Excellent Rectal Temperature Cooling Rates in the Polar Life Pod Consistent With Stationary Tubs

A randomized, crossover, repeated-measures design guided data collection for this study. The independent variables were cooling method (PLP or tub) and time (factor levels varied according to the dependent variable). The dependent variables were T_REC cooling rates, nadir T_REC, ESQ responses, and thermal sensation scores. We measured T_REC every 5 minutes during exercise and every 0.5 minutes during cooling. Participants' ESQ responses were recorded before exercise, near the end of exercise, and immediately after cooling. Thermal sensation was measured before exercise, near the end of exercise, halfway through cooling (ie, when T_REC was 38.75°C), and immediately after treatment. We also measured environmental chamber temperature, relative humidity, pre-exercise hydration status, exercise duration, and water temperature in the PLP and tub to ensure consistency of testing conditions between and within participants.

Authors⁵  using CWI with water temperatures of approximately 15°C have shown cooling rates up to 0.18°C/min, whereas the only published research¹⁷  on the PLP demonstrated an average T_REC cooling rate of 0.04°C/min. Thus, we estimated sample size a priori with this potential treatment effect size, an α level of .05, 80% power, and an SD of 0.08°C/min.⁵  Based on these assumptions, we needed 10 participants. To increase power and include a similar number of participants as other hyperthermia studies,^9,10,15,20  we tested a convenience sample of 17 healthy, physically active men and women. Unfortunately, 4 participants discontinued testing because of the difficulty of the exercise protocol; 13 participants completed testing (Table 1). Seven individuals used a tub on the first testing day, and 6 used the PLP on the first testing day.

Recruits were excluded from the study if they self-reported (1) an injury or illness that impaired their ability to exercise; (2) any neurologic, respiratory, gastrointestinal, esophageal, or cardiovascular disease; (3) taking any medications that may have affected fluid balance or temperature regulation; (4) a sedentary lifestyle (defined as exercising <30 minutes, 3 times per week)²¹; (5) a history of heat-related illness in the 6 months before data collection; (6) current pregnancy; (7) cold allergy; or (8) a positive COVID test result within 14 days of testing. All women were tested within the follicular phase of their menstrual cycle (ie, first 14 days after the onset of menstruation). All participants signed a written informed consent before testing, and all procedures were approved by Central Michigan University's Institutional Review Board.

Clinicians require ice for several purposes besides emergency use (eg, first aid, cryokinetics). We sought a balance between using ice for the PLP and respecting clinicians' other needs for ice. Therefore, before testing human participants, we conducted pilot testing to determine the water temperature inside several 37.9-L (10-gal) coolers over the course of 2 hours in a 22°C laboratory. We did this to ensure that the water in the coolers could be kept within the manufacturer recommendations¹⁸  (ie, 1.7°C–15°C) for the entirety of each trial (approximately 2 hours). We used the cooler's rough interior volumetric indicators and experimented with various combinations of crushed ice and 18.5°C tap water. Mixing 15.1 L (4 gal) of ice with 22.7 L (6 gal) of tap water consistently produced water temperatures between 2°C and 4°C. We believed that dedicating 24 gal (91.2 L) of ice across 6 coolers, approximately half of our ice machine's capacity, allowed for a valid comparison of the PLP and tub while respecting the other needs for ice clinicians might have to effectively do their jobs.

Procedures for this study followed those of other published laboratory studies whose authors^9,20,22  investigated the effectiveness of CWI for severe hyperthermia. Participants reported for 2 days of testing between 0800 and 1600. They were instructed to abstain from exercise for 24 hours and from stimulants (eg, caffeine) and depressants (eg, alcohol) for at least 12 hours. They were instructed to drink water regularly throughout the day before testing to ensure that their urine was clear or light yellow. Compliance with these instructions was self-reported before testing.

Participants voided their bladders completely, and we obtained a spot urine specific gravity measurement to assess hydration status (SUR-Ne refractometer; Atago USA, Inc). If the urine specific gravity indicated participants were hypohydrated (ie, >1.020),²³  they consumed approximately 1 L of water, and urine specific gravity was reassessed approximately 45 minutes later. If hypohydration persisted, they were rescheduled. If euhydrated, participants were weighed nude (scale model Defender 5000; Ohaus Corp). Then, they dressed in undergarments (including sports bras for women), shorts, socks, and T-shirts. We measured skinfolds at the chest, abdomen, and thigh (men) and the triceps brachii, abdomen, and thigh (women) in triplicate per Pollack et al²⁴  (Baseline skinfold caliper model 12-1110; Fabricated Enterprises, Inc). Skinfolds were averaged at each site and summed to estimate body density²⁵  and percentage of body fat.²⁶  Body surface area was estimated using the Du Bois and Du Bois²⁷  equation.

Participants donned a heart rate monitor (model T31; Polar Electro) and inserted a rectal thermistor (model 401; Advanced Industrial Systems) 15 cm past the anal sphincter.²²  They entered an environmental chamber, had their T_REC recorded, rated their thermal sensation on a 9-point scale (0 = unbearably cold, 1 = very cold, 2 = cold, 3 = cool, 4 = comfortable, 5 = warm, 6 = hot, 7 = very hot, 8 = unbearably hot),⁸  and completed their first ESQ. At this time, we measured environmental temperature and humidity (model 4400 Kestrel Heat Stress Tracker; Nielsen-Kellerman Co), and participants stood on a treadmill for 10 minutes to acclimate to the heat. Then they performed an incremental exercise protocol on a treadmill consisting of walking for 3 minutes at 3 miles/hour (5 km/hour) and running at approximately 80% to 90% of their age-predicted maximum heart rate for 2 minutes (0% incline).

Upon reaching a T_REC of approximately 39.45°C, participants rated their thermal sensation and completed a second ESQ. Once T_REC was 39.5°C, they stopped the treadmill and followed 1 of 2 protocols, depending on the day of testing.

On PLP days, we followed manufacturer recommendations for use.¹⁸  Thirty minutes before each participant's arrival, we filled six 37.9-L (10-gal) water coolers with 15.1 L (4 gal) of ice and 22.7 L (6 gal) of tap water and put the lids on the coolers. Each lid was numbered so we could average the temperatures of the water from the coolers used during treatment in the event we did not use the water from all 6 coolers. The coolers were kept sealed in our main laboratory (ambient temperature approximately 22°C) until participants' T_REC reached approximately 38.2°C. At that time, an assistant stirred the water in each cooler and measured the water temperature by placing a No. 401 thermistor approximately halfway (30.5 cm [12 in]) into the center of the cooler. Then, we moved the coolers inside the environmental chamber. Once participants' T_REC reached 39.5°C, they removed their shoes and lay inside the PLP. For shorter participants, we folded the end of the PLP closer to the participant's feet to minimize water accumulation at the end of the unit. One investigator poured four to six 37.9-L (10-gal) water coolers of preprepared ice and water (3.2 ± 0.6°C [38°F]) into the PLP so the individual's torso, arms, legs, and neck were covered. The participant's head rested on the pillow included with the unit to ensure airway patency during cooling. A separate No. 401 thermistor was placed next to the participant's neck in the water so we could monitor the water temperature in the PLP during cooling. The PLP zipper was closed, and the PLP straps were secured. We recorded the volume of water and ice initially added. The PLP was shaken continuously side to side during cooling (Figure 1). The PLP water temperature was also monitored continuously to ensure that the water temperature did not exceed 15°C per manufacturer instructions. Although the manufacturer recommends removing water and replacing it with more ice if the water temperature near the participant's neck exceeds 15°C,¹⁸  we did not find this step necessary. Thus, no additional water or ice was purposefully removed or replaced after the initial water was added to the PLP. Final water temperature was recorded once the participant's T_REC reached 38°C in both conditions.

For tubs, we followed the National Athletic Trainers' Association (NATA) recommendations for CWI.¹  Briefly, participants removed their shoes and immersed themselves up to the neck in a 1135.6-L capacity noncirculating water tub filled with approximately 568 L (approximately 150 gal) of water (160.7 cm [length] × 175.3 cm [width] × 63.5 cm [height]; model 4247; Rubbermaid). While the participant was exercising, we monitored tub water temperature and added ice as necessary to keep it close to 15°C (59°F). We were careful to ensure that any added ice had melted before participants entered the tub. The bath was stirred continuously with a plastic rod until participants' T_REC was reduced to 38°C. A separate No. 401 thermistor was secured approximately 20 cm from the bottom of the tub so we could monitor and record tub water temperature during cooling.

The T_REC was recorded every 0.5 minutes during cooling. A standard stopwatch was started the moment the participant's foot touched the water (tub) or when the first water cooler was poured on top of participants (PLP). The stopwatch was halted when the participant's T_REC reached 38°C so we could calculate T_REC cooling rates. Cooling rates for each condition were calculated by taking the difference in body temperatures between the end of exercise and the end of treatment and dividing it by the amount of time necessary to reduce T_REC to 38°C.

Participants self-reported shivering onset during cooling. Halfway through cooling (T_REC = 38.75°C), they identified thermal sensation a third time. Once T_REC was 38°C, participants exited the PLP or tub, towel dried their arms and legs, completed a third ESQ, and reported thermal sensation a fourth time. They sat in the environmental chamber for 10 minutes to recover, and we measured environmental conditions a second time. After recovery, participants exited the chamber, removed the rectal thermistor, towel dried, were weighed nude a second time, and were excused. No fluids were given once they entered the environmental chamber. They completed their second session at approximately the same time of day (±3 hours) and at least 48 hours after the first session.

It bears explanation why we used different water temperatures and water volumes for each cooling condition. The PLP cannot contain as much water as a tub. The manufacturer¹⁸  advises using 151 to 227 L (40–60 gal) of water. We added 570 L (150 gal) of water to the tub because it was a common size sold at various retailers and used in clinical practice. Regarding the different water temperatures in each device, 1 of the main benefits of portable CWI techniques, such as the PLP, is that less ice is needed to achieve the lower end of the recommended water temperatures for treating patients with EHS (ie, 1.7°C).¹  Consequently, devices such as the PLP can maximize the potential thermal gradient between the patient and water with fewer resources. In contrast, a major advantage of tubs is that they can hold a much higher volume of water than the PLP. However, to achieve similarly cold temperatures in a tub, we would have needed to almost completely empty our 189.3-L (50-gal) ice machine. Because we wanted to emulate the practical conditions under which clinicians use ice in their clinical practice, we elected to use similar amounts of ice (roughly 16–24 gal) for each device in our study. Consequently, this meant warmer water temperatures on tub days. Nevertheless, the tub water temperature was still within the NATA recommendation¹  for treating patients with EHS.

Because exercise and CWI durations differed among participants, we statistically compared T_REC only at times common to all. Means and SDs were calculated for each dependent variable and assessed for normality. Separate dependent t tests were conducted to examine differences in T_REC cooling rates, environmental conditions, pre-exercise urine specific gravity, and exercise durations because the data were normally distributed. Wilcoxon signed rank tests were computed when normality was violated, with the median and interquartile range reported (eg, nadir T_REC).

For ESQ responses, we summed the scores from the 16 items and created a new cumulative score.¹⁰  The cumulative scores were analyzed with a repeated-measures analysis of variance (ANOVA). Although thermal sensation is technically categorized as ordinal data, we elected to use a repeated-measures ANOVA to analyze this variable for 2 reasons. First, thermal sensation data are prone to large numbers of ties, and nonparametric models that use ranking to analyze for differences are not robust enough to overcome data with a large number of ties. Second, prior investigators^8,9,13,14  used parametric statistics to analyze thermal sensation data.

A separate repeated-measures ANOVA also examined differences in T_REC during exercise, cooling, and recovery between conditions. We assessed sphericity with the Mauchly test. Geisser-Greenhouse adjustments to P values and degrees of freedom were made if the sphericity condition was violated. Upon significant interactions or main level effects, Tukey-Kramer post hoc tests identified differences between cooling methods at each time point. Significance was accepted when P < .05 (version 2007; Number Cruncher Statistical Software; NCSS, LLC).

All participants self-reported compliance with testing instructions each day. They were well hydrated before exercise (t₁₂ = 1.4, P = .10; Table 1) and exercised for similar durations each day (t₁₂ = 0.5, P = .31; Table 2). The environmental chamber temperature was slightly warmer on PLP days (t₁₂ = 5.0, P < .001), but relative humidity was slightly higher on tub days (t₁₂ = 2.7, P = .01; Table 2).

The T_REC was comparable between conditions during exercise, and all participants discontinued exercise when T_REC was 39.5°C. Although T_REC was comparable between conditions in the first minutes of cooling, individuals were in the PLP for shorter durations than in the tub (Figure 2). This resulted in faster T_REC cooling rates in the PLP than in the tub (t₁₂ = 2.5, P = .01; Table 2). The T_REC continued to decline even during recovery and displayed a lower nadir in the PLP (z₁₂ = 3.2, P < .001; Table 2).

The ESQ scores (F_2,24 = 15.4, P < .001; Table 3) and thermal sensation (F_3,36 = 5.9, P = .002; Figure 3) differed between conditions over time. For the PLP, thermal sensation was lower at the halfway point (ie, T_REC at 38.75°C) and immediately postcooling than for the tub (P < .05). The ESQ scores differed at pre-exercise from postexercise and postcooling in both conditions. Although ESQ scores dropped postcooling compared with postexercise in the tub, this was not observed in the PLP. In fact, the PLP ESQ scores were higher than tub ESQ scores postcooling (P < .05).

Our main observation was that both the PLP and tub had excellent cooling rates in men and women with hyperthermia who did not experience EHS. The PLP cooling rate actually exceeded that of tubs and other portable CWI methods reported in the literature (eg, tarp-assisted cooling with oscillation [TACO], body bags).^19,28,29  Hosokawa et al²⁸  used TACO to cool participants with hyperthermia (T_REC = 39.73°C ± 0.27°C) by placing them in a tarp with 113.6 L (30 gal) of approximately 9°C water. Participants cooled at 0.17 ± 0.07°C/min. Others²⁹  demonstrated that 151 L (40 gal) of 2.1°C water using TACO cooled at 0.14 ± 0.06°C/min.

Our cooling rates were likely higher than those of TACO^28,29  for 2 reasons. First, we used a higher water volume and colder water. Second, the design of the PLP allows for more body surface area coverage and full-body immersion, whereas TACO often only covers the posterior half of the body. To our knowledge, the authors¹⁹  of only 1 study have used medical body bags, similar to the PLP, to treat heatstroke. Emergency room physicians cooled an elderly woman who collapsed with classic heatstroke by placing her in a body bag filled with ice and water up to the anterior axillary line.¹⁹  Based on her oral temperatures before and after cooling, she cooled at a rate of 0.16°C/min. However, it is unclear how much ice or water was used in resuscitation efforts. Given our cooling rates, we concluded that the PLP successfully reduced the body core temperature of individuals with severe hyperthermia and shows promise as an EHS treatment tool. Similarly, clinicians can continue to be confident using tubs for EHS treatment.⁵ 

We propose 3 reasons why we observed much faster cooling rates than Nye et al,¹⁷  who reported a PLP T_REC cooling rate of 0.04 ± 0.08°C/min. First, we strictly followed the manufacturer recommendations for use. Specifically, this included adjusting the size of the PLP based on each participant's height. Folding the end of the PLP was important to minimize the volume of water at the bottom of the unit and helped ensure that water covered as much body surface area as possible. Also, we oscillated the water inside the PLP during cooling. Shaking the PLP ensured conductive and convective cooling, thereby facilitating the cooling process. It is unclear how much water Nye et al¹⁷  used for each participant, though they described using an unknown volume of ice and 40 to 80 gal (151–303 L) of water. It is also unclear whether they oscillated the water during treatment. Second, our participants were immersed in colder water (3.2°C ± 0.6°C) than those in the Nye et al¹⁷  study. This is important because water temperature greatly influences cooling; Proulx et al³⁰  determined that cooling rates were 46% and 57% faster, respectively, when participants were immersed in 2°C than 8°C or 14°C water. Because Nye et al¹⁷  did not record the water temperature before participants were immersed in the PLP, this could explain the larger variability in the water temperatures reported (5°C–10°C). Third, our participants began cooling once T_REC was 39.5°C, as opposed to 38.4 ± 0.7°C.¹⁷  Our participants were hotter, so a larger temperature gradient existed, and our participants cooled faster.

Our study had several additional strengths. First, we standardized the water volumes and temperatures used on each testing day and recorded water temperature during cooling. Second, we monitored when participants self-reported shivering. As expected, fewer individuals shivered in a tub than in the PLP, and they experienced less time shivering during cooling. This was likely due to the differences in water temperatures between conditions and the fact that skin temperature is a primary driver in the onset of shivering.³¹  Third, our participants reported their perceptions of how they felt before, during, and after treatment. Despite having similar thermal sensation and ESQ scores before and after exercise, they felt much colder and had higher ESQ scores after PLP treatment. Although the people in tubs felt better postcooling than postexercise, this did not occur in the PLP. The extremely cold water temperatures used on the PLP day (approximately 3°C versus 15°C) were likely responsible. Anecdotally, some participants described feeling pain in their hands and feet during the PLP treatment that was not reported in the tub treatment. Previously, we¹⁵  noted that participants better tolerated and reported feeling more comfortable during and after water immersion when 21°C water was used to treat hyperthermia. However, the cooling rate with these warmer water baths was only acceptable (0.12 ± 0.05°C/min) and not ideal.¹⁵  Therefore, although the PLP was quite effective in reducing T_REC, the low water temperatures induced significant shivering, higher ESQ scores, and much lower nadir T_REC values. Consequently, clinicians who use the PLP at water temperatures comparable with what we used should heed the NATA's recommendation¹  to remove patients from the PLP when T_REC is 38.9°C and be prepared with rewarming strategies to help patients feel better and prevent hypothermic afterdrop.

We acknowledge our study's limitations. First, for safety reasons, none of our participants experienced EHS. Second, we used the 10-gal (37.9-L) coolers' rough volumetric markers to estimate how much ice and water to add to each cooler. Thus, variations in water temperatures occurred on the PLP days among participants. Third, despite our recording of the volume of water poured into the PLP for each individual, some water inevitably leaked out of the system during cooling or was lost during our attempt to pour water into the unit. Hence, the immersion volumes we identified must be considered rough estimates. Fourth, participants were not immediately fully immersed on the PLP day as they were on the tub day. We estimate it took approximately 2 minutes to pour all of the necessary water on the participants, close the zipper, and secure the straps on the PLP. If clinicians have more than 1 person pouring water on the participant, the PLP cooling rate may be even higher than the rates demonstrated here.

Finally, we acknowledge that some clinicians may have financial constraints or limitations on the equipment they are able to purchase in the case of emergencies. At the time of this study, the PLP cost approximately $425. In comparison, tarps for TACO, kiddie pools, or medical body bags may be purchased for less than $100, whereas a tub can cost up to $450, depending on size. Overall, each method of CWI has its own advantages and disadvantages. For example, TACO provides acceptable to ideal cooling^28,29  but requires help from several assistants to ensure that the patient's airway remains patent during treatment and that water covers as much body surface area as possible. Thus, it is vitally important that clinicians weigh and consider these points while also considering cooling effectiveness data when developing EHS policy and procedure documents.

In summary, despite using 2.5 to 3.0 times less water and ice than a tub, the PLP cooled hyperthermic men and women at ideal cooling rates for EHS. However, the intensity of the cold water induced more perceived stress during and after cooling. Thus, if clinicians cool individuals with the PLP using similar water temperatures, they should remove patients from the device at T_REC consistent with NATA recommendations¹  and have rewarming tools (eg, heated blankets) available. Future researchers may explore the use of the PLP in patients with EHS and whether these negative side effects can be mitigated by using warmer water in PLP while still maintaining its ideal cooling rate.

We thank Polar Products, Inc, for sponsoring this project and all the participants who volunteered their time and energy during data collection. We also thank Connor Shea, London Ross, and Andrea Willsey for their help with data collection.