In the present retrospective cohort study, we performed a multivariate Cox regression analysis of the first postoperative analgesia requests in surgical, gynecological, and urologic patients who underwent laparoscopic surgery. The first analgesia request after the end of anesthesia was set as the primary outcome, and the time to the first analgesic administration within 24 h was analyzed using objective data. Although the mean time was not significantly different between the two groups, the patients in the pre-RSB group showed a lower risk of first postoperative analgesic administration after PS matching. To the best of our knowledge, no previous studies examining the timing of RSB have examined the time to first analgesic administration.
Several studies have suggested that pre-RSB is more effective than post-RSB. Yagi et al. published a prospective observational cohort study of patients undergoing abdominal wall (transverse abdominal plane [TAP]) and RSB in total laparoscopic hysterectomy. They showed that nonsteroidal inflammatory drug usage in women who underwent a preoperative TAP and RS block was significantly lower than that in women who received a postoperative block within the first 12 h, with both groups receiving 40 mL of 0.25% ropivacaine in total (pre- versus post-, 54.4% versus 75.0%; P = 007) [4]. Jeong et al. reported a single-center randomized controlled trial (RCT) on the use of RSB in laparoscopic cholecystectomy. They showed that total rescue analgesic consumption at three points up to 24 h after surgery was significantly lower in the pre-RSB group than in the post-RSB group (1 h, P = 0.023; 9 h, P = 0.020; 18 h, P = 0.002), in which patients received 40 mL of 0.25% levobupivacaine [6]. In a meta-analysis of RSB in the setting of laparoscopic surgery by Hamid et al., which included 9 RCTs and 698 patients, pre-RSB showed better pain control than post-RSB in a subgroup analysis (P = 0.001) [3]. However, these previous studies, as well as the meta-analysis by Hamid et al. [3], evaluated specific procedures. In contrast, the present study included patients from many surgical departments. Furthermore, previous studies used quantitative pain assessments (i.e., pain scores, amount of total analgesics) as primary outcomes and assessed them over a certain period, whereas the present study assessed the primary outcome over time by using the Kaplan–Meier curve and Cox regression.
RSB in the pre-RSB group was performed approximately 2 h earlier than in the post-RSB group. Initially, we predicted that physicians who performed pre-RSB would intend to provide intraoperative analgesia rather than postoperative analgesia, and that physicians who performed post-RSB would intend to provide postoperative first analgesia later. However, the results of the present study contradicted these forecasts. The Kaplan–Meier curve was roughly parallel and showed that the ratio of analgesia-free patients changed approximately 2 h after surgery, suggesting that the intergroup differences in analgesia requests in the present study were earlier than those in previous studies. In previous studies, analgesia requests were only compared at fixed points, such as 1, 12, and 24 h after surgery, and differences in requests were examined only at these points [3, 4, 6]. However, the curves in the present study provide a clearer picture of the differences in the changes between the two groups over time.
Although some reports have suggested that pre-RSB is more effective for postoperative analgesia than post-RSB, we could not find any reports concluding that post-RSB is more effective [3, 4, 6]. We presumed that pre-RSB functioned as “preemptive analgesia” in the present and previous studies. Preemptive analgesia is a form of antinociceptive treatment that prevents the establishment of altered processing of afferent input, which amplifies postoperative pain [8]. In addition, although there are various definitions of preemptive analgesia, we defined it as analgesia that prevents central sensitization caused by incisional and inflammatory injuries (covering the period of surgery and the initial postoperative period) [8]. Animal studies have shown that preemptive analgesia can prevent the development of central hypersensitivity after nociceptive demands and reduce postoperative pain [13, 14]. Previous studies in patients undergoing laparoscopic surgery have also suggested that pre-RSB provides preemptive analgesia [3, 4, 6]. Although the concept of preemptive analgesia remains controversial [7, 9], it is supported by the results of the present study.
Notably, 0.2% ropivacaine was mainly used in this study. Ropivacaine (0.25% or higher) was usually used in previous studies that verified ultrasound-guided RSB with ropivacaine in laparoscopic surgery [1, 15, 16]. The pharmacokinetics of local anesthetics in RSB is not well understood, and the optimal analgesic concentration is also a topic of debate. Abdul et al. reported that 0.2% ropivacaine was as effective as 0.5% ropivacaine in postoperative analgesia with a TAP block in patients undergoing appendicectomy [17]. However, 0.2% ropivacaine is generally accepted as the optimal concentration required for postoperative analgesia in regional anesthesia techniques, except for the abdominal wall [18]. Therefore, we suppose that low-concentration local anesthesia (0.2% ropivacaine) is sufficiently effective for nociception of the umbilical region in laparoscopic surgery.
The patient characteristics in the present study were not well-balanced after PS matching, which may be attributable to the presence of an inappropriate covariate between the two groups before PS matching. Covariates that are strongly related to the intervention and unrelated to the outcome are suggested to increase the c-statistic (area under the receiver operating characteristic curve for the probability of intervention) and increase the non-overlapping distributions of PSs between the treated and the untreated groups, resulting in a reduction in the number of matched pairs for analysis [19]. The c-statistic in the present study was high (> 0.9), which suggested that such covariates had been included in the multivariate logistic regression model to calculate the PS. We suppose that “year” was an inappropriate covariate because the difference in the number of years was too large between the two groups before PS matching (absolute standardized [ASD] = 1.57). As a result, the number of PS distributions overlapped and the number of matched pairs decreased (see Additional file 1), and the balance of characteristics after PS matching was not good. In addition, some characteristics during surgery were not well-balanced. Intraoperative fentanyl and remifentanil dosages were lower in the pre-RSB group than in the post-RSB group (fentanyl, 49.5% versus 54.6%, ASD = 0.10; remifentanil, 80.4% versus 96.9%, ASD = 0.54), suggesting preoperative RSB may have reduced the need for analgesics during the surgery. In contrast, acetaminophen use was higher in the pre-RSB group (50.5% versus 43.3%, ASD = 0.17). In our institution, we tend to administer acetaminophen immediately before the end of surgery for postoperative analgesia. Before this study, we thought that postoperative administration of analgesics would be more effective for postoperative analgesia because of the longer duration of the effect of local anesthetics. Therefore, a higher percentage of patients in the pre-RSB group may have received acetaminophen as an adjunct to postoperative analgesia in advance. The duration of anesthesia was longer in the post-RSB group (168 versus 197 min, ASD = 0.44). We assumed that it was more difficult to perform RSB due to pneumoperitoneum and the presence of dressing material in the post-RSB group. Although the two groups were not well-balanced, we analyzed them with unbalanced covariates using multivariable Cox regression after PS matching to eliminate the effects of confounders. We showed that preoperative RSB was associated with a lower risk of postoperative first analgesia request (see Additional file 2). The number of postoperative analgesia requests did not differ significantly between the groups. However, the Kaplan–Meier curves for the two groups were roughly parallel, and a larger sample may have shown significant differences between the two groups. On the basis of our results, 508 cases in each group (two-sided alpha level of 0.05, power of 0.8) were required.
In addition to its retrospective nature and single-center design, this study had several other limitations. First, we were unable to compare the total postoperative analgesic requirements (frequency of analgesic use and dosage) between the two groups. We usually compare the total use of one type of analgesic as a surrogate for postoperative pain assessment [3, 4, 6]. However, various postoperative analgesics were administered in this study. Therefore, we did not assess the total use of analgesics because we could not compare the findings for different types of analgesics. The total frequency of analgesic use and dosage in the pre-RSB group may have been larger, but the low risk for first analgesia request during the first 24 h after surgery implied that many patients were at least free from pain commonly associated with postoperative discomfort, including nausea, vomiting, sore throat, and fatigue. This result may have contributed to an improvement in the quality of the patient’s recovery. Therefore, in determining the appropriate time to perform RSB, our findings showing that the proportion of patients who did not require analgesia within 24 h was larger in the pre-RSB group may be valuable. Second, an accurate time-specific quantitative pain assessment could not be performed. Although a pain score such as the visual analog scale (VAS) is typically used for this purpose, the VAS was presumed to be unreliable because nurses at our institution did not receive sufficient training to appropriately assess pain. Thus, to avoid incorrect results, we did not assess VAS scores. Third, anesthesiologists’ experience and skills were unknown. Although RSB was performed under ultrasound, our study was conducted during the transition toward performing RSB with ultrasound, and familiarity with ultrasound-assisted RSB was low for some time during the study period. Thus, there may have been some skill-related bias, and future studies should aim to eliminate this bias in their analyses. Fourth, background factors that influence pain are unknown. Anxiety, psychological distress, pre-existing pain, and others are predictive factors for postoperative pain [20], but we could not collect the data for these factors. We hope that future prospective studies will be conducted with these data.