Intravenous dexmedetomidine versus tramadol for treatment of shivering after spinal anesthesia: a meta-analysis of randomized controlled trials

Background Shivering is a frequent complication after spinal anesthesia. Increasing studies have compared the effect of intravenous dexmedetomidine and intravenous tramadol on shivering after spinal anesthesia, hence we performed a meta-analysis of randomized controlled trials to compare dexmedetomidine with tramadol on the treatment of post-spinal anesthesia shivering. Methods PubMed, Embase, Cochrane library, Web of Science and Google Scholar were searched to find the eligible studies comparing the effect of dexmedetomidine and tramadol on the treatment of shivering after spinal anesthesia. Mean difference (MD) or risk ratio (RR) along with 95% confidence interval (CI) was used to analyze the outcomes. I2 test was conducted to assess the heterogeneity of the included trials. We utilized Review Manager 5.3 to perform statistical analyses. Results Thirteen randomized controlled trials including 864 subjects were included. Dexmedetomidine had higher effective rate of shivering control (RR =1.03; 95%CI [1.01, 1.06], P = 0.01, I2 = 14%), shorter time to cease shivering (MD = -2.14; 95%CI [− 2.79, − 1.49], P < 0.00001, I2 = 98%), lower recurrent rate of shivering (RR = 0.45; 95%CI [0.27, 0.73], P = 0.001, I2 = 0%), lower incidences of nausea (RR = 0.10; 95%CI [0.05, 0.19], P < 0.00001, I2 = 48%), and vomiting (RR = 0.13; 95%CI [0.06, 0.30], P < 0.00001, I2 = 0%), higher incidence of sedation (RR = 2.48; 95%CI [1.32, 4.65], P = 0.005, I2 = 82%), hypotension (RR = 2.50; 95%CI [1.24, 5.03], P = 0.01, I2 = 0%) and bradycardia (RR = 4.78; 95%CI [1.76, 13.00], P = 0.002, I2 = 0%), compared with tramadol. Conclusions Dexmedetomidine is superior to tramadol for shivering treatment, due to higher effective rate of shivering control, earlier onset of action and lesser recurrence of shivering with higher incidence of sedation and lower incidences of nausea and vomiting. However, dexmedetomidine is also associated with higher incidences of hypotension and bradycardia than tramadol.


Background
Shivering is a common perioperative complication because of postanesthesia hypothermia [1]. Spinal anesthesia has impairment of shivering in the block area and greater heat loss than general anesthesia because of abnormal heat loss owing to vasodilatation [2,3]. Shivering can cause severe consequences, such as arterial hypoxia and myocardial ischemia by increasing oxygen consumption [4,5]. Tramadol is commonly used for the treatment of shivering in clinical practice. However, tramadol can lead to nausea and vomiting which is very distressing for the patient. Therefore, it is necessary to find a better drug with fewer side effects. Dexmedetomidine, an alpha 2adrenergic agonist, has been confirmed the effect on treatment and prevention of shivering in various surgeries by reducing the shivering threshold [6].
There are no large-sample clinical trials evaluating the advantages or disadvantages between dexmedetomidine and tramadol on post-spinal anesthesia shivering. Therefore, we conduct a meta-analysis of randomized controlled trials (RCTs) to compare the effect of intravenous dexmedetomidine and tramadol on postspinal anesthesia shivering.

Literature review
Relevant articles were found by searching PubMed, Cochrane library, Web of Science and Google Scholar by two investigators independently. The terms used for searching included: "Dex", "Dexmedetomidine", "Tramadol", "Anesthesia, Spinal", "Injections, Spinal" and "Shivering" through March 2020, without limits. Furthermore,

Inclusion criteria of studies
Inclusion criteria were as follows: 1) the patients underwent an operation under spinal anesthesia or combined spinal and epidural anesthesia; 2) the comparison was between intravenous dexmedetomidine and tramadol about the treatment effect of shivering; 3) the incidence of side effects was reported in both dexmedetomidine and tramadol groups; 4) the study was a RCT. Meeting papers, correspondences and editorials were excluded.

Data extraction
Data were collected independently by two researchers, including patient characteristics, types of surgery, anesthetic type, the drugs for spinal anesthesia, doses of the study drugs, shivering degree, efficacy of shivering treatment, incidence of recurrent shivering and adverse effects. Shivering was graded using a four point scale as per Wrench in all included papers [7]. Any disaccord was further settled by the third researcher.

Evaluation of risk of bias and the study quality
Two researchers independently evaluated the risk of bias and the qualities of all included studies according to Cochrane Handbook v5.0.2 and 5 point Jadad scale [8].
Each of the following items of risk of bias was graded as "high risk of bias", "uncertain risk of bias" or "low risk of bias": random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other bias. Disputes were settled by discussion, if necessary, a third investigator helped to make a decision.

Hypotension and bradycardia
The incidences of hypotension and bradycardia were recorded in all of the included RCTs, but one [13]. There were 402 patients receiving dexmedetomidine (24 experienced hypotension and 19 had bradycardia) and 402 patients receiving tramadol (9 experienced hypotension and 2 had bradycardia).
Sensitivity analysis was performed for the incidence of sedation by excluding single study sequentially, but no source of high heterogeneity was detected. There were no patients with over sedation reported in the included  Figure 11 showed that no publication bias was detected for recurrent rate of shivering.

Discussion
In this meta-analysis, we compare the efficacy of intravenous dexmedetomidine and tramadol on the treatment of shivering after spinal anesthesia in adult patients. Dexmedetomidine is associated with higher effective rate of shivering control, shorter time to cease shivering, lesser recurrence of shivering, lower incidences of nausea and vomiting, higher incidences of hypotension, bradycardia and sedation than tramadol.
In this meta-analysis, dexmedetomidine has shorter time to cease shivering and a higher incidence of sedation than tramadol. But these outcomes have high heterogeneities, which are possibly associated with the following: 1) Inclusion criteria for shivering degree are different among the included studies, shivering degrees of 2 to 4 are included in 4 RCTs [14,16,18,19] and shivering degrees of 3 or 4 in the other 9 RCTs [9-13, 15, 17, 20, 21]. 2) The types and doses of local anesthetics for spinal anesthesia are different among the studies; 3) The types and duration of the surgeries are different.
In this study, tramadol is associated with significantly higher incidences of nausea and vomiting than dexmedetomidine. Nausea and vomiting is very distressing for the patient. Moreover, vomiting may cause rare but serious consequences, such as aspiration, esophageal rupture, subcutaneous emphysema or pneumothorax [22]. However, dexmedetomidine has significantly higher incidences of hypotension and bradycardia compared to tramadol. Dexmedetomidine has an inherent property of postsynaptic activation of alpha 2adrenoceptors in the central nervous system to decrease heart rate and blood pressure. We can't assess the clinically significance of hypotension and bradycardia, because of lack of research data.
Tramadol is a well-established agent in treatment of shivering. The mechanism of anti-shivering action of tramadol may be its opioid or serotonergic and noradrenergic activity or both [23][24][25]. Dexmedetomidine, an alpha-2 adrenoceptor agonist, has antihypertensive, sedative, analgesic and anti-shivering properties [26]. The anti-shivering effects of alpha adrenoceptor agonists are mediated by binding to alpha receptors that mediate the vasoconstriction. In addition, it has hypothalamic thermoregulatory effects of reducing the vasoconstriction and shivering thresholds [27]. It suggests that dexmedetomidine acts on the central thermoregulatory system rather than preventing shivering peripherally [28]. The effect of dexmedetomidine on treatment of shivering has been confirmed in the previous studies [29][30][31]. Dexmedetomidine and tramadol are not only effective for shivering treatment, but also effective for shivering prevention [32][33][34].
The sedation achieved is better in patients receiving dexmedetomidine than patients receiving tramadol. None of patients experiencing over sedation or respiratory depression is reported in the included studies [9][10][11][12][13][14][15][16][17][18][19][20][21]. Since the surgery is done under spinal anesthesia, the sedation seen with dexmedetomidine is beneficial for the surgeon, anesthetist as well as the patient, because it provides comfort and amnesia to the patient, cardiorespiratory stability and good surgical conditions during  surgery. Therefore, dexmedetomidine may be a good choice for shivering control after spinal anesthesia because of its dual effects of anti-shivering and sedation.
There are two limitations in this study. First of all, there is a high heterogeneity regarding time to cease shivering and the incidence of sedation. Secondly, intravenous dexmedetomidine causes hypotension and bradycardia, but we haven't analyzed whether it is clinically significant, due to lack of research data. Therefore, more RCTs are required for further study.

Conclusions
Dexmedetomidine is superior to tramadol for shivering treatment, due to higher effective rate of shivering control, earlier onset of action and lesser recurrence of shivering with higher incidence of sedation and lower incidences of nausea and vomiting. However, dexmedetomidine is also associated with higher incidences of hypotension and bradycardia than tramadol.