Response to the ONS Consultation on Infant Mortality Outputs Review

Response to the ONS consultation on Infant mortality outputs review

23 October 2017

Contents

1 Introduction

2 Background

3 Summary of responses and our actions

4 Detailed responses and our actions

5 Getting in Touch

6 Accessibility

Annex A: Tables that are currently published

Annex B: Current data extracts, timing and linkage

Annex C: Organisations which responded

Annex D: Planned changes to data extracts, timing and linkage

Annex E: Planned changes to published tables providing infant mortality statistics for England and Wales

1 Introduction

This document summarises the responses received to the consultation on the proposed changes to Office for National Statistics (ONS) annual infant mortality statistics and sets out what actions ONS plans to take as a result.

The consultation ran from 20 April 2017 to 20 July 2017 and gave users an opportunity to provide feedback on the proposed changes to annual infant mortality statistics. The consultation document is available to download from the National Statistics website.

The consultation was conducted in line with the Code of Practice for Official Statistics.

2Background

ONS publishes annual infant mortality statistics for England and Wales in a number of separate statistical publications(Annex A).The timing of these outputs varies depending on when the relevant dataset is extracted from our system, and is also affected by the availability of final data from Scotland and Northern Ireland to compile UK data.

Due to the timings of the data extracts and the linkage (Annex B) that is carried out, there are different infant and child mortality figures across publications which can be confusing for users; these were outlined in the consultation document.

To ensure the publication of more timely and fit for purpose final figures, ONS is proposing to change the data extracts used in the creation of infant mortality publications; this will streamline the outputs. ONS will also change some of the tables published to ensure users have the most suitable tabulations available to them.

3Summary of responses and our actions

We are grateful to everyone who took time to respond to the consultation. There were a total of 45 responses received from a variety of organisations who monitor, formulate or influence policy and plan services. We also received responses from academics, media outlets, charities and private individuals. A full list of organisations that responded can be found in Annex C. The feedback we received is very valuable to us and will help us to develop infant mortality statistics in a way that will continue to meet your needs as much as possible.

The key points to note from the responses to the consultation are:

Users prefer figures based on deaths that have occurred in a calendar year rather than figures based on deaths registered in a calendar year but both are still required
Users prefer figures based on a birth cohort rather than a death cohort, but both are still required
Users require a range of the tables for various purposes utilising both numbers and the rates. Some tables are used more than others
Users have specified what other characteristics they require in the tables
Users would like to see more data at local authority level and would like figures at country level including UK
There was a larger proportion of users who do not use the ONS hierarchical cause groups
Users would prefer a single underlying cause of death for perinatal deaths

As a result of this consultation, we will take the following actions:

ONS will continue to produce figures based on both the year the death was registered and the year the death occurred
ONS will streamline the extracts used in the creation of publications (Annex D). This means we will produce statistics using:
a death cohort based on registrations in a calendar year (unlinked)
a death cohort based on occurrences in a calendar year (unlinked)
a death cohort based on occurrences in a calendar year (linked to birth registrations)
a birth cohort based on occurrences in a calendar year (birth registration linked to birth notification and death registration)
ONS will adapt the tables that are produced from these extracts to incorporate user feedback where possible and these can be found in Annex E
ONS will produce one set of tables for the birth cohort and this will merge together the Birth cohort tables for infant deaths (England and Wales) and the Pregnancy and ethnic factors influencing births and infant mortality (England and Wales) see Annex E Infant mortality (birth cohort), England and Wales
ONS will continue investigations into whether a single underlying cause can be derived for stillbirths and neonatal deaths; and will maintain the ONS hierarchical cause groups until investigations identify a better solution

Unfortunately we do not have the resources available to be able meet every single user requirement with our published tables providing infant mortality statistics. However, special extracts and tabulations of infant mortality data for England and Wales are available to order (subject to legal frameworks, disclosure control, resources and ourcharging policy,where appropriate). Enquiries should be made to Vital Statistics Outputs Branch via email by telephone on +44 (0)1329 444110. User requested data will be published ontoour website.

Infant and child mortality data based on the year the death was registered in and births data are also available in ourVirtual Microdata Laboratoryand the UK Data Service secure lab; this provides access to microdata and disclosive data, which have the potential to identify individuals. Access to such data requiresApproved Researcher accreditation.

Although not all the user needs expressed in the consultation can be met at present because of limited demand and a need to prioritise resource, ONS is continuously working to modernise its systems and produce statistical outputs more efficiently. The priority and feasibility of all the changes suggested will be kept under review and they may be implemented at a later date.

4Detailed responses and our actions

4.1 Users prefer figures based on deaths that have occurred in a calendar year rather than figures based on deaths registered in a calendar year but both are still required

From the responses, 47% of users prefer figures on occurrences in a calendar year, with 11% preferring figures on registrations in a calendar year, 33% preferring both and 9% having no opinion. ONS will continue to produce figures based on both the year the death was registered and the year the death occurred.

4.2 Users prefer figures based on a birth cohort rather than a death cohort, but both are still required

31% of respondents have a preference for deaths based on a birth cohort, with 9% preferring a death cohort, 31% preferring both and 27% having no opinion.

ONS will streamline the extracts used in the creation of publications (Annex D) to produce only one birth cohort extract instead of two.

This means we will produce statistics using:

a death cohort based on registrations in a calendar year (unlinked)
a death cohort based on occurrences in a calendar year (unlinked)
a death cohort based on occurrences in a calendar year (linked to birth registrations)
a birth cohort based on occurrences in a calendar year (birth registration linked to birth notification and death registration)

4.3 Users require a range of the tables for various purposes utilising both numbers and the rates where some tables are used more than others

40% of respondents use all of the childhood mortality tables with the most popular being live births, stillbirths and infant deaths (numbers and rates) by area of usual residence.

39% of respondents use all of the birth cohort tables for infant deaths with the most popular tables being live births, stillbirths and linked infant deaths (numbers and rates) by plurality and birthweight and, by mother’s country of birth.

42% of respondents use all of the pregnancy and ethnic factors influencing births and infant mortality tables with the most popular table being live births, neonatal and infant mortality (numbers and rates) by ethnic group and gestational age at birth.

ONS will continue to produce the majority of the current tables where some of these will be adapted to ensure consistency across publications. As there will only be one birth cohort extract, the tables from Birth cohort tables for infant deaths, England and Wales and the Pregnancy and ethnic factors influencing births and infant mortality, England and Wales will be merged together to produce one set of tables Infant mortality (birth cohort), England and Wales (Annex E); these will be adapted to ensure consistency across publications.

4.4 Users have specified what other characteristics they require in the tables

Some users requested additional variables in the tables such as place of delivery; causes mentioned on the death certificate; whether the death had a post-mortem/ was referred to a coroner; and parent(s) smoking status.

A table (Annex E) containing place of delivery currently exists for childhood mortality in England and Wales (CM(new)15) and new table has been planned for Infant mortality (birth cohort) which contains more detailed place of delivery variables (IMBC15) that are only available from the linkage carried out with the birth notification.

We asked users whether they required data based on causes mentioned on the death certificate.Of those that answered the question, 49% require this data and51% do not. At present we are unable to produce data for stillbirths and neonatal deaths with a single underlying cause of death; however we will continue to investigate the ability to do this with future software updates. We will continue to maintain the ONS cause groups anddata based on causes mentioned on the death certificate will not be published.

Three users requested data on whether the death had a post-mortem/ wasreferred to a coroner; due to limited demand and a need to prioritise resourcethis data will not be published.

However data based on causes mentioned on the death certificate and whether the death had a post-mortem/ was referred to a coroneris available (subject to legal frameworks and disclosure control) to order from Vital Statistics Outputs Branch or from the Virtual Microdata Laboratoryand the UK Data Service secure lab.

ONS does not hold data on the parent(s) smoking status; therefore we are unable to publish statistics on this.We will review whether any data sources could help identify this in the future.

4.5 Users would like to see more data at local authority level and would like figures at country level including UK

Over 45% of respondents require data at local authority level. Table (Annex E) CM(new)3 and IMBC2 will contain live births, stillbirths and infant deaths down to local authority level.

Individual users have requested data down to lower super output area (LSOA), lower tier local authority, neonatal network area and CCG. Due to resource limitations, data at these levels will not be published but are available (subject to legal frameworks and disclosure control)to order from Vital Statistics Outputs Branch or from the Virtual Microdata Laboratoryand the UK Data Service secure lab.

We also plan to include a table of live births, stillbirths and infant death by index of multiple deprivation(IMD) deciles for England and Wales separately.

54% of users have requested more detailed data forEngland only and 21% have requested more detailed data for Wales. However due to resource limitations, all tables will be produced for England and Wales combined. Separate data for England and Wales is available to order from Vital Statistics Outputs Branch or from the Virtual Microdata Laboratoryand the UK Data Service secure lab.

60% of users would like more detailed childhood mortality data at a UK level. Some data is currently published in ‘childhood mortality’ and will continue to be published in these tables. Due to the way in which data is provided by Scotland and Northern Ireland we are unable to produce UK statistics for more detailed characteristics. In England and Wales, stillbirths and neonatal deaths are registered using a special death certificate which enables reporting of relevant diseases or conditions of both the infant and the mother, this certificate is not used in either Scotland and Northern Ireland; therefore comparisons by cause of death are not possible. However we are continuing investigations to derive a single underlying cause of death for England and Wales.

4.6 There were a larger proportion of users who do not use the ONS hierarchical cause groups

58% of respondents do not use the ONS hierarchical cause groups for neonatal deaths and 62% do not use it for stillbirths. However from those who do use it (42% and 38% respectively), 40% found it useful. We will therefore continue to maintain the ONS hierarchical cause groups until investigations identify a better solution.

4.7 Users would prefer a single underlying cause of death for perinatal deaths

51% of respondents would prefer a grouping based on the underlying cause of death with 71% preferring perinatal data on a single underlying cause of death. ONS will continue investigations into whether a single underlying cause can be derived for stillbirths and neonatal deaths and whether this can be used to derive groupings.

5Getting in Touch

If you have any queries or comments about the consultation process, please email Simeon Bowen at or call 0845 601 3034.

You can also write to us at the following address:

Consultation Coordinator,

Room 1.101

Office for National Statistics,

Government Buildings,

Cardiff Road,

Newport,

South Wales,

NP10 8XG.

For further information on ONS consultations, please visit

FacebookTwitterLinkedInYouTubeStorify

6Accessibility

All material relating to this consultation can be provided in braille, large print or audio formats on request. British Sign Language interpreters can also be requested for any supporting events.

Annex A: Tables that are currently published

Childhood mortality in England and Wales
Table number / Table name / Time series / Table contains: / Planned changes
CM1 / Live births, stillbirths, infant deaths and childhood deaths under 15: numbers and rates, / 1980 onwards / Live births, stillbirths, early neonatal, neonatal, post neonatal, infant, 1-4. 5-9. 10-14, 1-15 / Merged with CM5
CM2 / Live births, stillbirths, infant deaths and childhood deaths under 15: country of occurrence and sex, numbers and rates, (UK and constituent countries) / Current year only / Live births, stillbirths, early neonatal, neonatal, post neonatal, infant, 1-4, 5-9, 10-14, 1-15, by sex country and UK
CM3 / Live births, stillbirths and infant deaths: area of residence, numbers and rates / Current year only / Live births, stillbirths, perinatal, neonatal, post neonatal, infant, regions
CM4 / Postneonatal and childhood deaths: broad underlying cause groups, age and sex / Current year only / Postneonatal deaths by underlying cause, age at months and years
CM5 / All infant deaths and linked infant deaths: numbers and rates / 1980 onwards / Live births, stillbirths, early neonatal, neonatal, post neonatal, infant, linked and unlinked / Merged with CM1
CM6 / Live births, stillbirths and linked infant deaths: ONS cause groups and birthweight / Current year only / Live births, stillbirths, early neonatal, neonatal, post neonatal, infant by ONS cause groups, birthweight groups
<2500,<1500,<1000,1000-1499,1500-1999,2000-2499,2500-2999,3000-3499,3500-3999,4000 and over, Not stated / Birthweight groupings changed
<1000,1000-1499,1500-1999,2000-2499,2500-2999,3000-3499,3500-3999,4000 and over, Not stated
CM7 / Live births, stillbirths and linked infant deaths: ONS cause groups and age of mother / Current year only / Live births, stillbirths, early neonatal, neonatal, post neonatal, infant by ONS cause groups, age of mother group
CM8 / Live births, stillbirths and linked infant deaths: ONS cause groups, marital status of mother/type of registration and number of previous children / Current year only / Live births, stillbirths, early neonatal, neonatal, post neonatal, infant by ONS cause group, plurality, marital status, type registration, number previous children
CM9 / Stillbirths: ONS cause groups and birthweight by gestation period / Current year only / Stillbirths by ONS cause groups, gestation groups, birthweight groups
<2500,<1500,<1000,1000-1499,1500-1999,2000-2499,2500-2999,3000-3499,3500-3999,4000 and over,Not stated
24-27, 28-31, 32-36, 37-39, 40 and over, not stated / Birthweight and gestation groupings changed
<1000,1000-1499,1500-1999,2000-2499,2500-2999,3000-3499,3500-3999,4000 and over, Not stated
Under 24, 24-27, 28-31, 32-36, 37-41, 42 and over
CM10 / Live births, stillbirths and linked infant deaths: birthweight by age of mother, numbers and rates / Current year only / Live births, stillbirths, early neonatal, neonatal, post neonatal, infant by mothers age groups, birthweight groups
<2500,<1500,<1000,1000-1499,1500-1999,2000-2499,2500-2999,3000-3499,3500-3999,4000 and over, Not stated / Birthweight groupings changed
<1000,1000-1499,1500-1999,2000-2499,2500-2999,3000-3499,3500-3999,4000 and over, Not stated
CM11 / Live births, stillbirths and linked infant deaths: birthweight and mother’s country of birth, numbers and rates / Current year only / Live births, stillbirths, early neonatal, neonatal, post neonatal, infant, birthweight groups, mothers country birth
<2500,<1500, 2500 and over,Not stated
CM12 / Live births, stillbirths and linked infant deaths: birthweight, type of registration and socio-economic classification, numbers and rates / Current year only / Live births, stillbirths, early neonatal, neonatal, post neonatal, infant by birthweight, type registration, NS-SEC