|Publication date:||17 April 2019|
|Who this is for:||Users and producers of statistics|
|Type:||Harmonisation guidance and principles|
The following guidance sets out how to collect and report statistics about general health to ensure statistics about this topic are as comparable as possible across the Government Statistical Service (GSS).
This section provides guidance on the survey questions to use when collecting information about activity restriction.
to ask to respondents if aged 16 and over, by proxy if under 16 and if a person is unfit to respond.
One question providing a measure of subjective self-reported health-related well-being for use in social surveys and suitable for data collection using CAPI, CATI and paper-based and online self completion forms.
Consistent with the question asked in the 2011 Census and with the EU-SILC and European Health Interview Survey Minimum European Health Module harmonisation principle, so direct comparison with these sources are possible.
How is your health in general? Is it…….
- Very Good
- Very Bad
(This question has replaced the question on General Health previously used in the 2001 Census and social survey data sources from 2008 onwards).
Presentation of outputs
This section provides guidance for outputting the survey questions to use when collecting information about general health.
|How is your health in general? Is it….||Code|
|General health (5 point scale)||Nominal scale|
|Dichotomous output categories|
|‘Good Health’||= 1 OR = 2|
|‘Poor Health’||= 3 OR = 4 OR = 5|
|Trichotomous output categories|
|‘Good Health’||= 1 OR = 2|
|‘Fair Health’||= 3|
|‘Poor Health’||= 4 OR = 5|
There is one question on general health covered in this principle which asks about self-assessed health which has been shown to be a good predictor of demands for health services and is used in the construction of health state life expectancy metrics such as Healthy Life Expectancy. The full five scaled output categories can be used in longitudinal sources to measure change in general health status among individuals; and in cross-sectional data sources to measure change in the proportion of the population classified to each category over time in time series cross-sections.
The variable can also be collapsed into a dichotomised scale to measure the proportion of the population in ‘Good’ or ‘Poor’ general health, and trichotomised into ‘Good’ ‘Fair’ and ‘Poor’ general health as undertaken in some Eurostat publications.
The Office for National Statistics dichotomises this variable to compute the Healthy Life Expectancy Metric; details of how this is calculated is available via the Method changes to life and health state expectancies – on the ONS website.
This question replaced a different question on general health using a three-category scale. An investigation into the change in the time series and the impact on estimates of healthy life expectancy was reported on in a Health Statistics Quarterly article, An Investigation into The Impact of Questions Change On Estimates of General Health Status and Healthy Life Expectancy.
There is a strong relationship between self-reported general health and prevalence of a long-standing illness which limits normal day to day activities. Estimates using General Lifestyle Survey data from 2005 and 2006 found more than half those reporting their general health as ‘Fair’ also reported a long-standing illness which limits normal day to day activities, while less than 10 per cent of those reporting their general health as ‘Good’ or ‘Very Good’ did so. These findings are available in an HSQ article, ‘An Investigation Into The Impact Of Questions Change On Estimates of General Health Status and Healthy Life Expectancy’, available for download from the ONS website at:
This page will be reviewed annually.