Original Paper Cerebrovasc Dis 2009;27:328–335 DOI: 10.1159/000202009
Received: July 10, 2008 Accepted: November 3, 2008 Published online: February 14, 2009
A Comparison of Acute and Long-Term Management of Stroke Patients in Barbados and South London Nigel C. Smeeton a David O.C. Corbin c Anselm J. Hennis c–e Ian R. Hambleton d Henry S. Fraser c–e Charles D.A. Wolfe b Peter U. Heuschmann a a
King’s College London, Division of Health and Social Care Research, and b NIHR Biomedical Research Centre, Guy’s and St Thomas’ NHS Foundation Trust and King’s College London, London, UK; c Queen Elizabeth Hospital, d Chronic Disease Research Centre, Tropical Medicine Research Institute, and e Faculty of Medical Sciences, The University of the West Indies, Bridgetown, Barbados
Key Words Stroke, acute management Stroke, caregivers Stroke, long-term management Stroke, statistics Stroke care Health services research Post-stroke rehabilitation
Abstract Background: To compare health care utilisation between stroke patients living in a middle-income country with similar patients in a high-income country in terms of the type and amount of health care received following a stroke. Methods: Data were collected from the population-based South London Stroke Register (SLSR) and the Barbados Register of Strokes (BROS) from January 2001 to December 2004. Differences in management and diagnostic procedures used in the acute phase were adjusted for age, sex, ethnic group, living conditions pre-stroke and socio-economic status by multivariable logistic regression. Comparison of subsequent management was made for 3 months and 1 year post-stroke. Results: Patients in BROS were less likely to be admitted to a hospital ward (OR 0.22; 95% CI 0.13–0.37), but the difference for the lower use of brain scans in BROS was smaller (OR 0.62; 95% CI 0.25–1.52). Additional adjustment for stroke severity (Glasgow Coma Score) showed that BROS patients
© 2009 S. Karger AG, Basel 1015–9770/09/0274–0328$26.00/0 Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/ced
were more likely to have a swallow test on admission (OR 2.95; 95% CI 1.17–7.45). BROS patients were less likely to be in nursing care at 3 months (OR 0.37; 95% CI 0.17–0.81), and less likely to be receiving speech and language therapy at 3 months (OR 0.10; 95% CI 0.03–0.33) and 1 year (OR 0.05; 95% CI 0.00–0.55). Conclusions: The lower use of hospital admission and nursing care at 3 months suggests that in Barbados, family and friends take greater responsibility for patient care around the time of the stroke and in the medium term thereafter. Copyright © 2009 S. Karger AG, Basel
Introduction
There were an estimated 5.7 million deaths from stroke worldwide in 2005, with 87% of these occurring in lowand middle-income countries [1]. Because of the ageing population and an increasing prevalence of vascular risk factors in low- and middle-income countries, the burden of stroke is expected to increase rapidly in those areas of the world least able to cope with the disastrous economic impact [2].
Nigel C. Smeeton, MSc King’s College London, Division of Health and Social Care Research Capital House, 42 Weston Street, London SE1 3QD (UK) Tel. +44 20 7848 6622, Fax +44 20 7848 6620 E-Mail
[email protected]
In order to reduce the burden of stroke in these societies, appropriate diagnostic tests, drug treatment, rehabilitation therapy and secondary prevention strategies are required [3]. Detailed guidelines for the management of stroke have been published [4, 5]. These recommendations include direct patient admission to a stroke unit and urgent brain scanning. There is considerable variation in the management of stroke throughout Western and Central Europe [6]. However, data on acute and long-term stroke management from low- and middle-income countries are scarce. The few studies available show that in some low- and middle-income countries, detection and management of hypertension is comparable to that observed in high-income countries [7, 8], while for others treatment rates for diagnosed hypertension are less than 50% [9]. However, it is difficult to generalise from most published studies as nearly all are hospital-based [3, 10]. Estimation of the occurrence of stroke and utilisation of health care facilities for the general population requires population-based studies of the relevant areas linked to an individual long-term follow-up [10, 11]. Methodology for the establishment of stroke registers in low- and middle-income countries has been developed [10]. Using data from existing population-based stroke registers, we compared health care utilisation in stroke patients in Barbados, a middle-income country, and the UK, a high-income country.
Methods Data Collection Data were collected in the South London Stroke Register (SLSR), an ongoing population-based stroke register recording all hospitalised and non-hospitalised first-ever strokes in patients of all age groups for a defined area of South London since January 1995 [12]. The source population of the SLSR are the 271,817 residents (UK Census 2001) of whom 63% were White, 9% Black Caribbean, 15% Black African, 4% Black Mixed, 5% Asian and 4% other ethnic origin. Acute stroke care facilities include two teaching hospitals (both with a stroke unit) within and three outside the study area. All hospitals have access to computed tomography (CT) scans and magnetic resonance imaging (MRI) units. The population-based Barbados Register of Strokes (BROS) collected information on all hospitalised and non-hospitalised first-ever stroke patients from all age groups in Barbados between October 2001 and March 2005 [13]. The BROS cases came from a population of 268,762 (Barbados Census 2000) comprising 93% Black Caribbean, 3% Black Mixed, 3% White, and 1% other. Health care in Barbados is provided through a government-sponsored public system consisting of primary, secondary and tertiary services, which are free at the point of delivery. There is a single tertiary care hospital and supporting network of eight mainly pri-
Management of Stroke Patients in Barbados and South London
mary care polyclinics which service specific geographical catchment areas. There is also a complementary private care system where services are provided for a fee by general practitioners and specialists, in addition to access to radiological, laboratory, pharmacy services and paramedical specialties. The only tertiary care general hospital on the island does not have a stroke unit, nor does the small private hospital. The main hospital has a CT scan unit while there is another CT scan unit and a single MRI unit in the private sector. The SLSR and BROS used identical data collection methods and standardised variable definitions for ensuring maximal comparability and case completeness [13]. Overlapping sources of notification were established to maximise case ascertainment. The main sources were: accident and emergency records; hospital wards; brain imaging requests; death certificates; coroner’s records; general practitioners; hospital medical staff; community therapists; bereavement officers; hospital-based stroke registries, and general practice computer records [12]. The studies are designed to have patients or relatives interviewed as soon as possible after the stroke, at 3 months, 1 year, and annually thereafter [14]. All patients with a first-ever stroke registered in SLSR and BROS between January 2001 and December 2004 were analysed. Health Care Utilisation For comparing acute management and diagnostic procedures in the acute stage of stroke, the following variables were documented: swallow test, brain imaging (CT scans and/or MRI); admission to a stroke unit; length of stay in the acute hospital (restricted to hospitalised patients who did not die in hospital or were not discharged to another hospital). Patients were assessed at 3 months and 1 year after the acute stroke event in terms of the appropriate therapy and management at the respective time points: physiotherapy (PT) and occupational therapy (OT) (for patients who had any paralysis, sensory impairment or visual field defect in the acute phase); speech and language therapy (SALT) (for patients who had dysphasia and/or dysarthria in the acute phase); antihypertensive medication (if the patients with a recorded diagnosis of hypertension received antihypertensive drugs by the corresponding time point); location of care (nursing care versus residence in private household or sheltered home). Variable Definition Stroke was defined according to the World Health Organisation (WHO) criteria [15]. Stroke was classified into cerebral infarction (CI), primary intracranial haemorrhage (PICH), and subarachnoid haemorrhage (SAH) based on the results from brain imaging within 30 days of stroke onset, cerebrospinal fluid analysis (SAH only) or necropsy examination; cases without or with unknown pathological confirmation of stroke were recorded as unclassified. Patient information was collected on ethnicity, socio-economic status, living circumstances pre-stroke (as a measure of dependence), prior to stroke risk factors, current smoking, alcohol intake, clinical signs and symptoms in the acute phase, and activities of daily living post-stroke (detailed definitions are given in Appendix 1). Case fatality rates were compared at 7, 28 and 90 days, and 1 year.
Cerebrovasc Dis 2009;27:328–335
329
Statistical Analysis Unadjusted differences between BROS and SLSR were investigated using the 2 test for categorical variables, the unpaired t test for age and the Mann-Whitney U test for the National Institutes of Health (NIH) 15-item stroke scale and length of admission. To obtain adjusted estimates of the impact of location, binary outcome variables were analysed using multiple logistic regression [16]. Multiple linear regression was used for continuous outcome variables. The impact of residence in BROS or SLSR on admission to hospital and the acquisition of brain scans in the acute phase was adjusted for age, sex, ethnic group, socio-economic status, living circumstances pre-stroke. Use of the swallow test, and the provision of care at 3 months and 1 year (nursing care, general practitioner consultations, PT, OT, SALT, prescriptions), were additionally adjusted for stroke severity as indicated by the Glasgow Coma Score (GCS) level of consciousness. Multivariable analyses were restricted to patients with White or Black Caribbean ethnic origin. Stata Version 9.2 was used for the statistical analyses [17]. Ethics Patients and/or their relatives gave written informed consent to participate in the study. The design of the study was approved by the ethics committees of Guy’s and St Thomas’ Hospital Trust, King’s College Hospital, Queens Square and St. Georges (London) and for BROS from the Medical Research Ethics Committee of the Ministry of Health (Barbados).
Results
Between 2001 and 2004, 1,058 patients with first-ever stroke were registered in the SLSR and, coincidentally, 1,058 in BROS. A comparison of socio-demographic factors and clinical severity characteristics for patients from BROS and SLSR is given in table 1. Patients in Barbados were older than those in South London (71.3 years (SD = 15.4) vs. 69.7 years (SD = 15.1), p = 0.013). They were less likely to be male (p = 0.001) and much more likely to be of non-manual socio-economic status (p ! 0.001). In Barbados, prior to stroke, more patients lived with some independence but not alone (p ! 0.001). The unadjusted case fatality rate in Barbados was significantly greater at 90 days (p = 0.003), but not at 7 days, 28 days and 1 year. Table 2 shows that in the acute phase, patients in Barbados were more likely to have a swallow test (p ! 0.001). The likelihood of having a brain scan was similar in the two locations (p = 0.101), although the use of MRI was greater in South London. Patients in Barbados were less likely to be admitted to hospital (p ! 0.001). Over one half of the South London patients received stroke unit care compared to none in Barbados, where there is no stroke unit. For both BROS and SLSR, hospital admission was 330
Cerebrovasc Dis 2009;27:328–335
not associated with age, sex or socio-economic status, but was more likely for patients with a severe stroke as indicated by a failed swallow test or incontinence. Similar proportions of patients consulted a general practitioner within 3 months of the stroke (p = 0.888). Appropriate patients in Barbados were more likely to receive PT (p = 0.029) but less likely to receive OT at 3 months (p ! 0.001), and SALT was less commonly used in Barbados (p ! 0.001) for patients with acute dysphasia and/or dysarthria. For both BROS and SLSR, most patients with a record of hypertension by 3 months poststroke were prescribed antihypertensive medication (p = 0.264). In Barbados, nursing care of the patient was less common at 3 months (p ! 0.001). At 1 year post-stroke, appropriate patients in Barbados were less likely to undergo OT (p = 0.001) and SALT (p ! 0.001). For both BROS and SLSR, more than 85% of patients with a record of hypertension by 1 year were prescribed antihypertensive medication at 1 year post-stroke (p = 0.193). As at 3 months, nursing care was less common in Barbados at 1 year (p = 0.007). Table 3 shows the relationship between service use and location adjusted for age, sex, ethnic group, socioeconomic class and living circumstances. Patients living in Barbados were more likely to have a swallow test in the acute stage of the stroke (p = 0.022) even allowing for the differences in the level of consciousness. However, they were less likely to be in nursing care at 3 months (p = 0.012), and less likely to receive SALT at both 3 months (p ! 0.001) and 1 year (p = 0.015).
Discussion
Aspects of care in the acute stage would be expected to differ between the two settings, due to the stroke unit care that has been available in South London since the start of the SLSR project and the lack of a stroke unit in Barbados. The rate of hospital admission suggests that in Barbados, family and friends take greater responsibility for patient care around the time of the stroke. However, the use of clinical and radiological methods of investigation is relatively high in Barbados, with a similar likelihood of a stroke patient having a brain scan in the acute stage and greater use of the swallow test. The greater importance of patient care by the family is also seen at 3 months post-stroke with patients in BROS less likely to be receiving nursing care. However, by 1 year post-stroke this difference had disappeared and BROS survivors were as likely as those in SLSR to be receiving Smeeton /Corbin /Hennis /Hambleton / Fraser /Wolfe /Heuschmann
Table 1. Socio-demographic, clinical severity characteristics and risk factors for stroke patients registered in
BROS and SLSR (2001–2004) Variable Socio-demography Age, mean (SD) Age group
