Abstract 69: Distinct Grouping of Systolic Blood Pressure Trajectories During the First 24 Hours After Stroke Admission and Their Outcomes
Background: Blood pressure of ischemic stroke patients is a potentially modifiable clinical prognostic factor during acute period. However, BP changes dynamically over time and its temporal variation during acute stage has not received much attention.
Methods: From a total of 3795 acute ischemic stroke patients who arrived within 24 hours after onset, we selected 2723 eligible patients who had more than 5 systolic blood pressure (SBP) measurements during 24 hours after arrival. To predict group SBPs for 8 time-points during the first 24 hours, a measured SBP reading was imputed to the nearest missing point. Trajectory grouping of acute stroke patients was estimated using PROC TRAJ, with delta BIC and prespecified modeling parameters. Early neurological deterioration (END) was captured during admission and recurrent vascular events was collected through a structured telephone interview at 1 years after.
Results: Of the included cases, mean age at onset was 68 ± 13 year-old. NIHSS score at arrival was median 4 [2, 10] and recanalization treatment was done in 598 (22%). Hypertension was diagnosed in 1930 (71%). Based on 48,445 SBP readings during the first 24 hours after arrival, stroke cases were grouped into 5 distinct SBP trajectories as shown in the Figure: Group 1 (low BP), 17%; Group 2 (stable BP), 41%; Group 3 (rapidly stabilized SBP), 11%; Group 4 (higher SBP), 23%; Group 5 (extremely high SBP without stabilization), 8%. Trajectory grouping was independently associated with END and recurrent vascular events (see Figure). Group 1 had low odds of having END (adjusted OR [95% CI]; 0.62 [0.44-0.87], but Group 4 and 5 showed higher probability of having END (1.34 [1.04-1.73] and 1.76 [1.22-2.51]) and recurrent vascular events until 1 year (1.28 [1.00-1.64] and 1.82 [1.29-2.55]). However, Group 3 had comparable risks with Group 2.
Conclusion: It was documented that SBP may successfully grouped into distinct trajectories, which are associated with outcomes after stroke.
Author Disclosures: B. Kim: None. M. Han: None. Y. Cho: None. K. Hong: None. J. Lee: None. J. Kim: None. T. Park: None. J. Lee: None. J. Lee: None. H. Bae: None.
- © 2017 by American Heart Association, Inc.