Ed Humpherson, Director General for Regulation, Office for Statistics Regulation writes to Professor Danny Dorling, Halford Mackinder Professor of Geography, Oxford University Centre for the Environment.
Dear Professor Dorling,
Thank you for your email to the Office for Statistics Regulation (summarised at Annex 1) expressing concerns about the official reporting of some health statistics. You raise three issues.
First, you note concerns about the accuracy of some Tweets that Public Health England (PHE) have made recently about life expectancy. I approached PHE with our concerns and welcome PHE’s cooperation in removing the Tweets and amending the associated Health Matters article.
Your second concern was about ONS’s independence, specifically related to the Quarterly Mortality Reports, the latest of which was published on 17 September 2018. ONS’s web landing page for the statistics notes ‘This publication was produced with support from Public Health England and the Department of Health and Social Care’ but no information is presented as to the nature of the support provided. We have no reason to believe that the independence of ONS statistics has been compromised by this collaboration, but we have asked ONS to publish details of how these two organisations feed into the publication so that ONS is transparent about how it demonstrates it complies with the Code of Practice principle T2 Independent decision making and leadership.
Finally, you had doubts about the possible attribution of changes in mortality to external temperature rises, as offered as an example in section 6 of the quarterly mortality release. Statistics producers are encouraged to provide insightful commentary in releases and to help users understand how the data may address questions of current interest. My view is that these weather events in the period under discussion were newsworthy across the UK. In its release, ONS is sensibly attempting to provide some insight as to what light the statistics may shed on the impact of these periods of warm weather. While ONS is clear that it cannot confirm that a heatwave is the cause of these deaths, it does want to provide commentary to help users understand the statistics. And I also note ONS commits to consider the question further, saying “The question of deaths which might be attributable to a heatwave will be addressed in our next quarterly release”.
Having looked at the release, I have therefore concluded that ONS’s comments about the impact of external temperature changes (both heat and cold) upon mortality are appropriate ones for a statistical producer to make.
Thank you very much for bringing these matters to our attention.
Yours sincerely,
Ed Humpherson
Annex 1: summary of concerns received from Professor Dorling
I raise a couple of issues over the official reporting of health statistics. The problem with the “Q2″ reporting is that as it says at the top of this web page: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/quarterlymortalityreports/apriltojune2018
The report is produced with Public Health England and the Department for Health and Social Care. The key question is why are ONS (supposedly largely independent of government) publishing text written by PHE and DHSC?
Hence why they talk of heatwave – in April and May and June! Even in June of this year the weather was just 1.8 degrees above the average for that time of year: https://www.metoffice.gov.uk/climate/uk/summaries/2018/june. It was later, in July, that it got really hot. But even then the effect of heat is mitigated by good social services, adult social workers and so on being available.
More recently my colleagues and I have noticed that PHE are tweeting that “average” life expectancy is rising? The most recent figures published by ONS was highest in 2014 and has been lower for both men and women ever since 2014.
Related Links
Ed Humpherson to Iain Bell (October 2018)