A New Stroke Unit: Implementation Issues

A New Stroke Unit: Implementation Issues

Research Report Hong Kong Physiother J 2009;27:7–10 A NEW STROKE UNIT: IMPLEMENTATION ISSUES Lorraine Sheppard,1,2 PhD; Kwai Fu Ko,3 PhD Abstract: ...

100KB Sizes 6 Downloads 88 Views

Research Report

Hong Kong Physiother J 2009;27:7–10

A NEW STROKE UNIT: IMPLEMENTATION ISSUES Lorraine Sheppard,1,2 PhD; Kwai Fu Ko,3 PhD

Abstract: Stroke units are geographically distinct units providing dedicated care for people who have suffered a stroke. This study aimed to understand the financial benefits of stroke unit care. The cost of acute beds and the average length of hospital stays were compared between the stroke unit group and the general ward group. The mean length of inpatient stay in the stroke unit group was 43.7 days; in the general medical wards, it was 69.3 days with a significant 95% confidence interval of 36.96–14.19. The reduced length of stay reduced the cost of care provided by the stroke unit. Stroke units provide more cost-effective care.

Key words: Chinese, financial, Hong Kong, implementation, stroke unit

Introduction The number of stroke patients is expected to increase as society ages [1], because stroke risk increases with age. In Hong Kong, stroke is the third highest leading cause of death [2–4]. Stroke occurs at approximately 1 to 2 cases per 1,000 of population [5] and accounted for 3,130 deaths out of 33,305 registered deaths (i.e. 9.4%) in 2001 [4]. Stroke also ranked fourth on the list of disease conditions with the highest hospitalization rates in 2001, with the number of hospitalizations at 24,702, a figure only surpassed by cancer, end-stage renal disease, and chronic lung disease [4]. As the incidence of stroke increases with age [6], it is also expected that elderly patients will have a poorer outcome compared with young stroke patients [7]. In Hong Kong, with the general ageing of the population, the aftermath of stroke is likely to be a major health care issue. Those aged 65 years and over increased from 482,040 (8.7%) in 1991 and 629,555 (10.1%) in 1996 to 747,052 (11.1%) in 2001 [8]. Furthermore, the expansion of life expectancy means that stroke will become a highly prominent worldwide problem [9]. On average, a baby boy born in Hong Kong in 2001 could expect to live 78.4 years and a baby girl to 84.6 years. However, there has been a steady rise in the life expectancy of the population in Hong Kong over the past two decades, and that of Hong Kong is among the highest in the world [10].

In the face of the growing population of elderly persons in Hong Kong, the burden of stroke is likely to increase substantially in the future. A stroke unit is a geographically discrete ward or location where a disease-specific service is provided by a team working exclusively in the care of stroke patients [11–13]. A systematic review of stroke unit trials published in 1993 [14] was complemented by further reviews of trials in 1997, with both showing mortality reductions with stroke unit care [15]. In 2000, the Stroke Unit Trialists’ Collaboration further examined outcomes of stroke units. On average, the mortality was reduced from 27% in general wards to 22% in a stroke unit. There was a similar reduction in the combined outcomes measure of death and dependency from 68% to 61% [16]. The comprehensive stroke unit (acute stroke services combined with early rehabilitation) of Kwong Wah Hospital was established in January 2001. The stroke unit consisted of 10 designated beds (six for males and four for females) for receiving patients with acute stroke. It was situated next to the rehabilitation ward so that rehabilitation of the patients could be conveniently followed. It was not a stroke intensive care unit (ICU). In addition to the stroke unit, the hospital also had one separate general ICU and one coronary care unit. In terms of medical treatment regimes, pharmacotherapy, nursing care and rehabilitation methods, the

1

School of Public Health, Tropical Medicine, and Rehabilitation Sciences, James Cook University, Queensland, School of Health Sciences, University of South Australia, South Australia, Australia, and 3Department of Medicine, Kwong Wah Hospital, Hong Kong SAR, China. Received: 18 November 2008 Accepted: 2 December 2009 Reprint requests and correspondence to: Dr Lorraine Sheppard, School of Health Sciences, University of South Australia, City East Campus, North Terrace, Adelaide, South Australia 5000, Australia. E-mail: [email protected]

2

Hong Kong Physiotherapy Journal • Volume 27 • 2009 ©2009 Elsevier. All rights reserved.

7

ward design and environment were similar in the stroke unit and the general medical wards. In fact, patients in both the stroke unit and the general wards were under the care of the same department. They were under the care of the same group of doctors, including neurologists, and the same group of physiotherapists, occupational therapists, neuropsychologists, speech therapists, and social workers. All patients in the stroke unit and in the general medical wards received a computerized tomography scan within 48 hours, an electrocardiogram, and routine blood tests on admission; other diagnostic procedures, such as Holter monitoring for detection of cardiac arrhythmias, were performed when indicated. The same procedures of the pathology, laboratory and radiology departments applied to all wards and units in the hospital. If respirators or continuous cardiological monitoring were required, the stroke patients could remain in the same stroke unit and in the same general medical wards, unless intensive care in the ICU was mandatory. One very important difference, however, was that the medical treatment in the stroke unit followed a programme that was standardized with regard to diagnostic evaluation, acute treatment and rehabilitation, while the treatment in the general medical wards did not. A randomized controlled trial was conducted at the Kwong Wah Hospital comparing those treated in a stroke unit and those treated in a general medical ward. Chinese subjects receiving comprehensive stroke unit care are associated with less mortality and shorter length of hospital stay than those having conventional care in general medical wards [17]. As part of a larger study, stroke unit treatment was analysed for the same patient group considered in this study. The study was previously reported by Ko and Sheppard [17]. It was found that stroke unit treatment significantly reduced both mortality and poor outcome (mortality and institutional placement) of patients with acute stroke after 28 days and 120 days. After 28 days, mortality was 3.3% and 17.2% for the stroke unit group and general ward group, respectively, whereas after 120 days, mortality was 5% and 24.7% for the stroke unit group and general ward group, respectively. While the stroke unit care was associated with better survival and less institutional placement than general medical ward care after 28 and 120 days, stroke unit care was further demonstrated by logistic regression analysis to have a contribution to reduction of mortality (p = 0.03) and reduction of poor outcomes (which includes mortality and institutionalization) (p = 0.03) at 120 days. At 28 days, there was just a trend for the stroke unit to have a contribution to the reduction of mortality (p = 0.05), but for poor outcome, the effect contributed by the stroke unit care at 28 days was not statistically significant (p = 0.48) [17]. This data and cohort were used to understand the financial implications of this change in service delivery.

8

The aim of this study was to understand the costeffectiveness of a new service, a dedicated stroke unit.

Methods A financial study of the stroke unit was conducted. The cost of an acute bed and the average length of hospital stay were compared between the stroke unit group and the general ward group. A cost estimate of the difference between the two groups in the study could not be performed because of the intrinsic design of the general wards, where mixed groups of inpatients with different diseases were treated. The cost of an acute bed and the length of hospital stay were compared between the two groups instead. The staffing was similar for both general wards and the stroke unit; however, the stroke unit coordinator was employed in the stroke unit and two consultants focused on the stroke unit, rather than being visiting consultants to the general wards. Kwong Wah Hospital serves an urban catchment population of approximately 635,570 (9.5% of total population) in the Yau Ma Ti, Mongkok, Tsim Sha Tsui, and Shamshuipo regions of Kowloon in Hong Kong [1] Thirteen percent of the population was 65 years or older [1]. The boundaries of the catchment areas were somewhat indistinct and overlap with other regional hospitals, in the east with Queen Elizabeth Hospital and in the west with Caritas Medical Centre. Patients with all types of stroke were included in the study, but transient ischemic attacks (focal neurological deficits with complete restoration of function within 24 hours), subarachnoid haemorrhages, and subdural haematomas were excluded. These patients were without metabolic disturbance, trauma or intoxication, which could account for the presentations. They were admitted to either the stroke unit or the general medical wards. There was no age limitation. However, those patients with known nonvascular neurological disease, recent head trauma or intoxication, or those requiring intensive or coronary care were admitted to general medical wards only. When the stroke unit was full, patients with suspected acute stroke could only be managed in the acute general medical wards. As there were approximately 65 patients with acute stroke admitted to the hospital each month (> 800 patients per year), the majority of stroke patients had to be treated and managed in the general wards, not in the stroke unit, because of the limiting factor of the number of beds available in the stroke unit.

Results From April 2001 to April 2002, 677 patients with acute stroke were admitted to the Department of Medicine Hong Kong Physiotherapy Journal • Volume 27 • 2009

and Geriatrics of Kwong Wah Hospital. A total of 188 patients in the stroke unit group and 177 patients in the general ward group were recruited. Group matching of the two groups of stroke patients was performed by requiring the study subjects to meet similar premorbid conditions (Barthel index ≥ 90) and stroke score (National Institutes of Health Stroke Scale ≥ 3) [18]. Patient data were then analysed for differences between the two groups. No significant differences existed concerning age, sex, and premorbid status between the two groups (Table 1). The age of the subjects ranged from 38 to 93 years, and the male-to-female ratio was approximately 3:2 in both study groups. Although the mean premorbid Barthel index was 99.41 in the stroke unit group and 98.88 for the general medical group (p = 0.04), 100% of all the patients in the two groups had a premorbid Barthel index of 90 or above and, therefore, were independent in daily living before the strokes. The mean length of inpatient stay was 43.7 days in the stroke unit group and 69.3 days in the general medical wards, with a significant 95% confidence interval of 36.96–14.19 (Table 2). The average cost of an acute bed day in the stroke unit was HK$1,995 in 2000–2001 (data from the Financial Department of Kwong Wah Hospital). The data for direct comparison with the costs of stroke patients who were managed in the general wards were not available, because these costs were not subdivided into diagnosis categories.

According to the hospital administration’s best estimate, the cost of an acute bed day (including stroke and nonstroke patients) in a general medical ward in 2000– 2001 was HK$1,885. However, this was an average figure for all patients in general medical wards. No isolated data for the costs of stroke patients in the general wards were available.

Discussion The mean length of hospital stay in the stroke unit was shorter than in the general medical wards. One study stated that cost savings were unlikely to accrue unless the average length of stay was reduced by at least 4 days or there were savings of more than 10% of the average total direct costs [19]. The major cost determinants of care of patients with acute stroke in a public hospital are related to supportive therapy and length of hospital stay, including convalescence. There was a slightly higher cost per bed day in the stroke unit than in the general wards because of the extra cost of a stroke nurse specialist and two consultants. It was not related to equipment, external and internal environments, or the accounting system, which were very similar to those in the general medical wards. The medical therapies were not considered to differ between the stroke unit care and general ward care. The cost of an

Table 1. Baseline characteristics in the two patient groups

Women, n (%) Age, mean ± SD (yr) Premorbid BI Patients with premorbid BI ≥ 90, n (%)

Stroke unit group (n = 188)

General ward group (n = 177)

83 (44.15) 70.29 ± 9.65 99.410 ± 1.808 188 (100)

62 (35.03) 71.62 ± 10.78 98.880 ± 2.599 177 (100)

95% CI

Standard error of difference

p

–0.009 to 0.189 –3.433 to 0.773 0.071 to 0.989

0.051 1.070 0.233

0.07 0.19 0.04

CI = confidence interval; SD = standard deviation; BI = Barthel index.

Table 2. Outcome after acute hospitalization

Death To old age home To convalescent hospital Discharge home Mean LOS in acute and convalescence hospital (d) Remained hospitalized in convalescent hospital

Stroke unit group (n = 188)

General ward group (n = 177)

95% CI

Standard error of difference

5 (2.66%) 4 (2.13%) 77 (40.96%) 93 (49.47%) 43.739

21 (11.86%) 5 (2.82%) 79 (44.63%) 70 (39.55%) 69.316

–0.150 to –0.040 –0.045 to 0.029 –0.137 to 0.064 –0.003 to 0.198 –36.962 to –14.192

0.027 0.016 0.052 0.052 5.789

81 (43.09%)

84 (47.46%)

–0.144 to 0.058

0.052

CI = confidence interval; LOS = length of stay.

Hong Kong Physiotherapy Journal • Volume 27 • 2009

9

acute bed day in the stroke unit was very close to that in the general medical wards. Hence, the stroke unit demonstrated a beneficial effect during the study period without a significant increase in expenditure. Hospital care (particularly during the acute phase of a stroke) represents a large proportion of stroke care cost; institutional care such as placement in governmentsubsidized homes for the elderly also contributes significantly to overall stroke care cost. The cost of stroke is further reduced by stroke unit care, which showed an increased number of patients discharged home (p = 0.052) and fewer institutional needs (p = 0.03 for poor outcome at 120 days in the logistic regression analyses) [17]. Costs in the longer term are largely attributable to the care of dependent individuals in hospitals or nursing homes [20]. Although hospital care represents a large proportion of the cost of stroke, institutional care also contributes significantly to overall stroke care costs [21]. Therefore, long-term (direct and indirect) costs are likely to be determined by the number of patients with longterm disability. It follows that stroke unit care will be more cost-effective than conventional general ward care if it reduces long-term disability without increasing the length of stay in a hospital. Because most of the direct hospital costs are attributed to nursing salaries and hospital overheads with relatively little being spent on investigation or specific treatment, the cost to the stroke unit service of managing a patient with stroke is highly dependent on the length of hospital stay. This assumes that the intensity of nursing input remains constant. Therefore, a policy of accelerated discharge is likely to reduce the per capita costs considerably, although, depending on funding arrangements, this is likely to be transferred to another budget, e.g. community care.

Conclusion Our estimations of the cost of an acute bed day and the average length of hospital stay demonstrate that stroke unit care is not more expensive than general medical ward care. The favourable length of stay and patient disposition (reduced institutional placement and increased number of patients discharged home) indicate that the stroke unit achieves savings at the institutional and societal levels. Stroke is a major cause of hospital resource consumption, not only in acute care hospitals but also in institutional long-term care. Although hospital care represents a larger proportion of the cost of stroke, institutional care also contributes significantly to overall stroke care costs. Reduced institutional placement and increased number of patients discharged home in the stroke unit group than in the general ward group were found. Because the medical therapies did not differ between the stroke unit care and general ward care, the

10

favourable length of stay and patient disposition suggest that the stroke unit care achieved savings at both the institutional and societal levels.

References 1. Census and Statistics Department, The Government of the Hong Kong SAR. 2001 Population Census: A Profile of Older Persons, 2001. Available at: http://www.censtatd.gov.hk/major_projects/ 2001_population_census/index.jsp 2. Thorvaldsen P, Kuulasmaa K, Rajakangas AM, et al. Stroke trends in the WHO MONICA Project. Stroke 2007;28:500–6. 3. Murray CJL, Lopez AD. Global burden of disease study. Lancet 1997;349:269–76. 4. Statistics and Research Unit, Professional Services and Medical Development Division, Hospital Authority. Hospital Authority Statistical Report 2001–2002. Hong Kong: Hospital Authority, March 2003. 5. Statistical Information Section, Hospital Authority. Hospital Authority Statistical Report 1995/96. Hong Kong: Hospital Authority, December 1996. 6. Bamford J, Dennis M, Sandercock P, et al. The frequency, causes and timing of death within 30 days of a first stroke: the Oxfordshire Community Stroke Project. J Neurol Neurosurg Psychiatry 1990;51:1373–80. 7. Ebrahim S. Clinical Epidemiology of Stroke. New York: Oxford University Press, 1990. 8. Census and Statistics Department, The Government of the Hong Kong SAR. 2001 Population Census: Summary Results. Available at: http://www.censtatd.gov.hk/FileManager/EN/ Content_41/01c_v1_s.pdf 9. Poungvarin N. Stroke in the developing world. Lancet 1998; 352(Suppl 3):SIII19–22. 10. Chan M. Annual Departmental Report 2001/2002. Hong Kong: Department of Health, 2003. 11. Zubair Kareem M. Guidelines for stroke center development. Stroke 2001;32:816–8. 12. Alberts MJ, Hademenos G, Latchaw RE, et al. Recommenda tions for the establishment of primary stroke centres. Brain Attack Coalition. JAMA 2000;283:3102–9. 13. Sinha S, Waraburton EA. The evolution of stroke units— towards a more intensive approach? QJM 2000;93:633–8. 14. Langhorne P, Williams BO, Gilchrist W, et al. Do stroke units save lives? Lancet 1993;342:395–8. 15. Stroke Unit Trialists’ Collaboration. Collaborative systemic review of the randomised trials of organised inpatient (stroke unit) care after stroke. BMJ 1997;314:1151–9. 16. Stroke Unit Trialists' Collaboration. Organised inpatient (stroke unit) care for stroke. Cochrane Database Syst Rev 2007;(4): CD000197. 17. Ko KF, Sheppard LA. The contribution of a comprehensive stroke unit to the outcome of Chinese stroke patients. Singapore Med J 2006;47:208–12. 18. Whisnant J. Stroke: Populations, Cohorts, and Clinical Trials. Oxford: Butterworth-Heinemann, 1993. 19. Smurawska LT, Alexandrov AV, Bladin CF, et al. Cost of acute stroke care in Toronto, Canada. Stroke 1994;25:1628–31. 20. DerSimonian R. Algorithm AS 221: maximum likelihood estimation of a mixing distribution. Appl Stat 1986;35: 302–9. 21. Bonita R. Epidemiology of stroke. Lancet 1992;339:342–4.

Hong Kong Physiotherapy Journal • Volume 27 • 2009