- Home » Data methodology
Health Inequalities
Data Methodology
Background
In seeking to better understand young people’s experiences of health inequalities, we looked to see which health data are available by deprivation breakdowns. Unfortunately, many data aren’t collected and reported by deprivation characteristics. Some data are, though often not for our age group of 10-24 year olds.
To overcome this barrier, we thought about the existing categories in which data are often broken down and presented. Certain data on young people are reported for on a Local Authority basis, such as Department for Education (DfE) statistics and population level data available through Nomis. We also knew that the English Index of Multiple Deprivation (IMD) is available for each Local Authority in England. So our analysis sought to match up datasets that are available by Local Authority with the IMD data. A similar approach is used in the Office of Health Improvement and Disparities’ Fingertips data tool.
Giving an IMD rating to Local Authorities
The IMD data are available to download online, via the Ministry of Housing, Communities and Local Government. We used the dataset ‘upper-tier local authority summaries’, which ranks all of the 151 Local Authorities in England by IMD (with a rank of 1 being the most deprived Local Authority).
There are two ranking options available: average rank and average score. These are both calculated by averaging the scores of the Lower Super Output Areas (LSOAs) within the larger area of the Local Authorities. The main difference between the two is that areas with high polarisation of deprivation will score higher within the ‘average score’, whereas areas that are more uniformly deprived will score highly on the ‘average rank’. For this reason, we selected the ‘average rank’ for our analysis.
We wanted to create five deprivation groupings (1=most deprived to 5=least deprived). In order to do this, we needed an even number of Local Authorities in each group. Given that there are a total of 151 Local Authorities in England, we decided on 30 Local Authorities per group with an additional Local Authority in the middle (third) group.
Looking at health outcome data by LA IMD ranking
Using this grouped ranking for Local Authorities, where they were sorted from most to least deprived in quinary bands, we could then explore various outcomes that are presented by LA in other data sets. For example, Wokingham is the least deprived Local Authority within the 2019 English IMD, so when looking at DfE data on Looked After Children we could cross reference to the total number of children in Wokingham who are looked after in 2020. This process was repeated for all Local Authorities. In this way we could see, for example, that the number of looked after children in an area is associated with the area’s deprivation score. Chart D3 below illustrates this.
As well as providing descriptive data in this way, we were able to conduct these analyses over time. The English IMD ranking is updated every four years, which means that the top and bottom 30 Local Authorities falling into the most and least deprived categories have changed slightly between 2015 and 2019 (as per the ‘average rank’). For our analysis of data before 2019 we used the 2015 rankings, whereas we used the 2019 rankings for data more recent than 2019, in order to remain as true as possible to the ranking of local authorities at the correct date in time. An example of a result of this kind of analysis is presented in Chart HO15. It would have also been possible to analyse the data for just one IMD dataset, we have tested this on all analyses we completed by matching all data to the 2019 IMD and it shows very similar trends to what we have presented here.
Missing data
The data we analysed are reported by Local Authority, we recognise that there may be data reporting or lag issues relating to how different Local Authorities report their data to central systems. This was most notable when looking at vaccine uptake data during the Covid-19 pandemic years. This meant that there were often gaps or small figures reported for some Local Authorities. To undertake this analysis we used raw datasets on Excel and did not have access to sophisticated data analysis software, such as Stata or SPSS. We were unable to perform weighting on the data that might alleviate and reduce the impact of missing data or small data figures. Given that much of the analysis produced trends expected and replicated in other data sources, we are confident that this has not impacted on our analysis and results.
Conclusion
We hope that this analysis offers a novel but straightforward approach to deprivation analysis relating to young people’s health inequalities. No transformations of data were undertaken; we just pegged the Local Authorities to the IMD ranked data. We have carried out this analysis on the following data:
- Looked After Children / care leavers
- Youth mortality rates
- Vaccine uptake within adolescence (HPV and MenACWY)
All data correct as of 1st May 2022