Four-Fold Increase in Direct Costs of Stroke Survivors With Spasticity Compared With Stroke Survivors Without Spasticity
The First Year After the Event
Background and Purpose—The prevalence of spasticity after first-ever stroke is approximately 20%, but there are no health economic studies on costs associated with spasticity after stroke. The objective of our study was to estimate direct costs of stroke with spasticity for patients surviving up to 1 year after the stroke event in comparison to costs of stroke without spasticity.
Methods—A representative sample of patients with first-ever stroke hospitalized at Uppsala University Hospital was eligible for our cross-sectional survey. All direct costs during 1 year were identified for each patient, including costs for hospitalization (acute and rehabilitation), primary health care, medication, and costs for municipality services. Swedish currency was converted to Purchasing Power Parities US dollar (PPP$).
Results—Median age (interquartile range) was 73 years (18), and the proportion of women was 48%. The majority of the direct costs (78%) was associated with hospitalization, whereas 20% was associated with municipality services during 1 year after a first-ever stroke. Only 1% of all direct costs were related to primary health care and 1% to medication. The level of costs for patients with stroke was correlated with the presence of spasticity as measured with the modified Ashworth scale (rs=0.524) and with the degree of disability as measured with modified Rankin Scale (rs=0.624). The mean (median, interquartile range) direct cost for stroke patients with spasticity was PPP$ 84 195 (72 116, 53 707) compared with PPP$ 21 842 (12 385, 17 484) for patients with stroke without spasticity (P<0.001).
Conclusions—Direct costs for 12-month stroke survivors are 4 times higher than direct costs for patients with stroke without spasticity during the first year after the event.
The costs for stroke are huge as is the variation of costs for stroke between and within countries.1 Reliable estimates of the costs are crucial for allocation of resources, and it is likely that specific estimations have to be made for different countries. Spasticity is a common motor disorder after stroke with a rate of 20% or higher.2–5 Although the clinical impact of spasticity after stroke as well as the exact definition of spasticity remain subject to debate,6–8 numerous physical and pharmacological interventions have been designed to target this motor disorder.9–11 There is increasing evidence that reversible, chemical denervation of spastic muscles by use of intramuscular injections of botulinum toxin12 can reduce focal limb spasticity.13–15 Previous studies,16–18 based on expert opinion according to Delphi panel surveys,19 indicate that treatment with botulinum toxin for spasticity might be cost-effective. In addition, there are 2 ongoing studies20,21 that are investigating the cost-effectiveness of botulinum toxin treatment for spasticity in patients with stroke.
Until now, there are no health economic analyses from a representative sample of patients with spasticity after stroke. The aim of the present study was to estimate the direct costs related to spasticity for patients who had a first-ever stroke and who had survived for 12 months. Therefore, we compared direct costs of stroke with spasticity with corresponding costs of stroke without spasticity in the same cohort of patients.
Materials and Methods
We followed a checklist developed by Evers et al.22 According to their checklist, our study can be described in the following terms.
The perspective of this study is a broad healthcare sector.
Study Type, Design, Estimation Procedures, Costs Included, and Unit of Measurement
The design was disease-specific, incidence-based, and bottom-up, including direct costs. In a bottom-up analysis, the costs are based on real costs. A group of patients with a specific disease is followed for a time period. Our patients were followed during 1 year after a first-ever stroke and all direct costs (both stroke- and nonstroke-related) were included. The Swedish currency (Swedish Krona [SEK]) was converted into Purchasing Power Parities23 US dollar (PPP$) at the exchange rate of 9.34 SEK for 1 PPP$ (2003 value).24 Purchasing Power Parities23 (PPPs) are rates of currency conversion that eliminate the difference in price levels between countries. It is recommended to convert local currencies into PPPs22 because it eliminates the variation in monetary levels between countries. The PPPs are exchange rates for a given basket of goods and services. We have used the PPP exchange rate given by the Organisation for Economic Co-operation and Development in which a PPP$ equals the purchasing power of $1 in the United States (Supplemental Tables I and II⇓). The Organisation for Economic Co-operation and Development describes PPP at their web site (www.oecd.org/std/ppp).
Data Source Population, Population Size, Inclusion and Exclusion Criteria, and Recruitment
Patients were recruited from the Swedish Stroke Register. Stroke was defined according to the World Health Organization criteria.25 In Sweden, >95% of the patients with stroke are admitted to a hospital.26–28 A detailed description of the study sample has been reported previously.2 Briefly, those eligible for the study were patients with stroke hospitalized at Uppsala University Hospital between January 2003 and April 2004. This hospital is responsible for the population in 4 municipalities with a total of 244 000 inhabitants. Inclusion criteria for our study were as follows: (1) resident in the catchment area; (2) age >18 years; (3) a first-ever stroke (cerebral infarction or intracerebral hemorrhage) with survival of at least 1 year; and (4) ability to give informed consent. Exclusion criteria were as follows: (1) any other neurological disorder that might affect muscle tone; (2) transient ischemic attack; and (3) subarachnoid hemorrhage. One hundred forty patients were included and investigated 1 year after the stroke.
The national healthcare system in Sweden is tax-financed. Patients have the possibility to visit private healthcare providers. In 2003, the cost for health care in our county was 440 million PPP29 with private health care providing 1.1% of the costs. There are no private hospitals in our country, but there are private primary healthcare centers.
Assessment Methods and Definitions
The following parameters and definitions were used: age, stroke type (ischemic versus hemorrhagic), gender, hypertension, diabetes, National Institutes of Health Stroke Scale, modified Rankin Scale (mRS), and the Modified Ashworth Scale (MAS).
Spasticity was assessed by MAS, a 6-point ordinal scale with documented reliability.30,31 Assessment included flexion and extension movements around upper (shoulder, elbow, wrist, and fingers) and lower extremity (hip, knee, and ankle) joints with the patient in a resting position. Spasticity was defined as a MAS score ≥1 at any of the passive movements.2–5
Included in the estimations of direct cost was the following:
Costs for Hospitalization
The patient administrative system at Uppsala University Hospital provided us the specific costs for each patient related to the hospitalization, for example, costs for medical personnel, diagnostic investigations, and medication and laboratory tests at all levels of care (Table 1). Costs for readmittances to the hospital during the first year after stroke were included.
Costs After Discharge
Primary Health Care
We scrutinized the medical records of the primary health care (both community and private care) for each patient after discharge from the hospital and recorded all outpatient visits at general practitioners, district nurses, physiotherapists, occupational therapists, dieticians, and speech therapists. A general practitioner consultation equals PPP$ 164, and the price for all other consultations and/or treatments was PPP$ 48.
All medications for each patient at 1 year after stroke were noted, and costs were calculated assuming that the same medication was used during the whole period.
The finance department of the 4 municipalities provided the total costs for each patient regarding home help service, residential and nursing home, adjustment of housing, transportation, food delivery, and home alarm (Table 1).
Before entering the study, the patients received written and oral information, and participation was based on informed consent. All participants also consented to have their resource use data abstracted from health finance systems. The study was approved by the Regional Human Ethics Committee (Ups. 03-595).
Numeric variables were explored for normal distribution. Age, National Institutes of Health Stroke Scale scores, and direct costs had a nonnormal distribution; hence, we used median (interquartile range) values for comparison. For future health economic calculations, the mean is also reported.
Direct costs of patients with spasticity were compared with direct cost of patients without spasticity in our stroke cohort by univariate analysis.
A sensitivity analysis was performed leaving out the hospital costs. Bootstrapping32 was also performed to give an alternative mean and confidence interval to that calculated from raw data.
We used χ2 for gender and stroke type and Mann-Whitney for age, National Institutes of Health Stroke Scale, mRS, MAS, and direct costs. In addition, we compared the direct costs for different patient groups regarding mRS and MAS with the Kruskal-Wallis test. We used Spearman correlation coefficient, rs, for bivariate correlation analyses between mRS and the direct costs and MAS and direct costs, respectively. A correlation coefficient between 0.5 and 0.75 was regarded as moderate to good. A probability value (2-tailed) <0.05 was set as significant. SPSS Version 16.0 for Macintosh (SPSS, Chicago, Ill) was used for statistical analyses.
Median age (interquartile range) of the study sample was 73 years (18), 75 (20) for women and 73 (18) years for men. The proportion of women was 48%. The stroke subtype was ischemic in 89% and hemorrhagic in 11% of the patients. The frequency of spasticity was 18%. Patients with stroke with spasticity did not differ from patients with stroke without spasticity with respect to gender, stroke type, hypertension, and diabetes (Table 2).
Distribution of Costs
The majority of the costs, 78%, were caused by hospital care, whereas 20% were due to services provided by the municipalities. Only 1% of all costs were generated by primary health care and another 1% by medication. During the hospital stay, patients had full access to the whole stroke team, including doctors, nurses, physiotherapists, occupational therapists, dieticians, and speech therapists. The mean costs for all patients were PPP $32 976 (Table 3). Six patients had extremely high costs for hospital care (PPP $100 000) due to long or complicated hospitalization, including intensive care.
Costs for Outpatient Care
The outpatient care at the primary health centers varied considerably; 38% of the patients had one or 2 visits to their general practitioner, and 27% had one or 2 visits to the district nurse during the first year after stroke. Only 4% visited a physiotherapist, 2% an occupational therapist, 1% a dietician, and no visits were recorded at a speech therapist after discharge from the hospital.
Costs for Stroke With Spasticity Compared With Stroke Without Spasticity
Direct costs for hospital care, primary care, and municipality services community health care as well as the sum of direct costs were higher for spastic patients than nonspastic patients (Table 3; Figure 1). The mean direct cost for patients with stroke with spasticity was PPP$ 84 195 compared with PPP$ 21 842 for other patients with stroke (P<0.001).
Costs were strongly associated with functional ability (Figure 2) with increasing costs for each mRS level (rs=0.624, P<0.001). Costs were also associated with spasticity (Figure 3) with higher costs for increasing MAS scores (rs=0.524, P<0.001).
A sensitivity analysis, by excluding hospital costs, showed 5-fold higher costs (mean values) for patients with spasticity. Bootstrapping, using data from the patients with spasticity (n=25), showed a somewhat higher mean value and wider confidence interval with PPP$ 99 329 (95% CI, 49 517 to 182 475).
Our study reports a 4-fold increase in direct costs for patients with stroke with spasticity compared with patients with stroke without spasticity. We found a substantial increase of both costs for hospital care and for municipality services in the patient group with spasticity. The costs for medication and primary health care, on the other hand, were low in relation to other costs and did not differ between patients with spasticity and patients without spasticity.
Reliable estimates of the costs of stroke are crucial for allocation of resources. Due to differences in national health financial systems, specific estimations are probably needed for different countries and have to be regularly updated as new treatments become available. In a recent review of the costs for stroke,1 the average costs ranged from PPP$ 7342 to PPP$ 146 149 for a 12-month follow-up between different countries. The large variation that was found was mainly attributed to differences in population, method, and in the costs that were included in the calculations. Differences in costs were also found within some countries; for example, in the United States, the costs varied 20-fold.
All patients in our study received standard medical care, including access to a multiprofessional team at the stroke unit and during early inpatient rehabilitation. The frequency of outpatient follow-up depended on local resources and contacts with therapists were surprisingly few. Although 20% (5 of 25) of patients with spasticity were prescribed benzodiazepines for various indications, no patient received baclofen, dantrolene, or tizanidine. Botulinum toxin was considered an exclusion criterion, but no patient was excluded for that reason.
It must be pointed out that our study does not provide evidence that spasticity as such is responsible for the increase of costs. However, our results indicate that spasticity reflects a more severe motor disorder that is clearly associated with disability, as illustrated by the MAS and the mRS running almost in parallel. Our data may provide a baseline for further studies of the cost-effectiveness of interventions, including botulinum toxin, and may be used to calculate the total costs for stroke with spasticity. The incidence rate of first-ever stroke in Sweden is 250 per 100 000 inhabitants,33 and two thirds of the patients survive the first year after the event.34 Assuming that 20% of these stroke survivors is afflicted by spasticity, direct costs for stroke with spasticity would amount to PPP$ 4209 750 per 100 000 inhabitants/year, including unavoidable cost (for initial intensive care) as well as potentially avoidable costs (for rehabilitation and municipality services).
The strength of the present study is that estimations are based on real costs using the bottom-up method. Previous studies of costs for stroke with spasticity16–18 have been based on Delphi panels, and this type of top-down estimate is subject to bias with the risk of inaccuracy. A limitation is that our sample only includes patients who survived the first year after first-ever stroke; hence, we excluded costs associated with care of patients who died during this period. It is likely that patients who died during the first year after stroke had a more severe stroke. Another limitation is that we omitted indirect costs such as loss of income by reduced work capacity. Although the majority of Swedish patients with stroke (80%)33 are >65 years of age, this is probably an important factor. Finally, the estimated costs were not separated with regard to costs derived from stroke versus nonstroke. Specification of disease-related costs is difficult to make and would have required closer monitoring of the patients.
In conclusion, our findings indicate that direct costs the first year after stroke are huge and approximately 4 times higher for patients who develop spasticity after stroke compared with stroke without spasticity. Our data may be used as a baseline for cost-effectiveness studies of therapeutic interventions of spasticity after stroke in Sweden.
We thank all the patients who participated in this study, the staff at the Stroke Unit at Uppsala University Hospital, and Oskar Åsberg, BBA.
Sources of Funding
This study was supported by grants from the Selanders Foundation at the Uppsala University Hospital and by Stroke-Riksförbundet.
- Received May 27, 2009.
- Revision received July 12, 2009.
- Accepted August 4, 2009.
Luengo-Fernandez R, Gray AM, Rothwell PM. Costs of stroke using patient-level data. A critical review of the literature. Stroke. 2009; 40: e18–e23.
Sommerfeld DK, Eek EU, Svensson AK, Holmqvist LW, von Arbin MH. Spasticity after stroke: its occurrence and association with motor impairments and activity limitations. Stroke. 2004; 35: 134–139.
Watkins CL, Leathley MJ, Gregson JM, Moore AP, Smith TL, Sharma AK. Prevalence of spasticity post stroke. Clin Rehabil. 2002; 16: 515–522.
Boyd RN, Ada L. Physiotherapy management of spasticity. In: Barnes MP, Johnson GR, eds. Upper Motor Neurone Syndrome and Spasticity: Clinical Management and Neurophysiology. Cambridge: Cambridge University Press; 2001: 96–121.
Charlton PT, Ferguson DW. Orthoses, splinting and casting in spasticity. In: Barnes MP, Johnson GR, eds. Upper Motor Neurone Syndrome and Spasticity: Clinical Management and Neurophysiology. Cambridge: Cambridge University Press; 2001: 142–187.
Ward AB, Ko Ko C. Pharmacological management of spasticity. In: Barnes MP, Johnson GR, eds. Upper Motor Neurone Syndrome and Spasticity: Clinical Management and Neurophysiology. Cambridge: Cambridge University Press; 2001: 165–187.
Wade D. Stroke rehabilitation: the evidence. In: Greenwood J, ed. Handbook of Neurological Rehabilitation. Hove, East Sussex: Psychology Press; 2003: 487–504.
Wallesch C-W, Maes E, Lecomte P, Bartels C. Cost-effectiveness of botulinum toxin type a injection in patients with spasticity following stroke: a German perspective. Eur J Neurol. 1997; 4: S53–S57.
Radensky PW, Archer JW, Dournaux SF, O'Brien CF. The estimated cost of managing focal spasticity: a physician practice patterns survey. Neurorehabil Neural Repair. 2001; 15: 57–68.
Rodgers H, Shaw L, Price C, van Wijck F, Barnes M, Graham L, Ford G, Shackley P, Steen N. Study design and methods of the BoTULS trial: a randomised controlled trial to evaluate the clinical effect and cost effectiveness of treating upper limb spasticity due to stroke with botulinum toxin type A. Trials. 2008; 9: 59.
Allergan. Botox Economic Spasticity Trial (BEST). 2008. Available at: http://clinicaltrials.gov/ct2/show/NCT00549783. Last accessed on January 9, 2009.
Evers SMAA, Struijs JN, Ament AJHA, van Genugten MLL, Jager JC, van den Bos GAM. International comparison of stroke cost studies. Stroke. 2004; 35: 1209–1215.
Vachris MA, Thomas J. International price comparisons based on purchasing power parity. Monthly Labor Review Online. 1999 122: 3–12. Available at: http://www.bls.gov/opud/mlr/1999/10/art1full.pdf. Last accessed on January 9, 2009.
Organisation for economic co-operation and development (OECD). Purchasing power parities for GDP, 2003. Available at: http://stats.oecd.org/wbos/Index.aspx?datasetcode=SNA_TABLE4. Accessed July 12, 2009.
Terént A. Increasing incidence of stroke among Swedish women. Stroke. 1988; 19: 598–603.
Landstingsförbundet. Statistik om hälso- och sjukvård samt regional utveckling 2003. Verksamhet och ekonomi i landsting och regioner. Statistics on health care and regional development for 2003. Activities and finances of county councils and regions [English translation]. 2003;2009. Available at: http://brs.skl.se/brsbibl/kata_documents/doc35987_1.pdf (in Swedish). Last accessed on May 13, 2009.
Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther. 1987; 67: 206–207.
Kirkwood BR, Sterne JA. Bootstrapping. Medical Statistics. Oxford: Blackwell Science; 2003: 350–353.
Hallström B, Jönsson AC, Nerbrand C, Norrving B, Lindgren A. Stroke incidence and survival in the beginning of the 21st century in southern Sweden: comparisons with the late 20th century and projections into the future. Stroke. 2008; 39: 10–15.
Terént A. Trends in stroke incidence and 10-year survival in Söderhamn, Sweden, 1975–2001. Stroke. 2003; 34: 1353–1358.
Cassel G. Abnormal deviations in international exchange. The Economic Journal. 1918.
Fregart K, Jonung L. Makroekonomi: Teori, politik och institutioner. Macroeconomics: Theory, Policy and Institutions [English translation]. Lund: Studentlitteratur; 2005.
The People’s Bank of China. Objective of the monetary policy. Available at: www.pbc.gov.cn/english/detail.asp?col=6630&ID=21. Accessed July 12, 2009.
US Department of Treasury. Major foreign holders of treasure securities holdings. Available at: www.treas.gov/tic/mfhhis01.txt. Accessed July 12, 2009.
Board of Governors of the Federal Reserve System. Foreign exchange rates (annual) 2003. Available at: www.federalreserve.gov/releases/g5a/20040102/. Accessed July 12, 2009.
World Bank. Relative prices and exchange rates. Available at: http://devdata.worldbank.org/wdi2005/Table5_7.htm. Accessed July 12, 2009.