Abstract
INTRODUCTION: Sepsis is a life-threatening condition caused by an exaggerated immune response to infection, causing damage to the body's own tissues and organ dysfunction. The elderly are at higher risk for mortality from sepsis compared to younger adults. Our objective in this study was to evaluate the use of a modified systemic inflammatory response syndrome (SIRS) criteria for patients ≥ 65 years of age including new criteria of reduced heart rate (> 75 rather than 90 beats per minute [bpm]) for patients taking atrioventricular nodal blocking drugs and altered mental status. METHODS: This was a retrospective observational study sampling patients ≥ 65 years of age diagnosed with sepsis. We compared our proposed modified SIRS criteria to the original criteria (heart rate, white blood cell count, respiratory rate, and temperature). Our primary outcome measure was comparing sensitivity and specificity of each model. We performed a regression analysis to evaluate the relationship of each individual criterion and its association with sepsis. Approximately half (47.1%) of the sampled population were taking an atrioventricular nodal blocking drug. RESULTS: Based on a 1:1 case-matched dataset, the modified SIRS criteria yielded a higher sensitivity (98.9%; 95% CI, 98.4-99.2%) compared to the original criteria (97.7%; 97.0-98.2%) in diagnosing sepsis and a lower specificity (14.1%, 12.8-15.5%) compared to the original criteria (20.5%; 18.9-22.1%). The modified model demonstrated an area under the curve (AUC) of 0.797 (95% CI, 0.785-0.809; P < .001), outperforming the original model (AUC 0.764; 0.751-0.778; P < .001). Altered mental status had the second highest individual specificity for sepsis (88.4%; 87.1- 89.6%), and third was the reduced heart rate > 75 bpm for patients using atrioventricular nodal blockers criterion (53.9%; 51.9-55.8%). Among 1,164 sepsis patients receiving atrioventricular nodal blockers, 83 additional cases (7.1%; 5.8-8.8%) were identified solely by the modified heart rate ≥ 75 bpm criterion. CONCLUSION: The modified SIRS criteria is associated with minimally higher but statistically significant rates of identifying sepsis at the cost of reduced specificity. These new criteria identify an additional 1.21% of septic patients in the vulnerable elderly population with a 6.4% reduction in specificity. Overall, sensitivity increased marginally at the expense of specificity with the modified criteria. However, the new criteria of altered mental status and 75bpm for patients taking atrioventricular nodal blocking medications had the second and third highest individual specificity for sepsis, respectively.