Utility of core to peripheral temperature gradient using infrared thermography in the assessment of patients with sepsis and septic shock in the emergency medicine department

Abstract Objective Sepsis is a disease affecting microcirculation, reflected in temperature changes between the core and the skin. This study explores correlation of this gradient using infrared thermography (IRT) with mortality and markers of hypoperfusion in patients admitted with sepsis and septic shock and its changes with resuscitation. Design We conducted a prospective, single center observational study on patients admitted in the Department of Emergency Medicine of a tertiary care center in Karnataka, India. These patients were enrolled based on the inclusion criteria and infrared thermography was performed and cases were followed up after 28 days. Adults presenting to the emergency medicine department with clinically suspected sepsis or septic shock were enrolled and infrared thermography was performed. A final sample size of 187 cases was analyzed after retrospectively excluding patients with any exclusion criteria. Interventions Patients underwent thermal imaging of all four limbs on arrival and after 3  hours of resuscitation. Core temperature was measured using a tympanic thermometer. Infrared thermography was performed, and limb temperature was extracted from the images. Other parameters including mean arterial pressure and lactate were recorded and SOFA score was calculated. Outcome measure(s) The temperature gradients were correlated with 7 and 28-day mortality along with markers of hypoperfusion including mean arterial pressure and serum lactate levels. Results A total of 187 patients were included, with a mean SOFA score of 5. Forty four patients (23.5%) died within 7-days. 28-day mortality was 31%. Temperature gradients of core to knee > 8.85°F (p = 0.003) and core to great toe > 12.25°F (p = 0.020) on arrival were found to be correlated with 7-day mortality. Core to knee temperature gradient was found to correlate with 48-hour mortality(p < 0.013). Core to index finger gradient on arrival correlated with vasopressor requirement within 48h (p = 0.020). Core to index finger temperature gradient had a negative correlation with mean arterial pressure (spearman coefficient − 0.286, p = < 0.001), and a positive correlation with lactate (0.281, p = < 0.001), SOFA score (0.242, p = 0.001), qSOFA score (0.167, p = 0.023). Conclusions Core-to-knee and core-to-toe temperature gradients using IRT significantly correlate with 7-day mortality. IRT can be a useful adjunct to predict clinical courses in patients with sepsis and septic shock.