Assessment of Antishivering Medication Requirements During Therapeutic Normothermia: Effect of Cooling Methods

Abstract

Shivering during targeted temperature management (TTM) should be minimized because it can cause cerebral and metabolic stress. It has been proposed that surface cooling (SC) may result in more shivering than endovascular cooling (EC) methods. The purpose of this study was to compare antishivering medication requirements and degree of shivering in these groups during TTM to Normothermia (NT). This was a retrospective single-center cohort study of patients treated with protocolized TTM through SC and EC methods to achieve NT (37.0–37.5°C). The number of interventions and daily dose of antishivering medications, per institutional protocol, were compared between the two groups. The intensity of shivering was assessed with the Bedside Shivering Assessment Scale. Patients in the EC group (n = 23) had more antishivering interventions per patient day than those in the SC group (n = 43) (3.28 vs. 2.67, p = 0.002). Acetaminophen (81% vs. 59%, p < 0.001), buspirone (75% vs. 53%, p < 0.001), and magnesium infusions (52% vs. 36%, p = 0.012) were used on more patient days in the EC group. Patients treated with SC required more patient days of propofol (35% vs. 19%, p = 0.006) and higher average dexmedetomidine dosing per patient-day (0.70 vs. 0.56 μg/[kg·h], p = 0.03). Dosing of other medications was similar. There were no observed differences in degree or intensity of shivering. In our cohort, patients in EC group required more antishivering interventions, but less sedation, during TTM than patients in SC group. Optimizing nonsedating medications, such as acetaminophen, buspirone, and magnesium infusions, may decrease the requirement for sedatives to control shivering in both SC and EC.

Introduction

Targeted temperature management (TTM) is a therapeutic intervention in which core body temperature (T_c) is intentionally manipulated to either induce therapeutic hypothermia (TH, T_c <36°C, usual 33–35°C) or normothermia (NT, T_c 36–37.3°C, usual 36–37°C) (Nunnally et al., 2011). The goal of TTM is to provide neuroprotection, control of intracranial hypertension and cerebral edema, and mitigate the effects of fever (Polderman and Herold, 2009). Although TTM to NT may be used to primarily control for fever, recent studies suggest that this approach may also have a neuroprotective role (Nielsen et al., 2013).

Shivering is a common adverse effect seen during TTM. Shivering is the body's normal response to decreased T_c when vasoconstriction alone is unable to maintain homeostasis. It is an involuntary, oscillatory skeletal muscular activity that increases the metabolic activity, heat production, and oxygen consumption (Pozos et al., 1987; Sessler et al., 1988; Badjatia et al., 2008, 2009; Weant et al., 2010). During TTM, uncontrolled shivering may impede temperature lowering due to heat transfer from the core to the periphery (Badjatia et al., 2008, 2009). In addition, shivering causes cerebral metabolic stress, increases intracranial pressure, and decreases cerebral oxygenation, all of which should be avoided in patients with brain injury (Oddo et al., 2010). Because skin temperature contributes to ∼20% of the vasoconstriction and shivering response (Cheng et al., 1995), it has been proposed that surface cooling (SC) may result in more shivering than endovascular cooling (EC), but direct comparison has been limited (Polderman and Callaghan, 2006).

The purpose of this study was to compare the antishivering medication requirement and the degree of shivering in patients undergoing TTM to NT through SC and EC. We hypothesized that SC would be associated with more shivering and antishivering medication requirements compared to EC.

Methods

Study design, setting, and patient population

This is a retrospective cohort study at a single tertiary academic referral center. Our Institutional Review Board approved the study (Control No. 14D.67) and waived the need for informed consent. We reviewed the electronic medical records (EMR) of critically ill neurological patients undergoing TTM to NT at our Neuroscience Intensive Care Unit (NICU) from February 2012 to April 2014. Patients were identified based on nurse documentation of both a cooling device and the Bedside Shivering Assessment Scale (BSAS) (Badjatia et al., 2008) in the electronic medical record. Antishivering management during TTM and collection of the BSAS at our institution were protocolized in July 2010. We deliver protocolized TTM through FDA-cleared Arctic Sun^® 5000 System (Medivance, Inc., Louisville, CO), noninvasive SC system, or the Thermoguard XP Temperature Management System^® with endovascular Quattro or Cool Line catheters (Zoll Medical Corporation, Chelmsford, MA). Goal temperature for NT at our institution is 36.0–37°C. Selection of cooling device was based upon the clinician's discretion. It has been demonstrated that both modalities are equally effective at achieving temperature goals (Hoedemaekers et al., 2007; Oh et al., 2015).

Our institutional antishivering protocol follows published widely accepted recommendations (Choi et al., 2011). Briefly, stepwise interventions are used to achieve shivering control with the least sedating therapies up front. Patients receive around the clock (ATC) skin counter warming, acetaminophen, and buspirone, along with optional hands, face, and neck air warming. First-tier interventions consist of intravenous (i.v.) opioid agents (meperidine or fentanyl), i.v. dexmedetomidine, and continuous i.v. magnesium infusions. Persistent shivering is managed with second-tier interventions such as i.v. propofol, neuromuscular blockers, and/or i.v. dantrolene.

Eligible patients were identified based on EMR documentation. Ineligible patients were excluded if they were treated with both the SC and EC devices, if documentation of BSAS or device applied was unclear, or treated by our TTM protocol to achieve TH. Patient records were then tracked from the first utilization of NT to 7 days or until the cooling device was removed. We ascertained and collected the following data: demographics, antishivering medication (dose and duration), daily range of BSAS, cooling method, duration of TTM to NT, daily serum electrolyte values during TTM to NT, hospital length of stay, and hospital discharge status (dead or alive).

Exposure and outcome measures

The primary endpoint was the number of antishivering medications required per patient day. Secondary endpoints included antishivering medication dosing requirements and the degree of daily shivering incidence as measured by the BSAS.

Statistical analyses

Continuous variables were analyzed using accepted standards for parametric and nonparametric data as means and standard deviations (SD) or medians and interquartile ranges. Categorical variables were reported as count and proportions in each group. Bivariate comparisons were made using the t-test or Mann–Whitney U test for continuous variables and χ² or Fisher's exact test for categorical variables and statistical significance was set at p < 0.05. Our reporting of observational data conforms with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (von Elm et al., 2007).

Results

Seventy-seven patients were identified who were treated with NT during the study period. Six patients were excluded due to treatment with concurrent or sequential SC and EC. Five patients were excluded due to unclear documentation of cooling method and/or BSAS. Data were collected for a total of 66 patients who underwent 249 patient days of NT. Baseline demographics were similar in both groups (Table 1). The mean patient age was ∼56 years. The EC group had a higher percentage of males compared to the SC group, but this was not significantly different (56.5% vs. 41.9%). The type of neurologic injury was similar in both groups, with the exception that two patients with status epilepticus were treated with SC and none were cooled endovascularly. Mortality was numerically higher in the SC group, but this was not significantly different (30.2% vs. 17.4%) (Table 1).

Table 1.

Patient Demographics

Characteristic	SC (n = 43)	EC (n = 23)	p
Age (years), mean (SD)	55.8 (12.8)	56.3 (8.8)	0.876
Male, n (%)	18 (41.9)	13 (56.5)	0.306
Duration of NT (hours), mean (SD)	78.0 (40.0)	94.9 (54.4)	0.155
Patient days of NT	152	97	—
Neurologic injury
Subarachnoid hemorrhage, n (%)	16 (37.2)	12 (52.2)	0.300
Intracranial hemorrhage, n (%)	15 (34.8)	7 (30.4)	0.789
Ischemic stroke, n (%)	10 (23.3)	4 (17.4)	0.755
Status epilepticus, n (%)	2 (4.7)	0 (0)	0.539
Hospital length of stay	20.4 (15.8)	23.6 (13.7)	0.153
Mortality, n (%)	13 (30.2)	4 (17.4)	0.377

EC, endovascular cooling; NT, normothermia; SC, surface cooling; SD, standard deviation.

Patients in the EC group were treated with 3.3 antishivering medications per patient day, compared to 2.7 medications per patient day in patients who were managed with SC (p = 0.002, Table 2). Both ATC medications were used more commonly in the EC group (p < 0.001). Patients treated with EC received acetaminophen 81% of patient days and buspirone on 75% of patient days. SC patients received these ATC medications on 59% and 53% of patient days, respectively (Table 2).

Table 2.

Antishivering Medication Requirements

Total interventions	SC (n = 152)	EC (n = 97)	p
Number of shivering medications per patient day–mean (SD)	2.66 (1.65)	3.28 (1.21)	0.002^a
ATC medications^b
Acetaminophen, n (%)	89 (58.6)	79 (81.4)	<0.001^a
Buspirone, n (%)	81 (53.3)	73 (75.3)	<0.001^a
Step 1 medications^b
Dexmedetomidine, n (%)	83 (54.6)	57 (58.8)	0.601
Magnesium infusion, n (%)	55 (36.2)	50 (51.5)	0.012^a
Any fentanyl, n (%)	41 (27.0)	38 (39.2)	0.051
Bolus, n (%)	26 (17.1)	24 (24.7)	0.148
Infusion, n (%)	20 (13.2)	17 (17.5)	0.365
Step 2 medications^b
Propofol, n (%)	53 (34.9)	18 (18.6)	0.006^a

No patients were treated with meperidine (step 1), cisatracurium, or dantrolene (step 2).

Statistically significant.

Data reported as the number (percent) of patient days requiring individual medications.

ATC, around the clock.

Dexmedetomidine was used for a similar percent of patient-days (59% in the EC group and 51% in the SC group, p = 0.6). Overall fentanyl use was not significantly different between the two groups at 39% of EC patient days and 27% of SC patient days (p = 0.051). However, patients in the EC group had more patient days on continuous magnesium infusions (51% vs. 36%, p = 0.01). No patients were treated with meperidine, a tier-1 medication. Of the tier-2 medications, propofol was more commonly utilized in the SC group (35% vs. 19%, p = 0.006), and no patients received cisatracurium or dantrolene (Table 2).

Medication dosing per patient day was similar, with the exception that SC patients required higher dexmedetomidine doses (0.70 vs. 0.56 mcg/kg/h, p = 0.03) (Table 3). Magnesium infusion dosing could not be assessed due to inconsistent documentation, but infusion rates were titrated to serum magnesium of 3–4 mEq/L in both groups.

Table 3.

Antishivering Medication Dosing

Medication	SC (n = 152)	EC (n = 97)	p
Acetaminophen, g/day	2.93 (1.22)	3.10 (1.23)	0.365
Buspirone, mg/day	54.5 (25.9)	53.0 (15.6)	0.688
Dexmedetomidine, mcg/(kg·h)	0.700 (0.330)	0.564 (0.351)	0.030^a
Propofol, mcg/(kg·min)	48.6 (22.2)	44.5 (19.7)	0.484
Fentanyl
Bolus, mcg/day	120.2 (102.5)	111.5 (80.0)	0.740
Infusion, mcg/h	111.2 (87.6)	90.3 (54.7)	0.401

Doses expressed as mean (SD) per patient day.

Statistically significant.

No differences in amount or intensity of shivering could be detected (Table 4). An episode of shivering was documented on 47.4% of patient days in the endovascular group and 40.8% of patient days in the surface group (p = 0.359). Moderate shivering was observed on 16.5% and 13.8% of patient days, respectively (p = 0.587). Severe shivering was documented on 4.1% and 3.3% of patient days (p = 0.739). No patients experienced hypomagnesemia, while treated with NT (Table 5). Patients in the EC group experienced more days with hyponatremia (16.1% vs. 3.5%, p = 0.001).

Table 4.

Shivering Assessment

Shivering assessment	Surface (n = 152)	Endovascular (n = 97)	p
Any shivering, n (%)	62 (40.8)	46 (47.4)	0.359
BSAS 1, n (%)	61 (40.1)	45 (46.4)	0.359
BSAS 2, n (%)	21 (13.8)	16 (16.5)	0.587
BSAS 3, n (%)	5 (3.3)	4 (4.1)	0.739
BSAS 2 or 3, n (%)	24 (15.8)	18 (18.6)	0.605

Data reported as the number (percent) of patient days with any shivering or specified BSAS documented.

BSAS, Bedside Shivering Assessment Scale.

Table 5.

Serum Electrolyte Abnormalities During Normothermia

Electrolyte	SC (n = 152)	EC (n = 97)	p
Hypomagnesemia (magnesium <1.3 mEq/L)	0/144 (0)	0/144 (0)	—
Hyponatremia (sodium <135 mM)	5/143 (3.5)	15/93 (16.1)	0.001^a

Data reported per patient day in which electrolyte values were obtained.

Statistically significant.

Discussion

Our investigation is the first direct comparison of antishivering medication requirements in patients treated with NT by SC or EC. In this observational study, patients cooled with advanced EC methods used more antishivering medications per patient day than patients cooled by advanced SC. This was primarily driven by the use of nonsedating medications, including acetaminophen, buspirone, and magnesium infusions. Patients cooled with advanced SC methods used more patient days of propofol and higher dexmedetomidine dosing. No difference in the frequency or degree of shivering could be detected between these two groups, suggesting differences in escalation of measures rather than on the degree of shivering accounted for our results.

A previous study of comatose patients after cardiac arrest treated with TH found similar rates of shivering between intravascular and SC (22/92 and 20/75, respectively, p = 0.683) (Tomte et al., 2011). Of note, patients in both groups were deeply sedated and nearly half received paralytics, which may have obscured any difference in shivering. Aside from paralytic use, antishivering medication requirements were not reported. Although some practitioners more readily use neuromuscular blockade for shivering control, in our NICU, paralytics and sedative medications are minimized.

The observation that more antishivering medications were used in patients cooled endovascularly in our study appears to contradict what others have proposed (Polderman and Callaghan, 2006). However, because ATC medications were more commonly used in the EC group, comparisons of other medication requirements between the two cohorts are limited. It is possible that use of less sedating medications, such as acetaminophen, buspirone, and magnesium, decreased the dexmedetomidine dose required and decreased the need to prescribe propofol to control shivering in the EC group.

Although patients cooled endovascularly more commonly received minimally sedating ATC medications, the use of these medications was suboptimal in both groups. Ultimately, choice of antishivering regimen is to clinical judgment based on individual patient circumstances. However, the goal of using ATC medications in every patient is to prevent progression to Tier-1 and Tier-2 medications, which may result in prolonged sedation. Ideally, in future investigations the stepwise protocol should strictly be followed to standardize ATC medication utilization, allowing for more accurate assessment of Tier-1 and Tier-2 medication needs. If a more balanced use of ATC medications reflects differences in Tier-1 and Tier-2 requirements, the hypothesis that SC induces more shivering than EC can be tested.

In addition to less than optimal utilization of ATC medications, dosing of both acetaminophen and buspirone was below the maximum daily limit. Mean acetaminophen dose per day was ∼3 g, and mean buspirone dose was just over 50 mg per day. If tolerated, the maximum dose of acetaminophen 4 g/day and buspirone 60 mg/day should be used, as stated in our protocol.

This study does not reveal if there is a difference in the generation of shivering between advanced cooling techniques. Because shivering is associated with metabolic and cerebral stress, aggressive shivering control is warranted in NT. Thus, it is possible that no difference was seen in the amount or degree of shivering because medications are titrated to a goal BSAS of 0, regardless of cooling method.

Risk factors that have been identified for shivering during TTM are young age, male gender, hyponatremia, and hypomagnesemia (Badjatia et al., 2007). The mean age in both groups was ∼56, so differences in antishivering medications cannot be attributed to age. Although a higher percentage of patients cooled by advanced EC methods were male compared to those cooled by SC, this difference was not of statistical significance. The EC group experienced more patient days of hyponatremia, which could have affected our findings. However, only one study could be identified that cites hyponatremia as a risk factor for shivering (Badjatia et al., 2007). More evidence exists to support low magnesium levels as a risk factor for shivering (Badjatia et al., 2007, 2008). In fact, magnesium infusions are effective interventions to treat shivering (Kizilirmak et al., 1997; Badjatia et al., 2008). There were no episodes of hypomagnesemia in either group.

There are several limitations of this study that should be acknowledged. First, this was a retrospective study limiting the inferences that could be made about causal relationships. Second, data collection was limited by documentation of medication administration and hourly shivering assessments. Third, our study may have suffered from selection bias as practice variation among providers caring for patients undergoing TTM, rather than cooling modality, could have influenced the use of antishivering medications. Fourth, our data are representative of a single institution, making the generalizability of the results impractical. Fifth, sedating medications are avoided in our NICU unless deemed necessary for a therapeutic reason, such as prevention or treatment of shivering. Thus, the observation that propofol was more commonly used and dexmedetomidine dosing was higher in the SC group likely reflects the need for shivering control. It is possible that these medications were used strictly for sedation and not for shivering. Finally, differences in the frequency of hyponatremia and gender distribution could have affected results.

Conclusions

In our cohort, patients in the EC group required more antishivering interventions, but less sedation, during TTM than patients in the SC group. Although no differences in shivering were observed between the two groups, patients undergoing TTM by EC were more likely to receive Tier-1 and Tier 2 antishivering interventions. Optimizing nonsedating medications, such as acetaminophen, buspirone, and magnesium infusions, may decrease the requirement for sedatives to control shivering in both SC and EC.

Footnotes

Acknowledgments

Dr. Rincon has received salary support from the American Heart Association (AHA 12CRP12050342). Andrew Kirk, Cara McDaniel, and Dorota Szarlej have no sources of funding to declare.

Author Disclosure Statement

Dr. Rincon: Consultant for Bard Medical, Inc. Andrew Kirk, Cara McDaniel, and Dorota Szarlej have no conflicts of interest to declare.

References

Badjatia

, Kowalski

, Schmidt

, et al. Predictors and clinical implications of shivering during therapeutic normothermia. Neurocrit Care, 2007; 6:186–191.

Badjatia

, Strongilis

, Gordon

, et al. Metabolic impact of shivering during therapeutic temperature modulation: the Bedside Shivering Assessment Scale. Stroke, 2008; 39:3242–3247.

Badjatia

, Strongilis

, Prescutti

, et al. Metabolic benefits of surface counter warming during therapeutic temperature modulation. Crit Care Med, 2009; 37:1893–1897.

Cheng

, Matsukawa

, Sessler

, et al. Increasing mean skin temperature linearly reduces the core-temperature thresholds for vasoconstriction and shivering in humans. Anesthesiology, 1995; 82:1160–1168.

Choi

, Ko

, Presciutti

, et al. Prevention of shivering during therapeutic temperature modulation: the Columbia anti-shivering protocol. Neurocrit Care, 2011; 14:389–394.

Hoedemaekers

, Ezzahti

, Gerritsen

, van der Hoeven

. Comparison of cooling methods to induce and maintain normo- and hypothermia in intensive care unit patients: a prospective intervention study. Crit Care, 2007; 11:R91.

Kizilirmak

, Karakas

, Akca

, et al. Magnesium sulfate stops postanesthetic shivering. Ann N Y Acad Sci, 1997; 813:799–806.

Nielsen

, Wetterslev

, Cronberg

, et al. Targeted temperature management at 33°C versus 36°C after cardiac arrest. N Engl J Med, 2013; 369:2197–2206.

Nunnally

, Jaeschke

, Bellingan

, et al. Targeted temperature management in critical care: a report and recommendations from five professional societies. Crit Care Med, 2011; 39:1113–1125.

10.

Oddo

, Frangos

, Maloney-Wilensky

, Andrew Kofke

, Le Roux

, Levine

. Effect of shivering on brain tissue oxygenation during induced normothermia in patients with severe brain injury. Neurocrit Care, 2010; 12:10–16.

11.

, Oh

, Kim

, et al. An observational study of surface versus endovascular cooling techniques in cardiac arrest patients: a propensity-matched analysis. Crit Care, 2015; 19:85.

12.

Polderman

, Callaghan

. Equipment review: cooling catheters to induce therapeutic hypothermia?. Crit Care, 2006; 10:234.

13.

Polderman

, Herold

. Therapeutic hypothermia and controlled normothermia in the intensive care unit: practical considerations, side effects, and cooling methods. Crit Care Med, 2009; 37:1101–1120.

14.

Pozos

, Israel

, McCutcheon

, Wittmers

Jr. , Sessler

. Human studies concerning thermal-induced shivering, postoperative “shivering,” and cold-induced vasodilation. Ann Emerg Med, 1987; 16:1037–1041.

15.

Sessler

, Israel

, Pozos

, Rubinstein

. Spontaneous post-anesthetic tremor does not resemble thermoregulatory shivering. Anesthesiology, 1988; 68:843–850.

16.

Tomte

, Draegni

, Mangschau

, Jacobsen

, Auestad

, Sunde

. A comparison of intravascular and surface cooling techniques in comatose cardiac arrest survivors. Crit Care Med, 2011; 39:443–449.

17.

von Elm

, Altman

, Egger

, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med, 2007; 147:573–577.

18.

Weant

, Martin

, Humphries

, Cook

. Pharmacologic options for reducing the shivering response to therapeutic hypothermia. Pharmacotherapy, 2010; 30:830–841.