Fluctuations in core body temperature beyond a narrow range are of concern as they indicate changed homeostasis. Thermoregulation remains a key aspect of neonatal care. New ‘non touch’ method of measuring temperature may reduce infection rate and discomfort of neonates. Body temperature of neonates admitted in Neonatal Intensive Care Unit was measured using axillary digital thermometer and a handheld infrared non touch thermometer. The two methods did not agree well (mean difference = -0.5, 95% limits of agreement: [-2.3, 1.2]). The agreement was similar with a negligible difference when patients in open care warmers were excluded (mean difference = -0.6, 95% limits of agreement: [-2.3, 1.1]). As the accuracy is unsatisfactory, the said technology needs further improvement before it can be adapted for routine patient care.

Keywords: Measurement, Methods, Neonate, Temperature.

Accurate temperature recording remains an essential component of neonatal care. The best method to measure temperature should accurately reflect core body temperature and should not be affected by external factors. Axillary temperature measurement is recommended by American Academy of Pediatrics and National Association of Neonatal Nurses [1]. Mercury glass thermometer has been replaced by digital thermometer that is safer and more convenient.

Temperature readings were taken from newborns kept under radiant warmers as well as newborns kept on cot with their mothers. A skin probe was used with a set temperature of 36.5ºC for neonates in NICU who were kept under open care servo-controlled radiant warmer. A room temperature of 25ºC-30ºC was maintained for neonates kept in intermediary care unit on cot. A single reading by each method was recorded 3 hourly in neonates for 7 days. A total of 520 readings were recorded in Celsius unit from 12 neonates. To study the degree of agreement between the two methods, standard methods suggested by Bland and Altman [3] were used and the mean difference with 95% confidence limits was reported for clinical consideration. A difference of 0.5ºC was considered clinically acceptable.

The body temperature measurements by axillary and forehead methods did not agree well (mean difference =
-0.5 0C, 95% limits of agreement: [-2.3, 1.2]) (Fig. 1). The agreement was similar with a negligible difference when patients in open care warmers were excluded (mean difference = -0.6ºC, 95% limits of agreement: [-2.3, 1.1]) (Fig. 2).

Fig. 1 Bland–Altman plot showing comparison of body temperature measurements by axillary and no touch (0 C) of all neonates, and (b) excluding neonates kept in warmer.

Infrared forehead thermeter was reported to be a simple, noninvasive instrument for measuring temperature accurately. Chiappini, et al. [2] reported a good agreement (mean difference = 0.07 0C, 95% limits of agreement: [-0.62, 0.76]) between Infrared forehead thermometry (IRFT) and axillary thermometry using glass mercury thermometer in pediatric population. In contrast, Fortuna, et al. compared IRFT to rectal thermometry in 200 children aged 1 month to 4 y with a mean age of 1.4 y and reported a broader 95% prediction band (on the order of 4ºF) which is clinically not useful. Moreover, the infrared measurements overestimated rectal readings in hypothermia and underestimated rectal measurements in fever [4]. In the current study, we found a broader band for 95% limits of agreement and higher mean difference, Sener, et al. [5] reported similar result with slightly narrower band for 95% confidence limits (mean difference = 0.2ºC, 95% limits of agreement: [-1.2, 1.6]) in adult population [5]. Many factors may contribute to this discrepancy of findings from different studies. Previous studies have been conducted in different settings like NICUs, wards as well as in populations of different ages.