Table 3.

Risk of death by initial dosing strategy of thromboprophylaxis

Initial dosing strategy of thromboprophylaxis	No. of patients	Events/person-days	IR per 1.000 person-days (95% CI)	HR (95% CI) of death ≤ 28 days
				Univariable model	Multivariable modela	Multivariable imputed modelb
High dosec	37	5/923	5.4 (2.3–13.0)	0.31 (0.12–0.82)	0.33 (0.13–0.87)	0.30 (0.11–0.81)
Medium dosed	48	12/1182	10.2 (5.8–17.9)	0.59 (0.30–1.16)	0.88 (0.43–1.83)	0.87 (0.42–1.82)
Low dosee	67	26/1453	17.9 (12.2–26.3)	1.00 (Ref.)	1.00 (Ref.)	1.00 (Ref.)

Risk of death during the first 28 days among 152 patients admitted to the intensive care unit due to COVID-19 at Södersjukhuset, Stockholm, March 6 to April 30, 2020, by initial dosing strategy with tinzaparin/dalteparin as thromboprophylaxis

CI, confidence interval; IR, Incidence Rate; HR, Hazard Ratio

^aAdjusted for sex, age (continuously), body mass index (</≥ 30 kg/m² and missing [n = 6]), invasive mechanical ventilation (yes/no), and Simplified Acute Physiology Score III (continuously)

^bAdjusted like the multivariable model but with body mass index imputed due to missing values (n = 6), and flexibly modeled with restricted cubic splines at three knots over the percentile (10^th, 50^th, and 90^th) distribution of body mass index in the population

^cTinzaparin, ≥ 175 IU/kg of body weight per daily, or dalteparin, ≥ 200 IU/kg of body weight daily

^dTinzaparin, > 4500 IU daily to < 175 IU/kg of body weight daily, or dalteparin, > 5000 IU daily to < 200 IU/kg of body weight daily

^eTinzaparin, 2500–4500 IU daily, or dalteparin, 2500–5000 IU daily