Stroke Presentation and Hospital Management
Comparison of Neighboring Healthcare Systems With Differing Health Policies
Background and Purpose— Acute stroke care is shaped by healthcare policies. Differing policies in similar populations allow for assessment of policy impact on health and healthcare outcomes. The purpose of this study was to compare stroke presentation and hospital care in two adjacent healthcare systems with differing healthcare policies.
Methods— Interviews and chart review of consecutive acute stroke admissions in Northern Ireland (n=103) and the Republic of Ireland (n=100).
Results— Marked regional contrasts were evident for key aspects of hospital care. Northern Ireland performed significantly better on 15 of 16 quality of care (Sentinel Audit) items. Delivery on standards was significantly better in Northern Ireland for early assessment (Northern Ireland 72%; Republic of Ireland 54%, P<0.01), multidisciplinary review (Northern Ireland 69%; Republic of Ireland 31%, P<0.001), medications review (Northern Ireland 54%; Republic of Ireland 19%, P<0.001), and for discharge-rehabilitation planning (Northern Ireland 83%; Republic of Ireland 8%, P<0.001). Preadmission prescription of advised cardiovascular medications was similar between regions for antihypertensives and anticoagulants but significantly higher in Northern Ireland for antiplatelets (Northern Ireland 65%; Republic of Ireland 38%, P=0.001) and lipid-regulating medication (Northern Ireland 44%; Republic of Ireland 26%; P=0.006). Prescribing levels increased in both regions and all medication categories by discharge but with significantly lower levels in Northern Ireland for antihypertensives (Northern Ireland 60%; Republic of Ireland 75%, P=0.025). Northern Ireland patients were more functionally dependent (mean Barthel Index 10.5 versus 12.7 [Republic of Ireland], P=0.013) and less aphasic (mean Frenchay Aphasia Screening Test 17.8 versus 16.8 [Republic of Ireland], P=0.022).
Conclusions— In similar neighboring acute stroke populations, differing healthcare policies were associated with significant differences in processes of patient care. Policy reform is an important tool in ensuring optimal stroke care delivery.
Health management strategies for stroke can significantly influence early detection and short- and long-term outcomes. Comparison within and across systems allows for benchmarking and continuing quality improvement, for instance, the UK Sentinel Audit of Stroke.1 Current evidence suggests major variations between Western European countries in outcome after acute stroke,2 with the United Kingdom and Ireland scoring poorly in comparison to other European countries. Whereas there are marked variations between European countries in acute stroke facilities available,3 there is little focus on the impact of primary and other care systems on presentation with stroke, prevention strategies, and outcome.
The present study compared stroke patients in the Republic of Ireland and Northern Ireland. Northern Ireland has a population of 1.7 million4 and the Republic of Ireland has 4.2 million inhabitants5; both have an estimated 230 strokes per 100000 people annually.
These 2 systems on 1 island have differing health policies, service structures, and funding arrangements. Northern Ireland has universal free healthcare coverage, with free primary care physician and hospital healthcare to all. For those aged 65 years and over, medications are free. Private health insurance is unusual in Northern Ireland (10% in 2005).6 The Republic of Ireland has a mix of public and private healthcare coverage. Approximately 26% of the population is eligible for universal free healthcare coverage.7 This is income or age related with all people in lower income categories and all aged 70 years and over were eligible for universal health care during the period of this study; this includes medications. Others have state-provided partial coverage and pay for primary care physician and some hospital services. These include Emergency Department attendance charges and a fee payable per bed-night in hospital up to a small maximum amount annually. Almost half the population of the Republic (52% in 2005) have privately funded health insurance,8 mainly to cover the cost of access to private hospital healthcare. Three percent of those entitled to universal healthcare also have private insurance. A further 25% have state-provided partial coverage only. The Northern Ireland system sets specific targets and financial incentives to primary care physicians for preventive and chronic disease management assessments and intervention. This is not the case in the Republic of Ireland, although a recent national pilot project—the HeartWatch scheme—has provided payment for management of patients at high cardiovascular risk to a minority of primary care practices. Northern Ireland has participated in the UK Sentinel Audit of Stroke since it began in 1998 and has consistently out-performed England and Wales in all stroke audits:1 there has also been a formal policy to develop stroke units in the NHS since 2001. In the Republic of Ireland, there is no formal policy on stroke. The first national audit of stroke was published in 2008.9 These health system differences provide a unique opportunity for evaluation of stroke management of two groups of stroke patients on one island with differing healthcare systems.
The aim of this study was to compare the presentation and early management of stroke patients in 2 neighboring healthcare systems with differing policy and service organization, ie, the Republic of Ireland and Northern Ireland.
Participants were identified as consecutive adult hospital admissions with a diagnosis of stroke in 4 hospitals (2 large teaching hospitals in Northern Ireland, and 1 large teaching hospital and 1 small district general hospital in the Republic of Ireland), admitting largely from their surrounding urban areas. All interviews took place within 2 to 6 weeks of stroke and were carried out between September 2004 and September 2005.
Eligibility criteria were:
Confirmed diagnosis of stroke (defined in accordance with ICD-10 diagnostic criteria as symptoms of rapid onset lasting more than 24 hours of presumed vascular origin, reflecting a focal disturbance of cerebral function while excluding isolated impairment of higher function) supported by computerized brain tomographs, and MRI in a minority of cases. Neurological assessment of the patient was conducted by the consultant-led clinical team on the ward, within 48 hours of admission for a possible stroke.
Ability to participate in interview assessment—criteria used were Abbreviated Mental Test (AMT) score ≥810,11 and Frenchay Aphasia Screening Test (FAST) Score ≥14,12 both to be achieved within 2 to 4 weeks of admission for stroke. Where a patient was unable to complete the AMT during this time period, information was sought from the next-of-kin. This included patients with impaired consciousness and patients with aphasia (those with expressive and receptive dysphasia were unable to complete the AMT or the FAST).
Exclusion criteria were transient cerebral ischemic attacks and related syndromes; traumatic intracranial hemorrhage; vascular dementia or subarachnoid hemorrhage. In exceptional cases, patients could be deemed unsuitable for invitation to participate by the consultant/medical team (eg, where patient was critically ill over the 2 to 4 weeks postadmission period).
Eligible patients, or next-of-kin where AMT could not be achieved as described, were invited to participate. Two researchers carried out the data collection, 1 in Northern Ireland and 1 in the Republic. The researchers were trained jointly to a common protocol for conducting interviews. To ensure compatibility across centers, quality checks were carried out during data collection and with the completed dataset. Because of the nature of stroke and the fatigue experienced by patients in the acute phase, the patient interviews in certain instances were administered in 2 separate interviews. Patients records were also examined in all cases.
Ethical approval for this study was obtained from the Office of Research Ethics Committees (Northern Ireland) and the specific hospital research ethics committees (Republic of Ireland).
The research interview included questions about demographic factors, prestroke and current health status, functional and cognitive status, and discharge profile. Functional status was assessed by the Barthel Index (BI; score 0 to 20 where higher scores indicate more functional independence)13; cognitive status by the Abbreviated Mental Test (AMT; score 0 to 10 where ≥8 suggests normal cognitive function); communication by the Frenchay Aphasia Screening Test (FAST; score 0 to 20 where ≤13 indicates dysphasia); stroke severity by the Orpington Prognostic Score (OPS; score 1.6 to 6.0, higher scores indicate more severe deficit prognosis, 1.6 to 6.0; <3.2 the deficit prognosis is mild-moderate, 3.2 to 5.2, moderate-severe, and >5.2 severe-very severe)14; and Instrumental activities of daily living by the Nottingham Extended ADL Scale (NEADL; 22 items, score 0 to 66 where higher scores reflect more independence).15,16 Hospital management of the patient in both regions was determined using select questions from the Sentinel Audit proforma of stroke. Questions selected measured early physical assessment, multidisciplinary review, medications review, and discharge-rehabilitation planning.
Analyses were conducted using SPSS for Windows (Version 14.0). Descriptive statistics were used throughout. Comparisons were made using Pearson Chi-square and independent t tests where appropriate.
A total of 431 consecutive admissions were assessed for eligibility, with similar numbers eligible in Northern Ireland and the Republic of Ireland (208 Republic of Ireland, 213 Northern Ireland). This yielded 303 eligible patients (166 Northern Ireland or 77% of total assessed; 137 Republic of Ireland or 66% of total assessed). Of those eligible, 67% were included in the study (n=203, age range 20 to 98 years, n=103 Northern Ireland). The remainder declined (22%), were too ill with no proxy or a declined proxy (3%), or were not invited as a resulted of early discharge or lost to follow-up (8%). Forty-eight next-of-kin were recruited as “proxy” interviewees where patients were unable to take part (n=28 Northern Ireland).
Table 1 describes the demographic and clinical characteristics of included patients. The gender and age profile of patients was similar across the health systems. The proportions of patients reporting various past events or conditions were also similar in both regions with one exception, Northern Ireland, having a significantly larger percentage of patients with diabetes mellitus (24.3% versus 15%, χ2=7.2, P=0.027). The stroke type or location was similar in both regions, except the absence of brain stem lesions in Northern Ireland.
Functional and Cognitive Status on Admission
Comparison of scores on physical function, language, and cognition between the 2 patient populations showed a complex pattern. The regions had similar OPS scores (mean 3.3, SD 1.6 Northern Ireland; mean 3.3, SD 1.4 Republic of Ireland, t=−0.056, P=0.955) and AMT scores (mean 7.4, SD 3.6 Northern Ireland; mean 7.9, SD 2.8 Republic of Ireland, t=1.086, P=0.279), indicating that patients had similar severities of stroke and cognitive ability. However, results from the Barthel Index (mean 10.5, SD 6.0 Northern Ireland; mean 12.7, SD 6.6 Republic of Ireland, t=2.494, P=0.013) and FAST (mean 17.8, SD 2.1 Northern Ireland; mean 16.8, SD 2.8 Republic of Ireland, t=−2.317, P=0.022) indicate Northern Ireland patients to be significantly more dependent and less aphasic than those in the Republic. Both regions had a mean FAST score greater than 13, reflecting in part the effect of screening for significant aphasia. NEADL scores (mean 33.1, SD 21.3 Northern Ireland; mean 24.4, SD 20.9 Republic of Ireland, t=−2.925, P=0.004), however, indicated patients in Northern Ireland to be significantly less dependent than patients in the Republic in terms of their instrumental activities of daily living.
Admission and Discharge Medication Profile
Table 2 shows the pre- and posthospital use of recommended cardiovascular medications relevant to stroke. On admission there were differences in prescribing patterns across the 2 healthcare systems. Northern Ireland patients with self-reported cardiovascular conditions such as ischemic heart disease and diabetes were more likely to report being prescribed relevant medication than their Republic of Ireland counterparts. Significantly more Northern Ireland patients were prescribed antiplatelets (65% versus 38%, χ2=14.2, P=0.001) and lipid-regulating medication (44% versus 26%, χ2=7.4, P=0.006). There were no differences in preadmission prescribing patterns of antihypertensives or anticoagulants. This pattern shifted on discharge with patients in Northern Ireland and the Republic of Ireland equally likely to be prescribed antiplatelets, anticoagulants, and lipid-regulating medications (see table 2) but with those in the Republic more likely to be discharged on antihypertensive medications (75% versus 60% Northern Ireland, χ2=5.0, P=0.025).
Table 3 shows the results from the Sentinel Audit for the patients in the study. Striking differences between the 2 regions were evident. Northern Ireland performed significantly better than Republic of Ireland for almost all areas of patient management measured by the audit, except the prescription of warfarin where there was no significant difference and the prescription of aspirin within 48 hours of stroke, which was significantly better in the Republic of Ireland.
The difference between the 2 regions was 18% for audit items concerned with early assessment (72% versus 54%: χ2=8.0, P<0.005), 38% for those measuring multidisciplinary review (69% versus 31%: χ2=49.2, P<0.001), 35% for review of medications (54% versus 19%: χ2=17.3, P<0.001), and 75% for discharge/rehabilitation planning (83% versus 8%: χ2=250.0, P<0.001). This latter area of discharge/rehabilitation planning showed consistently large differences between the 2 healthcare systems with Northern Ireland performing better in each comparison. Although the area of early assessment, multidisciplinary review, and medications review were similar when all items were considered, individual items within each section are important to note. For example in the Republic of Ireland, 45% fewer patients were assessed by a physiotherapist within 72 hours; 80% fewer by an occupational therapist at 7 days; and 45% fewer by a multidisciplinary team.
The presenting clinical profile of the patients included in this study was similar in Northern Ireland and in the Republic. Likewise the OPS score, which stratified the patients into different severity groups, showed the 2 regions had almost identical profiles. Given these similarities, interesting differences were found with respect to their physical function, with other instruments measuring specific attributes of physical function, showing significant differences between patient groups.
Although aphasia was not a widespread problem according to FAST scores, Northern Ireland patients were significantly less aphasic than their counterparts in the Republic. However, results from the Barthel Index indicated Northern Ireland patients to be significantly more dependent than those in the Republic. Conversely, results from the NEADL indicated more independence in Northern Ireland patients. The Barthel measures physical function by scoring a patient’s actual ability to perform various physical tasks. NEADL measures similar tasks but by asking patients their view on their ability to perform them. Thus Northern Ireland patients had a greater self-belief in their physical ability while actually being less able than their Republic of Ireland counterparts. Results from a parallel study by this team showed that levels of functional impairment were significantly higher among a large sample (n=2033) of community-dwelling older people in Northern Ireland than in the Republic of Ireland.17 This difference was mirrored in the current study in the populations with acute illness entering hospital care.
Past medical history of these patients showed similar frequencies of hypertension and high cholesterol reported in both regions. Interestingly, the population study also showed that those in the Republic of Ireland were more likely to see a primary care physician and to see him or her more often in the previous year.18 Thus differences in preadmission drug prescribing reported in this study could not be explained by lower levels of contact with primary care services in the Republic. There were also no differences between regions with respect to patient satisfaction with primary care services.17 Differences in drug prescribing are thus likely to be related to other factors influencing primary care practice. The Northern Ireland system sets targets and provides financial incentives to primary care physicians for chronic disease management activities such as hypertension, diabetes, and asthma screening and management. This does not occur in the Republic. The poorer primary care cardiovascular management profile in the Republic may be a consequence of the difference in prioritization and reimbursement. For example, the higher prevalence of diabetes in Northern Ireland and the greater use of prevention medications may be a direct consequence of better detection as a result of financial incentives.
Regarding hospital differences in care, although results indicated substantial scope for improvement on some dimensions in Northern Ireland as well as the Republic, for instance in assessment of cognition and incontinence in the nonacute phase, the overall profile highlighted the relative dearth of multidisciplinary input in the Republic of Ireland setting and the particularly low level of documented preparation of patients for discharge and postdischarge management.
There is now an international evidence base that stroke units are a cost-effective means of providing high-quality stroke care which significantly reduce mortality and improves outcome.19,20 In 2006, Northern Ireland had 92% stroke unit provision within their hospitals. Conversely, the Republic of Ireland had only one hospital with a designated stroke unit, representing 3% of relevant hospitals. The 4 hospitals in the current study were typical: both Northern Ireland hospitals and neither of the Republic hospitals had acute stroke units. The UK National Sentinel Audit, which includes Northern Ireland in its coverage, has shown over its 3 cycles since 1998 that services for stroke can be significantly improved in a reasonable time by the introduction of stroke units. In the Republic of Ireland, by extrapolating the UK data, it is estimated that 350 to 500 deaths per year could be prevented by the introduction of stroke unit care alone.10 The present study is the first to compare the hospital management of stroke patients in 2 different but neighboring healthcare systems. The results confirm a marked contrast between them with respect to availability of multidisciplinary specialist services. Services for the acute/emergency medical aspects of care closest to professional recommendations are those involving multidisciplinary team work, rehabilitation, and community liaison issues; these most differentiate the 2 systems. The recent audit is likely to accelerate plans for stroke unit and other service needs in the Republic.21
The present study also establishes a “baseline” of functional health status, preadmission cardiovascular medication profile, and hospital management targets in acute stroke patients in 2 differing health systems in Ireland. While the preadmission physical capacity of patients was poorer in Northern Ireland, their prehospital cardiovascular disease management profile was more complete than in the Republic and their in-hospital stroke care was more in keeping with international best practice. The policy environment of established treatment targets and audits and reimbursement/investment on the basis of these is likely to explain the more developed primary care for those experiencing stroke in Northern Ireland. The challenge for the Republic, as exemplified by this comparison of hospitals in capital cities less than 100 miles apart, is to consider the system and structural changes needed to ensure all patients with stroke on the island receive equivalent and optimal care.
This study shows the value of comparing different health care systems in the management of stroke and the fact that lessons can be learnt from such comparisons. To our knowledge, it is the first prospective study in Europe comparing the impact of different healthcare systems on stroke prevention and management. Further international comparisons can inform service developments to optimize stroke care for the widest patient and family benefit.
We thank other Healthy Aging Research Programme (HARP) staff and Steering Group members who contributed in this research: Professor Ronan Conroy, Rebecca Garavan, Dr Anne Hickey, Dr Frances Horgan, Dr Karen Morgan, Dr Emer Shelley (RCSI), Dr David Hevey (TCD), and Professor Richard Layte (ESRI). We also sincerely thank research participants for their time and cooperation. We are also grateful to Professor Paul Baltes (deceased; Max Planck Institute for Human Development, Germany) and Professor Marie Johnston (University of Aberdeen, Scotland) who were external advisors to the HARP Programme.
Sources of Funding
This research was supported by a Programme Grant from the Irish Health Research Board (HRB) to: Professor Hannah McGee (Royal College of Surgeons in Ireland; Principal Investigator), Professor Desmond O'Neill (Trinity College, Dublin), Dr Tony Fahey (Economic and Social Research Institute), and Professor Bob Stout (Queen’s University Belfast; Coinvestigators).
- Received December 15, 2008.
- Accepted December 23, 2008.
Royal College of Physicians of London. National Sentinel Stroke Audit. Phase I, Organisational audit 2006 Phase 2 (clinical audit) 2006, Report for England, Wales and Northern Ireland, 2007.
Gray LJ, Sprigg N, Bath PMW, Sørenson P, Lindenstrøm E, Boysen G, De Deyn PP, Friss P, Leys D, Marttila R, Olsson J-E, O'Neill D, Ringelstein B, van der Sande J-J, Turpie AGG. Significant variation in mortality and functional outcomes: data from the tinzaparin in acute ischaemic stroke trial (TAIST). J Neurol Neurosurg Psychiatry. 2006; 77: 327–333.
Leys D, Ringelstein B, Kaste M, Hacke W. Facilities available in European hospitals treating stroke patients. Stroke. 2007; 38: 2985.
Northern Ireland Statistics and Research Agency, 2006. http://www.nisra.gov.uk/publications.
Central Statistics Office Ireland, 2006. http://www.cso.ie/statistics/population.htm.
Appleby J for Department of Health and Social Services Northern Ireland Independent review of Health and Social Care Services in Northern Ireland. Belfast: DHSS 2005.
Hurst J, Jee-Hughes M. Performance Measurement and Performance Management in OECD Health Systems. Labour Market and Social Policy - Occasional Papers, 2001;47. Paris, OECD.
Health Insurance Authority. The Private Health Insurance Market in Ireland. Dublin: Insight Statistical Consulting, 2005.
Qureshi KN, Hodkinson HM. Evaluation of a 10 question mental test in the institutionalized elderly. Age Ageing. 1974; 3: 152–157.
Jitapunkul S, Pillay I, Ebrahim S. The abbreviated mental test: its use and validity. Age Ageing. 1991; 20: 332–336.
Nouri FM, Lincoln NB. An Extended Activities of Daily Living Index for stroke patients. Clin Rehabil. 1987; 1: 301–305.
Gladman J. Validity of the Extended ADL Scale in stroke patients. Age Ageing. 1993; 22: 3–4.
McGee H, O'Hanlon A, Barker M, Hickey A, Garavan R, Conroy R, Layte R, Shelley E, Horgan F, Crawford V, Stout R, O'Neill D. One Island – Two Systems: A comparison of health status and health and social service use by community-dwelling older people in the Republic of Ireland and Northern Ireland. Dublin & Belfast: The Institute of Public Health, 2005.
Layte R, McGee H, O'Hanlon A. Do consultation charges deter general practitioner use among older people? A natural experiment. Economic and Social Research Institute, Ireland, 2007; working paper No 194.
Jarman B, Aylin P, Bottle A. Acute stroke units and early CT scans are linked to lower in-hospital mortality rates. BMJ. 2004; 328: 369.
Horgan F, Hickey A, McGee H, O'Neill D. National Audit of Stroke Care, Dublin. Irish Heart Foundation, April 2008; 1(62).