Dear Iain,
HEALTH STATE LIFE EXPECTANCIES STATISTICS, UK
We recently carried out a short review of compliance with the Code of Practice for Statistics of the Health State Life Expectancies (HLE) statistics for the UK. This was prompted by the changes in methods used to estimate HLE introduced in the December 2018 bulletin.
We reviewed the processes that your team undertook to develop the new methods, and its consultation with users. The changes made have clearly improved the value of these statistics for users and I am therefore pleased to confirm that these should continue to be designated as National Statistics. The rest of this letter highlights the main findings of our review.
HLE was previously calculated using estimates from the Annual Population Survey (APS). However, due to the sample size this resulted in volatile estimates for some local areas, causing concern for users. The new method uses imputation and modelling of APS and census data to produce more stable estimates. The positive aspects of this work include:
- Good communication and user engagement: plans for revising the method were announced in November 2016, long before implementation. A report presenting options was published in December 2017, followed by a consultation phase that reported in April 2018. The statistics based on the new method were published in December 2018. All these documents are easy to find on the ONS website.
- Good cross-national working: statisticians from National Records of Scotland (NRS) and ONS have been working together to enable NRS to fully replicate ONS’s methods.
- Improved methods: the new approach addresses the problem of volatile local estimates without introducing excess complexity.
- Increased harmonisation and coherence: NRS and ONS now use the same method to produce HLE statistics so consistent estimates are now available for Scotland and the rest of the UK.
Although only a small number of users responded to the formal consultation stage, we are aware that important users of the statistics contributed their views. We are content that the information provided in the reports noted above provided users with sufficient opportunities to contribute their views about the proposed changes. The recent steps that your team has taken to build its stakeholder relationships with users of ONS’s Avoidable Mortality statistics might helpfully be applied more widely to develop your stakeholder network for HLE statistics.
We welcome the fact that your team has plans to adapt the methods further if required, for example if the data needed is no longer available after the 2021 Census. As with the most recently introduced changes, it will be important to provide users with sufficient notice of any planned revisions and suitable opportunities to be involved in any decisions taken. I would welcome updates from you on any future developments. The cross-UK estimates of HLE that are now available will be valuable to users, we encourage you to continue collaborating with NRS to maintain comparability wherever possible. We will be encouraging ONS, NRS and the Northern Ireland Statistics and Research Agency to work together to ensure that harmonised measures of health and disability continue to be available after the next Census.
I have also written to NRS today confirming that their HLE statistics should continue be designated as National Statistics.
Yours sincerely
Mark Pont
Assessment Programme Lead
Related Link:
Mark Pont to Amy Wilson (March 2019)