) on the day of and one day after surgery was evaluated. Postoperative adverse events including nausea, vomiting, and pruritus were also noted. Additionally, the patients’ duration of hospital stay was recorded.
The results of routine blood examinations before and 24 hours after the surgery were retrospectively collected for all patients. The leukocyte and neutrophil counts were analyzed. At 1 week and 4 weeks after surgery, patients were interviewed by telephone and asked the following questions from a standardized questionnaire: Was there an existing pain? Was the location of pain at the incision, viscera, both, or none? Was sleep affected? Were they able to perform the activities of daily life with full autonomy, partial dependency, or absolute dependence?
Sample size determination
The sample size was calculated according to the design of chi-square test for four-fold table data in a cohort study. Since previous studies reported the incidence of moderate to severe pain under postoperative analgesia for primipara as approximately 50% [4, 5], the current study hypothesized that the relative risk (RR) value for multiparas was 1.5 compared to that for primiparas. The anticipated incidence for multiparas was 75%. Therefore, based on a significance level of 0.05, power of 0.9, and an estimated ratio between the number of multiparas and primiparas of 1.5, according to the retrospective analysis based on the data from our hospital Electronic Medical Records System of the past 1 year, and considering about 3% loss of follow-up, the total required minimum sample size was determined to be 168 individuals using the sample size calculation software PASS, version 11.0 (NCSS, Kayesville, UT).
Statistical Analysis
Statistical analysis was performed using SPSS for Windows version 19.0 (SPSS Inc., Chicago, IL). A two-tailed P-value less than 0.05 was considered statistically significant. The mean ± standard deviation (SD), median (interquartile range), and number (frequency) were used to summarize the variables. The patients who were scheduled to undergo a primary cesarean section were designated as group P in the final analysis, while the patients scheduled to undergo repeat cesarean section were designated as group R. The primary outcomes (postoperative inadequate analgesia on incision or visceral pain) were respectively described and analyzed. Logistic regression analysis using enter model was performed to evaluate the role of group P or group R in the prediction of postoperative inadequate analgesia. The presence of postoperative inadequate analgesia on incisional and visceral pain was considered as the outcome variable. BMI, age, gestational age, surgery time, preoperative complications (yes/no), depression (yes/no), and anxiety (yes/no) were also considered in the model. Odds ratios (OR) with 95% confidence intervals (CIs) were determined based on the logistic regression analysis.
An independent-sample t test was used to compare the differences in demographic and preoperative data between group P and R. Due to abnormal distribution, HAD scale, incision pain VAS at rest, and visceral pain VAS were compared using a Mann-Whitney U test. Propensity score matching (PSM) analysis was performed using STATA version 12 (Stata Corp, College Station, TX). Group P and group R were matched by propensity scores, and factors used to generate the propensity scores were those preoperative factors which had significant difference between the two groups. These factors included age, gestational age, and preoperative complications. Patients were matched in a 1:1 ratio without replacement. The caliper was defined as 0.2. The absolute standardized difference was calculated, and the absolute standardized difference less than 10% was considered to support the assumption of balance between the two groups. Then, other postoperative outcomes including the start time to feel pain, early walking time, hospital stays and PCIA administration were compared between groups P and R.
Differences in the incidence of postoperative inadequate analgesia, sleep quality, adverse events, and long-term pain status between the two groups were analyzed using Pearson’s chi-squared test. Furthermore, RR values and 95% CI for the probability of the occurrence of inadequate analgesia on incision pain and visceral pain during the postoperative 48-hour follow-up were calculated, as well as the postoperative pain status at 1 and 4 weeks. Subgroup analysis according to age group (≤30 years or >30 years) was performed. Two-way repeated analysis of variance (ANOVA) with post hoc LSD testing was used to compare the preoperative and postoperative systolic pressure, diastolic pressure, heart rate, respiratory rate, and leukocyte and neutrophil counts between the two groups.