You are here

Article Text

Article menu

Download PDFPDF

Original articles
Axillary temperature measurement during hypothermia treatment for neonatal hypoxic-ischaemic encephalopathy
  1. Marc-Antoine Landry1,
  2. Lex W Doyle1,2,3,4,
  3. Katherine Lee5,
  4. Susan E Jacobs1,2,4
  1. 1Neonatal Services, Royal Women's Hospital, Melbourne, Victoria, Australia
  2. 2Department of Obstetrics Gynaecology, University of Melbourne, Victoria, Australia
  3. 3Department of Paediatrics, University of Melbourne, Melbourne, Victoria, Australia
  4. 4Critical Care and Neurosciences, Murdoch Children Research Institute, Melbourne, Victoria, Australia
  5. 5Clinical Epidemiology and Biostatistics Unit, Murdoch Childrens Research Institute, Melbourne, Victoria, Australia
  1. Correspondence to Susan E Jacobs, Royal Women's Hospital, Neonatal Services, 20 Flemington Rd, Parkville, Victoria 3052, Australia; sue.jacobs{at}thewomens.org.au

Abstract

Objective To determine the accuracy of axillary temperature relative to core rectal temperature during whole-body therapeutic hypothermia for moderate-to-severe hypoxic-ischaemic encephalopathy.

Design Retrospective audit.

Setting Single tertiary neonatal intensive care unit at The Royal Women's Hospital in Australia.

Patients Fifty-eight term newborn infants with moderate-to-severe hypoxic-ischaemic encephalopathy. Forty infants were treated with whole-body hypothermia between February 2001 and May 2010, 16 of whom were enrolled in the Infant Cooling Evaluation (ICE) trial, and 18 control infants randomised to normothermia in the ICE trial.

Intervention Comparison of simultaneous axillary and rectal temperatures measured between 0 and 84 h post randomisation or induction of cooling.

Results During the initiation of hypothermia (0–<6 h) axillary and rectal temperatures were similar (mean difference rectal-axillary =0.07°C), but with large variability (95% limits of agreement −1.18 to 1.33°C). There was larger variability in measurements between 6 and <72 h in the hypothermic infants (total SD 0.44) than in the normothermic group (total SD 0.24, p<0.001). In the hypothermic infants, the mean difference between the measurements during the rewarming phase (72–<84 h) was −0.19°C (95% limits of agreement −0.95 to 0.57°C).

Conclusion As there is wide variability in the difference between axillary and rectal temperatures at all stages of whole-body cooling, our data do not support the use of axillary temperature as a surrogate for core rectal temperature during therapeutic hypothermia.

https://doi.org/10.1136/archdischild-2011-301066

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    What is already known on this topic

    • In term newborns with moderate-to-severe hypoxic-ischaemic encephalopathy, hypothermia treatment reduces mortality and neurodevelopmental disability.

    • Therapeutic hypothermia, with continuous temperature monitoring, is now standard care in tertiary neonatal intensive care units.

    • The correlation between axillary and rectal (core) temperature in normothermic term and preterm infants is well described, but is uncertain during whole-body therapeutic hypothermia.

    What this study adds

    • There is large variability between axillary and rectal temperature measurements during whole-body hypothermia treatment (±1°C).

    • Intermittent digital axillary temperature measurement is not an accurate surrogate for continuous core (rectal) temperature during whole-body therapeutic hypothermia (induction, maintenance or rewarming phases).

    Introduction

    Therapeutic hypothermia is now standard care for moderate-to-severe hypoxic-ischaemic encephalopathy (HIE), due to evidence that it reduces both mortality and major neurodevelopmental disability in early childhood.1,–,6 During whole-body therapeutic hypothermia, the core (rectal or oesophageal) temperature is maintained between 33 and 34°C. In most published protocols, a thermistor (probe) is inserted 2–5 cm through the anus into the rectum, or into the oesophagus, to measure the baby's core temperature continuously.1 ,3,–,5

    Axillary temperature in normothermic term and preterm infants has been shown to correlate well with rectal temperature.7,–,16 In a study of systemically cooled adults undergoing cardiac surgery, axillary temperatures (mean difference (SD) 0.05 (0.46)) were closer to ‘core’ pulmonary artery temperature than rectal temperatures (−0.80 (0.27)).17 In a cohort study of 39 term newborns with HIE undergoing hypothermia induction in non-tertiary hospitals and transport to specialist cooling centres, there was a high correlation between 61 paired axillary and rectal temperatures (mean difference of 0.1°C; 95% limits of agreement from −1.1°C to 1.3°C).18

    Therapeutic hypothermia protocols include continuous temperature monitoring during cooling and rewarming. Core (rectal) temperature is not routinely measured in newborn infants and so rectal thermistor probes are not always available outside neonatal intensive care units (NICUs). As the majority of newborns requiring therapeutic hypothermia for HIE are born outside tertiary centres and require transport to cooling centres after birth, it may be desirable to commence controlled therapeutic hypothermia within strict protocols under guidance of the NICU or transport staff before the transport team arrive. However, there must be an accurate and safe method to measure the temperature during induction of hypothermia to ensure it reaches and is maintained in the target range of 33–34°C before a rectal probe is available; axillary temperature may be a candidate for this role. However, information regarding the correlation between axillary and rectal temperature measurement during whole-body therapeutic hypothermia in newborn infants is limited. The aim of this study was to determine the accuracy of axillary temperature relative to core (rectal) temperature measurement during whole-body therapeutic hypothermia.

    Methods

    We reviewed the axillary and rectal temperature data of all term and near-term newborns with moderate-to-severe HIE treated with whole-body therapeutic hypothermia according to the infant cooling evaluation (ICE) trial protocol at The Royal Women's Hospital (RWH) from February 2001 to May 2010.3 ICE was a multicentre randomised controlled trial of hypothermia versus normothermia for the treatment of HIE. The results of this study have been reported elsewhere.3 The current study includes infants enrolled in the ICE trial between February 2001 and July 2007, and infants who received hypothermia as standard treatment for moderate-to-severe HIE using the same protocol between August 2007 and May 2010. Infants allocated to normothermia in the ICE trial (between February 2001 and July 2007) were used as a comparison group.

    All infants had continuous core temperature measurement with a rectal temperature probe (Mon-a-thermTM 9 Fr, Mallinckrodt Inc, St Louis, Missouri, USA) inserted at least 5 cm into the rectum. The target rectal temperature range was 33–34°C for 72 h in infants treated with hypothermia (including a 6 h induction period, maintenance phase between 6 and 72 h, followed by 8–12 h of rewarming) and 36.8–37.3°C for the normothermic controls. Rectal temperature was recorded hourly according to the ICE protocol, with intermittent digital axillary temperature measurements documented 4–6 hourly; surface skin temperature was not measured. To be eligible for the study presented here, the infant had to have at least one axillary temperature measurement recorded at the same time as a rectal temperature between 0 and 84 h post initiation of cooling or randomisation.

    All infants received care on a radiant warmer, with heat applied according to management with hypothermia or normothermia. Infants treated with hypothermia to 33.5°C had the radiant warmer turned off to expose the infant to the ambient temperature. If the core temperature was over 35.5°C, two refrigerated gel packs were placed across the chest and under the head and shoulders and sequentially removed when the core temperature was below 35.0°C and then 34.5°C. If the core temperature was under 33.5°C, the radiant warmer heater output was manually adjusted every 15–30 min. If the core temperature was above 34.0°C during the 6–<72 h maintenance phase, gel packs were reapplied. Infants were rewarmed by 0.5°C every 2 h between 72 and 84 h. Control infants had the radiant warmer turned on with their rectal temperature maintained at 37°C by manual adjustment of the radiant warmer heater output every 15–30 min as needed.

    This analysis includes all available axillary and rectal temperature measurements that were recorded at the same time from the initiation of hypothermia (time 0) to the end of rewarming (84 h) in hypothermia-treated infants and all available measurements 0–84 h post randomisation for infants in the control (normothermic) group.

    The RWH Research and Ethics Committees approved the ICE study, for which parents gave written informed consent. The additional data collection was approved as an audit by the same committee.

    To explore whether axillary temperature can be used as a surrogate for rectal temperature, we need to know if axillary temperatures systematically under or overestimate the temperture compared with rectal temperatures, and whether axillary temperatures are more imprecise (ie, more variable) than rectal temperatures. We produced scatterplots of axillary versus rectal temperatures, as well as Bland–Altman plots of the difference in the two measurements plotted against the average to explore whether the difference was constant across the range of temperatures, presented separately for the induction (0–<6 h), maintenance (6–<72 h) and rewarming (72–<84 h) phases of hypothermia. The nature of the relationship between the difference and the average was explored using linear regression fitted using random effects to allow for multiple observations within an individual. Regression lines are presented with 95% limits of agreement (calculated as 95% prediction intervals) to demonstrate the variability between the two measurements across different average temperatures (again allowing for the correlation between measurements within an individual using random effects). For comparison, the relationship between the difference and the average temperatures was also explored in the normothermic group, and compared to the relationship in hypothermic patients by fitting a single model to all of the 6–<72 h measurements including an interaction term to explore whether the relationship was different during periods of normothermia and hypothermia. We focus on measurements taken 6–<72 h post randomisation (or post cooling in infants not in the ICE trial) as this represents a period of fairly stable temperatures which are comparable between the two groups. Finally, we compared the variability in the difference between the rectal and axillary temperatures taken 6–<72 h post randomisation or cooling in the normothermia and the hypothermia groups using the estimates of the within and between individual variability from a random effects model allowing separate within and between components of variability in the two groups.

    All analyses were carried out using Stata 11, Statacorp, College Station, Texas, USA.

    Results

    Thirty-four newborn babies of at least 35 weeks' gestational age and less than 6 h old were enrolled in the ICE trial and managed at the RWH in Melbourne, Australia between February 2001 and July 2007: 16 neonates were randomised to therapeutic hypothermia (one file was not available and one baby had no documented axillary temperature measurements), with the remaining 18 randomised to normothermia. A further 31 babies were treated with therapeutic hypothermia between July 2007 and May 2010 of whom five had no axillary temperature measurements. Thus this analysis is based on data from 40 hypothermia and 18 normothermia infants.

    From the 40 hypothermic infants, there were 486 paired axillary and rectal measurements performed between the initiation of cooling (time 0) to the end of rewarming (84 h). There were 267 paired measurements available between 6 and <72 h from the 18 newborns in the normothermia group (table 1).

    View this table:
    Table 1

    Summary of participants and paired temperature measurements

    Measures during the induction phase (0–<6 h) of hypothermia

    There were 25 paired measurements documented between 0 and <6 h in the hypothermic infants (figure 1). There was no evidence that the difference between the two measures varied with the average temperature (estimate of slope=−0.13, 95% CI −0.35 to 0.09, p=0.26). Assuming a constant difference across the average temperature, the rectal temperature was on average 0.07°C higher than the axillary temperature, with 95% limits of agreement from −1.18 to 1.33°C (figure 2).

    Figure 1
    Figure 1

    Scatter plot of rectal versus axillary temperatures in measurements taken 0–<6 h post initiation of hypothermia (n=25). The line represents the line of equality. Pearson's correlation (r)=0.87.

    Figure 2
    Figure 2

    Bland–Altman plot for measurements taken 0–<6 h post initiation of hypothermia (n=25). Lines represent the mean difference (rectal-axillary) and 95% limits of agreement assuming a constant difference across the average temperatures.

    Measures during the maintenance phase (6–72 h) of hypothermia

    There were 358 measurements in the maintenance hypothermia phase and 267 in the normothermic infants taken between 6 and <72 h post randomisation (or post initiation of hypothermia in infants not in the ICE trial). Exploring the relationship between the difference in the rectal and axillary measurements and the average of the two measurements found evidence that this relationship was different in the hypothermic and normothermic groups (interaction p=0.04). In hypothermic infants, the rectal temperatures were higher than axillary temperatures at the lower temperatures (with a difference of approximately 0.05°C when the average temperature was 33°C) but lower than the axillary at the higher temperatures (with a difference of −0.23°C when the average temperature was 34°C, regression coefficient −0.28, 95% CI −0.37 to −0.20, p<0.001). In contrast, there was little evidence that the difference in measurements varied with the average temperature in normothermic infants (regression coefficient −0.11, 95% CI −0.25 to 0.02, p=0.10). Figure 3 shows a plot of the relationship between the average and difference in temperatures along with a 95% prediction interval for the difference, allowing for different relationships in the hypothermia and normothermia groups.

    Figure 3
    Figure 3

    Bland–Altman plot for measurements taken 6–<72 h post randomisation or initiation of cooling. Fitted lines show the relationship between the difference and the average allowing for a different relationship in the hyopthermic and normothermic groups and the 95% prediction interval for the difference.

    There was more variability in the difference between rectal and axillary temperatures both within and between individuals for measurements 6–<72 h post randomisation or cooling in the hypothermic group compared with the normothermic group (total variability in the hypothermic group 0.44 compared with 0.24 in the normothermic group, p<0.001 from the likelihood ratio test, compared with the model assuming a single estimate of variance within and between individuals).

    Measures during the rewarming phase (72–<84 h) of hypothermia

    There were 103 paired measurements documented between 72 and <84 hours in the hypothermic infants. There was no evidence that the difference between the two measures varied with the average temperature (estimate of slope=0.007, 95% CI −0.06 to 0.07, p=0.84). Assuming a constant difference across the average temperature, the rectal temperature was on average 0.19°C lower than the axillary temperature, with 95% limits of agreement from −0.95 to 0.57°C.

    Discussion

    The results from this study suggest that there is variability in the difference between axillary and rectal temperatures during therapeutic hypothermia for moderate-to-severe HIE which is present during all phases of cooling (induction, maintenance and rewarming). On average, rectal temperatures were 0.07°C higher than axillary temperatures during induction of hypothermia and axillary temperatures were 0.19°C higher than rectal temperatures during the rewarming phase. Despite these reasonably small average differences, the width of the 95% prediction limits for the difference between rectal and axillary temperatures exceeded +/− 1°C and +/− 0.5°C during the cooling induction and rewarming phase, respectively, which suggests that in some cases the difference between the two temperatures was fairly large. During the maintenance phase of hypothermia, the 95% confidence limits showed differences of nearly +/− 1°C between the two measurements, with evidence that the relationship between rectal and axillary temperatures changed systematically with differences in body temperature. Importantly, this study shows that the variability of the difference between axillary and rectal temperatures during the maintenance phase was larger in the hypothermic group than in the normothermic group suggesting that axillary temperatures are more unreliable measures of rectal temperatures during hypothermia than in normothermic conditions. The only other study that explored axillary and rectal temperatures in newborns with HIE during induction of passive hypothermia (which used a similar method of temperature control to the ICE trial) reported almost identical results, with large variability in the difference between rectal and axillary temperatures (95% limits of agreement from −1.1 to 1.3°C).18 However, we disagree with the authors' conclusion that axillary temperature can be used as a surrogate for rectal temperature measurement during induction of hypothermia, because both studies demonstrate that axillary temperature measurement is imprecise for use during therapeutic hypothermia.

    The main limitation of this study is that it was retrospective, and hence we cannot guarantee that the axillary and rectal temperature measurements were taken at exactly the same time. The ICE and subsequent hypothermia protocols stipulate hourly documentation of rectal temperature, but the ICE protocol did not include axillary temperature measurement. Therefore, there may have been slight variation in timing between the rectal and axillary temperatures, which could result in more variation between the two measures than if they had been taken exactly at the same time. Any additional variation in measurements related to differences in timing of axillary and rectal temperatures would be expected to affect infants equally when they were either hypothermic or normothermic. The observation that there was more variation in differences in temperatures in the hypothermic group than the normothermic group suggests that these differences are real, and that timing of axillary and rectal temperature measurements is not the explanation for this difference.

    The major strength of this study is the relatively large number of infants and paired axillary-rectal temperature measurements from each infant treated with therapeutic hypothermia according to a strict protocol. Data have been analysed using a random effects model to allow for clustering between and within individuals which adjusted the estimates and CI for the repeated measures structure of the data (table 1).

    In conclusion, the magnitude of the differences between the axillary and rectal temperature measurements in the current study suggests that intermittent axillary digital temperature measurement may not be an appropriate surrogate for core rectal temperature during therapeutic hypothermia. Until a prospective trial shows that it is safe to use axillary temperature measurement, rectal temperature should be measured, preferably continuously by a rectal thermistor, to induce and maintain core temperature during therapeutic hypothermia.

    References

    1. ↵
      1. Azzopardi DV,
      2. Strohm B,
      3. Edwards AD,
      4. et al
      .; TOBY Study Group. Moderate hypothermia to treat perinatal asphyxial encephalopathy. N Engl J Med 2009;361:1349–58.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Gunn AJ,
      2. Gluckman PD,
      3. Wyatt JS,
      4. et al
      . Selective head cooling after neonatal encephalopathy. Lancet 2005;365:1619–20.
      OpenUrlPubMedWeb of Science
    3. ↵
      1. Jacobs SE,
      2. Morley CJ,
      3. Inder TE,
      4. et al
      .; Infant Cooling Evaluation Collaboration. Whole-body hypothermia for term and near-term newborns with hypoxic-ischemic encephalopathy: a randomized controlled trial. Arch Pediatr Adolesc Med 2011;165:692–700.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Shankaran S,
      2. Laptook AR,
      3. Ehrenkranz RA,
      4. et al
      .; National Institute of Child Health and Human Development Neonatal Research Network. Whole-body hypothermia for neonates with hypoxic-ischemic encephalopathy. N Engl J Med 2005;353:1574–84.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Simbruner G,
      2. Mittal RA,
      3. Rohlmann F,
      4. et al
      .; neo.nEURO.network Trial Participants. Systemic hypothermia after neonatal encephalopathy: outcomes of neo.nEURO.network RCT. Pediatrics 2010;126:e771–8.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Zhou WH,
      2. Cheng GQ,
      3. Shao XM,
      4. et al
      .; China Study Group. Selective head cooling with mild systemic hypothermia after neonatal hypoxic-ischemic encephalopathy: a multicenter randomized controlled trial in China. J Pediatr 2010;157:367–72, 372.e1–3.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Akinbami FO,
      2. Sowunmi A
      . Body temperature in the Nigerian neonate–comparison of axillary and rectal temperatures. Afr J Med Med Sci 1991;20:49–52.
      OpenUrlPubMed
    8. ↵
      1. Bliss-Holtz J
      . Comparison of rectal, axillary, and inguinal temperatures in full-term newborn infants. Nurs Res 1989;38:85–7.
      OpenUrlPubMedWeb of Science
    9. ↵
      1. Kunnel MT,
      2. O'Brien C,
      3. Munro BH,
      4. et al
      . Comparisons of rectal, femoral, axillary, and skin-to-mattress temperatures in stable neonates. Nurs Res 1988;37:162–4, 189.
      OpenUrlPubMedWeb of Science
    10. ↵
      1. Lodha R,
      2. Mukerji N,
      3. Sinha N,
      4. et al
      . Is axillary temperature an appropriate surrogate for core temperature? Indian J Pediatr 2000;67:571–4.
      OpenUrlCrossRefPubMed
    11. ↵
      1. Mayfield SR,
      2. Bhatia J,
      3. Nakamura KT,
      4. et al
      . Temperature measurement in term and preterm neonates. J Pediatr 1984;104:271–5.
      OpenUrlPubMedWeb of Science
    12. ↵
      1. Moen JE,
      2. Chapman S,
      3. Sheehan A,
      4. et al
      . Axillary versus rectal temperatures in preterm infants under radiant warmers. J Obstet Gynecol Neonatal Nurs 1987;16:348–52.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Rekha S,
      2. Lewin S,
      3. Rao SD,
      4. et al
      . Comparison of rectal and axillary temperatures in neonates admitted to a neonatal unit. Indian Pediatr 1993;30:37–9.
      OpenUrlPubMed
    14. ↵
      1. Schiffman RF
      . Temperature monitoring in the neonate: a comparison of axillary and rectal temperatures. Nurs Res 1982;31:274–7.
      OpenUrlPubMedWeb of Science
    15. ↵
      1. Seguin J,
      2. Terry K
      . Neonatal infrared axillary thermometry. Clin Pediatr (Phila) 1999;38:35–40.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Weiss ME,
      2. Richards MT
      . Accuracy of electronic axillary temperature measurement in term and preterm neonates. Neonatal Netw 1994;13:35–40.
      OpenUrlPubMed
    17. ↵
      1. Robinson J,
      2. Charlton J,
      3. Seal R,
      4. et al
      . Oesophageal, rectal, axillary, tympanic and pulmonary artery temperatures during cardiac surgery. Can J Anaesth 1998;45:317–23.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Kendall GS,
      2. Kapetanakis A,
      3. Ratnavel N,
      4. et al
      .; Cooling on Retrieval Study Group. Passive cooling for initiation of therapeutic hypothermia in neonatal encephalopathy. Arch Dis Child Fetal Neonatal Ed 2010;95:F408–12.
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Funding None.

    • Competing interests None.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    Linked Articles

    • Fantoms
      Highlights from this issue
      Martin Ward Platt
      Archives of Disease in Childhood - Fetal and Neonatal Edition 2012; 98 F1-F1 Published Online First: 18 Dec 2012. doi: 10.1136/fetalneonatal-2012-303440