Brachial artery reactivity in patients with severe sepsis: an observational study

Table 4 Hyperemic velocity in survivors versus nonsurvivors: stratified analysis^a

Covariable	Survivors	Nonsurvivors	P ^b
Age
≤ 62	38 (29-52, n = 45	16 (14-22), n = 4	0.005
> 62	39 (32-48), n = 33	25 (21-28), n = 13	0.003
Gender
Men	37 (29-46), n = 43	19 (15-22), n = 6	0.001
Women	42 (30-52), n = 35	26 (17-28), n = 11	0.009
Hypertension
Yes	38 (27-49), n = 44	25 (16-28), n = 15	0.001
No	39 (30-51), n = 34	17 (12-21), n = 2	0.027
Diabetes mellitus
Yes	38 (26-53), n = 18	17 (15-22), n = 9	< 0.001
No	39 (30-48), n = 60	28 (26-34), n = 8	0.054^c
Hyperlipidemia
Yes	36 (26-46), n = 26	22 (16-28), n = 8	0.024
No	39 (30-51), n = 52	25 (16-27), n = 9	< 0.001
Active smoking
Yes	38 (30-53), n = 20	16 (12-28), n = 3	0.022
No	39 (30-48), n = 58	25 (17-28), n = 14	< 0.001
Coronary artery disease
Yes	39 (32-46), n = 13	19 (14-24), n = 4	0.007
No	38 (29-51), n = 65	25 (17-28), n = 25	0.001
Pressors
Yes	36 (25-44), n = 19	19 (13-28), n = 8	0.03
No	39 (30-51), n = 59	25 (21-28), n = 9	0.001
Blood pressure
MAP ≤ 80 mm Hg	37 (29-46), n = 37	17 (14-27), n = 11	0.001
MAP > 80 mm Hg	41 (30-51), n = 41	26 (22-29), n = 6	0.018
Charlson index
≤ 3	37 (29-51), n = 47	25 (17-34), n = 8	0.020
> 3	42 (30-48), n = 31	22 (15-27), n = 9	0.001

MAP, mean arterial pressure at the time of brachial artery reactivity measurements; n/a, not applicable. ^aContinuous variables were dichotomized according to their median value (in the severe sepsis patients alone) for this analysis. ^bP value refers to the comparison of HV between survivors and nonsurvivors within the specified subgroup. The borderline statistical significance in this group suggested an interaction between diabetes mellitus and HV in predicting hospital mortality was possible (see Methods). However, the test for interaction [42] was insignificant (P > 0.20).

ISSN: 1364-8535