Status
Conditions
Treatments
About
The HOME Project evaluates the effects of structured health dialogues with individuals aged 67-84 years in the municipality of Borgholm, Sweden. A combination of registry data and survey responses will be used to monitor quality of life, morbidity, healthcare needs, and lifestyle factors over a six-year period. Outcomes will be compared between randomized groups within Borgholm municipality and a matched control group from seven other municipalities in Region Kalmar.
The project also includes an analysis of cost-effectiveness and the reach of the intervention. A qualitative interview study will explore participants' perceptions of their health, their motivations for health improvement, and their experiences of how the health dialogues may influence these aspects.
In a subproject, machine learning models will be developed to predict functional decline and high healthcare needs among older adults. These models will be validated against established risk assessment tools such as the Adjusted Clinical Groups (ACG) system and the Charlson Comorbidity Index. Digital motion analysis using Skeleton Avatar Technology will be employed both independently and in combination with other variables to support model development.
Full description
BACKGROUND: The world faces a demographic shift with an aging population and increasing numbers of older adults experiencing frailty and complex care needs [1, 2]. Effective preventive strategies are requested from many stakeholders [3]. Health dialogues have been introduced in several Swedish regions to promote healthier aging, yet evidence for their effectiveness and cost-efficiency remains limited. While systematic reviews have not confirmed significant effects on morbidity or mortality [4, 5], some cohort studies suggest benefits in cardiovascular outcomes [6, 7] and a few primary preventive interventions have been directed at older individuals[8] . The results are disputed, and several authors argue that health dialogues and health checks tend to reach individuals with lower cardiovascular risk rather than those at highest risk [9]. There is a lack of studies linking the outcomes of health dialogues to different risk levels.
One strategy to identify individuals in the population at highest risk of morbidity is to use risk assessment instruments. In primary care populations, Adjusted Clinical Groups (ACG) and Charlson Comorbidity Index (CCI) are the most studied [10]. In recent years, several predictive instruments/models using existing health and medical data have been developed to identify older individuals at risk of future functional decline and morbidity [11]. However, the relatively moderate precision of these instruments limits their clinical usefulness. Digital motion analysis using the SAT (Skeleton Avatar Technology) technique has shown potential in assessing physical activity levels, mobility, and balance in older individuals [12, 13]. Still, data from broader populations of older individuals with a wide range of diseases and functional levels are lacking, and the method has not been tested for its predictive ability regarding functional decline and extensive care needs.
Among older individuals, the wide variation in health and functional levels creates a greater need for individualized advice and interventions [14]. Holistic interventions targeting frail older individuals are well-studied, and there is some evidence of positive effects, although results are also conflicting [15, 16]. International recommendations point to a multifactorial causal relationship behind frailty/functional decline, and there is consensus that increased physical activity and reduced malnutrition can counteract these issues, and that an active lifestyle is associated with a reduced risk of frailty [17].
A current question is which health outcomes are most important for the older population. Studies show that many older individuals value good quality of life and high independence more than maximum lifespan[18, 19] . As a complement to traditional measurements of health-related quality of life using EQ-5D, which focuses on symptoms and function, it is proposed to measure "Capability": the ability to perform activities that are meaningful and important to the individual [20]. ICECAP-O is a quality-of-life instrument that has also been used for health economic evaluations [21].
Research on communication shows that conversations focusing on individualized advice for older individuals can have a motivating effect regarding preventive health efforts and lifestyle changes [22]. Older individuals are often aware of lifestyle-related problems but may need support to implement changes[23]. To understand what older individuals need to make changes, further research is needed on which factors are essential to the patients themselves to carry out lifestyle changes.
AIM:
We will evaluate the short - and long term effects of preventive health dialogues forolder individuals and the cost-effectiveness . The project also aims to evaluate and further develop models to predict the risk of future illness.
Research questions:
We will also examine which residents choose to participate and how participation affects both self-rated health and quality of life, as well as health and social care needs. This is important from an equity perspective: can health dialogues help reduce health disparities, or is there a risk that they reinforce existing inequalities in health between groups? For the same reason, we are conducting a cost-effectiveness analysis of health dialogues in an older population, examining the balance between cost and benefit.
The qualitative part of the project aims to increase understanding of what happens in the communication during health dialogues and what older individuals themselves perceive as influencing their motivation to take personal responsibility for their health.
Hypotheses:
STUDY DESIGN This is a mixed-methods study combining quantitative and qualitative approaches.The quantitative component includes a controlled intervention study with three groups: intervention, randomized control, and matched control. The qualitative component explores participants' experiences and perceptions through interviews and thematic analysis.
DATA COLLECTION Outcomes 1,2,11 and 12 will be collected by a postal/web-based questionnaire. All other outcomes will be collected from healthcare registries.
ANALYSIS
Effects of Health Dialogues:
The following comparative analyses will be conducted:
A: Differences between the intervention arm and the passive control arm (overall and stratified by risk level).
B: Differences between the two randomized arms and the matched controls (overall and stratified by risk level).
C: Differences between participants and non-participants within the intervention arm (overall and stratified by risk level).
Differences in primary outcomes between intervention and control groups will be estimated using 95% confidence intervals. Appropriate statistical methods will be applied based on the type of outcome measure. Subgroup estimates will be calculated similarly, and relevant statistical tests for heterogeneity will be used. Risk stratification will be based on Adjusted Clinical Groups (ACG), Charlson Comorbidity Index (CCI), frailty status, and a predictive model for hospital admissions.
The aim of these analyses is to obtain unbiased estimates of statistically significant differences in outcomes between groups. Given the randomized design, statistically significant differences will be interpreted as effects of the intervention. Analyses will follow the intention-to-treat (ITT) principle, which maintains group assignment throughout the study and is appropriate for this design.
Cost-Effectiveness Analysis:
The cost-effectiveness evaluation will estimate the total costs of the intervention using established health economic methods. The analysis will adopt a societal perspective. Costs will be compared to selected outcome measures and benchmarked against alternative resource use scenarios, such as no health dialogue or other similar interventions.
Development of Risk Prediction Models:
A combination of variables from electronic health records (diagnoses, healthcare contacts), surveys (self-rated health, symptoms, lifestyle factors), and digital motion analysis (SAT) will be collected. Machine learning models will be trained on these data to predict functional decline (ADL limitations and care needs) and morbidity (hospitalizations, healthcare visits, mortality). Techniques such as logistic regression, random forest, and neural networks will be used to optimize prediction accuracy. Model performance will be evaluated using AUC/ROC (Receiver Operating Characteristic Curve) and MAE (Mean Absolute Error).
Motivational Factors in Health Dialogues:
The qualitative sub-study aims to explore whether and how health dialogues influence motivation for adopting health-promoting behaviors. Interviews will focus on changes in participants' views on personal responsibility for health, lifestyle choices, and behavioral change. Data will be analyzed using qualitative content analysis, without predefined categories or themes.
REFERENCES:
Enrollment
Sex
Ages
Volunteers
Inclusion criteria
Exclusion criteria
Primary purpose
Allocation
Interventional model
Masking
2,952 participants in 2 patient groups
Loading...
Data sourced from clinicaltrials.gov
Clinical trials
Research sites
Resources
Legal