Abstract Background There is conflicting evidence on association between quick sequential organ failure assessment (qSOFA) and sepsis mortality in ICU patients. The primary aim of this study was to determine the association between qSOFA and 28-day mortality in ICU patients admitted for sepsis. Association of qSOFA with early (3-day), medium (28-day), late (90-day) mortality was assessed in low and lower middle income (LLMIC), upper middle income (UMIC) and high income (HIC) countries/regions. Methods This was a secondary analysis of the MOSAICS II study, an international prospective observational study on sepsis epidemiology in Asian ICUs. Associations between qSOFA at ICU admission and mortality were separately assessed in LLMIC, UMIC and HIC countries/regions. Modified Poisson regression was used to determine the adjusted relative risk (RR) of qSOFA score on mortality at 28 days with adjustments for confounders identified in the MOSAICS II study. Results Among the MOSAICS II study cohort of 4980 patients, 4826 patients from 343 ICUs and 22 countries were included in this secondary analysis. Higher qSOFA was associated with increasing 28-day mortality, but this was only observed in LLMIC ( p < 0.001) and UMIC ( p < 0.001) and not HIC ( p = 0.220) countries/regions. Similarly, higher 90-day mortality was associated with increased qSOFA in LLMIC ( p < 0.001) and UMIC ( p < 0.001) only. In contrast, higher 3-day mortality with increasing qSOFA score was observed across all income countries/regions ( p < 0.001). Multivariate analysis showed that qSOFA remained associated with 28-day mortality (adjusted RR 1.09 (1.00–1.18), p = 0.038) even after adjustments for covariates including APACHE II, SOFA, income country/region and administration of antibiotics within 3 h. Conclusions qSOFA was independently associated with 28-day mortality in ICU patients admitted for sepsis. In LLMIC and UMIC countries/regions, qSOFA was associated with early to late mortality but only early mortality in HIC countries/regions. Graphical Abstract
Regional citrate anticoagulation (RCA) is the preferred mode of anticoagulation for continuous renal replacement therapy (CRRT). Conventional RCA-CRRT citrate dose ranges from 3 to 5 mmol/L of blood. This study explored the effectiveness of an RCA protocol with lower citrate dose and its impact on citrate-related complications.This prospective observational study compared two RCA-CRRT protocols in the intensive care unit. RCA Protocol 1 used an initial citrate dose of 3.0 mmol/L while Protocol 2 started with 2.5 mmol/L. The citrate dose was titrated by sliding scale to target circuit-iCa 0.26-0.40 mmol/L. Calcium was re-infused post-dialyzer and titrated by protocol to target systemic-iCa 1.01-1.20 mmol/L.Two hundred RCA-CRRT sessions were performed (81 Protocol 1; 119 Protocol 2). The median age was 65.4 years and median APACHE-II score was 23. Citrate dose for Protocol 1 was significantly higher than Protocol 2 in the first 12 h. The circuit clotting rate was similar in both arms (Protocol 1: 9.9%; Protocol 2: 9.2%; P = 0.881). With Protocol 2, circuit-iCa levels were 2.42 times more likely to be on target (P = 0.003) while the odds of hypocalcaemia was 4.67 times higher with Protocol 1 (P < 0.001). There was a wider anion gap was noted with Protocol 1, which suggests a propensity for citrate accumulation with higher citrate exposure.The RCA protocol with a lower initial citrate dose of 2.5 mmol/L blood had less citrate-related complications with no loss of efficacy. A more precise RCA prescription at the start of treatment avoids unnecessary citrate exposure and improves safety.
Continuous renal replacement therapy (CRRT) is the treatment of choice for acute renal failure in critically ill patients. Our study compared the efficacy and safety of regional citrate anticoagulation (RCA) with acid citrate dextrose A (ACD-A) against Prismocitrate 10/2 (Gambro). Use of Prismocitrate 10/2 eliminates the need for a separate infusion pump for ACD-A as it doubles as replacement fluid and anticoagulant. This removes a source of error in the calculation of fluid balance. The combination of tri-sodium citrate and citric acid in Prismocitrate 10/2 avoids the metabolic alkalosis associated with ACD-A [1,2].