Strategy to calculate magnesium sulfate dose in obese patients: A randomized blind trial

BACKGROUND Magnesium sulfate has analgesic properties in the postoperative period. Among obese patients, there is a gap in the knowledge of its pharmacology related to the use of real, ideal, or corrected ideal body weight in calculating its dose. This trial compared postoperative analgesia using actual and corrected ideal body weight. METHODS Seventy-ve obese patients scheduled to undergo laparoscopic cholecystectomy under general anesthesia were randomly assigned to three groups. Patients in the control group received no magnesium sulfate; patients in the other two groups received magnesium sulfate 40 mg·kg − 1 of actual body weight or corrected ideal body weight. A ten nonobese patients group helped us as a model of the expected blood magnesium concentration after magnesium sulfate administration in general population. (p and pain scores (p = 0.006) in the postoperative period compared to the control group. There was no signicant difference in the consumption of morphine (p = 0.323) or pain scores (p = 0.082) between these groups. There was no difference in the total duration of neuromuscular block induced by cisatracurium among the three groups (p = 0.181) or in the blood magnesium concentrations throughout the study. groups magnesium sulfate calculated weight.

Worldwide growth in obesity prevalence 19 is associated with an increase in the frequency of obese patients in surgery rooms. This population has also taken the advantage of using MS in many situations 20 . It is necessary to adjust the dosage of some drugs in obese populations due to pharmacokinetic changes caused by increased fat tissue [21][22][23][24] . However, we found no studies analyzing the best way to calculate MS dose in obese patients, using actual, ideal, or corrected ideal body weight.
This trial tested the hypothesis that administering MS in obese patients calculating its dose using real body weight, compared to corrected ideal body weight, results in lower morphine consumption and postoperative pain scores.
We also investigated the level of interference of MS over cisatracurium action and blood magnesium concentration. The outcomes were onset and total duration of neuromuscular blockade 25 and blood magnesium concentrations over time.

Materials And Methods
This study was approved by the Institutional Review Board of the Universidade de Taubaté, SP, Brazil (IRB number 09006119.2.0000.5501) and written informed consent was obtained from all subjects participating in the trial. The trial was registered prior to patient enrollment at clinicaltrials.gov (NCT04003688, Principal investigator: Sebastião Ernesto da Silva Filho; Date of registration: June 24, 2019). This is a prospective, controlled trial, with random distribution covering the participants, medical staff, and evaluators, carried out at the hospital of the Sociedade de Bene cência Portuguesa de Santos, SP, with data collected from August 26, 2019, to November 12, 2020. This manuscript adheres to the applicable CONSORT guidelines.

Study Population
The inclusion criteria were patients aged 18 to 60 years, with American Society of Anesthesiologists physical status I or II and body mass index (BMI) > 30 kg.m − 2 and scheduled for video laparoscopic cholecystectomy. The exclusion criteria included history of allergy to any component of the study protocol, refusal to participate or sign the informed consent form, neuromuscular disorders, cardiac conduction block other than rst-degree atrioventricular block, use of illicit drugs, psychiatric disorders that compromise the assessment of symptoms, use of calcium channel blockers, and renal failure.
The sample size calculated via the ANOVA effect for 95% con dence level and 80% statistical power resulted in 14 individuals for each of the three groups. We increased it to 25 participants per group to compensate for losses, with a total of 75 participants. Individuals from the population of interest were invited to participate voluntarily, and after signing the informed consent, the 75 selected individuals were divided electronically, using the resource of the website www.random.org, into three groups. In the control group (CG), 25 patients received only general anesthesia. In the real body weight group (RWG), patients received general anesthesia and MS at a The electronic drawing allowed 75 envelopes with information about the related groups and procedures that were performed by a professional blinded to the study protocol. Another team member prepared concealed solutions.
The study also included a group of 10 patients with BMI 20-30 kg.m 2 to receive magnesium sulfate 40 mg·kg − 1 and undergo the same protocol as the participants of the real body weight group. They helped us to create a model or base of the expected blood magnesium concentration after magnesium sulfate administration in general population to be compared to obese patients.

Anesthetic Technique
Participants were monitored with continuous electrocardiography, pulse oximetry, noninvasive blood pressure on a multiparameter monitor (Mindray, model IPM-9800, China), and hypnosis level (patient state index, SedLine® Sedation Monitor, USA), before receiving any medication. Patients were also attached to a neuromuscular function monitor (TOF-Watch SX, Ireland). Immediately after venipuncture, we collected the rst blood sample (2 mL) to measure the blood magnesium concentration. Next, the infusion of the concealed solution was performed and completed in 10 min. The concealed solution, depending on the group, comprised of a 100 mL saline solution or saline solution with MS for a total of 100 mL.
Patients were also administered dipyrone 15 mg·kg − 1 , clonidine 2 µg.kg − 1 , cefazolin 2 g, dexamethasone 4 mg, ketoprofen 100 mg, and lidocaine 1.5 mg·kg − 1 , following the institutional protocol to prevent infection, and reduce pain, nausea, and vomiting presentation. At the end of the concealed solution, they received preoxygenation/denitrogenating with fraction of inspired oxygen = 1 for 3 min and propofol in target-controlled infusion to reach a concentration of 4 µg.mL − 1 guided by a hypnosis monitor. After appropriate hypnosis (Patient State Index [PSI] < 50), calibration of the neuromuscular function monitor was performed using train of four (TOF) monitoring, followed by 0.15 mg·kg − 1 cisatracurium intravenously and remifentanil infusion through target-controlled infusion until 5 ng.mL − 1 effect target concentration was reached. The anesthesia was maintained with propofol in target-controlled infusion to maintain PSI of 25 to 50, remifentanil in targetcontrolled infusion (target concentration of 3 a 5 ng.mL − 1 ), and cisatracurium 0.03 mg·kg − 1 if TOF > 2. The administration of cisatracurium was avoided in the last 20 min of surgery. At the end of surgery, patients with TOF > 2 were administered 20 µg.kg − 1 atropine and 40 µg.kg − 1 neostigmine. Blood samples were collected to measure the magnesium blood concentration in the arm contralateral to the arm receiving the medication, through another venous catheter put in place after anesthetic induction and kept in place throughout surgery.
Before extubation, all patients were administered morphine 0.05 mg·kg − 1 and dipyrone 15 mg·kg − 1 intravenously. Five minutes after anesthetic recovery and at every 30 min, patients received another morphine dose if the pain score was higher than 3 on the verbal numeric scale (VNS: 0 [no pain] to 10 [the highest possible pain]). In the ward, they received dipyrone 1 g intravenously (every 6 h), nalbuphine hydrochloride 10 mg every 8 h, and morphine 0.05 mg·kg − 1 if the pain score was higher than 3 in the VNS. All patients were discharged from the hospital on the rst morning after surgery.

Outcomes Measured
The analgesic effectiveness of different doses of MS was assessed through the following outcomes: VNS at 5, 15, 30, 60, 120 and 240 minutes after awakening, and after that at 1 hour interval if the participant was awake, until the next morning, when they were all discharged. It was also recorded the highest pain score in the perioperative period during hospital stay (VNS), and morphine consumption during hospital stay.

Page 5/17
The effects of MS on cisatracurium pharmacology were evaluated by means of onset time (time from cisatracurium administration to TOF = 0) and total duration time (time from cisatracurium administration to TOF reaching T4/T1 = 0.9). The differences in blood magnesium concentrations between the groups were compared by the concentrations veri ed at the collection times de ned in the study protocol. The 10-patient group (BMI 20 to 30 kg.m 2 ), called here non-obese group, was used as a reference to show how close or far the average magnesium concentration in obese patients receiving different doses of MS was from the average concentration in non-obese patients at those times.

Statistical Analysis
The hypotheses of interest were tested using a parametric test ANOVA or repeated measures ANOVA when the observations were taken over time. Samples without normal distribution were compared using the Kruskal-Wallis test. All results with a descriptive level of less than 5% (p < 0.05) were considered signi cant.
Results CG, control group; RWG, real weight group; CWG, corrected ideal weight group.
One patient was excluded from the study (CONSORT owchart, Fig. 1). Table 1 shows similar ASA physical status, duration of anesthesia, weight, height, and BMI. The pain scores at awakening were similar between all participants. Of note, the real and the corrected ideal body weight groups had one patient referring pain score 3, and four patients in control group referred pain level 2. Regarding the highest postoperative pain scores during their hospital stay, the real body weight group and the corrected ideal body weight group did not differ from each other, and both had lower pain scores than the control group. Pain scores and morphine consumption did not show a normal distribution. Comparison among groups showed a difference (p = 0.006, Kruskal-Wallis test).
The post hoc test for multiple comparisons revealed a statistically signi cant lower pain score between the real body weight group and the control group (p = 0.005, Bonferroni) and between the corrected ideal body weight group and the control group (p = 0.016, Bonferroni), but there was no statistical difference between the real weight group and the corrected ideal body weight group (p = 0.082, Bonferroni) ( Table 2). The real body weight group and the corrected ideal body weight group showed no signi cant difference in postoperative morphine consumption, and both groups showed less morphine consumption than the control group.
The multiple post hoc comparisons adjusted by the Bonferroni correction showed similarity in the consumption of morphine in the real body weight group and the corrected ideal body weight group (p = 0.108; corrected p = 0.323). The real body weight group (p ≤ 0.001; corrected p ≤ 0.001) and the corrected ideal body weight group (p = 0.013; corrected p = 0.040) had signi cantly lower morphine consumption than the control group ( Table 2).
The real body weight group and the corrected ideal body weight group showed higher pain scores when compared to control group 30 minutes after awakening, but they were similar to each other (Table 3 -One-way ANOVA). At 60 minutes after awakening the corrected ideal body weight group was similar to control and real body weight group, but the real body weight group showed lower pain scores, when compared to control group (Table 3). Pain scores were similar between groups 120 and 240 minutes after awakening. The latency and total duration of action of cisatracurium were analyzed independently using the one-way ANOVA test, with no statistical differences among the groups in any of the two variables studied (Table 4). Greenhouse-Geisser method was used. The analysis showed similar blood concentrations of magnesium in all measurements performed in the control group, as expected. The real body weight group and the corrected ideal body weight group showed similar baseline blood magnesium concentrations to the control group, but an increase at 15 min, with a progressive decay in the subsequent moments. There was no signi cant difference in blood magnesium concentrations between the real body weight group and corrected ideal body weight group at times of collection times provided by the study protocol ( Fig. 2 and Table 5  We compared the mean blood concentrations of magnesium in the obese groups with the mean concentrations generated in the nonobese group and the mean blood concentrations of magnesium evolved with values similar in all groups that received MS (Fig. 2 and Table 5).
In the corrected ideal body weight group, we compared the participants' average real body weight (92.54 kg) and the average corrected ideal body weight (73.54 kg). There was an average difference of 18.72 kg. Thus, these participants received a dose of MS 21.6% lower than the one they would have received if we had calculated this dose based on their actual weight.

Discussion
The effect of MS dose calculated using the actual body weight was similar to that calculated using the corrected ideal body weight in obese patients in terms of the blood concentration of magnesium, postoperative analgesia and recovery of neurmomuscular block.
The increase in adipose tissue and muscle mass results in changes in the pharmacokinetics of many medications [21][22][23][24] . Moreover, diseases associated with obesity reduct the physiological reserves this population 27 .
Despite the bene ts of MS in various areas of medicine 1-8 it has side effects 11,27,28 , such as delayed recovery of neuromuscular function and orotracheal extubation 29 .
The low-level pain at awakening has bias related to multimodal analgesic strategy used.
In this trial, patients in groups receiving MS showed lower mean postoperative pain scores and morphine consumption. This result, already reported in other studies [9][10][11][12] , is attributed to the action of magnesium in the calcium channels and N-methyl-D-aspartate (NMDA) receptors 4,5 .
There was a difference of 21.6% between the actual body weight and the corrected ideal body weight in the corrected ideal body weight group participants in the present study. Although receiving proportionally less MS these participants showed similar analgesic outcomes, compared to the real body weight group. That lower dose might have decreased the risk of adverse effects.
The onset of the cisatracurium effect in this trial was not altered by MS. Germano et al. 30 did not nd difference in latency of rocuronium 0.6 mg·kg − 1 after MS. Czarnetzki et al. 15 found a signi cant reduction in the latency (average of 77 versus 120 s) rocuronium 0.6 mg·kg − 1 after MS at higher doses than in Germano's and in the present study (60 mg·kg − 1 ). The difference in the time gap between the administration of MS and the administration of rocuronium might have interfered in the results. In the aforementioned studies rocuronium was the neuromuscular blocker agent, whereas cisatracurium was the neuromuscular blocker agent in our study. Thus, the absolute times in this study cannot be compared to those found by them.
In the current study, the groups showed no differences in the total duration of the neuromuscular block. This result is different from that reported by Czarnetzki 16 , who found that the average total recovery time was 73.2 min (SD = 22 min) with previous administration of MS 57.8 min (SD = 14.2 min) in the control group. However, a MS dose of 60 mg·kg − 1 was used before administration of rocuronium 0.6 mg·kg − 1 . The absolute times in this study could not be compared to those found in ours.
The 21.6% difference in the MS dose administered in the groups did not result signi cant difference in the resulting magnesium blood concentrations. These concentrations were always within the safe values for patients in the study 31 .
This study included a nonobese group of 10 patients, who received MS at a dose of 40 mg·kg − 1 , as a reference for the average magnesium concentration in the non-obese population. The comparison between the magnesium concentrations in obese patients and the nonobese group did not show any signi cant difference.
The concentrations were also similar to those reported by Taheri et al. 17 . The average body content of magnesium is 24 g in an individual weighing 70 kg 35,36 . Only about 0.3% of this content is distributed in the plasma 35,36 . This is a possible cause for the rapid balance in the concentration and similar analgesia between the groups that received MS. The patients had an average increase in body magnesium of 1.2% in the corrected ideal body weight group and 1.55% in the real body weight group. Pascoal et al. 37 compared two groups of 31 patients undergoing treatment with MS to prevent pre-eclampsia. After an initial dose of 6 g of MS patients received a continuous infusion of 1 g.h − 1 or 2 g.h − 1 . The initial concentration was statistically equal between the groups (3.7 mEq.L − 1 ; p = 0.96). Thereafter, concentrations increased in the group that received an infusion of 2 g.h − 1 and decreased in the group that received 1 g.h − 1 . The authors concluded that infusion of 1g.h − 1 can be as effective as infusion of 2 g.h − 1 , with a small reduction in side effects. This knowledge may be transferred to the use of MS for analgesic purposes.
Finally, there is a huge variety of volume of distribution among patients as shown in the literature. This fact might have been an important bias affecting every outcome assessed in this study 8 . But, because of the safety concerns, the present study included relatively low BMI range of the participants. Studies with participants with a higher BMI range are needed to assess behavior with even more different doses between groups.
Sugimoto et al. 38 recorded a reduction in the production of in ammatory cytokines (tumor necrosis factor and interleukin 6) in pregnant women submitted to MS, a mechanism that needs to be investigated in the context of the use of the substance for analgesic purposes.
Future research may clarify advantages and disadvantages of MS infusion association, the main mechanism of magnesium elimination, the role of reduction of in ammatory cytokines induced by MS in analgesia and the therapeutic window of this medication.
In conclusion, MS decreased postoperative pain and morphine consumption without affecting the recovery time of cisatracurium in obese patients undergoing laparoscopic cholecystectomy. Compared to dose based on actual weight in obese patient, the dose of MS based on corrected ideal weight induces similar analgesia. On the other hand, the resultant magnesium blood concentration is not different with both strategies.  Blood magnesium concentration throughout the study