Clinical outcomes during and beyond different COVID-19 critical illness variant periods compared with other lower respiratory tract infections

Table 3 Long-term outcomes in the cohorts

Outcome	Cohort	Outcome description^a	Adjusted estimate (95% CI)^b,c
180-day post-discharge mortality Cox proportional hazards model regression	All COVID-19	2.0 (31/1573)	0.26 (0.17–0.38)
	Wild-type	1.4 (14/989)	0.18 (0.10–0.31)
	Alpha	2.7 (10/368)	0.36 (0.19–0.68)
	Delta	3.1 (4/127)	0.47 (0.17–1.27)
	Omicron	3.4 (3/89)	0.39 (0.12–1.21)
	LRTI	9.2 (149/1613)	Reference
180-day hospital readmission Cox proportional hazards model regression	All COVID-19	21.0 (331/1573)	0.52 (0.45–0.60)
	Wild-type	20.0 (198/989)	0.46 (0.39–0.55)
	Alpha	21.5 (79/368)	0.53 (0.41–0.68)
	Delta	18.9 (24/127)	0.42 (0.28–0.65)
	Omicron	33.7 (30/89)	0.68 (0.47–0.99)
	LRTI	44.2 (713/1613)	Reference
180-day hospital readmission Fine-Gray subdistribution hazard model regression	All COVID-19	21.0 (331/1573)	0.52 (0.45–0.60)
	Wild-type	20.0 (198/989)	0.46 (0.39–0.55)
	Alpha	21.5 (79/368)	0.53 (0.41–0.69)
	Delta	18.9 (24/127)	0.42 (0.27–0.66)
	Omicron	33.7 (30/89)	0.68 (0.48–0.99)
	LRTI	44.2 (713/1613)	Reference
180-day days alive and at home after minus before the critical illness hospitalization Difference-in-differences analysis	All COVID-19	− 21 [− 50, − 9]	33.85 (29.70–38.00)
	Wild-type	− 24 [− 54, − 10]	33.54 (29.04–38.03)
	Alpha	− 19 [− 42, − 7]	36.70 (30.60–42.81)
	Delta	− 13 [− 26, − 5]	40.44 (31.25–49.62)
	Omicron	− 24 [− 107, − 8]	16.07 (2.07–30.08)
	LRTI	− 45 [− 171, − 16]	Reference

CI Confidence interval, COVID-19 Coronavirus disease 2019, LRTI Lower respiratory tract infection
^aPercentages and number of individuals are presented for 180-day post-dicharge all-cause mortality and 180-day all-cause hospital readmission. Median and interquartile number of days are presented for 180-day days alive and at home after minus before the critical illness hospitalization
^bCause-specific Cox-proportional hazards regression modelling was used to model 180-day post-discharge all-cause mortality. 180-day all-cause hospital readmission was analysed with both Cause-specific Cox-proportional hazards regression modelling and Fine-Gray subdistribution hazard regression modelling. Difference-in-differences analyses was used to model 180-day days alive and at home after minus before the critical illness hospitalization
^cModels were adjusted for age, sex, region of birth, yearly disposable income quartile, and all studied comorbidities (cancer, cardiac or cerebrovascular disease, chronic kidney failure, chronic liver disease, chronic lung disease, diabetes, hypertension, immunocompromised state, mental health disorder, neurological disease, and obesity). For 180-day post-discharge mortality, adjustments were only made for age and sex due to the low number of events observed

ISSN: 1364-8535