The Geography of Alzheimer's Disease in New Zealand: A Spatial Epidemiology Authors Hamish Robertson Nick Nicholas A/Prof Joanne Travaglia A/Prof Tuly Rosenfeld
Jul 16, 2015
The Geography of Alzheimer's Disease in New Zealand:
A Spatial Epidemiology
AuthorsHamish Robertson
Nick NicholasA/Prof Joanne TravagliaA/Prof Tuly Rosenfeld
Contents
• Introduction
• The dementias in New Zealand
• A spatial model
• Mapping the dementias
• Spatial visualisation
• Dementia sub-types
• Service infrastructure
• Conclusion
• Future developments
Introduction
• We keep on saying this – location matters and more so as populations age!
• Information systems need to reflect the world people live, work and survive in
• Geography is central to understanding social policy predicaments because nothing is uniformly distributed – quantity and quality
• Spatial science goes beyond geography to include a variety of approaches centered on space/place relations
• The distribution of health, illness, people and health systems will always be spatially patterned
The Dementias in New Zealand
• Population ageing – NZ still fairly young population compared to many countries -> specific groups, differential ageing and locational issues
• Context - growing international knowledge but still far from complete – ADI, WHO etc
• Signs of dynamic variation in prevalence rates e.g. Denmark, the UK and Australia
• Limited information base in New Zealand but this will improve in time
• 2008 paper by Tobias et al – Burden of Alzheimer's disease: population-based estimates and projections for New Zealand, 2006-2031
• 2012 report update by Access Economics for Alz NZ
The Role of Geographic Knowledge• Situations vary be location because populations (social and biological) and
environments vary by location
• Geography supports physical/social system complexity and applied technology (e.g. GIS, GPS, virtual earth, simulation etc)
• Scale is an important factor often missed in modelling activities e.g. often assume sameness up and down in complex systems but this is very problematic (also Boje on systemicities)
• Need to consider interdisciplinarity for coping better with and understanding ageing – not just medicine or health sciences exclusively -> meta-science of ageing
• Most service providers need to understand ageing better – health, finance, social services, legal, police, transport etc
• Ageing is both personal and collective, highly local and globally important – geography helps link these conceptually and practically
• Ageing is (also) a space-place experience – as personal experience will attest
A Spatial Model
• Prior developmental work presented in 2012
• Modelling updated with 2013 NZ Census, AD estimate data and GIS software
• Maptitude GIS software – NZ (maps) geography and population data in the one package -> low learning curve and cheap too!
• Illustration of these issues using basic prevalence estimates and 2013 Census data
• NB - not just technology for its own sake…
Official Geographic Boundaries in New Zealand
• meshblock boundaries
• area unit boundaries.
• general and Māori electoral district boundaries
• regional council boundaries
• territorial authority boundaries
• ward boundaries - in these examples mostly
• community boards and local board boundaries
• BUT you can also create your own geographies as well – map community or group perceptions
Estimated Prevalence 2001By Ward
Estimated Prevalence 2006By Ward
Estimated Prevalence 2013By Ward
Other Forms of Visual EngagementTree Mapping the Same Data
Scale Factors for Different Audiences
• Keynes said governments don’t like too much information because it makes their decisions harder (!)
• The experience of ageing will differ by location e.g. access to appropriate/quality services, quality/experience/availability of staff, choice, family, community etc
• Sometimes larger places are better, sometimes smaller ones – varies by factor e.g. formal services versus informal care and support
• Scale is central to mapping because the results people perceive change with scale – e.g. global versus neighbourhood
Localised Prevalence Changes Over Time
Costing Shifts in Changing Epidemiology
• Shifting dynamics of public versus private service provision (NFP, personal, group?)
• Composition and management vary in significant ways globally
• Philosophical and political debates about who pays for what (if you can buy it)
• Implicit rationing in much of the health and social support system – who gets access to what?
• Costs can be dynamic over time – not just linear
• Impacts of different services can vary over time
Estimated Costs by Ward
Some Service Infrastructure Issues
• Demand will differ by location – geography and scale will matter
• Will we have enough facilities, places and people to service demand now and into the future?
• Where will these issues be lesser or greater and what patterns are we likely to see?
• What will be the downstream impacts on services and suppliers?
• What will we do in places where more skilled people won’t live and work?
• What should we be doing now for those future events?• What options do we need to plan for now and trial/test for
future scenarios?• What will we use the facilities/people for when population
ageing peaks?
The Dynamics of Service Provision and Demand
Source: NZ Ministry of Health 2004 report via Joyce De La Torre on Academia.edu
Impact on Acute Hospitals by Distance
Spatial Visualisation
• Visualisation is increasingly central to information sharing and access
• Broad audiences and the public may not share the same understanding of an issue – visualisation adds value to these often complex situations
• Dashboards and other visual formats are increasing in health informatics
• Spatial data representation methods are rising rapidly e.g. qualitative software, Tableau, data mining packages etc
• No longer an expert domain – open source etc
Dementia Sub-Types
• We can estimate and map (spatially model) sub-types – AD, VaD, DLB, mixed dementias and so on
• Ageing is likely to produce new/emerging conditions just because of the sheer numbers of very old people
• Service issues associated with sub-types can then be modelled e.g. acute, sub-acute, specialist etc
• As data improves assumptions can be tested and revised to better support what is actually happening
Conclusion
• Dementia and sub-types represent a highly dynamic aspect of the epidemiology of ageing and flow-on effects
• High investment socially, economically and politically
• Spatial technology is moving very fast and supports complexity work – not a replacement but an addition
• Visuo-spatial methods can inform and support the many people and professions involved in population ageing and its consequences
• Also these techniques are increasingly accessible, interesting and useful
• Good science makes use of what is available and works
Future Developments
• Mapping incidence by address/location• Refining and combining prevalence estimates and
incidence data -> spatial data mining applications• Expand options for visualisation and access by a
broad audience• Building systems for knowledge integration – not
just more data collection in silos• Advance ‘what if’ modelling for trends and
options• Ethics of knowledge and care will expand i.e. if we
hold/possess knowledge and don’t act or advocate in the interests of the community
New Zealand Maptitude Bundle