Accuracy Of A Heart Rate Based Estimate Of Core Temperature In Construction Workers In The Summer

Core temperature (Tcore) is a key safety indicator in occupations involving heavy workloads in hot environments. Traditionally, Tcore is measured in the field using an ingestible pill, which is invasive, expensive, and poses risks, limiting widespread use. To overcome these limitations, entities have sought to develop approaches to accurately estimate Tcore. One such approach, developed by researchers at the U.S. Army Research Institute of Environmental Medicine (USARIEM), involves continuous measurement of heart rate (ECTemp). ECTemp provides a reasonably accurate measure of Tcore in agriculture workers. However, the accuracy of ECTemp has not been examined in other workforces at high risk of heat illness, such as outdoor construction. PURPOSE: Test the hypothesis that ECTemp provides an accurate measure of Tcore in construction workers. METHODS: Over eight days in two consecutive summers, 98 workers (39 ± 12 y, 5 women) employed at large commercial construction sites in the Midwest U.S. were instrumented with a heart rate monitor (Polar H10) and ingested a Tcore pill (eCelsius, BodyCAP) before work. Workers then worked a normal workday (547 ± 125 min in duration). ECTemp was calculated post hoc using the most recent USARIEM ECTemp algorithm. Inter-worker accuracy of ECTemp vs. Tcore was independently assessed using mean and peak values. RESULTS: Mean - ECTemp (37.5 ± 0.2 °C) was higher than Tcore (37.3 ± 0.2 °C, p < 0.001). ECTemp was significantly positively correlated with Tcore (r = 0.374, p < 0.001). Bland-Altman plot analyses indicated relatively wide 95% limits of agreement (0.3 to -0.7 °C) and a bias towards ECTemp overestimating Tcore by 0.2 ± 0.2 °C. Peak - ECTemp (37.8 ± 0.3 °C) and Tcore (37.8 ± 0.3 °C) did not differ (p = 0.393). ECTemp was significantly positively correlated with Tcore (r = 0.452, p < 0.001). Bland-Altman plot analyses indicated relatively wide 95% limits of agreement (0.6 to -0.6 °C) yet no systemic bias for ECTemp estimating Tcore (0.0 ± 0.3 °C). CONCLUSIONS: ECTemp slightly overestimated mean Tcore but provided a reasonably accurate estimate of peak Tcore. Before implementing this technology in the construction sector, future work should examine the intra-worker validity of ECTemp. Supported by: R01OH011528, Center for Construction Research and Training, & Turner Construction Foundation