Mortality in sepsis and septic shock in Europe, North America and Australia between 2009 and 2019- results from a systematic review and meta-analysis

Abstract

Background: Sepsis and septic shock remain drivers for mortality in critically ill patients. The heterogeneity of the syndrome hinders the generation of reproducible numbers on mortality risks. Consequently, mortality rates range from 15 to 56%. We aimed to update and extend the existing knowledge from meta-analyses and estimate 30- and 90-day mortality rates for sepsis and septic shock separately, stratify rates by region and study type and assess mortality rates across different sequential organ failure assessment (SOFA) scores.

Methods: We performed a systematic review of articles published in PubMed or in the Cochrane Database, between 2009 and 2019 in English language including interventional and observational studies. A meta-analysis of pooled 28/30- and 90-day mortality rated separately for sepsis and septic shock was done using a random-effects model. Time trends were assessed via Joinpoint methodology and for the assessment of mortality rate over different SOFA scores, and linear regression was applied.

Results: Four thousand five hundred records were identified. After title/abstract screening, 783 articles were assessed in full text for eligibility. Of those, 170 studies were included. Average 30-day septic shock mortality was 34.7% (95% CI 32.6-36.9%), and 90-day septic shock mortality was 38.5% (95% CI 35.4-41.5%). Average 30-day sepsis mortality was 24.4% (95% CI 21.5-27.2%), and 90-day sepsis mortality was 32.2% (95% CI 27.0-37.5%). Estimated mortality rates from RCTs were below prospective and retrospective cohort studies. Rates varied between regions, with 30-day septic shock mortality being 33.7% (95% CI 31.5-35.9) in North America, 32.5% (95% CI 31.7-33.3) in Europe and 26.4% (95% CI 18.1-34.6) in Australia. A statistically significant decrease of 30-day septic shock mortality rate was found between 2009 and 2011, but not after 2011. Per 1-point increase of the average SOFA score, average mortality increased by 1.8-3.3%.

Conclusion: Trends of lower sepsis and continuous septic shock mortality rates over time and regional disparities indicate a remaining unmet need for improving sepsis management. Further research is needed to investigate how trends in the burden of disease influence mortality rates in sepsis and septic shock at 30- and 90-day mortality over time.

Keywords: Meta-analysis; Mortality; SOFA; Sepsis; Septic shock; Trend.

Fig. 1

PRISMA flowchart of studies. PRISMA flowchart describing the process of selecting studies eligible for meta-analysis

Fig. 2

Joinpoint estimation of 30-day septic shock mortality. Results from the estimation of pooled 30-day septic shock mortality rates per year. The Joinpoint analysis shows the time point when trends in mortality rates change. A trend change of 30-day sepsis mortality was found in 2011

Fig. 3

30-day/90-day septic shock/sepsis mortality stratified by study type. Comparison of pooled mortality rates derived from RCTs and prospective and retrospective cohort studies, showing rates and 95% confidence intervals. Mortality rates per study type are reported for a 30-day septic shock, b 90-day septic shock, c 30-day sepsis and d 90-day sepsis

Fig. 4

30-day/90-day septic shock/sepsis mortality stratified by geographic region. Comparison of pooled mortality rates per region, showing rates and 95% confidence intervals. Studies were assigned to the regions Europe, North America or Australia based on primary study site. Mortality rates per region are reported for a 30-day septic shock, b 90-day septic shock, and c 30-day sepsis

Fig. 5

Joinpoint estimation of 30-day sepsis mortality. Results from the estimation of pooled 30-day sepsis mortality rates per year. Joinpoint analysis shows the time point when trends in mortality rates change. One trend change of 30-day sepsis mortality was found in 2015

Fig. 6

Linear regression of SOFA score and mortality rates. A change in average mortality rates with increasing SOFA scores was found for a 30-day septic shock, b 90-day septic shock, c 30-day sepsis and d 90-day sepsis. Graphs show positive correlations of mortality rates and SOFA scores for a 30-day septic shock, b 90-day septic shock and c 30-day sepsis