Gender Differences in Stroke Examined in a 10-Year Cohort of Patients Admitted to a Canadian Teaching Hospital
Background and Purpose— Studies suggest that women with stroke are investigated less aggressively and receive tissue plasminogen activator less frequently than men. We tested whether gender differences in the investigation, treatment, and outcome of stroke are due to confounding factors.
Methods— Gender differences in the use of investigations, trial enrollment, treatment with intravenous tissue plasminogen activator, and in-hospital outcomes were examined in data from our prospective registry using multivariate analysis to adjust for age, prestroke functional status, stroke subtype and severity, and atrial fibrillation.
Results— Of 2725 consecutive hospitalized patients (1996 to 2006), 88% had ischemic stroke and 48% were women. Women were older (median age, 77 versus 70 years), had more severe strokes, and were less likely to be independent prestroke (78% versus 87%) compared with men (all P<0.001). The proportion of women, but not men, aged ≥80 years, increased significantly between 1996 to 1997 and 2005 to 2006. After adjustment for confounding, women were less likely to have infratentorial strokes (OR, 0.78; 95% CI, 0.62 to 0.97), be able to walk unaided on admission (OR, 0.69; 95% CI, 0.54 to 0.87), be treated with tissue plasminogen activator (OR, 0.51; 95% CI, 0.35 to 0.72), experience pneumonia (OR, 0.38; 95% CI, 0.26 to 0.55), achieve a discharge Barthel Index of ≥95 (OR, 0.75; 95% CI, 0.61 to 0.94, and were more likely to experience a urinary tract infection (OR, 2.06; 95% CI, 1.61 to 2.64). There was no gender difference in adjusted use of investigations.
Conclusion— The majority of the gender differences in stroke were explained by confounding. More research is required to understand gender differences in stroke pathophysiology and the utilization of thrombolytic therapy.
Some studies of gender differences in the management of acute stroke have shown that women are investigated less thoroughly1,2 and treated less frequently with antiplatelet agents3 and tissue plasminogen activator (tPA).4,5 The explanation for these findings is not clear but may relate to gender differences at presentation such as age, comorbidity, prestroke functional status, and stroke severity. Other studies, however, demonstrate no gender differences in the use of investigation or treatment with tPA or antiplatelet agents.2,6 Stroke incidence rates are lower for women across all age groups,7 although women experience worse outcomes,1,3,6 comprising 61% of stroke-related deaths in the United States.8 Given these observations, we sought to determine among patients admitted to our stroke program whether there were any gender differences in the use of investigations, use of intravenous tPA, enrollment in clinical trials, and outcomes that were independent of confounding variables.
The acute stroke program based at the Halifax Infirmary, Queen Elizabeth II Health Sciences Centre has previously been described.9 In brief, the Halifax Infirmary is a large (approximately 1000 beds) teaching hospital providing primary and secondary care for approximately 300 000 people and tertiary care for an additional 500 000. Longstanding local practice has been for emergency physicians to refer patients with acute neurological disorders to the on-call neurology team. The acute stroke program (directed by S.P.) provides inpatient care for patients admitted with a diagnosis of ischemic stroke (IS) or transient ischemic attack. Care is also provided for the majority of patients with intracerebral hemorrhage, ie, those who do not undergo neurosurgical treatment and those who require postoperative medical and rehabilitation therapy. Periodic surveys by the program coordinator before 2000 and continuous surveillance by a research assistant since then show that the program provides care for virtually all such patients.
Information about all patients admitted to the program is captured prospectively in a registry,9 which includes data on stroke subtype (either hemorrhagic stroke or Oxfordshire Community Stroke Project ischemic stroke subtype, ie, total or partial anterior circulation stroke; posterior circulation stroke (POCS); lacunar stroke),10 stroke severity (using a stroke severity score that grades severity as mild [1 to 4] moderate [5 to 7], and severe [8 to 10] on the basis of symptoms, signs, and functional impairment),11 location of stroke (infra- or supratentorial, left or right hemisphere) determined from cranial imaging (CT or MRI), prestroke functional status (Oxford handicap score, a measure of disability; 0 to 2 independent, 3 to 5 dependent),12 patient-reported prior transient ischemic attack or stroke, electrocardiographic evidence of atrial fibrillation (AF), use of diagnostic investigations (eg, CT, MRI, echocardiography) and interdisciplinary services (eg, physiotherapy, occupational therapy), complications (eg, pneumonia), discharge disposition, Oxford handicap score, and Barthel Index (BI)13 at discharge, and participation in research (clinical trials and other projects). An excellent outcome was defined as a discharge BI ≥95, and discharge independence defined as a discharge Oxford handicap score ≤2. From 2001 on (n=1503 of 2725 or 55% of patients), the following 4 predictive baseline variables derived from a 6 simple variable stroke model14 were recorded in addition to age and prestroke functional status: living alone before stroke, verbal component of the Glasgow Coma Score,15 being able to walk independently, and able lift both arms off the bed. This subgroup (ie, from 2001 on) was used to test whether the variable living alone or not before stroke affected the probability of receiving tPA or being enrolled in a clinical trial. Some patients with very severe strokes with or without multiple comorbidities resulting in a very poor prognosis received palliative care (comfort measures only).
In this study, we analyzed data pertaining to patients admitted because of IS or intracerebral hemorrhage between May 1996 (when we started using tPA) to February 2006. For all patients treated with intravenous tPA, chart review was performed to obtain additional data on medical history, blood pressure, and blood glucose level at presentation and the times of stroke onset, presentation to the emergency department, CT scan, and tPA administration. The Capital Health Research Ethics Board approved collection of the registry data without consent from each individual patient for the purposes of quality assurance and monitoring of practice.
Descriptive statistics include means (SD) or median (interquartile range), and relative frequencies for continuous and categorical data, respectively. Continuous variables were compared using Student t test or Wilcoxon rank sum test. Chi-squared analyses were used to compare categorical data.
We compared stroke presentation (eg, stroke severity, subtype), use of investigation (eg, cranial imaging, Holter monitoring), intravenous tPA use, participation in research, and in-hospital outcomes between women and men. Multivariate logistic regression or general linear modeling was used to adjust for confounding factors identified in univariate analyses. These factors included age, prestroke functional status based on the Oxford handicap score, a history of AF, stroke severity, and Oxfordshire Community Stroke Project classification of stroke subtype. A probability value of <0.05 was considered statistically significant in the analyses. Gender was forced into the multivariate models regardless of statistical significance. Cases missing data for the variables of interest were excluded from the analyses. SAS software (SAS Institute Inc, version 9.1) was used for all analyses.
Between May 1996 and February 2006, 2725 patients (1414 men, 1311 women; ratio 1.08) were admitted with a diagnosis of stroke (88% IS). Only 0.27% of data points were missing (range, 0% to 1.5% per variable), the majority (79%) relating to complications (pneumonia, deep venous thrombosis/pulmonary embolus, and urinary tract infection).
Women were older (Table 1⇓) with a higher proportion aged ≥80 years compared with men (40% versus 18%, P<0.0001). The proportion of women with stroke aged ≥80 years increased from 35% in 1996 to 1997 to 48% in 2005 to 2006 (chi-square for an annual trend P<0.01, Figure), whereas the proportion of men aged ≥80 years did not significantly increase (16% to 19%, P=0.47) over the same period. There was no gender difference in the proportion of patients presenting to hospital within 3 hours of symptom recognition (39% versus 37% for women versus men).
In univariate analysis, women were more likely to live alone prestroke, have AF, TACS Oxfordshire Community Stroke Project subtype, and severe stroke (Table 1⇑). Women were less likely to be independent prestroke and less likely to have POCS Oxfordshire Community Stroke Project subtype or infratentorial stroke localization (IS and intracerebral hemorrhage). At the time of first medical assessment after stroke onset, women were less likely to have a normal verbal component of the Glasgow Coma Score, to be able to walk unaided, or lift both arms off the bed. In view of the baseline gender differences in stroke characteristics and demographics identified in univariate analysis, the following variables were considered, a priori, to be possible confounding variables: age, prestroke functional status (Oxford handicap score), AF, stroke severity, and Oxfordshire Community Stroke Project subtype. After adjustment for these 5 confounding variables, only the following gender differences in stroke characteristics remained significant: men more commonly had POCS subtype and infratentorial stroke localization and women were more commonly unable to walk unaided at the time of first medical assessment after stroke onset (Table 2).
Use of Investigations and Research Study Enrollment
There was no gender difference in the use of investigations (eg, CT, echocardiography) after adjustment for confounding variables (Tables 1 and 2⇑⇑). Men were more likely to be enrolled in a clinical trial (all trials of pharmacological agents), a finding that was independent of the 5 confounding variables. However, inclusion of the variable living alone or not prestroke (available in 55% of patients) in the multivariate analysis reduced the gender difference in likelihood of being enrolled in a trial (OR, 0.73; 95% CI, 0.45 to 1.29; P=0.31). There was no gender difference in the proportion of patients enrolled in noninterventional studies, eg, the Registry of the Canadian Stroke Network.9 There was no gender difference (women versus men) in the proportion of patients seen by occupational therapy (85% versus 85%), physiotherapy (83% versus 82%), social work (58% versus 56%), or speech–language pathology (56% versus 58%).
Gender Differences in Use of Tissue Plasminogen Activator in Stroke
Of patients with IS (n=2393), women were less likely than men to receive tPA (61 of 1149 [5%] versus 109 of 1244 [9%], P=0.0011; OR, 0.57; 95% CI, 0.41 to 0.79). Adjusting for 5 confounding variables (Table 2) or limiting the adjusted analysis to patients with IS presenting within 3 hours of symptom recognition (n=899; adjusted OR, 0.55; 95% CI, 0.38 to 0.80; P=0.002) did not alter this gender difference. To examine whether living alone prestroke was a confounding variable for tPA use, we analyzed the subgroup of patients with IS from 2001 on for whom residential status was known (n=1305). Adjustment for living alone prestroke did not alter the gender difference in likelihood to receive tPA (OR, 0.51; 95% CI, 0.36 to 0.70; P=0.0002). The only notable difference comparing tPA-treated women and men was that women were older (median, 75 [67–81] versus 70 [61–77] years; P=0.02). There was no interaction between age and gender in the analysis of tPA treatment (data not shown).
Analysis of the timing of tPA use among patients admitted directly to the emergency department (excluding patients who were first triaged in a separate hospital [n=7] or who had a stroke after hospitalization for another reason [n=28]) showed no gender difference in onset to arrival, onset to treatment, or onset to CT scan times (data not shown). However, both CT to treatment (median, 73 [56–93] versus 56 [37–82] minutes; P=0.014, Wilcoxon rank sum test) and door to treatment (107 [88–121] versus 94 [64–120] minutes; P=0.039) times were significantly longer in women (Table 3). Only CT to treatment time remained significantly longer in women compared with men after adjustment for the 5 confounding variables (adjusted means 72.6±7.0 versus 61.8±6.1 minutes, P=0.044, analysis of covariance).
Outcomes and Complications After Stroke
There were several significant gender differences in outcomes in the univariate analyses (Table 1⇑); however, after adjustment for age and other confounding variables, the only persisting gender difference in outcome was that an excellent outcome at hospital discharge (BI ≥95) was less likely for women (Table 2). After adjusting for the 5 confounding variables, urinary tract infection was more common in women and pneumonia in men.
The main findings of this study are that after adjustment for confounding variables, women were less likely to be treated with intravenous tPA, to be entered into a clinical trial, and to experience an excellent outcome. These differences are noteworthy given the rising proportion of women with stroke aged ≥80 years being admitted to our hospital over the last decade. The risk of stroke increases with age7 and the proportion of the population in Nova Scotia aged ≥75 years is forecast to increase by 39% between 2001 and 2021.16 Therefore, it may be anticipated that the healthcare system will have to manage an increasing number of elderly patients, particularly women, with disabling strokes that result in longer hospital stays, more complications, and an increased need for long-term institutional care.
Our observation that men more commonly had a POCS subtype or infratentorial stroke localization is supported by a previous study demonstrating that men were more likely to present with cerebellar and brain stem symptoms and to have POCS than women.6 Although AF is more common in women13 and more frequently associated with anterior circulation IS,17 adjusting for AF in our data did not abolish the higher proportion of posterior circulation strokes in men, suggesting that there is an alternative explanation for this observation.
We report no gender difference in the age-adjusted proportion of patients receiving investigations for stroke as seen in one study6 but not others.1,2 We observed no gender difference in the majority of outcomes after adjusting for confounding variables. However, despite controlling for age, prestroke functional status, stroke subtype, AF, and stroke severity, women were less likely to have an excellent outcome (BI ≥95), confirming the findings of others.1,3,6 Women were less likely to be discharged home and more likely discharged to a long-term care facility; however, these differences are less apparent after accounting for the confounding variables. The lower proportion of women being enrolled in clinical trials might be explained by exclusion criteria in many studies (eg, patients >80 years or who are dependent), producing selection bias toward men. However, adjusting for these variables did not negate the gender difference in trial enrollment. Adjusting for whether a patient lived alone prestroke did, however, reduce the gender difference in trial enrollment, suggesting the presence of a surrogate decision-maker (ie, relative/partner) increases the chance of being enrolled in a research study. The gender bias in clinical trials is well documented,18 but effort is clearly required to design new pragmatic trials that are more relevant to the entire stroke population.
We demonstrate that men with IS are more likely to receive tPA than women even after adjustment for possible confounding variables. Several studies that provide data on the proportion of men and women in their total stroke population (thus allowing calculation of gender-specific likelihood of tPA use) show higher proportions of men receiving tPA compared with women.4,5,19,20 The US National Hospital Discharge Survey of almost 1 800 000 patients with IS showed that 0.41% of women and 0.86% of men were treated with tPA.5 Men outnumbered women (3:2) in a European registry of more than 6000 tPA-treated patients21 and in the randomized trials of tPA.26 Also, some studies report that the likelihood of tPA treatment decreases with age and increases with stroke severity.4,22 Clinicians might be disinclined to use tPA in elderly patients (who are more commonly women) for several reasons, including the paucity of randomized evidence in the elderly, the prevalence of preexisting disability and cognitive impairment, and a reported higher risk of symptomatic intracranial hemorrhage.23 Evidence from a self-administered survey suggests that women are less likely to consent to tPA therapy than men.24 In our study, tPA was less frequently administered to women than men, even after adjustment for confounding variables, including living alone prestroke. Clearly, additional studies are required to further explain gender differences in the use of tPA.
We demonstrated that tPA-treated women tend to have longer CT to treatment times than men. Conceivably, there could have been increased deliberation over the balance of risk and benefit among frail, elderly women; however, this result persisted after adjustment for age and prestroke functional status. Given that there was no gender difference in “door to needle” tPA treatment times, the difference in CT to treatment time may be a chance finding. Against this explanation is the finding in another study25 that women with stroke arriving in the emergency department waited significantly longer than men to see an emergency physician, but not a neurologist. It remains unclear whether the apparent gender differences in the use and timing of tPA are due to bias or to the influence of confounding variables that were not examined in our study. Because the benefit of tPA is time-dependent and women may receive more benefit from tPA than men,26 it is imperative to expedite the administration of tPA to maximize its effect in both women and men.
We saw no gender difference in the involvement of the interdisciplinary team or in discharge destination after adjustment for baseline differences, as previously noted.3 Although we did not record smoking use, the higher rate of poststroke pneumonia among men may relate to a higher prevalence of smoking and preexisting pulmonary disease.1,3,6
Our study has somewhat limited generalizability given that it relates only to selected aspects of acute care in a single tertiary care center with specific stroke expertise. Other limitations of our study are that we did not have detailed data on comorbidity (eg, cognitive impairment), type of prestroke residence (eg, living in a care home), and contraindications to tPA use. Strengths of this study are high case ascertainment in a single hospital providing organized stroke care9 and prospective data collection over a decade.
In summary, we highlight the rising proportion of elderly women hospitalized because of stroke. After adjusting for confounding variables, we show that women were less likely to have an excellent outcome, were less likely to be entered into a clinical trial, and less likely to be treated with tPA. The reasons for lower tPA use in women requires further study, but the observation may be explained, at least in part, by a gender difference in the likelihood to consent to treatment.
J.M.R. is supported by the Dalhousie University Internal Medicine Research Foundation and unrestricted educational grants from Hoffmann-LaRoche and Merck Frosst. M.K.K. holds a New Investigator Award from the Canadian Institutes for Health Research and receives support from the University Health Network Women’s Health Program and the Canadian Stroke Network.
- Received May 29, 2007.
- Revision received August 18, 2007.
- Accepted September 13, 2007.
Di Carlo A, Lamassa M, Baldereschi M, Pracucci G, Basile AM, Wolfe CD, Giroud M, Rudd A, Ghetti A, Inzitari D. European BIOMED Study of Stroke Care Group. Sex differences in the clinical presentation, resource use, and 3-month outcome of acute stroke in Europe: data from a multicenter multinational hospital-based registry. Stroke. 2003; 34: 1114–1119.
Holroyd-Leduc JM, Kapral MK, Austin PC, Tu JV. Sex differences and similarities in the management and outcome of stroke patients. Stroke. 2000; 31: 1833–1837.
Bateman BT, Schumacher HC, Boden-Albala B, Berman MF, Mohr JP, Sacco RL, Pile-Spellman J. Factors associated with in-hospital mortality after administration of thrombolysis in acute ischemic stroke patients: an analysis of the nationwide inpatient sample 1999 to 2002. Stroke. 2006; 37: 440–446.
Kapral MK, Fang J, Hill MD, Silver F, Richards J, Jaigobin C, Cheung AM. Investigators of the Registry of the Canadian Stroke Network. Sex differences in stroke care and outcomes: results from the Registry of the Canadian Stroke Network. Stroke. 2005; 36: 809–814.
Ayala C, Croft JB, Greenlund KJ, Keenan NL, Donehoo RS, Malarcher AM, Mensah GA. Sex differences in US mortality rates for stroke and stroke subtypes by race/ethnicity and age, 1995–1998. Stroke. 2002; 33: 1197–1201.
Phillips SJ, Eskes GA, Gubitz GJ; Queen Elizabeth II Health Sciences Centre Acute Stroke Team. Description and evaluation of an acute stroke unit. CMAJ. 2002; 167: 655–660.
The EC/IC Bypass Study Group. The International Cooperative Study of Extracranial/Intracranial Arterial Bypass (EC/IC Bypass Study): methodology and entry characteristics. Stroke. 1985; 16: 397–406.
Counsell C, Dennis M, McDowall M. Predicting functional outcome in acute stroke: comparison of a simple six variable model with other predictive systems and informal clinical prediction. J Neurol Neurosurg Psychiatry. 2004; 75: 401–405.
http://www.gov.ns.ca/finance/stats.div/papers/demograf/demo4.htm. Last accessed on February 11, 2008.
Hajat C, Dundas R, Stewart JA, Lawrence E, Rudd AG, Howard R, Wolfe CD. Cerebrovascular risk factors and stroke subtypes: differences between ethnic groups. Stroke. 2001; 32: 37–42.
Deng YZ, Reeves MJ, Jacobs BS, Birbeck GL, Kothari RU, Hickenbottom SL, Mullard AJ, Wehner S, Maddox K, Majid A. Paul Coverdell National Acute Stroke Registry Michigan Prototype Investigators. IV tissue plasminogen activator use in acute stroke: experience from a statewide registry. Neurology. 2006; 66: 306–312.
Johnston SC, Fung LH, Gillum LA, Smith WS, Brass LM, Lichtman JH, Brown AN. Utilization of intravenous tissue-type plasminogen activator for ischemic stroke at academic medical centers: the influence of ethnicity. Stroke. 2001; 32: 1061–1068.
Wahlgren N, Ahmed N, Davalos A, Ford GA, Grond M, Hacke W, Hennerici MG, Kaste M, Kuelkens S, Larrue V, Lees KR, Roine RO, Soinne L, Toni D, Vanhooren G; SITS-MOST investigators. Thrombolysis with alteplase for acute ischaemic stroke in the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST): an observational study. Lancet. 2007; 369: 275–282.
Reed SD, Cramer SC, Blough DK, Meyer K, Jarvik JG. Treatment with tissue plasminogen activator and inpatient mortality rates for patients with ischemic stroke treated in community hospitals. Stroke. 2001; 32: 1832–1840.
Tanne D, Kasner SE, Demchuk AM, Koren-Morag N, Hanson S, Grond M, Levine SR. Markers of increased risk of intracerebral hemorrhage after intravenous recombinant tissue plasminogen activator therapy for acute ischemic stroke in clinical practice: the Multicenter rt-PA Stroke Survey. Circulation. 2002; 105: 1679–1685.
Menon SC, Pandey DK, Morgenstern LB. Critical factors determining access to acute stroke care. Neurology. 1998; 51: 427–432.
Kent DM, Price LL, Ringleb P, Hill MD, Selker HP. Sex-based differences in response to recombinant tissue plasminogen activator in acute ischemic stroke: a pooled analysis of randomized clinical trials. Stroke. 2005; 36: 62–65.