Objective
Dexamethasone is a corticosteroid with minimal side effects that may improve quality of recovery. We sought to evaluate standard use of this medication prior to vaginal reconstructive surgery.
Study Design
This was a double-blind, randomized, placebo-controlled trial of women undergoing vaginal reconstructive surgery for pelvic organ prolapse. Patients scheduled for an intraperitoneal vaginal vault suspension, with general anesthesia and an overnight stay, were enrolled. The intervention arm received dexamethasone 60 minutes prior to surgery, and controls received placebo. Postoperative pain medications, antiemetics, and voiding trials were standardized. Our primary outcome was the difference in Quality of Recovery (QoR-40) scores on postoperative day 1. Secondary measures included Postoperative Nausea and Vomiting Intensity scores, and visual analog scales for nausea/vomiting, and pain. Our power calculation demonstrated 31 subjects in each group would be necessary to document difference in QoR-40 scores; to allow for attrition, a goal of 74 subjects was set.
Results
Seventy-four women were enrolled and randomized. Two withdrew, 9 were excluded, and 63 were analyzed (36 placebo, 27 dexamethasone). The mean age was 63 years. No significant differences were noted among demographics other than American Society for Anesthesiologists class; there were greater numbers of dexamethasone subjects that were class 3 (5 vs 11; P = .030). Postoperatively, more patients in the placebo group required promethazine as a rescue antiemetic for control of their nausea/vomiting (11 vs 2; P = .029). Placebo subjects also failed their voiding trials more frequently, which remained following a logistic regression controlling for suburethral sling (30 vs 15; P = .037). Regarding the QoR-40 following surgery, the emotional state domain declined less in dexamethasone patients (–14.3, interquartile range [IQR], 16.8 vs –4.6, IQR, 20.1; P = .042), indicating better symptoms. Whereas pain scales were similar, the visual analog scales for nausea/vomiting was lower in dexamethasone subjects (0.7; IQR, 4.1 vs 0.4; IQR, 1.4; P = .042). Postoperative Nausea and Vomiting Intensity scores were not significantly different; nevertheless, twice as many placebo subjects had severe range symptoms (4 vs 2; P = .47). No adverse effects from the dexamethasone were noted.
Conclusion
Use of dexamethasone prior to vaginal reconstructive surgery was associated with less nausea/vomiting and need for antiemetics as well as greater success with voiding trials. Furthermore, quality of recovery was enhanced, suggesting use of dexamethasone should be considered for these patients.
Reducing postoperative complications and improving quality outcomes are important goals for surgeons and health care systems. Quality of recovery (QOR) is a quantification of various postoperative dimensions such as physical and emotional well-being. Poor QOR has been associated with longer hospital stays, delayed convalescence, and lower patient satisfaction. Two of patients’ most feared postoperative complications that have an impact on QOR are uncontrolled pain and postoperative nausea and vomiting (PONV). Strategies to reduce these consequences can improve outcomes and reduce morbidity.
Dexamethasone is a potent corticosteroid, which has been used in combination with other antiemetics to augment their effect. The mechanism of action is not completely understood, but it appears to act synergistically with serotonin receptor antagonists to reduce nausea while also providing analgesic benefits. In fact, the Enhanced Recovery After Surgery program recommends that patients with a history of PONV receive dexamethasone during anesthesia to reduce this complication.
Because of the robust impact of dexamethasone, it has been suggested that all patients undergoing surgery should be considered candidates for this treatment. Indeed, dexamethasone has been evaluated in conjunction with both regional and general anesthesia prior to cesarean delivery, laparoscopic cholecystectomy, and ambulatory gynecology surgery, with improvements in PONV, opioid consumption, and QOR. The administration of dexamethasone is inexpensive, with minimal side effects and risks. In conjunction with other medications, dexamethasone may be part of a multimodal treatment plan that could enhance symptom control.
Vaginal reconstructive surgery for pelvic organ prolapse is a common and minimally invasive procedure, with most patients being discharged on postoperative day one (POD#1). However, poor QOR may delay advancing care. Moreover, female patients have been shown to be at risk for inferior QOR, rendering it important to seek treatment modalities that benefit them.
Dexamethasone has the potential to be efficacious in this regard, yet such application has not been investigated. Therefore, the purpose of this randomized controlled trial was to evaluate whether standard administration of dexamethasone prior to general anesthesia induction would result in improved patient QOR following vaginal reconstructive surgery as compared with placebo. Secondary outcomes were to document symptoms of PONV and pain control.
Materials and Methods
This was an institutional review board–approved, randomized, double-blind, placebo-controlled trial. All subjects were patients of a tertiary care referral-based urogynecology practice and planning surgery at 1 of the 2 hospitals in the health care system, TriHealth Inc (Cincinnati, OH). Study enrollment was between January 2013 and June 2014. The study was registered with the National Clinical Trials Registry (NCT no. 02073734).
Subjects were eligible if they consented to major vaginal reconstructive surgery for symptomatic pelvic organ prolapse of stage 2 or greater involving general anesthesia and an overnight stay. Surgery was to include an anterior and posterior colporrhaphy with an intraperitoneal vaginal vault suspension to the uterosacral ligaments. Other concomitant procedures such as hysterectomy and antiincontinence surgery could be performed.
Subjects were excluded if their surgical procedure involved the removal of transvaginal mesh, placement of transvaginal mesh, robotic or abdominal surgical treatment, anal sphincteroplasty, or anal fistula repair. Other criteria for exclusion included an inability to perform an intraperitoneal vaginal vault suspension, daily use of steroids or antiemetics in the month before surgery, history of PONV (which would allocate them to receive dexamethasone per our institutional standard of care), planned regional anesthesia, chronic pain requiring daily opioid treatment, history of allergy to the study medication, severe renal or liver disease, uncontrolled diabetes, pregnancy, non-English speaking or inability to complete the questionnaires, and evidence of systemic fungal or other infections. Patients were approached and enrolled at the time of surgical consent or in the preoperative area before surgery.
Our randomization schedule was conducted using a computer-generated table into 2 groups. All subjects were assigned a number and the hospital pharmacy was provided a confidential table to allocate each patient. On the day of surgery, the study number was communicated to the pharmacy, and only the pharmacist was aware of the group allocation. The patient, physicians, anesthesia personnel, nursing, data collection staff, and statistician were all blinded.
The intervention group received 8 mg dexamethasone intravenously 60 minutes prior to surgery, whereas the placebo group received an identical syringe with normal saline solution. The dexamethasone is concentrated at a dose of 4 mg/mL, and thus, all subjects received 2 mL of solution containing either study drug or saline. The drugs were provided before the patient’s induction of anesthesia in a similar design to prior research. The purpose of this timing was to optimize the impact of the dexamethasone on the stress response of the surgical start. The peak effect of the medication is within 45 minutes to 1 hour.
All surgeries were performed by board-certified surgeons in female pelvic medicine and reconstructive surgery using a similar technique. As part of the academic institutional affiliation, fellows also participated in portions of the surgery with direct supervision.
Lidocaine with epinephrine (1%) was injected into the vaginal mucosa for vasoconstriction and hydrodissection. After the peritoneal cavity was entered, packing was placed in the pelvis using moist laparotomy sponges to aid in visualization. A high uterosacral ligament suspension was performed in a previously described manner using 2 delayed-absorbable sutures through each uterosacral ligament, at or above the ischial spines bilaterally, and brought through the vaginal cuff. All patients also underwent an internal McCall’s culdoplasty (permanent suture reefed across the posterior cul-de-sac peritoneum incorporating the uterosacral ligaments bilaterally and tied intraperitoneally).
Baseline data included demographics, medical history, and physical examination findings. Surgical information recorded included anesthesia medications, surgery duration, estimated blood loss, and all other intraoperative medications. Postoperative data were collected regarding postoperative pain medications, antiemetics, complications, nursing verbal pain scores from 0 to 10 (0 indicating no pain and 10 indicating maximum pain), and voiding trial results. Complications were recorded including postoperative fever (defined as temperature elevation greater than 100.4°F), blood transfusion, anemia (as diagnosed by the treating physician, or a hemoglobin level <10 g/dL), neuropathy, and ileus.
All patients received a standardized regimen for pain control: patient-controlled analgesia (PCA), with intravenous (IV) hydromorphone for breakthrough pain as needed. IV ketorolac was available on request every 6 hours (30 mg for patients younger than 65 years and 15 mg for patients older than 65 years of age). All patients were transitioned to oral medications on POD#1 and were provided both oxycodone 5-10 mg/acetaminophen 325-650 mg every 4 hours and ibuprofen 600 mg every 6 hours as needed.
A standardized antiemetic regimen instructed nursing to administer ondansetron 4 mg IV every 8 hours as first line, followed by metoclopramide 10 mg IV every 6-8 hours and promethazine 25 mg IV every 4-6 hours for refractory PONV.
The Quality of Recovery-40 (QoR-40) is a validated scale with 5 domains. These measure physical comfort, emotional state, physical independence, psychological support, and pain. Each domain is scored to a maximum global score of 200. Higher scores indicate better QOR. QoR-40 scores have been found to be associated with both quality-of-life scales and patient satisfaction indices as well as postoperative pain.
Patients completed the QoR-40 index at baseline and in the morning of POD#1. On POD#1, subjects also completed a PONV Intensity Scale, and 2 visual analog scales (VAS) from 0 to 10 cm. One VAS queried about their level of pain, with 0 indicating no pain and 10 indicating severe pain. The other VAS asked about nausea and/or vomiting symptoms in a similar manner. VAS scales are widely used in behavioral science and previous studies report their usefulness and validity.
The PONV scale is designed to measure clinically important nausea and vomiting. Four questions are asked, with a total score of greater than 50 indicating intense or clinically important symptoms. It has also been validated in gynecological patients.
Our sample size calculation showed that a sample size of 31 subjects in each group would have 90% power to detect a mean difference of 22 points in the global QoR 40, SD 26 (n-Query Advisor; Statistical Solutions Ltd, Los Angeles, CA). This difference was noted in a prior study that evaluated preoperative dexamethasone after ambulatory gynecological surgery. To allow for patient attrition, a sample size of 74 subjects was set (6 additional subjects per group, or 20%).
Statistical analysis was performed using SPSS (version 22; SPSS, Inc., Chicago, IL). Data were reported using χ 2 and Fisher exact tests. Nonparametric tests were applied when data did not meet assumptions of normality. To adjust for nonnormally distributed raw scoring data, the QoR-40 domain and global scores were analyzed by first computing the percent change between preoperative and postoperative scores and then comparing the mean or median percentage changes across study groups using either an unpaired Student t test or a Mann-Whitney rank sum test.
Logistic regression was performed to control for suburethral sling procedure on outcome of voiding trials postoperatively. Statistical evaluation was performed by advisers in the Hatton Institute for Research and Education, a research arm for the affiliated institution, and all statisticians were blinded until data analysis was complete.
Results
Seventy-nine women were approached to participate and 5 declined. Seventy-four were enrolled and randomized. Two withdrew, and 9 were excluded (1 did not have an intraperitoneal vaginal vault suspension, 1 canceled her surgery, 4 developed new medical problems and had to cancel surgery, 2 did not receive the study drug because of a pharmacy error, and 1 required hydrocortisone preoperatively). Thus, 63 were analyzed; 36 received placebo and 27 received dexamethasone ( Figure ). The mean age was 63 (SD 9) years and the mean body mass index was 29 (SD 5) kg/m 2 .
No significant differences were noted in the demographics between the 2 groups at baseline other than American Society for Anesthesiologists (ASA) class. Dexamethasone subjects were more likely to be class 3 than class 2 when compared with placebo subjects, indicating higher morbidity in the dexamethasone cohort ( P = .030; odds ratio [OR], 4.40; 95% confidence interval [CI], 1.26–15.35). The type of anesthesia medications provided, amount of lidocaine injected (milliliters) for surgical dissection, surgical procedures, estimated blood loss, and surgical duration were the same. However, there was slightly more subjects who underwent a suburethral sling in the placebo group, which was not statistically significant ( Table 1 ).
| Characteristic | Placebo (n = 36) | Dexamethasone (n = 27) | P value a |
|---|---|---|---|
| Age, mean (SD), y | 62.0 (9.6) | 63.2 (8.3) | .61 b |
| BMI, median (IQR) | 26.8 (5.1) | 28.4 (6.3) | .40 c |
| Comorbidities d | |||
| Motion sickness | 2 (5.6) | 0 (0.0) | .50 |
| Depression | 6 (16.7) | 4 (14.8) | 1.00 |
| Irritable bowel syndrome | 3 (8.3) | 1 (3.7) | .63 |
| Fibromyalgia | 1 (2.8) | 2 (7.4) | .57 |
| Chronic pain syndrome | 1 (2.8) | 0 (0.0) | 1.00 |
| Diabetes | 5 (13.9) | 1 (3.7) | .23 |
| Endometriosis | 1 (2.8) | 0 (0.0) | 1.00 |
| Surgical history | |||
| Hysterectomy | 5 (13.9) | 7 (25.9) | .23 e |
| Hysterectomy with oophorectomy | 3 (8.3) | 4 (14.8) | .45 |
| Prior prolapse repair | 2 (5.6) | 3 (11.1) | .64 |
| Other abdominal surgery | 8 (22.2) | 10 (37.0) | .20 e |
| Gynecological history | |||
| Postmenopausal | 31 (86.1) | 23 (85.2) | 1.00 |
| Using hormone replacement therapy | 5 (13.9) | 4 (15.4) | 1.00 |
| Obstetrical history | |||
| Gravidity, median (IQR) | 3.0 (2.0) | 3.0 (2.0) | .87 c |
| Parity, median (IQR) | 3.0 (1.0) | 2.0 (2.0) | .90 c |
| Vaginal deliveries, median (IQR) | 2.0 (1.0) | 3.0 (2.0) | .42 c |
| Social history | |||
| Smoker | 3 (8.3) | 5 (19.2) | .26 |
| Alcohol use | 12 (33.3) | 11 (42.3) | .47 e |
| Illicit drug use | 0 (0.0) | 1 (3.8) | .42 |
a P values are for the Fisher exact test unless otherwise noted
d Values are expressed as n (percentage) unless otherwise noted
Postoperative use of narcotics via PCA and other means were not different among the 2 groups. However, as a third-line antiemetic, more required promethazine in the placebo arm than the dexamethasone ( P = .029; OR, 5.28; 95% CI, 1.06–26.35). Although not statistically significant, subjects treated with placebo also received greater amounts of ondansetron and metoclopramide. Furthermore, subjects in the placebo group also failed their voiding trial on POD#1 more frequently and required discharge with a catheter ( P = .01) ( Table 2 ). This difference remained following logistic regression controlling for sling procedure, supporting that patients receiving dexamethasone were less likely to fail their voiding trials ( P = .037; OR, 3.79; 95% CI, 1.08–13.26).
| Variable | Placebo n = 36 | Dexamethasone n = 27 | P value a |
|---|---|---|---|
| Type of surgery b | |||
| Vaginal hysterectomy | 27 (75.0) | 16 (59.3) | .18 c |
| Unilateral oophorectomy | 0 (0.0) | 1 (3.7) | .43 |
| Bilateral oophorectomy | 2 (5.6) | 4 (14.8) | .39 |
| Anterior repair | 33 (91.7) | 24 (88.9) | 1.00 |
| Posterior repair | 33 (91.7) | 27 (100.0) | .25 |
| Suburethral sling procedure | 16 (44.4) | 6 (22.2) | .11 |
| Anesthesia data | |||
| Anesthesia class d | |||
| I | 4 (11.4) | 1 (4.0) | .55 |
| II | 26 (74.3) | 13 (52.0) | n/a |
| III | 5 (14.3) | 11 (44.0) | .03 |
| General | 36 (100.0) | 27 (100.0) | — |
| Local lidocaine, mean (SD), mL | 29.4 (9.7) | 24.8 (12.3) | .10 e,f |
| Anesthesia medications used, n (%), dosed | |||
| Inhalational | 35 (97.2) | 27 (100.0) | 1.00 |
| Fentanyl | 36 (100.0) | 27 (100.0) | — |
| Morphine | 2 (5.6) | 2 (7.4) | 1.00 |
| Versed | 30 (83.3) | 22 (81.5) | 1.00 |
| Diprivan | 35 (97.2) | 26 (96.3) | 1.00 |
| Naloxone | 0 (0.0) | 1 (3.7) | .43 |
| Glycopyrrolate | 30 (83.3) | 20 (74.1) | .37 c |
| Intraoperative antiemetics, n (%), dosed | |||
| Ondansetron | 35 (97.2) | 26 (96.3) | 1.00 |
| Metoclopramide | 1 (2.8) | 3 (11.1) | .31 |
| Promethazine | 5 (13.9) | 4 (14.8) | 1.00 |
| Other intraoperative data | |||
| Surgery duration, median (IQR), min | 165.5 (56.0) | 160.0 (50.0) | .32 g |
| Estimated blood loss, median (IQR), mL | 125.0 (50.0) | 150.0 (100.0) | .31 g |
| IVF, mean (SD), mL | 1977.8 (556.1) | 1937.9 (741.3) | .81 d |
| Postoperative pain medication used | |||
| Total hydromorphone in PACU, median (IQR), mg | 0.2 (1.0) | 0.0 (1.0) | .22 g |
| Total hydromorphone upon transfer to patient’s room, median (IQR), mg | 0.7 (1.13) | 0.4 (0.64) | .11 g |
| Ketorolac, n (%), dosed | 26 (74.30) | 21 (77.80) | .44 d |
| Total ketorolac, median (IQR), mg | 30.0 (37.5) | 30.0 (30.0) | .77 g |
| PCA utilization of hydromorphone, mg | |||
| POD 0, median (IQR) | 1.4 (1.5) | 1.4 (1.6) | .88 g |
| POD 1, median (IQR) | 1.0 (1.3) | 1.0 (1.6) | .62 g |
| Total, median (IQR) | 2.4 (2.8) | 2.6 (3.2) | .69 g |
| Postop antiemetics used | |||
| Ondansetron, n (%), dosed | 21 (58.3) | 13 (50.0) | .52 c |
| Ondanestron, median (IQR) doses/patient | 1 (1.0) | 1 (2.0) | .50 g |
| Metoclopramide, n (%), dosed | 10 (27.80) | 5 (19.20) | .44 c |
| Metoclopramide, median (IQR) doses/patient | 1 (2.5) | 2 (2.5) | .82 g |
| Promethazine, n (%), dosed | 11 (30.60) | 2 (7.70) | .03 h,c |
| Promethazine, median (IQR) doses/patient | 1 (0.0) | 1 (0.0) | .89 g |
| Postop complications i | |||
| Anemia, n (%) | 0 (0.0) | 1 (3.7) | .43 |
| Voiding trial results POD 1 | |||
| Failed | 30 (85.7) | 15 (57.7) | .01 h,c |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
