Hospital-onset sepsis and community-onset sepsis in critical care units in Japan: a retrospective cohort study based on a Japanese administrative claims database

Table 3 Sensitivity analyses of in-hospital mortality

	Hospital-onset sepsis	Community-onset sepsis	Unadjusted difference [95% CI]	Adjusted difference [95% CI]
Analysis #1^a	n = 3912	n = 122,364
In-hospital mortality	1610 (41.2)	30,219 (24.7)	16.5 [15.1–17.8]	15.5 [13.8–17.1]
Analysis #2^b	n = 45,054	n = 463,940
In-hospital mortality	16,783 (37.3)	89,017 (19.2)	18.1 [17.7–18.5]	17.2 [16.6–17.9]
Analysis #3^c	n = 118,039	n = 398,085
In-hospital mortality	28,826 (24.4)	78,712 (19.8)	4.6 [4.4–4.9]	3.6 [3.2–4.0]
Analysis #4^d	n = 26,266	n = 372,168
In-hospital mortality	9470 (36.1)	69,662 (18.7)	17.3 [16.8–17.8]	16.0 [15.1–16.8]

Data are presented as n (%). 95% CI indicates confidence interval
^aAnalysis of the subset of patients with ICD-10 codes that directly indicated “sepsis”
^bAnalysis with the definition of hospital-onset as critical care unit admission after the fourth day of hospitalization and community-onset as critical care unit admission on the first or second day of hospitalization
^cAnalysis with the definition of hospital- and community-onset according to whether the ICD-10 diagnoses of infectious diseases were coded at the time of hospitalization
^dAnalysis with the definition of hospital- and community-onset by the date of critical care unit admission and ICD-10 codes at the time of hospitalization

ISSN: 1364-8535