Validation and enhancement of two predictive models evaluating physiological strain during physical work while wearing personal protective equipment.

The prediction of physiological strain is essential for the safety of personnel in high-risk environments especially when wearing personal protective equipment (PPE). This study aimed to develop a usability-enhanced variant of the Heat Strain Decision Aid (HSDA), named HSDA-FL (FAME Lab), by implementing automatic workload estimations, alternative environmental inputs, and a revised initialization delay logic. The second aim was to evaluate whether these modifications preserved the original model's validity. The third aim was to evaluate the performance of HSDA, HSDA-FL, and the PHS-FL models while wearing chemical, biological, radiological, and nuclear (CBRN) PPE. Eleven acclimatized participants completed a 40-min simulated CBRN reconnaissance scenario involving walking and inspection tasks. The HSDA-FL was tested with both the original and the revised delay logic, demonstrated significantly lower residual errors than the original HSDA (p<0.001), with the revised delay logic variant achieving the highest accuracy (MAE=0.151°C, RMSE=0.191°C, bias=0.011°C). In contrast, the PHS-FL performed significantly worse than all HSDA-based models (p<0.001). These findings confirm that HSDA-FL maintains and slightly exceeds the predictive validity of the original HSDA while improving usability, whereas PHS-FL was not deemed sufficiently reliable for use with CBRN clothing.