Skip to main content

PhD Reflections: A study of the impact of an ageing population on local health care demand (Stephen Clark)

Category
health
microsimulation
phd
Date

Author: Stephen Clark.s.clark

My PhD began in January 2012 and, nearly 4 years later, the end is in sight. The topic area is close to epidemiology but with a strong geographic context. My task was to try and forecast, for the medium term, the health condition of the older population of England at a geography that would be useful for health care planning purposes. This was a new area of study for me, previously I had worked as a transport researcher and planner, so much of the first year was spent trying to define the terms within my thesis title (“How to measure health?” “Is health care delivery also important?” “What is a suitable geography”). In the end I decided to use information on the presence of a doctor diagnosed morbidity, and chose two of the three morbidities that incur most hospital care and cause most deaths in England, cardio vascular disease and respiratory illness. Looking at the literature however I realised that I needed to also include diabetes since this is a common co-morbidity with cardio vascular disease. Since April 2013, much public health planning in England is coordinated at the local authority district level, so this was my chosen level of geographic forecasting.

The overall methodology and links to thesis chapters

The overall methodology and links to thesis chapters

My methodology to produce these forecasts brought together a number of methodological elements that are represented here. Firstly I needed a spatial microsimulated population of those aged 50 or older in each local authority within England for 2011. This is nearly 16 million people, each of which is individually represented in the model. I then aged this population in two year time steps. During each step I needed to update individual’s morbidity status using the predicted probabilities from a series of hazard models. I also needed to re-shape the population along projected lines for its gender, age and ethnic composition, all of which are important determinants of health. Some nifty interpolation was also needed to make sure that over time my population of interest still covered the age range 50 and older.

So, what do my results show? The maps below show the number of people aged 50 and older forecast to have each morbidity in 2031 (a circle) and the percentage of these people (the colouring) in each local authority. Firstly, the good news. Both the number and percentage of the population with cardio vascular disease disease will decline over the period 2011 to 2031. This is attributable to a number of factors mainly related to changes in occupation and lifestyles and will occur in all local authorities to varying degrees. The equally good news is that whilst the numbers with a respiratory illness will stay fairly level over the 20 years the percentage in the population with a respiratory illness will fall. The most import factor to influence this result is the reduction in smoking that occurred in younger cohorts working itself into this age group. Now the bad news. Both the number and the percentage with diabetes will increase. Many studies have shown that diabetes is more prevalent in certain ethnic communities, particularly the south Asian community. As these communities age and become a more common in the 50 years and older age group, they will carry this heightened prevalence with them. There are also some lifestyle factors than are contributing to this increase, such as greater obesity in the population.

Prevalences and number of individuals with morbidities, LADs, 2031. Top left: CVD. Top right: diabetes or high blood sugar. Bottom left: respiratory illness. Botton right: CVD and diabetes or high blood sugar

Prevalences and number of individuals with morbidities, LADs, 2031. Top left: CVD. Top right: diabetes or high blood sugar. Bottom left: respiratory illness. Botton right: CVD and diabetes or high blood sugar

As I hinted above, my viva was last week and after some tough questioning I was awarded a pass subject to making some changes to the thesis. This result was such a relief and my heartfelt thanks go out to those colleagues here in CSAP who have helped me along the way. I was pleased to be joined by colleagues and examiners at a celebratory meal that evening.

After Stephen's post-viva meal

After Stephen's post-viva meal

Author: Stephen Clark.