Abstract

Aims: The overall aims of this thesis were: a) to compare stroke unit (SU) care and its continuum with care on general wards (GW) for elderly patients concerning resource use and costs for care and rehabilitation focusing on assistive technology in a one-year follow-up and, b) to explore the use and cost of assistive devices (ADs), ability in daily activities and self-rated health-related quality of life (HRQL) longitudinally. Methods: Two hundred and forty-nine persons ≥70 years were followed in a randomized, prospective study evaluating stroke unit care after acute stroke. The patients were interviewed and observed four times during the first year by two registered occupational therapists. In two of the studies the patients were followed longitudinally. Results: There were no statistically significant differences in total resource use and costs for care and rehabilitation between the SUs and the GWs during the first year after stroke. Costs in relation to the severity at stroke onset showed statistically significant differences, indicating that patients with severe stroke were treated at a higher cost. Informal care widely exceeded the care provided by the community. The total cost of ADs was 2% of the total costs of care and rehabilitation during the first year. One third of the patients had ADs before stroke. One year after stroke the majority of the patients used ADs. There was a statistically significant difference between the SU group and the GW group in the proportion of patients who had supplementary ADs prescribed between 0-3 months after stroke; the patients at the SU had a higher number of uncomplicated and cheap ADs prescribed. The ADs had a large impact and were prescribed at low cost. Different types of ADs were needed at different stages in the rehabilitation process. Constant routine evaluation of elderly patients with stroke is recommended during the first year after acute stroke. High concordance was found between the assessments in the Functional Independence Measure (FIM?) and the Barthel Index (BI). The assessments in the 7-level FIM? polarize, and the intermediary levels are rarely used, suggesting that a 5-level FIM is sufficient. There were no statistically significant differences between the SU group and the GW group regarding dependence or HRQL as assessed longitudinally. There were statistically significant differences in daily activities and HRQL as assessed with the 5-level FIM and the Nottingham Health Profile (NHP) in patients who used at least one AD and those who did not. Conclusions: Since the majority of the patients with stroke used ADs one year after acute stroke but at relatively low cost, assistive technology must be considered to be one of the best tools for maintaining the highest possible level of daily activity in this patient group. Regular routine evaluations are required both for best use of resources and adaptation to individual needs. The 5-level FIM proved to be useful, but further research into its clinical utility is required. More attention should be paid to the role of spouses as caregivers in stroke survivors, since informal care carries a far greater burden than the care provided by the community. Support schemes of all types are required to relieve the spouses.     

Cost of informal care for stroke victims in a non-institutionalized general population

BACKGROUND: Stroke has a strong social impact since it causes disability, leading to dependency and the need for informal care. Although awareness of the importance of dependency is increasing, registries of the cost of informal care are lacking and consequently the real value of this activity to society is still unknown. OBJECTIVES: To calculate the cost of informal care of stroke victims in a general population, evaluate these costs according to patients' degree of dependency, and perform a one-way sensitivity analysis with variable unit costs from diverse sources. MATERIAL AND METHODS: Of all the patients with stroke diagnosed at 12-31-2004 (n = 95) among the population within a district health service of Navarre (Spain), 40 (44.4%) required informal care. Dependency for activities of daily life was evaluated by means of the Barthel (basic activities of daily life [BADL]) and Lawton-Brody indices (instrumental activities of daily life [IADL]). Time of informal care was evaluated following a bottom-up approach and diary survey method. RESULTS: The cost of the informal care of patients with stroke was 21,551.28 euros per year. According to the sensitivity analysis the range varied from 6,490.80 to 31,436.72 euros per year. Statistically significant differences in the cost of informal care were found according to patients' degree of dependency (BADL: 24,865.2 euros per year; IADL: 10,442.9 euros per year). CONCLUSIONS: The cost of informal care in ictus is high and is directly related to the degree of dependency.     

Health care resource use and stroke outcome. Multinational comparisons within the GAIN International trial

BACKGROUND AND PURPOSE: Outcome in patients hospitalized for acute stroke varies considerably between populations. Within the framework of the GAIN International trial, a large multicenter trial of a neuroprotective agent (gavestinel, glycine antagonist), stroke outcome in relation to health care resource use has been compared in a large number of countries, allowing for differences in case mix. METHODS: This substudy includes 1,422 patients in 19 countries grouped into 10 regions. Data on prognostic variables on admission to hospital, resource use, and outcome were analyzed by regression models. RESULTS: All results were adjusted for differences in prognostic factors on admission (NIH Stroke Scale, age, comorbidity). There were threefold variations in the average number of days in hospital/institutional care (from 20 to 60 days). The proportion of patients who met with professional rehabilitation staff also varied greatly. Three-month case fatality ranged from 11% to 28%, and mean Barthel ADL score at three months varied between 64 and 73. There was no relationship between health care resource use and outcome in terms of survival and ADL function at three months. The proportion of patients living at home at three months did not show any relationship to ADL function across countries. CONCLUSIONS: There are wide variations in health care resource use between countries, unexplained by differences in case mix. Across countries, there is no obvious relationship between resource use and clinical outcome after stroke. Differences in health care traditions (treatment pathways) and social context seem to be major determinants of resource use. In making comparisons between countries, great care should be exercised in using outcome variables as indicators of quality of stroke care.     

Improving the sensitivity of the Barthel Index for stroke rehabilitation

The Barthel Index is considered to be the best of the ADL measurement scales. However, there are some scales that are more sensitive to small changes in functional independence than the Barthel Index. The sensitivity of the Barthel Index can be improved by expanding the number of categories used to record improvement in each ADL function. Suggested changes to the scoring of the Barthel Index, and guidelines for determining the level of independence are presented. These modifications and guidelines were applied in the assessment of 258 first stroke patients referred for inpatient comprehensive rehabilitation in Brisbane, Australia during 1984 calendar year. The modified scoring of the Barthel Index achieved greater sensitivity and improved reliability than the original version, without causing additional difficulty or affecting the implementation time. The internal consistency reliability coefficient for the modified scoring of the Barthel Index was 0.90, compared to 0.87 for the original scoring.     

The Sunnaas Index of Activities of Daily Living: Responsiveness and Concurrent Validity in Stroke

The aim of this study was to evaluate the responsiveness and concurrent validity of the Sunnaas Index of activities of daily living (ADL) as an instrument for measuring primary and instrumental ADL functions after stroke. Fifty-five consecutive stroke patients were scored using the Sunnaas ADL Index and the Barthel ADL Index on admission and at discharge after median 21 days inpatient rehabilitation. The results of the two indices were compared, and the Barthel Index was used as a gold standard for validity tests. The neurological impairments and motor functions of the patients were scored using the Scandinavian Stroke Scale and the Modified Motor Assessment Scale. The total score of the Sunnaas ADL Index and the subscores reflecting primary and instrumental ADL functions increased significantly (p < 0.0001) during rehabilitation. The Sunnaas score correlated significantly with the scores of the Barthel Index (Spearman correlation coefficient r = 0.83 on admission and 0.88 at discharge), the Scandinavian Stroke Scale (r = 0.81 on admission) and the Modified Motor Assessment Scale (r = 0.79 on admission and 0.76 at discharge). The Sunnaas ADL Index seems able to detect clinically important improvements of primary and instrumental ADL functions after stroke, i.e. its responsiveness is good. These results provide further evidence for the concurrent validity of the Sunnaas Index for measuring functional recovery after stroke.     

Systematic review of economic evidence on stroke rehabilitation services

OBJECTIVES: Given the resource-intensive nature of stroke rehabilitation, it is important that services be delivered in an evidence-based and cost-efficient manner. The objective of this review was to assess the evidence on the relative cost or cost-effectiveness of three rehabilitation services after stroke: stroke unit care versus care on another hospital ward, early supported discharge (ESD) services versus "usual care," and community or home-based rehabilitation versus "usual care." METHODS: A systematic literature review of cost analyses or economic evaluations was performed. Study characteristics and results (including mean total cost per patient) were summarized. The level of evidence concerning relative cost or cost-effectiveness for each service type was determined qualitatively. RESULTS: Fifteen studies met the inclusion criteria: three on stroke unit care, eight on ESD services, and four on community-based rehabilitation. All were classified as cost-consequences analysis or cost analysis. The time horizon was generally short (1 year or less). The comparators and the scope of costs varied between studies. CONCLUSIONS: There was "some" evidence that the mean total cost per patient of rehabilitation in a stroke unit is comparable to care provided in another hospital ward. There is "moderate" evidence that ESD services provide care at modestly lower total costs than usual care for stroke patients with mild or moderate disability. There was "insufficient" evidence concerning the cost of community-based rehabilitation compared with usual care. Several methodological problems were encountered when analyzing the economic evidence.     

Costs and disability among stroke patients

BACKGROUND: The high costs of health and social care support for stroke survivors, and the development of new service arrangements, have concentrated growing attention on economic issues. However, there are few data on costs and their association with levels of disability. METHODS: Secondary analyses of data from the OPCS (Office of Population Censuses and Surveys) Surveys of Disability conducted in the mid-1980s were used to examine service utilization and costs for more than 1000 people who have had a stroke. Costs were estimated for all health and social care services. Regression analyses examined the cost-disability association in the context of other covariates for people living in private households. RESULTS: Disability problems were common among stroke survivors, particularly in relation to locomotion, self-care and holding. Among people living alone, the major contributors to costs were in-patient care (Pound Sterling 27 per week) and home help (Pound Sterling 30 per week). Among people living with others, in-patient hospital care was also a major cost (Pound Sterling 28 per week). Other services costing more than Pound Sterling 5 per week were general practitioner consultations, hospital out-patient care and day centre attendances. Resource use patterns varied considerably. Costs were associated with severity of disability, time since stroke and whether the person was living alone. Looking at the overall balance of care, a greater proportion of stroke survivors with severe disability were resident in communal establishments. CONCLUSION: The analyses provide a baseline from which more recent local studies and evaluations can be compared. Key issues for economic studies of stroke are the inclusion of a broad range of services, a reasonable duration of follow-up and consideration of the impact of the substitution of informal for formal services.     

Future costs of stroke in the Netherlands: the impact of stroke services

OBJECTIVES: In the next decades, the number of stroke patients is expected to increase. Furthermore, organizational changes, such as stroke services, are expected to be implemented on a large scale. The purpose of this study is to estimate the future healthcare costs by taking into account the expected increase of stroke patients and a nationwide implementation of stroke services. METHODS: By means of a dynamic multistate life table, the total number of stroke patients can be projected. The model calculates the annual number of patients by age and gender. The total healthcare costs are calculated by multiplying the average healthcare costs specified by age, gender, and healthcare sector with the total number of stroke patients specified by age and gender. RESULTS: In the year 2000, the healthcare costs for stroke amounted to euro 1.62 billion. This amount is approximately 4.4 percent of the total national healthcare budget. Projections of the total costs of stroke based on current practice result in an increase of 28 percent (euro 2.08 billion) in the year 2020. A nationwide implementation of stroke services in 2020 would result in a substantial reduction of the costs of stroke (euro 1.81 billion: 13 percent cost reduction) compared with the regular care scenario. CONCLUSIONS: A nationwide implementation of stroke services is a strong policy tool for cost containment of health care in an aging population like that in the Netherlands. Policy makers should optimize the organization of stroke care.     

Diabetes in Canada: Direct Medical Costs of Major Macrovascular Complications

Objectives: To estimate direct medical costs of managing major macrovascular complications in diabetic patients.
Methods: Costs were estimated for acute myocardial infarction (AMI) and ischemic stroke by applying unit costs to typical resource use profiles. Data were obtained from many Canadian sources, including the Ontario Case Cost Project, provincial physician and laboratory fee schedules, provincial formularies, government reports, and peer-reviewed literature. For each complication, the event costs per patient are those associated with resource use specific to the acute episode and any subsequent care occurring in the first year. State costs are the annual costs per patient of continued management. All costs are expressed in 1996 Canadian dollars.
Results: Acute hospital care accounts for approximately half of the first year management costs ($15,125) of AMI. Given the greater need for postacute care, acute hospital care has less impact (28%) on event costs for stroke ($31,076). The state costs for AMI and stroke are $1544 and $8141 per patient, respectively.
Conclusions: Macrovascular complications of diabetes potentially represent a substantial burden to Canada's health care system. As new therapies emerge that may reduce the incidence of some diabetic complications, decision makers will need information to make critical decisions regarding how to spend limited health care dollars. Published literature lacks Canadian-specific cost estimates that may be readily translated into patient-level cost inputs for an economic model. This paper provides two key pieces of the many needed to understand the scope of the economic burden of diabetes and its complications for Canada.     

Social and economic costs and health-related quality of life in stroke survivors in the Canary Islands, Spain

ABSTRACT: BACKGROUND: Cost-of-illness analysis is the main method of providing an overall vision of the economic impact of a disease. Such studies have been used to set priorities for healthcare policies and inform resource allocation. The aim of this study was to determine the economic burden and health-related quality of life (HRQOL) in the first, second and third years after surviving a stroke in the Canary Islands, Spain. METHODS: Cross-sectional, retrospective study of 448 patients with stroke based on ICD 9 discharge codes, who received outpatient care at five hospitals. The study was approved by the Research Ethics Committee of Nuestra Senora de la Candelaria University Hospital. Data on demographic characteristics, health resource utilization, informal care, labor productivity losses and HRQOL were collected from the hospital admissions databases and questionnaires completed by stroke patients or their caregivers. Labor productivity losses were calculated from physical units and converted into monetary units with a human capital-based method. HRQOL was measured with the EuroQol EQ-5D questionnaire. Healthcare costs, productivity losses and informal care costs were analyzed with log-normal, probit and ordered probit multivariate models. RESULTS: The average cost for each stroke survivor was [euro sign]17 618 in the first, [euro sign]14 453 in the second and [euro sign]12 924 in the third year after the stroke; the reference year for unit prices was 2004. The largest expenditures in the first year were informal care and hospitalizations; in the second and third years the main costs were for informal care, productivity losses and medication. Mean EQ-5D index scores for stroke survivors were 0.50 for the first, 0.47 for the second and 0.46 for the third year, and mean EQ-5D visual analog scale scores were 56, 52 and 55, respectively. CONCLUSIONS: The main strengths of this study lie in our bottom-up-approach to costing, and in the evaluation of stroke survivors from a broad perspective (societal costs) in the first, second and third years after surviving the stroke. This type of analysis is rare in the Spanish context. We conclude that stroke incurs considerable societal costs among survivors to three years and there is substantial deterioration in HRQOL.     

三级单位卒中单元模式对脑卒中疾病经济负担的影响

目的:探讨三级单位卒中单元在降低医疗费用、实现医疗资源合理配置方面的作用效果。方法:建立三级单位卒中单元,包括急性卒中病房、康复中心和家庭护理单位。通过完备的专人随诊体系,对卒中病人提供卒中管理,随诊1年。于急性期、康复期、3个月末、6个月末和12个月末共进行5次评价,统计各阶段的医疗费用。结果:三级单位卒中单元内患者,脑出血及脑梗塞急性期例均住院费用分别为7547元和6078元,仅为全国同类省级医院平均住院费用的57.30%及59.30%;平均住院日分别为14日和11日,均较全国同类省级医院为短。脑卒中患者发病1年内例均疾病负担为15074元,其中,入院急性期费用占45.57%,康复期占21.86%,出院康复期及间接费用占32.57%。结论:三级单位卒中单元是治疗脑卒中的一种有效方法,能明显减少住院时间和医疗费用,将有限的医疗资源合理分配到疾病全程,有助于调整医疗费用的结构,实现医疗资源的合理配置。     

Self ratings of health predict functional outcome and recurrence free survival after stroke

STUDY OBJECTIVE: To measure stroke victims' self rated health (SRH) status and SRH transition, and to compare how the two are prospectively associated with disability and recurrence free survival. DESIGN: Prospective case registry study with face to face follow up interviews at three months, one, two, and three years. Ascertained were SRH status and SRH transition using single question assessments, Barthel Index (BI), Frenchay Activities Index (FAI), and Mini Mental State Examination (MMSE). SETTING: A multiethnic inner city population of 234 533. PARTICIPANTS: Patients surviving the initial three months after a first in a lifetime stroke in 1995 to 1998. RESULTS: Of 690 stroke survivors 561 (81.3%) could complete the self report items. Answers to the item on SRH status did not vary significantly between the four follow up interviews. However, responses to the item on SRH transition changed significantly during follow up with three months ratings being more negative than all subsequent ratings. SRH transition, but not SRH status, showed a prospective association with long term outcome in multivariate analyses controlling for the BI, FAI, and MMSE. Compared with all other patients, patients reporting "Much worse health" at three months were more likely to be disabled ( = BI<20) at one year (OR 6.29, 95% CI 2.26 to 17.52) and their combined risk of stroke recurrence and death was increased over five years (HR 1.72, 95% CI 1.25 to 2.38). CONCLUSIONS: Items on SRH should be used with caution in populations with high rates of disability and language problems, as many participants are unable to complete them. SRH transition may be a better predictor of disability and recurrence free survival after major medical events than SRH status.     

'Neurodevelopmental treatment' following stroke: no beneficial effect shown in measurements taken after 1 year; comparative study

OBJECTIVE: To determine whether neurodevelopmental treatment (NDT) in the care of stroke patients is effective with respect to the functional status and quality of life (QoL) during one year after stroke onset. DESIGN: Prospective, non-randomised, comparative parallel group design. METHODS: 324 consecutive stroke patients from 12 Dutch hospitals were divided into 2 groups: an experimental group (n=223), in which nurses and physiotherapists used the NDT approach, and a control group who received conventional therapy (n=101). Functional status was assessed with the Barthel Index. Primary outcome was considered poor when the Barthel Index <12 after 1 year or when the patient had died. QoL was assessed with the 'Stroke adapted sickness impact profile'-30 and on a visual analogue scale. RESULTS: At 12 months, 59 patients in the NDT group (26%) and 24 patients in the control group (24%) had a poor outcome (corresponding adjusted odds ratio: 1.7; 95% CI: 0.8-3.5). At point of discharge and after 6 months, the adjusted odds ratio was 0.8 (95% CI: 0.4-1-5) and 1.6 (95% CI: 0.8-3.2) respectively. The adjusted mean differences of the QoL measurements did not show statistically significant differences between the 2 study groups at 6 and 12 months after stroke onset. CONCLUSION: The NDT approach was not an effective method in the care of stroke patients. Health care professionals need to reconsider the use of the NDT approach.     

Budget Impact Analysis of Thrombolysis for Stroke in Spain: A Discrete Event Simulation Model

Objective:  Thrombolysis within the first 3 hours after the onset of symptoms of a stroke has been shown to be a cost-effective treatment because treated patients are 30% more likely than nontreated patients to have no residual disability. The objective of this study was to calculate by means of a discrete event simulation model the budget impact of thrombolysis in Spain.
Methods:  The budget impact analysis was based on stroke incidence rates and the estimation of the prevalence of stroke-related disability in Spain and its translation to hospital and social costs. A discrete event simulation model was constructed to represent the flow of patients with stroke in Spain.
Results:  If 10% of patients with stroke from 2000 to 2015 would receive thrombolytic treatment, the prevalence of dependent patients in 2015 would decrease from 149,953 to 145,922. For the first 6 years, the cost of intervention would surpass the savings. Nevertheless, the number of cases in which patient dependency was avoided would steadily increase, and after 2006 the cost savings would be greater, with a widening difference between the cost of intervention and the cost of nonintervention, until 2015.
Conclusion:  The impact of thrombolysis on society's health and social budget indicates a net benefit after 6 years, and the improvement in health grows continuously. The validation of the model demonstrates the adequacy of the discrete event simulation approach in representing the epidemiology of stroke to calculate the budget impact.     

脑卒中医院-社区-家庭三级连续康复治疗的效果评价

观察脑卒中急性期后在“医院一社区一家庭三级康复治疗”、持续在大医院康复治疗和家庭康复三种方案的治疗效果及医疗费用的差异。方法：随机选取101例脑卒中患者分为三组,A组：30例,医院一社区一家庭三级康复治疗;B组：30例,持续在大医院康复治疗者。c组：41例,从综合医院内科直接进入家庭康复治疗者;5三组分别在康复治疗前与治疗后3个月、6个月采用Barthel指数、简化的Fugl—Meyer评定量表、社区康复肢体残疾功能评定表（CEDA）进行评估,比较三组同期之间ADL、运动功能改善以及回归社会情况的差异和治疗费用差异。结果：三组患者康复治疗前ADL评分、简化的Fuel—Meyer评分、CEDA评分无显著差异（P〉0．05）;康复治疗后3个月、6个月评估,A与B组比较无显著差异（P〉0．05）。A与c比较差异具有统计学意义（P〈0．05）;A与B组治疗费比较,A组明显少于B组。结论：“医院一社区一家庭三级康复治疗”与大医院康复治疗同样可以提升脑卒中患者的ADL、运动功能,减轻肢体残疾情况,治疗费用方面前者具有价廉优势。     

Community support for young physically handicapped people

This study examined the cost of home and hospital care for individuals who are severely physically handicapped and who may have many years to live. In addition to quantifying the overall level of these services and their costs, factors which might influence the level of service were identified and their use related to the diagnosis, duration and severity of disability, and number of unpaid carers. Costs of health and social service community support increased with severity of disability, and were mostly incurred by individuals with a Barthel score of less than 50 (more severe handicap). Costs, except for those individuals with cerebral palsy, were independent of the number of careers, and for those of school and university age the costs of special educational facilities were significant.     

乳腺常见肿瘤的良恶性彩超鉴别诊断

目的：探讨彩超在乳腺常见良恶性肿瘤鉴别诊断方面的应用价值。方法：选取120例经手术病检为乳腺良恶性肿瘤患者,依据肿块声像学特点和血流参数来鉴别诊断。结果：通过二维超声像图鉴别乳腺常见肿瘤的检出率明显低于二维超声像图结合多普勒超声血流特点鉴别疾病的检出率。与金标准手术病检结果相比,A 组二维超声像图鉴别结果的灵敏度为76.1%明显低于 B 组的92.5%,差异有统计学意义。结论：应用二维超声像图来鉴别乳腺常见肿瘤的灵敏度和特异度不高,但是应用彩色多普勒超声,结合二维超声像图和血流特点来鉴别常见乳腺肿瘤,灵敏度和特异度高,是早期乳腺常见肿瘤诊断的有效方法。     

An economic evaluation of a perindopril-based blood pressure lowering regimen for patients who have suffered a cerebrovascular event

Objectives  Cerebrovascular disease (or stroke) is one of the main causes of long-term disability and the second leading cause of death worldwide. The economic impact of stroke is clearly seen, as it is the largest single cause of bed occupancy in hospitals in England and accounts for 6% of hospital costs. This analysis is the first to quantify the economic consequences of a blood pressure lowering regimen based on the PROGRESS study (perindopril-based regimen), for reducing future cardiovascular events. Design  A Markov decision analytical model was used to estimate the cost per quality adjusted life year (QALY) of blood pressure lowering in the treatment of patients presenting with a cerebrovascular event. The health states are based upon Barthel indices for which resource utilisation and health benefits have previously been estimated. Setting  The participants for the economic analysis were obtained from the PROGRESS study database. 6,105 clinical study participants were recruited through both primary and secondary care centres. Participants  The mean age was 64 years; 70% were male in the original study. Interventions  In the PROGRESS study, blood pressure lowering by a perindopril-based regimen was compared to standard care. Main outcome measures  Cost per quality adjusted life year for the duration of the study (4 years) and for a time span of 20 years. Results  Using only direct hospital medical costs, the cost per QALY for a perindopril based regimen is ￡6,927 for the base study period and ￡10,133 for a 20-year time period. These results are sensitive to the cost of perindopril, the cost of the stroke unit, length of stay, and to a lesser extent, the cost of indapamide. Conclusions  This analysis demonstrates a cost-effective treatment for patients suffering a cerebrovascular event with a blood pressure lowering regimen. The findings of this study are in line with current decisions and guidance by the national institute for health and clinical excellence (NICE) in England.     

Cost function estimation: the choice of a model to apply to dementia

Statistical analysis of cost data is often difficult because of highly skewed data resulting from a few patients who incur high costs relative to the majority of patients. When the objective is to predict the cost for an individual patient, the literature suggests that one should choose a regression model based on the quality of its predictions. In exploring the econometric issues, the objective of this study was to estimate a cost function in order to estimate the annual health care cost of dementia. Using different models, health care costs were regressed on the degree of dementia, sex, age, marital status and presence of any co-morbidity other than dementia. Models with a log-transformed dependent variable, where predicted health care costs were re-transformed to the unlogged original scale by multiplying the exponential of the expected response on the log-scale with the average of the exponentiated residuals, were part of the considered models. The root mean square error (RMSE), the mean absolute error (MAE) and the Theil U-statistic criteria were used to assess which model best predicted the health care cost. Large values on each criterion indicate that the model performs poorly. Based on these criteria, a two-part model was chosen. In this model, the probability of incurring any costs was estimated using a logistic regression, while the level of the costs was estimated in the second part of the model. The choice of model had a substantial impact on the predicted health care costs, e.g. for a mildly demented patient, the estimated annual health care costs varied from DKK 71 273 to DKK 90 940 (US$ 1 = DKK 7) depending on which model was chosen. For the two-part model, the estimated health care costs ranged from DKK 44714, for a very mildly demented patient, to DKK 197 840, for a severely demented patient.