This single-centric, prospective, double-blinded, interventional, randomized controlled trial was conducted in the Department of Neuroanaesthesia and Neuro-critical Care at Paras Hospital, Gurugram, India, from October 2022 to April 2024 after securing Institutional Ethical Committee approval (PHPL/HR-Comm/2022/453). This study was registered in the Clinical Trial Registry of India (CTRI/2022/12/048425).

This study included patients aged 18 years and older, with American Society of Anesthesiologists (ASA) physical status I/II, and undergoing elective open thoracolumbar spinal fusion surgery. This study excluded patients with known allergy to study drugs; presence of severe uncontrolled renal, liver, cerebrovascular, or cardiovascular disease; on chronic beta-blocker treatment with heart rate of <50/min; patients with psychiatric illness; low Glasgow Coma Scale (<13/15) or those unable to understand pain scores; pregnant women; and patients who refused to give consent. Written informed consent was obtained from all individual partcipants included in the study.

Patients were randomly allotted (using computer-generated randomization) into groups R, RD, and RM. Group R received 38 mL of 0.2% ropivacaine and 2 mL of normal saline, group RD received 38 mL of 0.2% ropivacaine and 1 mcg/kg ideal body weight (IBW) (in a 2 mL volume) of dexmedetomidine, and group RM received 38 mL of 0.2% ropivacaine and 3 mg of morphine sulfate (in a 2 mL volume). Allocation and concealment were performed employing the sealed envelope technique. Further, an anesthesiologist who was not involved in the study prepared the study drug.

This study primarily aimed to compare the total opioid consumption in the first 24-hour postoperative period between the groups. Further, we compared the time of the first request for rescue analgesia, the postoperative analgesia, and the side effect profile between the groups.

Sample size for the present study is calculated to be 22 in each groups (a total of 66) based on a 95% confidence interval, 90% power, standard deviation of doses of fentanyl consumption of 30.85 mcg and 39.22 mcg in groups A and B, respectively, from previous study by Deshwal et al. [4], and expected difference in the mean value between the groups of ±50 mcg.

Upon arrival in the operating theatre, ASA standard monitors were attached, a BIS-CONOX sensor (Fresenius Kabi, Bad Homburg, Germany) was placed on the forehead, and baseline parameters were recorded. Fentanyl of 2 mcg/kg and propofol were used for anesthesia induction, and rocuronium was administered to facilitate intubation. Anesthesia was maintained with oxygen and medical air, total intravenous anesthesia with target-controlled infusion (TCI) of propofol (2.5–4 mcg/mL) using an Alaris BD infusion pump (BD, Franklin Lakes, NJ, USA; Schneider Model with Ce) and fentanyl infusion (1 mcg/kg/hr) using a Graseby 2000 (Smiths Medical, Minneapolis, MN, USA) infusion pump. Additional propofol doses were administered at the anesthesiologist’s discretion. A BIS of 30–40 was targeted. End-tidal carbon dioxide of 35–45 mm Hg was maintained.

Fentanyl infusion was stopped at the start of closure, propofol TCI was tapered, and BIS was maintained between 40 and 60. Patients received paracetamol infusion (20 mg/kg) during surgical wound closure and every 8 hours thereafter as part of the multi-model analgesic regimen. The operating surgeon performed multilayer and multilevel infiltration of the study drug according to group allocation. Propofol was tapered and stopped at skin closure. Extubation was performed following standard criteria. All patients were kept in the neurosurgical intensive care unit for overnight observation. Fentanyl infusion (1 mcg/kg/hr) was administered for breakthrough pain when the patients complained of a Visual Analog Scale (VAS) score of ≥3. Ondansetron was administered for postoperative nausea and vomiting (PONV).

Pain score was assessed immediately after extubation and at 30 minutes, 1, 2, 3, 4, 8, 12, 16, 20, and 24 hours postoperatively. The time of the first request for rescue analgesia, the total opioid consumption, and side effects were recorded and compared between the groups.

Data were entered in Microsoft Excel version 2016 (Microsoft Corp., Redmond, WA, USA) and analyzed using the IBM SPSS ver. 21.0 (IBM Corp., Armonk, NY, USA) statistical program for Microsoft Windows.

Categorical data were expressed as frequencies (number of cases) and relative frequencies (percentages), whereas continuous data were presented using the most suitable measures of central tendency (mean with standard deviation for normally distributed data or median with interquartile range for data not distributed normally). The Kolmogorov-Smirnov test was used to identify whether the data were normally distributed.

The chi-square (χ²) test of significance or Fisher’s exact modification was used to compare categorical variables across the groups. The Kruskal-Wallis test (analysis of variance test) with Tukey’s post-hoc test was used to compare continuous variables across the groups. A probability value (p-value) of less than 0.05 indicated statistical significance.

Fig. 1 illustrates the enrolment, exclusion, and randomization of participants. The study enrolled 66 patients, comprising 22 in each group. Demographic parameters (age, sex, weight, and IBW) and clinical parameters (ASA physical status and number of segments treated) were similar across the groups (Table 1).

Compared to group R, the mean VAS was significantly less in group RD at extubation, 30 minutes, 1–3 hours, and 20 hours postoperatively (Table 2). Similarly, the comparison of group R with RM revealed a significant difference in VAS score at extubation, 30 minutes, and 1–24 hours postoperatively. Lastly, the comparison of group RD with RM revealed the significant difference in VAS score at extubation, 1–8 hours, 20 hours, and 24 hours postoperatively. Further, the comparison of the mean total VAS score revealed a statistically significant difference between groups R and RD (27.32±4.89 vs. 20.36±4.35, p<0.001), groups R and RM (27.32±4.89 vs. 12.18±4.88, p<0.001), and groups RD and RM (20.36±4.35 vs. 12.18±4.88, p<0.001).

The intraoperative opioid consumption was similar and comparable across the groups (Table 3). However, during the postoperative period, opioid consumption was significantly higher in group R compared with group RD (403.18±132.18 mcg vs. 307.73±73.84 mcg, p=0.002), in group R compared with group RM (403.18±132.18 mcg vs. 202.5±67.5 mcg, p<0.001), and in group RD compared with group RM (307.73±73.84 mcg vs. 202.5±67.5 mcg, p=0.001). Similarly, the time to first request of rescue analgesia was longer in group RD than in group R (220.45±44.59 minutes vs. 137.68±51.53 minutes, p=0.003), in group RM compared with group R (369.75±139.78 minutes vs. 137.68±51.53 minutes, p<0.001), and in group RM compared with group RD (369.75±139.78 minutes vs. 220.45±44.59 minutes, p<0.001) (Fig. 2). Subgroup analysis revealed a similar trend of opioid consumption, time to first request of rescue analgesia, and VAS score among patients undergoing single-level or two–three-level spine fusion surgery (Figs. 3, 4).

In group R, one patient experienced vomiting, whereas another patient had abdominal distension and constipation (Table 4). The gastrointestinal side effects observed in group R were mainly due to fentanyl. When compared across the groups, the incidence was not statistically significant (p=0.389).

The postoperative fentanyl consumption significantly increased from the group receiving morphine (group RM) to dexmedetomidine (group RD) to plain ropivacaine (group R). Our results align with a meta-analysis where dexmedetomidine with LA led to lower rescue analgesia consumption after lumbar spine fusion [5]. Similar results were noted when dexmedetomidine was combined with ropivacaine for spine fusion [6–8], lumbar discectomy [4,9], laminectomy, laminectomy [10], and discectomy [11]. One study using morphine with bupivacaine for posterior spinal surgery revealed significantly reduced postoperative morphine consumption compared with the control group [12]. Other studies that compared intrathecal morphine with or without LA in lumbar spine surgeries revealed significantly reduced 24–48-hour opioid consumption in the morphine group [13–16]. However, intrathecal morphine spreads cephalad, adding to respiratory depression or central nervous system complications, which are undesirable in neuroanesthesia practice [17,18]. Our results do not align with the findings of Li et al. [19], who revealed no difference in 48-hour hydromorphone consumption among groups receiving ropivacaine plain or with dexmedetomidine or dexamethasone, or both for lumbar spine fusion.

The time to first rescue analgesia significantly reduced from group RM to group RD to group R. Comparable results are observed with ropivacaine and dexmedetomidine in lumbar spine fusion surgery [5,7,8], lumbar discectomy [9], lumbar laminectomy [10], and lumbar discectomy [11]. Further, a similar result was observed when morphine was combined with bupivacaine in posterior spinal surgery [12]. However, our study is not congruent with one study comparing ropivacaine with dexmedetomidine and/or dexamethasone versus ropivacaine for two-segment lumbar fusion [19].

In our study, the patients receiving morphine (group RM) demonstrated significantly lower mean total VAS scores compared with patients receiving dexmedetomidine (group RD) or ropivacaine (group R). Similar results were observed when dexmedetomidine was combined with ropivacaine for infiltration [4,5,7–9,11]. VAS was similarly significantly lower in a study using morphine for wound instillation [17]. However, one study revealed no difference in pain scores using dexmedetomidine or dexamethasone, or both with ropivacaine [19].

Two patients (3.03%) receiving plain ropivacaine (group R) developed vomiting and constipation. We presume both these complications are associated with opioid fentanyl, as our regimen includes total intravenous anesthesia (TIVA) using propofol, which is a known antiemetic, and definite emetogenic agents, such as nitrous oxide, were not part of our study. Further, various meta-analyses report significantly lower PONV incidence among patients receiving dexmedetomidine [5,20–22], whereas some studies report no difference in the incidence [4,23]. Similarly, morphine PCA-related side effects were lower when morphine was used for wound instillation [12].

Each group received a total volume of 40 mL of the LA mixture based on the study by Deshwal et al. [4], Trekajonsak et al. [12], Li et al. [19], and Shrestha et al. [24]. Dexmedetomidine was used as per IBW to avoid overdosing in the obese or heavy-built, leading to bradycardia or postoperative sedation and delaying immediate postoperative neurological examination [25–27]. Based on previous studies demonstrating the analgesic potency of epidural morphine at a 3 mg dose, 3 mg of morphine was used for local infiltration. With increasing doses, the beneficial effect was accompanied by adverse effects [28–30]. Further, a study using an LA cocktail containing morphine (3 mg) for posterior spinal surgery revealed significantly reduced 24-hour morphine consumption compared with the normal saline group [12].

Pain relief after spine fusion depends on multimodal analgesia; thus, the infiltration technique can add to the armamentarium of postoperative analgesia as an alternative or addition to traditional techniques such as spinal or epidural analgesia. The addition of an adjuvant, such as morphine, may provide a cost-effective alternative to potent agents, including liposomal bupivacaine. Moreover, avoiding higher morphine doses as an adjuvant and using TIVA can not only decrease PONV but also reduce postoperative opioid requirement.

The study had a small sample size, which may limit both internal and external validity, thereby restricting the ability to draw definitive conclusions regarding drug safety. It was conducted at a single center. Different classes of drugs were used as adjuvants to LA. We have used dexmedetomidine dosage as per IBW, whereas most studies have used it as per total body weight, leading to the superiority of morphine over dexmedetomidine. Only static VAS and not dynamic VAS for pain assessment.

Pain was assessed until 24 hours postoperatively, whereas some studies have observed patient well-being until 48 hours postoperatively. A fixed amount of ropivacaine was used for all patients, whereas some studies have used LA according to body weight. Similarly, a fixed morphine dose was used for all patients. Fentanyl infusion was used to control postoperative pain, whereas many studies have used fentanyl (or opioid) PCA.