Impact of timing of preoperative gabapentin administration on postoperative somnolence

Methods

After obtaining approval from the Mayo Clinic Institutional Review Board (IRB number: 20-003782), a retrospective cohort chart review was completed. The electronic health records (EHRs) of patients ≥18 years old who received preoperative gabapentin and underwent major gynecologic surgery (benign, oncologic, and urogynecologic procedures) between March 2018 and November 2020 were identified. Major gynecologic surgery was defined as a surgical procedure that required entrance into the peritoneal cavity (laparotomy, laparoscopy, and robotic and vaginal surgeries). Patients were grouped based on the timing from gabapentin administration to surgical incision (less than 4 h vs. four or more hours). Medical, surgical, and anesthesia records were reviewed. Preoperative, intraoperative, and postoperative data were abstracted and compared.

The demographic data abstracted included age, weight, height, menopause status, race/ethnicity, chronic narcotic use, chronic gabapentin use, and chronic pain status. Race/ethnicity was self-designated by patients utilizing the hospital-predetermined classifications of: American Indian or Alaska Native; Asian; Black or African American; Hispanic or Latino; Native Hawaiian or Other Pacific Islander; White; or Other. Race/ethnicity were assessed solely to describe the patient demographics in this study. The surgical data abstracted included surgical indication, surgical approach, procedure performed, time of first incision, time of closure, surgical duration, and estimated blood loss. The time of the gabapentin administration was recorded. Per institutional ERAS protocol, patients aged 18–59 were given 600 mg of gabapentin, and patients aged 60 years and older received 300 mg of gabapentin. Preoperative, intraoperative, and postoperative narcotic consumption were abstracted. Narcotic administration was converted to IV morphine milligram equivalents (MME) [19]. The PACU data abstracted included variables that were utilized collectively as surrogate markers of somnolence. These included vital signs, RASS scores, oxygen administration, Glasgow Coma Scale (GCS) scores, unplanned hospital admission, dizziness, nausea/vomiting, urinary retention, and duration of PACU stay. Postoperative nausea and vomiting were recorded if they were subjectively reported in the patient’s medical record.

A sample size of 254 patients with 127 patients in each group was estimated to provide power of 80% or greater, assuming a mean somnolence recovery time of 90 min in the more than or equal to 4 h gabapentin group, 150 min in the less than 4 h gabapentin group, a standard deviation of 150 min, and a significance level of 5%. This sample size allowed for a missing data rate of 10%. Patients were identified starting in March 2018 and continued until 127 patients were reached in each group. Patients were excluded if they did not have both a gynecologic surgery and preoperative gabapentin administration.

The data were analyzed to determine if there was a difference in patient-oriented outcomes among individuals taking gabapentin within 4 h of surgery or more than 4 h of surgery. This cutoff was based on the typical time intervals with patients self-administering gabapentin at home vs. receiving gabapentin in the preoperative unit. Associations were tested utilizing two-sample equal-variance t-tests or linear model ANOVA for continuous variables, or Fisher’s exact tests for categorical variables. Associations between the time from taking gabapentin to the time of first incision and the time until RASS score of 0 were modeled utilizing linear regression, adjusting for age, initial PACU RASS score, and postoperative narcotics (MME). Statistical analysis was performed utilizing R version 4.0.3.

Results

A total of 254 patients were included. Gabapentin was administered less than 4 h prior to skin incision for 127 patients (50.0%) (<4 h group) and four or more hours in 127 patients (50.0%) (≥4 h group). Demographics were similar between the groups with the exception of age (<4 h group mean=44.2; ≥4 h group=mean 40.5; p=0.021), chronic pain (<4 h group=17.6%, n=22; ≥4 h group=43.3%, n=55; p<0.001), and surgical indication (<4 h group=pelvic pain [29.1%, n=37]; ≥4 h group=pelvic pain (51.2%, n=65); p=0.007). Surgical approach was minimally invasive in 78.0% (n=99) of patients in <4 h group and 88.2% (n=112) of patients in ≥4 h group. Hysterectomy was the most frequent procedure performed (<4 h group=50.4%, n=64; ≥4 h group=49.6%, n=63), followed by excision of endometriosis (<4 h group=27.6%, n=35; ≥4 h group=55.1%, n=70). Details of demographics and surgical characteristics are summarized in Table 1.

Narcotic administration preoperatively (p=0.111) and intraoperatively (p=0.237), and postoperatively (p=0.185) were similar between the two groups. Oxycodone and hydromorphone were administered less frequently in the <4 h group (31.5%, n=40 and 29.9%, n=38, respectively) than in the ≥4 h group (48.8%, n=62 and 44.9%, n=57) (p=0.007 and 0.019 respectively). Preoperative, intraoperative, and postoperative medications are outlined in Table 2.

The minutes from surgical closure until the time the patient received a RASS score of 0 (<4 h group mean [SD]=163.5 [117.9] min, ≥4 h group mean [SD]=143.9 [100.4] min, p=0.168) and initial PACU pain score (VAS) were similar for both groups (p=0.245). The initial PACU oxygen administration volume (p=0.678), hours from surgical closure until the patient transitioned to room air (p=0.679), and initial PACU respiratory rate (p=0.913) were similar between the two groups. The mean PACU stay duration did not show a statistical difference with the less than 4 h group duration of stay of 250.1 min compared to 240.1 min in the more than 4 h group (p=0.449). Unplanned admission secondary to excessive somnolence was uncommon in both groups, with one instance (0.8%) in the less than 4 h group and two instances (1.6%) in the more than 4 h group (p=1.00). The initial PACU GCS score (p=0.473) showed no statistically significant difference between the two groups. Postoperative dizziness had no significant difference between the two groups; however, postoperative nausea/vomiting (<4 h group=24.4%, n=31; ≥4 h group=13.4%, n=17; p=0.036) and urinary retention (<4 h group=14.2%, n=18; ≥4 h group=5.5%, n=7; p=0.033) were both decreased in the 4 h or more group. Table 3 compares the postoperative clinical outcomes by time of taking gabapentin to time of first incision.

After controlling for age, initial PACU RASS score, and postoperative narcotics use, patients took on average 0.105 min less (95% CI: 0.232 min less to 0.021 min more) to achieve a RASS score of 0 for every increase of 1 min between taking gabapentin to the time of first incision (p=0.10). However, this finding was not statistically significant. Controlling for time from taking gabapentin to time of first incision, initial PACU RASS score, and postoperative narcotics use, patients took on average 2.983 min more (95% CI: 2.675 min less to 8.640 min more) to achieve a RASS score of 0 for every five-year increase in age, although this difference was not statistically significant (p=0.30). Results for the linear regression for the outcome of time until RASS score of 0 are outlined in Table 4.

Minutes from closure until time achieved a RASS score of 0 stratified by indication of surgery showed the indication of abnormal bleeding took the greatest amount of time (median, 155). This was followed by pelvic pain (median, 134), cancer (median, 129), other (median, 128), and finally, prolapse (median, 95.50) (p=0.85). Table 5 reports the summary of time until RASS score of 0 by surgical indication.

Discussion

This retrospective chart review demonstrated that the timing of preoperative gabapentin administration as part of the ERAS protocol prior to gynecologic surgeries does not have a significant effect on postoperative somnolence, GCS scores, or pain VAS scores. Further, this study found that gabapentin does not have a significant effect on respiratory depression when given as part of the ERAS protocol for gynecologic surgeries. This finding is important in light of a recent study stating that some ERAS protocols are discontinuing the use of gabapentin due to the risk of respiratory depression [17].

This study shows that the timing of gabapentin does not affect postoperative somnolence. Previous studies have found an increased risk of sedation with preoperative gabapentin [9, 15]. It has been hypothesized that receiving gabapentin four or more hours prior to surgery would decrease the risk of postoperative sedation. However, this study shows no significant differences in RASS scores and duration of PACU stay between the two groups. Further, both groups achieved a RASS score of 0 within 2.4–2.7 h, on average, after surgical closure. Both groups were discharged from the PACU within 4–4.17 h, on average, of surgical closure. The protocol at the study institution is to monitor patients following a hysterectomy for a minimum of 4 h in the PACU prior to discharge, indicating that, on average, patients were discharged as early as the protocol allowed. Between both groups, only three patients had an unplanned admission secondary to excessive somnolence. These findings suggest that the risk for postoperative somnolence may not be as high in patients receiving preoperative gabapentin in gynecologic surgeries as previous studies have shown [9, 15].

One study of 8,567 patients who underwent laparoscopic surgeries lasting 90 min or longer found that 15.3% of the patients had episodes of respiratory depression when receiving gabapentin preoperatively [16]. In this study, respiratory depression was evaluated through respiratory rate, liters of oxygen needed, and hours from surgical closure until the patient was transitioned to room air. This study did not find evidence of complicated respiratory depression in either group, indicating that the timing of gabapentin does not cause differences in respiration postoperatively and further illustrates that the risk of postoperative respiratory depression may be smaller in the gynecologic surgery population than other studies have previously shown [16, 20].

This study found that the adverse outcomes of nausea and vomiting may be decreased when gabapentin is given more than 4 h prior to surgery. Other studies have found the incidence of postoperative nausea and vomiting to be decreased in patients receiving gabapentin [9, 21, 22]. A meta-analysis found a 40% relative risk reduction in postoperative nausea and vomiting in patients who received preoperative gabapentin [22]. This may be attributed to the innate anti-emetic effects [20] of gabapentin and the longer time for gabapentin to be distributed throughout the body. Gabapentin decreases nausea and vomiting symptoms by regulating efferent and afferent neural pathways involved in their production. It is increasingly being utilized in ERAS protocols and for chronic nausea [23]. Furthermore, this study found that incidents of urinary retention may be decreased when gabapentin is given more than 4 h preoperatively. A meta-analysis of 827 patients found that urinary retention was decreased in patients who received gabapentin preoperatively compared to those who did not, with a 0.57 relative risk [24]. With this understanding and based on the results of this study, the optimal timing of gabapentin administration for nausea, vomiting, and urinary retention is more than 4 h prior to the start of surgery.

Conclusions

Administration of gabapentin less than 4 h or greater than 4 h preoperatively does not significantly affect postoperative somnolence or respiratory depression. Those who took gabapentin four or more hours prior to surgery had significantly less adverse outcomes of nausea/vomiting and urinary retention as compared to those who took it within 4 h of surgery.