IPN 2017 Case Competition

 

To download the case competition data + additional resources and access the users’ forum, navigate to the member’s area in the left hand panel 

 

 

Aging is normally accompanied by a mild decline in cognitive functions, such as memory and decision-making. However, dementia is characterized by a more severe decline, as well as significant changes in mood and behaviour. The most common form of dementia is Alzheimer’s disease, which accounts for ~60% of all dementia cases. Other potential causes of dementia include vascular dementia, Parkinson’s disease, Lewy body dementia, fronto-temporal degenerations, traumatic brain injuries, and excessive alcohol consumption^[1]. Dementia can severely impact the ability of a person to live independently, and consequently adds burden to healthcare systems.

 

There are currently over 6 million Canadians (or 16.5% of the population) aged 65 years and older^[2], and this is expected to increase to up to 11 million (23.2%) over the next 20 years^[3]. With a prevalence rate of 9.5% for adults aged 65 years and older, there are currently an estimated 571,000 Canadians living with dementia, with that number projected to grow to 1,153,000 by 2037^[4]. The projected costs for dealing with this increased burden to the healthcare system are high, and researchers still have much to learn.

 

Objective

Identify a brain health-related problem currently facing Canada’s aging population and make recommendations based on your analysis of the data.

 

Submissions should consist of three sections:

First round submissions should be presented in this linear order, and each section will be assessed separately.

 

Part 1 – Motivation

What is the problem or question that your team is proposing to address? It must relate to aging and brain health. Please provide a brief introduction describing the problem and an explanation of how the dataset can answer the question.

 

Part 2 – Data Analytics

Describe the details of the analytics work you have done to identify the problem and/or the solution. You do not have to analyse all variables in the dataset, but you will be evaluated based on your ability to select appropriate variables to answer your problem or question. In keeping with the theme of the conference, you must include some variables related to brain health in your analysis (e.g., neuropsychological assessments, incidence of dementia, etc.), but you may also include non-brain related variables as well. You may use multivariate statistical modelling techniques. Your conclusions from this section may have the form of something like (but not exclusive to):

 

• “PCA/Clustering/Classification/Regression/Graphical model analyses on variables X predicts outcome Y”
• “Analyses + other information + common sense indicate a causal relationship of variables X on outcome Y. Therefore we propose to target X to improve Y”

 

You will be marked on the quality and performance of your applied techniques, appropriateness and correctness of usage, applicability to the problem, documentation, and provision of code.

 

Part 3 – Recommendation

Propose a solution to the identified problem based upon the results of the data analyses. This may refer to existing policies, infrastructure, or technologies, and may be addressed to patients/families, government, healthcare providers, or industry.

 

Description of Dataset

The Canadian Study of Health and Aging (CSHA) was a longitudinal study conducted nationally to investigate patterns and risk factors of health and disease in aging, including dementia and Alzheimer’s disease, as well as health service utilization. The CSHA sample includes responses from 10,263 people aged 65 and over. The sample was drawn from 39 urban and rural areas across Canada, equally from five geographic regions (British Columbia, Prairie provinces, Ontario, Quebec, and the Atlantic region). The sample included 9,008 participants from community settings, and 1,255 from long-term care settings^[5]. More information about sampling methods, and how the data should be weighted to properly estimate population parameters, can be found in McDowell et al (2001)^[6].

 

CSHA participants were assessed at three 5-year intervals (1991, 1996, 2001); however, for the IPN2017 Case Competition, you will be working only with the 1991 dataset.

 

Included in the CSHA 1991 dataset are 1,724 variables, which were culled from community and caregiver interviews, clinical examinations, neuropsychological examinations, physical measurements, and provincial health care utilization records. For a high level description of available variables see: http://csha.ca/r_data_collected.asp. Note that some variables are only available in the 1996 and the 2001 dataset, but only the variables within the 1991 dataset will be provided.

 

For more information about the CSHA see: http://csha.ca/default.asp

 

References

1. Ng, R., Maxwell, C.J., Yates, E.A., Nylen, K., Antflick, J., Jette, N., & Bronskill, S.E. (July, 2015). Brain Disorders in Ontario: Prevalence, Incidence and Costs from Health Administrative Data. Toronto, Ontario. Institute for Clinical Evaluative Sciences.
2. Statistics Canada. Population estimates (year 2016): http://www.statcan.gc.ca/tables-tableaux/sum-som/l01/cst01/demo10a-eng.htm. Accessed 2017.
3. Statistics Canada. Population estimates (year 2036): http://www.statcan.gc.ca/tables-tableaux/sum-som/l01/cst01/demo23g-eng.htm. Accessed 2017.
4. Canadian Study of Health and Aging Working Group (1994). Canadian Study of Health and Aging: study methods and prevalence of dementia. Canadian Medical Association Journal, 150(6): 899-913.
5. Canadian Study of Health and Aging Working Group. Study methods. http://csha.ca/r_study_methods.asp. Accessed 2017.
6. McDowell, I., Aylesworth, R., Stewart, M., Hill, G., & Lindsay, J. (2001). Study Sampling in the Canadian Study of Health and Aging. International Psychogeriatrics, 13(Suppl 1): 19-28.