Objective Sepsis Identification Tool to Study Sepsis Epidemiology

Globally, sepsis affects an estimated 31.5 million people and accounts for 5.3 million deaths each year. The incidence of sepsis is increasing and is now a more common cause for hospital admission than acute myocardial infarction and stroke.

However this apparent improvement in sepsis outcomes may also be due to changes in definition and reporting of sepsis. The latest Sepsis-3 definition outlines sepsis as life-threatening organ dysfunction due to dysregulated host response to infection. Trends on sepsis incidence and outcome were based on previous definitions of sepsis. Furthermore, changes in coding practices and documentation for hospital claims may cause misleading changes in incidence and outcomes. Electronic health record clinical data provides more reliable data on sepsis epidemiology than diagnostic coding accounts for. Our understanding of sepsis epidemiology is based on studies from high income western countries. Even within these countries, health expenditure and resources vary which may affect sepsis outcomes. Current data on the burden of sepsis from the rest of the world is insufficient. Hong Kong is well poised to provide these lacking data on the epidemiology of sepsis. More than 90% of inpatient care is provided by the public healthcare system and would be captured by the electronic Clinical Management System. The purpose of this study is to develop an objective sepsis identification tool using electronic health records to study the epidemiology of sepsis in Hong Kong.

Inclusion criteria

• January 2018 to February 2018 adult inpatients at Prince of Wales Hospital
• 1st microbiological culture tests during hospital admission

Exclusion criteria

• duplicated microbiological sampling
• missing values for sequential organ failure assessment score calculation
• microbiological tests were for screening only (methicillin-resistant staphylococcus aureus swabs, vancomycin resistant Enterococci rectal swab)
• antibiotics not started 2 days before or after first microbiological sampling
• antibiotics duration less than 4 days unless death
• not given antibiotics on discharge if given less than 4 days antibiotics in hospital