M-PALS Guidelines

Evidence for gabapentin was limited to three randomized trials (one high RoB) including 357 patients. Two studies compared gabapentin, 300 mg; or 1200 mg versus placebo. An additional study compared a preoperative regimen of acetaminophen and celecoxib alone versus acetaminophen and celecoxib plus the addition of a single dose of gabapentin (600 mg). Two trials excluded patients with chronic pain or psychiatric comorbidity. Huynh et al (high RoB) included patients with history of chronic pain but did not conduct a subgroup analysis and did not report on chronic opioid use or psychiatric comorbidity among the included patient sample.

Overall, two studies reported no significant difference in pain scores at 24 hours and one study reported no difference in pain scores at two weeks (insufficient evidence). No other pain scores were reported. Two studies reported inconsistent results for total postoperative opioid consumption at 24 hours (insufficient evidence); one study reported a significant difference favoring the addition of gabapentin but a second study demonstrated no difference in opioid consumption. One study reported no difference in, 2-week opioid consumption and no difference in hospital length of stay.

With respect to harms, the evidence was insufficient based on three studies. Two studies found no difference in nausea and vomiting between the gabapentin versus the control groups while one study reported less nausea and vomiting in the gabapentin group (insufficient).

Overall, there was limited evidence on pain scores and postoperative opioid consumption at 24 hours and no data on 48 hour or discharge pain scores or opioid consumption. Therefore, we conclude that the evidence is insufficient for the effectiveness, comparative effectiveness and harms of preoperative single dose gabapentin versus placebo or when added to celecoxib and acetaminophen in adults undergoing abdominal laparoscopic surgery. See Table 1

For intraperitoneal or preperitoneal instillation of local anesthesia, 10 studies (9 intraperitoneal) including 1091 patients (3 high RoB) informed this clinical question. Two RCTs included 232 patients with a history of chronic pain or chronic opioid use but this represented only <10% of 233 randomized patients (n=250 patients; k=2 RCTs). None of the RCTs explicitly excluded 234 patients with a psychiatric comorbidity, but only one RCT included patients with any psychiatric comorbidity (<5% out of 127 patients included in the RCT).

Overall, for patient-reported pain outcomes at 24 hours, there may be a trivial reduction in pain scores that were not clinically meaningful. Nine studies reported on 24 hour pain scores, and five were meta-analyzed (low strength of evidence) demonstrating a trivial reduction in pain (-240 0.47; 95% CI -0.88 to -0.07). One study found no difference in pain scores at 48 hours (insufficient).

For cumulative postoperative opioid consumption at 24 hours, there was insufficient evidence based on three studies that were meta-analyzed (-20.71 OME; 95% CI -40.32 to -1.10). Four studies reported a reduction in opioid rescue medications (insufficient). There was insufficient and conflicting evidence on hospital length of stay; four studies found no difference and one study favored the intervention. One study found no difference in patient satisfaction (insufficient evidence).

With respect to harms, five studies reported on nausea and vomiting: and only one 251 study demonstrated significantly less nausea and vomiting in the intervention arm.

Overall, there was no information on pain or opioid prescriptions at discharge and limited data showing non-clinically meaningful differences in 24 hour pain scores and insufficient evidence for 48 hour pain scores. Information on length of stay and opioid rescue medications was deemed insufficient. There were no studies on pain scores at 72 hours or discharge or opioid consumption at 48 hours, 72 hours, or hospital discharge. Thus, we concluded that there was insufficient evidence on the role of intraperitoneal or preperitoneal instillation of local anesthesia compared to usual care on postoperative opioid consumption and pain outcomes. See Table 2

Twenty-one RCTs evaluated regional anesthesia (peripheral nerve blocks of varied types and approaches) for a wide spectrum of abdominal and pelvic laparoscopic procedures. Studies comparing regional anesthesia with usual care (k=9) and regional anesthesia versus sham (k=13) were included in the review.

Regional anesthesia compared with usual care (9 RCTs, n=1226 patients). Of 9 RCTs evaluating regional anesthesia versus usual care, five RCTs did not report baseline information regarding chronic pain or chronic opioid use, and three RCTs did not report about any psychiatric comorbidity. Three RCTs excluded patients with a history of chronic opioid use, four RCTs excluded patients with chronic pain, and five excluded those with history of psychiatric comorbidity. One RCT reported excluding patients with a history of substance use disorder. One RCT included patients with a psychiatric comorbidity (anxiety or depression), representing 36% of their study sample.

Overall, results were inconsistent for patient reported pain scores at 24 hours, 48 hours, and 72 hours, with effect sizes that did not meet the threshold for minimal clinically important differences. For 24 hour pain scores, there may be no clinically meaningful difference based on seven RCTs (low strength of evidence). Pooled mean difference from four RCTs was -1.73 (95% CI -4.22 to 0.77). Two studies, using a different scale, reported a statistically significant difference favoring regional anesthesia, and one study showed no difference. For 48 hour pain scores, there may be no difference based on six RCTs (low strength of evidence). The pooled mean difference across three RCTs was -0.99 (95% CI -5.25 to 3.27) and one RCT reported differences in pain scores favoring regional anesthesia while two studies reported no difference. For 72 hour pain scores, there was insufficient evidence; two studies found no statistically significant difference while one study reported a difference (insufficient evidence).

There was insufficient evidence on total postoperative opioid consumption at 24 hours (2 RCTs), and 48 hours (3 RCTs). For opioid consumption at 72 hours (1 RCT), there may be no difference in mean opioid consumption and 7 days (1 RCT). Two RCTs showed conflicting results at 24 hours and a trivial reduction at 48 hours. For opioid consumption at 72 hours, there may be no difference (low strength of evidence) in postoperative opioid consumption. Two RCTs reported on opioid rescue medications and demonstrated conflicting results. One study (high ROB) reported on discharge prescription amounts/refills (one person in the control arm and two persons in the intervention arm requested refills) (insufficient strength of evidence). For hospital LOS, results were inconsistent across 6 RCTs with three studies demonstrating a pooled mean difference of -0.25 days (95% CI -1.61 to 1.12) favoring the intervention, and three studies reporting conflicting results for median LOS (insufficient strength of evidence). Two studies suggested there may be improvement in patient satisfaction with regional anesthesia (low strength of evidence), but it was not clear if this was related to pain/opioid use.

With respect to harm, evidence was judged to be insufficient. Three studies noted no adverse effects, one reported no significant difference between groups, two reported no difference in nausea between arms, and four reported no difference in postoperative nausea and vomiting; while two studies showed a difference favoring the intervention.

Overall, the evidence demonstrated that there may be no clinically meaningful differences in pain scores at 24 and 48 hours (low strength of evidence) and insufficient evidence for pain scores at 72 hours. No information was available for pain at discharge. There was insufficient and inconsistent evidence for postoperative opioid consumption at 24 hours, 48 hours, and at discharge, with no difference in opioid consumption at 72 hours. There was insufficient evidence for hospital length of stay and harms but there may be higher patient satisfaction with regional anesthesia. Therefore, across all outcomes, the evidence was insufficient regarding the effect of regional anesthesia compared to usual care on postoperative opioid consumption and pain outcomes. See Table 3.

Regional anesthesia compared with sham block (13 RCTs, n=2193 patients) Of the 13 RCTs, six did not report on chronic pain or opioid use, and nine RCTs did not report about psychiatric comorbidity. Five RCTs reported excluding patients with a history of chronic opioid use, three RCTs excluded patients with chronic pain, and three RCTs excluded those with any history of psychiatric comorbidity. Two RCTs excluded patients with substance use disorder.

Overall, for patient reported pain scores at 24 hours and 48 hours, effect sizes did not meet the threshold for minimal clinically important differences (low strength of evidence). For 24 hour pain scores, based on data from 12 RCTs (7 low ROB, 4 some concerns, 1 high ROB) there may be no clinically meaningful difference in pain scores across most studies. The pooled mean difference from three RCTs was -0.43 (-1.16 to 0.29). For 48 hour pain scores, three RCTs showed there may be no difference (low strength), while one study reported no difference in pain scores at discharge (low strength).

There was insufficient evidence based on conflicting data on total postoperative opioid consumption at 24 hours (7 RCTs). Five studies provided data for meta-analysis demonstrating a trivial to small reduction -5.0mgs OME (95% CI -16.99 to 7.0). The other two studies showed conflicting results. One additional RCT reported less rescue tramadol in the regional anesthesia arm compared with sham. No studies reported on opioid prescription amounts at discharge but one study reported on frequency of refill request. There was probably no difference in hospital length of stay based on four RCTs (moderate strength). Two RCTs showed there may be improvement in patient satisfaction (low strength).

There may be no differences in intervention associated harms based on 12 RCTs (low strength of evidence). Four demonstrated no adverse events from regional anesthesia. Six RCTs reported no difference in postoperative nausea and vomiting, and one found no difference in pruritus frequency. One RCT found no significant difference in combined adverse events (nausea, vomiting, hypotension, bradycardia, and others). One RCT showed less postoperative nausea and vomiting in the intervention arm. Overall, there was no difference in intervention associated harms (low SOE).

Overall, the evidence demonstrated that there may be no clinically meaningful differences in pain scores at 24 and 48 hours and at discharge (low strength of evidence). No information was available for pain at 72 hours. There was insufficient evidence for postoperative opioid consumption at 24 hours. There were no studies for opioid consumption at 48 hours, 72 hours, and at discharge. There was probably no difference in hospital length of stay but there may be higher patient satisfaction with regional anesthesia. There may be no difference in intervention associated harms. Thus, across all outcomes, the evidence was insufficient regarding the effect of regional anesthesia compared to sham on postoperative opioid consumption and pain outcomes. See Table 4.