Quality of Life After Intracerebral Hemorrhage
Results of the Factor Seven for Acute Hemorrhagic Stroke (FAST) Trial
Background and Purpose— Neurological impairment and physical disability are frequent and important complications of stroke with serious consequences for health-related quality of life (HRQOL). Little data exist, however, on the risk factors for poor HRQOL after intracerebral hemorrhage, the deadliest and most disabling form of stroke.
Methods— Factor Seven for Acute Hemorrhagic Stroke (FAST) was an international, randomized, double-blind, placebo-controlled trial conducted between May 2005 and February 2007 at 122 sites in 22 countries. All patients were followed for 3 months after stroke onset and HRQOL was assessed using the EuroQoL. Multivariate stepwise logistic regression was used to identify predictors of poor HRQOL based on demographic and clinical baseline characteristics and in-hospital complications.
Results— Six hundred fifty-seven patients survived until 3 months after stroke onset, and 621 (95%) completed the EuroQoL. Two percent had a utility score <0 (HRQOL worse than death), 15% a utility score <0.2, 32% a utility score <0.5, and 87% a score <0.87 (average score in the general population). At the other end of the scale, 13% had a utility score of 1 (perfect HRQOL). Independent predictors of poor HRQOL were advanced age (OR, 1.80; P<0.0001), higher baseline National Institutes of Health Stroke Scale score (OR, 1.11; P<0.0001), higher systolic blood pressure (OR, 1.05; P=0.0039), higher baseline intracerebral hemorrhage volume (OR, 1.11; P=0.015), deep (versus lobar) hematoma location (OR, 3.05; P=0.003), and increase in neurological deficit in first 72 hours after ICH onset (Δ Glasgow Coma Scale ≥2 or Δ National Institutes of Health Stroke Scale ≥4; OR, 2.04; P=0.006). The model explained a large amount of the variation in the utility score (C-statistic 0.77).
Conclusion— The vast majority of survivors after intracerebral hemorrhage have very poor HRQOL. Critical care interventions designed to control blood pressure or prevent neuroworsening may improve HRQOL in intracerebral hemorrhage survivors.
Neurological impairment and physical disability are frequent complications of stroke with serious consequences for the long-term health-related quality of life (HRQOL).1–4 HRQOL after stroke can be drastically changed given its sudden impact and the inability of stroke victims and their families to cope with its complications.5 Unlike the clinical outcomes typically measured in stroke trials, HRQOL reflects the impact of the disease from the perspective of the patient and thus provides a more holistic picture of disease impact and recovery. From a clinical point of view, it is important to understand stroke survivors’ perception of their HRQOL and its influencing factors to develop strategies that can improve HRQOL for patients with stroke.
Studies on the long-term HRQOL in stroke survivors have identified age,2,4,6 gender,2,4 depression3,4,6–8 baseline cognitive impairment,9 disability,3,4,6–8 aphasia,7 and poor social network3,8 as determinants of poor HRQOL. These studies, however, have certain limitations. The study populations considered have typically been restricted to hospitalized subjects,3,9–12 stroke rehabilitation units,6,8 specific age groups,7,12,13 or specific stroke types.9,10 To our knowledge, no studies have yet examined the risk factors for poor HRQOL specifically after intracerebral hemorrhage (ICH), the most disabling form of stroke. Furthermore, the design of many prior studies has been cross-sectional, assessing patients at various time points after their stroke; hence, these studies evaluated HRQOL at different stages of stroke recovery. Finally, in most studies, determinants of poor HRQOL have only been examined in univariate analyses, preventing independent determinants of HRQOL to be identified. Only 2 fairly recent studies have performed multivariate analyses of predictors of poor long-term HRQOL after stroke, and both studied a mixture of patients with ischemic stroke and patients with ICH.2–4 One of these studies identified advanced age, female gender, and cognitive impairment as risk factors for poor HRQOL one year after stroke onset,4 whereas the other found initial National Institutes of Health Stroke Scale (NIHSS) score, spatial neglect, and low socioeconomic status to be significant predictors in addition to age and gender.2
In this article, we report the results of a prospective assessment of HRQOL in a recent multinational, clinical trial investigating the effect of recombinant Factor VIIa for the treatment of spontaneous ICH.14 In this trial, HRQOL was assessed at 3 months after the ICH according to the EuroQoL questionnaire.15 Because no statistically significant effect of recombinant Factor VIIa was identified on HRQOL, we report the HRQOL in all ICH survivors and examine its predictors at 3 months after ICH onset.
Subjects and Methods
The Factor Seven for Acute Hemorrhagic Stroke (FAST) trial was a randomized, multicenter, double-blind, placebo-controlled trial conducted between May 2005 and February 2007 at 122 sites in 22 countries.14 Informed consent was obtained from the patient or a legally acceptable surrogate in all cases. The trial was approved by local Institutional Review Boards and reviewed by national regulatory authorities as applicable.
The study population was defined as patients aged ≥18 years with spontaneous ICH documented by CT scan within 3 hours of symptom onset. Patients with any of the following characteristics were excluded from the study: Glasgow Coma Scale score ≤5; surgical hematoma evacuation planned within 24 hours of admission; secondary ICH due to aneurysm, arteriovenous malformation, trauma, or other causes; known oral anticoagulant use or thrombocytopenia; history of coagulopathy; acute sepsis, crush injury, or disseminated intravascular coagulation; pregnancy; prior disability (pre-ICH modified Rankin Scale [mRS] score >2); and known thrombotic or vaso-occlusive disease (ie, angina, claudication, deep vein thrombosis, or cerebral or myocardial infarction) within the last 30 days before ICH onset.
HRQOL was assessed using the EuroQoL questionnaire.15 The EuroQoL has been shown to be a valid measure of HRQOL after stroke,15 can be converted into a single numeric score, and has been validated in numerous countries.16 In addition to a Visual Analog Scale (VAS) score (which takes a value of 100 for perfect health and 0 for dead), the EuroQoL consists of 5 subscales that assess mobility, self-care, usual activities, pain, and anxiety/depression. Each subscale is scored based on 3 levels of severity (no problems, moderate problems, extreme problems). The EuroQoL index integrates the ratings on the 5 subscales into a single score. A utility score can be calculated by using population-based preference weights for each subscale, and in this study, we used the weights obtained from the US population.17,18 Utility scores express HRQOL quantitatively as a fraction of perfect health with a score of 1 representing perfect health, a score of 0 representing death, and negative scores (minimum score −0.109) representing health states considered worse than death. In our analysis of predictors of HRQOL, we focused on the utility score because it directly inquires about dimensions of health relevant to the patients with ICH, accounts for patient preferences, and incorporates health states considered worse than death.
Univariate and multivariate analysis was performed to identify predictors of poor HRQOL at 3 months after ICH onset. Predictors were individually tested in a univariate model using 5 blocks of prognostic variables: (1) demographics (age, race, and gender); (2) comorbidity (diabetes, hypertension, heart disease, depression); (3) clinical status at admission (neurological deficit as measured by Glasgow Coma Scale and the NIHSS, baseline systolic blood pressure [mm Hg], baseline blood glucose [mmol/L], body temperature [°C], level of white blood cells [109/L], and fibrinogen [g/L]); (4) ICH severity as measured by CT scan (ICH volume, intraventricular hemorrhage volume, edema volume, total lesion volume, and hematoma location [categorized into nonlobar versus lobar]); and (5) clinical status during hospital stay (increase in neurological deficit defined as a decrease in Glasgow Coma Scale score ≥2 or increase in NIHSS score ≥4 from baseline assessment to 24, 48, or 72 hours after ICH onset), development of infections (reported as serious adverse events), management of edema by mannitol or hypertonic saline, and support by mechanical ventilation. Poor HRQOL was defined as a dichotomous outcome. Given the multiethnic nature of the trial population that we studied, we used the distribution of the utility scores for the entire trial population to identify a threshold for poor HRQOL as opposed to an average utility score reported in any specific population. To examine independent predictors of poor HRQOL, we undertook multivariate regression analysis. We sequentially examined the 5 blocks of prognostic variables starting with the demographic block only and then adding the 4 additional blocks one at a time. Variables significant at a 10% significance level were included in the next model. When more than one variable measuring the same construct (ie, lesion volume, admission clinical status) within a given block showed a significant univariate association, only the variable with the strongest association with poor HRQOL in the multivariable model was retained. The 3 demographic variables (age, gender, and race) were included in all 5 blocks regardless of their individual significance. The C-statistic was computed to estimate the model’s goodness of fit. For all analyses, we used the entire intention-to-treat population because no significant treatment effect of recombinant Factor VIIa was observed on HRQOL at Day 90.
Baseline characteristics of the patients are summarized in Table 1. A total of 841 patients with ICH were enrolled and randomized, and 821 received study drug. A total of 657 patients survived until 3 months after stroke onset, and 621 (95%) completed the EuroQoL and were included in the present analysis. Among the 621 completers, 593 (90% of survivors) completed the VAS score and 621 (100%) completed the 5 subscales to estimate the utility score. Mean age was 64 years (range, 23 to 97 years); 60% were male; and 68% were white, 21% were Asian, and 9% were black. The 36 survivors who failed to complete the EuroQOL were more often male and nonwhite than those who were included in the analysis (Table 1). As previously reported, the main outcome measure of the trial, the proportion of patients who were dead or severely disabled according to the mRS at 90 days, was not significantly different among the 3 treatment groups in the trial.14 Similarly, the distribution of outcomes on the mRS and median Barthel Index scores were similar among the 3 treatment groups in the trial.
At Day 90±7 after ICH onset, the mean VAS score was 62.2 (SD ±25) and the mean utility score was 0.62 (SD ±0.3). The distribution of utility scores revealed 2 peaks, one above and one below a score of 0.5 (Figure 1). For this reason, a utility score of 0.5 was used as the threshold to define poor HRQOL in subsequent univariate and multivariate analyses. Two percent had a utility score <0 (HRQOL worse than death), 15% a utility score <0.2, 32% a utility score <0.5, and 87% a score <0.87 (average score in general population).18 In the other end of the scale, 13% had a utility score of 1 (perfect HRQOL).
Patients with a utility score <0.5 were significantly older; had worse NIHSS and Glasgow Coma Scale scores at baseline, higher baseline systolic blood pressure, larger baseline ICH, edema, and total lesion volumes; more often experienced an increase in neurological deficit during first 72 hours after ICH onset; more frequently developed infections; or underwent edema management or mechanical ventilation (Table 2). The relationship between HRQOL and disability at Day 90 was particularly pronounced. The mean VAS score decreased from 86.9 (SD 14.8) at mRS level 0 to 30.4 (SD 19.0) at mRS level 5, whereas the mean utility score decreased from 0.9 (SD 0.1) at mRS 0 to 0.1 (SD 0.1) at mRS 5 (Figure 2). The correlation coefficient between level of functional outcome on the mRS and VAS (r=−0.53) and utility scores (r=−0.76) reported at Day 90 were all statistically significant (P<0.0001).
When adjusting for potentially relevant predictors of poor HRQOL, the multivariate analysis identified advanced age, higher baseline NIHSS score, higher baseline systolic blood pressure, higher baseline ICH volume, deep hematoma location, and neuroworsening as independent significant predictors of poor HRQOL (Table 3). All variables were consistently identified in the 5 models as the blocks in which they were assigned were added to the multivariate model. Advanced age was a significant predictor in all models. The final model explained a very large proportion of the variation in utility score (C-statistic=0.77).
This study provides the first large, global assessment of HRQOL after ICH and its clinical predictors using the EuroQoL, which is a well-validated and widely used HRQOL instrument in clinical practice. Patients in our trial were recruited in 22 countries across Europe, Asia, Australia, North America, and South America. We found the overall quality of life observed after ICH to be substantially lower than that observed in the general population.19,20 Approximately one third of ICH survivors had a utility score <0.5, 16% a utility score <0.2, and 2% a utility score worse than death. We found baseline demographic and clinical characteristics (age, initial neurological deficit, systolic blood pressure, ICH volume, and deep ICH) and neuroworsening during the acute phase of treatment to all be independent predictors of poor HRQOL.
The association of poor quality of life with deep as opposed to lobar ICH is novel. This may reflect the ability of basal ganglia hemorrhages to cause more severe motor deficits or to undercut and impair multiple cortical functions such as language, sensation, and spatial perception. Alternately, this observation may reflect the fact that lobar hemorrhage is more often related to amyloid angiopathy as opposed to hypertensive small vessel disease, which may in some way be associated with more severe local tissue injury.
The association of elevated admission blood pressure with poor quality of life is also novel. Severe hypertension is common during the acute phase of stroke as part of a generalized stress response, and increased blood pressure on admission has been linked to an increased risk of death and disability after ICH.21–24 To our knowledge, however, hypertension has not previously been associated with poor quality of life among ICH survivors. The association between elevated blood pressure and poor outcome after ICH suggests that uncontrolled hypertension may exacerbate ICH-related tissue injury, leading to more severe deficits and worse outcome. Alternately, this association may be noncausal. At the very least, our findings suggest that quality of life should be used as an end point for studies of blood pressure management in ICH.
The association of early neuroworsening with poor quality of life was somewhat surprising given the fact that the severity of the acute neurological deficit was accounted for in the model. Neurological deterioration after ICH is most often attributed to ICH growth in the acute phase (<6 hours) and evolving perihematomal tissue injury, edema, and mass effect thereafter.25 It would seem most likely that neuroworsening results in worse quality of life by producing a more severe neurological deficit. This finding supports the notion that early in-hospital worsening, regardless of the specific mechanism of injury, should receive more emphasis as a target for critical care interventions.
Advanced age, severity of initial neurological deficit, and various measures of CT lesion volume were also associated with poor HRQOL in our study and have previously been identified as risk factors for poor HRQOL in patients with ischemic stroke. In the 2-year follow-up assessment of HRQOL in the Northern East Melbourne Stroke Incidence Study trial,2 the authors identified handicap (measured by the London Handicap scale), neurological impairment (measured by the NIHSS), anxiety and depression, disability (measured by Barthel Index), institutionalization, dementia, and age to be associated with poor HRQOL. All of these factors were evaluated concurrently with the assessment of HRQOL. When considering baseline characteristics in relation to HRQOL reported 2 years after stroke onset, the authors found age, female gender, initial NIHSS score, neglect, and low socioeconomic status to be significant independent predictors of poor HRQOL. In the 1- and 3-year follow-up assessment of HRQOL in stroke subjects recruited from the South London Stroke Register,4 the authors identified female gender, manual workers, diabetes, right hemispheric lesions, urinary incontinence, and cognitive impairment to be significant independent predictors of poor physical health and being Asian, age (< 65 years), ischemic heart disease, and cognitive impairment to be independent predictors of poor mental health.
Given the frequent use of the Barthel Index and the mRS in clinical practice of patients with stroke, these scales are frequently used as outcome measures and proxies for HRQOL.26,27 Measures of functional outcome may also be more sensitive to differences in health among patients with stroke and development in health status over time. In the ICH population, it is therefore of interest to examine the extent to which the mRS represents a reasonable proxy for HRQOL. Our analysis of the relationship between the level of functional outcome observed at Day 90 and the VAS score and the utility score revealed a consistent decrease in overall quality of life with higher levels of disability. Given these observations, it would appear reasonable to use the mRS as a proxy for HRQOL when direct measures of HRQOL are missing.
A major strength of our study is the prospective assessment of HRQOL in a large global ICH population with a high level of follow-up. By evaluating potential risk factors before HRQOL assessment and analyzing these in multivariate regression, it was possible to identify independent predictors of poor HRQOL rather than simple associations. We selected a range of predictors easily observable at baseline and in the acute phase of treatment to illustrate the potential for the treating physicians to affect HRQOL in the months after ICH. The independent significance of clinical variables at baseline and acute phase of treatment indicates the potential for the treating physician to affect HRQOL in the acute phase of the hospital stay.
There are certain limitations to our study that must be considered. First, the patients in the trial represent a selected group of patients with ICH, in particular given the short timeline from symptom onset to CT scan, and hence it is questionable to what extent our results can be generalized to the entire ICH population. Second, we used a generic measure of HRQOL designed to measure quality-of-life outcomes for any disease or treatment as opposed to disease-specific dimensions. Although the EuroQoL has been shown to be a valid measure of HRQOL after stroke,15 the instrument may nevertheless not reveal the full spectrum of symptoms and impairments associated with stroke, in particular psychological complications.28 However, in the conduct of international stroke trials, there are few scales better suited for quality-of-life evaluation than the EuroQOL. Third, there are important potentially relevant nonclinical risk factors that were not assessed in our study, including socioeconomic status of the patients, the social support available, the role of caregivers, and the overall quality of the stroke care delivered.
In summary, we found very poor HRQOL to be a common outcome after ICH. Quality of life deserves more attention as an end point in ICH epidemiological studies and clinical trials, and critical care interventions focused on control of blood pressure and the prevention of neuroworsening hold promise as potential strategies to improve HRQOL among ICH survivors.
Source of Funding
This trial was funded by Novo Nordisk A/S.
M.C.C. and J.-M.F. are employees of Novo Nordisk A/S. S.M. was principal investigator of the FAST Trial and has received research support, consulting fees, and speaking honoraria from Novo Nordisk.
- Received September 30, 2008.
- Accepted October 27, 2008.
Herrmann N, Black SE, Lawrence J, Szekely C, Szalai JP. The Sunnybrook Stroke Study: a prospective study of depressive symptoms and functional outcome. Stroke. 1998; 29: 618–624.
Sturm JW, Donnan GA, Dewey HM, Macdonell RA, Gilligan AK, Srikanth V, Thrift AG. Quality of life after stroke: the Northern East Melbourne Stroke Incidence Study (NEMESIS). Stroke. 2004; 35: 2340–2345.
King RB. Quality of life after stroke. Stroke. 1996; 27: 1467–1472.
Patel MD, McKevitt C, Lawrence E, Rudd AG, Wolfe CD. Clinical determinants of long-term quality of life after stroke. Age Ageing. 2007; 36: 316–322.
Ahlsiö B, Britton M, Murray V, Theorell T. Disablement and quality of life after stroke. Stroke. 1984; 15: 886–890.
Niemi M, Laaksonen R, Kotila M, Waltimo O. Quality of life 4 years after stroke. Stroke. 1988; 19: 1101–1107.
Astrom M, Asplund K, Astrom T. Psychosocial function and life satisfaction after stroke. Stroke. 1992; 23: 527–531.
Greveson GC, Gray CS, French JM, James OF. Long-term outcome for patients and carers following hospital admission for stroke. Age Ageing. 1991; 20: 337–344.
Indredavik B, Bakke F, Slørdahl SA, Rokseth R, Håheim LL. Stroke unit treatment improves long-term quality of life: a randomized controlled trial. Stroke. 1998; 29: 895–899.
Wilkinson PR, Wolfe CD, Warburton FG, Rudd AG, Howard RS, Ross-Russell RW, Beech R. Longer term quality of life and outcome in stroke patients: is the Barthel index alone an adequate measure of outcome? Qual Health Care. 1997; 6: 125–130.
Dorman PJ, Waddell F, Slattery J, Dennis M, Sandercock P. Is the EuroHRQOL a valid measure of health-related quality of life after stroke? Stroke. 1997; 28: 1876–1882.
Brooks R, Rabin R, de Charro F, eds. The Measurement and Valuation of Health Status Using EQ-5D: A European Perspective: Evidence From the EuroHRQL BIO MED Research Programme. Dordrecht, The Netherlands: Kluwer Academic Publishers; 2003.
de Haan RJ, Limburg M, Van der Muelen JHP, Jacobs HM, Aaronson NK. Quality of life after stroke: impact of stroke type and lesion location. Stroke. 1995; 26: 402–408.
Ohwaki K, Yano E, Nagashima H, Hirata M, Nakagomi T, Tamura A. Blood pressure management in acute intracerebral hemorrhage: relationship between elevated blood pressure and hematoma enlargement. Stroke. 2004; 35: 1364–1367.
Willmot M, Leonardi-Bee J, Bath PMW. High blood pressure in acute stroke and subsequent outcome: a systematic review. Hypertension. 2004; 43: 18–24.
Terayama Y, Tanahashi N, Fukuchi Y, Gotoh F. Prognostic value of admission blood pressure in patients with intracerebral hemorrhage. Stroke. 1997; 28: 1185–1188.
Fogelholm R, Avikainen S, Murros K. Prognostic value and determinants of first-day mean arterial pressure in spontaneous supratentorial intracerebral hemorrhage. Stroke. 1997; 28: 1396–1400.
Mayer SA, Sacco RL, Shi T, Mohr JP. Neurologic deterioration in non-comatose patients with supratentorial intracerebral hemorrhage. Neurology. 1994; 44: 1379–1384.
Duncan PW, Jorgensen HS, Wade DT. Outcome measures in acute stroke trials: a systematic review and some recommendations to improve practice. Stroke. 2000; 21: 1429–1438.
Sulter G, Steen C, de Keyser J. Use of Barthel Index and modified Rankin Scale in acute stroke trials. Stroke. 1999; 30: 1538–1541.
Higginson IJ, Carr AJ. Using quality of life measures in the clinical setting. BMJ. 2001; 322: 1297–1300.