Identifying adults with sepsis, bedside tools versus administrative data: a cohort study

Abstract Background Our objective was to calculate how well the bedside severity scores qSOFA, NEWS, and SIRS predict 30-day mortality from onset of infection compared to the Sepsis-3 recommended diagnostic criteria of an increase in SOFA score of ≥2 as a consequence of infection. We then assessed the ability of routinely collected administrative data (ICD-10 codes and blood culture sampling) to identify patients with clinical sepsis. The overall purpose is to inform development of a robust proxy measure for sepsis surveillance at scale.Methods This single centre retrospective case note review was set in a district general hospital in Scotland. Adult admissions between 1 st October 2015 and 31 st March 2016 with a blood culture were matched to admissions without a blood culture. The performance characteristics of SOFA, qSOFA, NEWS and SIRS were calculated to predict 30 day mortality. The ability of routinely collected administrative data to identify people with sepsis was assessed using receiver operating characteristic curves.Results This cohort of 958 admissions comprised 479 patients with a blood culture sampled and 479 without. There were 269 (28%) patients with sepsis as per the Sepsis-3 definition, and 361 (37.7%) with infection. 30-day mortality from onset of infection was 19.0% and 7.2% in the sepsis and infection groups respectively (p<0.001). NEWS ≥7 (AUROC 0.71) was a more accurate predictor of 30-day mortality from onset of infection compared to SOFA ≥2 (or Δ2) (AUROC 0.63), qSOFA ≥ 2 (AUROC 0.64) and SIRS≥2 (AUROC 0.65). ICD-10 sepsis codes (A40, A41 & R57.2) were recorded in only 26 (9.4%) sepsis admissions. Blood culture sampling performed better at identifying patients with sepsis (AUROC 0.63) compared to ICD-10 sepsis codes (AURO 0.54) or positive blood cultures (AUROC 0.53).Conclusions NEWS ≥7 was a better predictor of 30-day mortality in patients with infection than SIRS, qSOFA or SOFA. ICD-10 sepsis codes lack sensitivity for reliable sepsis surveillance. Blood culture sampling showed potential for inclusion as part of a clinically relevant proxy marker for sepsis surveillance.