From paper to practice: the use of cortisol in ambulatory field research to examine associations between police lethal force decision-making and mental health

Abstract

Acute and chronic operational and organizational stress dysregulate associated neuroendocrine responses and have been shown to increase health risks and occupational performance errors. The current paper addressed a research gap comparing biopsychosocial stress measures and its interaction on police lethal force errors during realistic critical incident (CI) scenarios. A combined sample (n=101) of frontline and tactical police officers completed externally valid, live-action scenarios requiring lethal force decisions paralleling field experiences. Officers provided diurnal and reactive (pre-to-post CI) salivary cortisol samples, and self-reported mental health and stress symptoms. Higher cortisol awakening response significantly predicted increased odds of committing a lethal force error by 6.45%. However, further regressions deemphasizing influential outliers via logarithmic transformations (as per Psychoneuroendocrinology expert consensus guidelines⁠—Stalder et al., 2016) revealed a no longer significant association. Reactive cortisol did not predict lethal force error across all subsamples. Consistent with previous literature, several significant relationships between mental health symptoms and cortisol were observed. The use of cortisol in field research presents a number of collection, control, and cost challenges. Furthermore, outliers are necessary to include in analyses to represent the potential cortisol dysregulation in police officers deemed ‘fit for duty’ by their respective organizations. The current findings suggest that diurnal cortisol is heavily influenced by outlier values, and reactive cortisol may be too subtle to predict lethal force errors during acute stress. While purposeful for predicting long-term health outcomes, pursuit of alternative measures may be more suitable for assessing acute operational errors in field research.