Objective: Although a degree of irreversible damage occurs at the time of primary brain injury, secondary insults during the critical care phase can contribute to further injury that may affect long term outcome. This secondary injury is potentially preventable. Continuous simultaneous monitoring of arterial blood pressure (ABP) and intracranial pressure (ICP) allows for calculation of dynamic indices reflecting cerebral perfusion pressure (CPP) and cerebrovascular autoregulation, two processes that may contribute to secondary brain injury. The purpose of this study was to relate CPP and the Pressure Reactivity Index (PRx), an index of autoregulation, to outcome at 6 month followup.
Design: The study design was prospective, longitudinal, correlational, and predictive.
Population, Sample, Setting, Years: The study population was individuals with acute brain injury, including traumatic brain injury (TBI) or subarachnoid hemorrhage (SAH), undergoing continuous invasive ABP and ICP monitoring in intensive care units at a university affiliated medical center. Participants were 169 adults 16 through 89 years of age (Mean=44.20, SD=18.07), of whom 56.2% (95) were men and 43.8% (74) were women. 42.6% (72) had a diagnosis of SAH, and 57.4% (97) had a diagnosis of TBI. Post-resuscitation Glasgow Coma Scale (GCS_PR) scores ranged from 3-15 (Mean=8.95, SD=4.21). Mortality at 6 months was 20.1%. Participants were enrolled from March 2000 through January 2002.
Concept or Variables Studied Together or Intervention and Outcome Variables:
A. OUTCOME VARIABLES:
A.1) Glasgow Outcome Scale - Extended (GOSE).
A.2) Survival/mortality at 6 months post-injury.
B. PREDICTOR VARIABLES:
B.1) Percent time CPP (ABP-ICP) falls below thresholds (80,75,70, 65, 60, 55 mmHg).
B.2) PRx level. PRx is a correlational measure of the dynamic coupling between the ABP and ICP.
C. COVARIATES:
C.1-5) Age, GCS_PR, diagnosis, ventriculostomy, and cranial bone flap removal.
Methods: Continuous collection of ABP and ICP data was initiated within 24 hours of insertion of monitoring lines, and typically continued for several days. Follow up assessment of outcome was carried out at 6 months using the structured interview form of the GOSE, generally by telephone. Statistical analysis was based on multivariate linear regression (for the GOSE) and multivariate logistic regression (for survival/mortality). The analysis strategy was to first fit a base model regressing outcome on covariates reflecting personal characteristics and initial injury (Age, GCS_PR, diagnostic subgroup, ventriculostomy, and bone flap removal subgroups). Then the CPP and PRx variables were each added to the model to explore the extent to which they improved the predictive ability beyond the base model.
Findings:
A. GOSE AT 6 MONTHS:
A.1) Impact of CPP levels on GOSE:
A.1.a) %CPP < 55 (% time CPP < 55 mmHg), partial beta=-0.202, overall model R-square=0.359, p=0.000, change in R-square=0.036, p=0.005.
A.1.b) %CPP < 60, partial beta=-0.157, overall model R-square=0.344, p=0.000, change in R-square=0.021, p=0.034.
A.2) Impact of PRx on GOSE:
A.2.a) PRx (dichotomized at 0.60), partial beta=-0.192, overall model R-square=0.357, p=0.000, change in R-square=0.034, p=0.006.
B. SIX MONTH SURVIVAL:
B.1) Impact of CPP levels on 6 month survival (logistic regression, statistically controlling for variables in the base model, odds ratios (OR) with respect to 10% difference in time that CPP is less than threshold level):
B.1.a) %CPP<55: OR=0.548 (0.362-0.828), p=0.004.
B.1.b) %CPP<60: OR=0.627 (0.461-0.852), p=0.003.
B.1.c) %CPP<65: OR=0.699 (0.551-0.888), p=0.003.
B.1.d) %CPP<70: OR=0.761 (0.622-0.932), p=0.008.
B.1.e) %CPP<75: OR=0.806 (0.664-0.977), p=0.028.
B.1.f) %CPP<80: OR=0.835 (0.691-1.009), p=0.062.
B.2) Impact of PRx on 6 month survival:
B.2.a) PRx (dichotomized at 0.60): OR=0.051 (0.012- 0.217), p=0.000.
Conclusions: After statistical control of age, diagnosis, GCS_PR, ventriculostomy, and bone flap removal, both PRx and % time CPP < threshold are significant independent predictors of outcome at 6 months. There is clear evidence that episodes of CPP dropping below 60 mmHg are related to more negative outcomes and greater mortality. However, mortality appears to be related to excursions of the CPP below thresholds as high as 75 or 80 mmHg. Participants whose PRx indices were in the top quartile were significantly less likely to survive, and had worse functional outcome.
Implications: Continuous monitoring of the dynamic relations of ABP and ICP, as expressed in the CPP and PRx indices, can help nurses identify and respond to episodes of inadequate perfusion and impaired autoregulation that may be affecting outcome in TBI and SAH patients.
Back to Rehabilitation: Assessment Continued
Back to The Advancing Nursing Practice Excellence: State of the Science