The Scottish Health Survey Past, present and future Catherine Bromley Scottish Centre for Social Research.

The Scottish Health SurveyPast, present and futureCatherine Bromley

Scottish Centre for Social Research

Aims of the Scottish Health Survey

• Scottish Government’s stated aims for the survey:• To estimate the prevalence of particular health conditions in

Scotland• To estimate the prevalence of certain risk factors associated with

these health conditions and to document the pattern of related health behaviours

• To look at differences between regions and between subgroups of the population in the extent of their having these particular health conditions or risk factors, and to make comparisons with other national statistics for Scotland and internationally

• To monitor trends in the population’s health over time• To make a major contribution to monitoring progress towards health

targets

Examples of uses of the Survey

• Scottish Government annual reports & web tables• Reports on most of the survey’s aims (national estimates, trends,

differences between groups, etc)

• Scottish Government topic reports• UK comparisons, Glasgow effect, obesity, older people (29/11/11)

• Journal articles / conference papers• Prospective analysis of: CVD episodes, alcohol-related hospital

admissions, psychiatric admissions, mortality• Trend analysis of: socio-economic patterning of CVD risk factors,

diabetes prevalence and risk factors• International comparisons of: self-reported health, diabetes

management• Regional differences within Scotland• Estimates of screening burden for CVD or diabetes risk

Methods and design

• General population, multi-stage PAF sample• All adults in selected households & up to 2 children

• Two stage design (1995-2011)• Main interview with height & weight measurement• Follow-up visit from a nurse

• Topics• Main focus: cardiovascular disease and its risk factors• Key data on: smoking, alcohol, phys activity, diet, BMI• Plus general health and wellbeing, respiratory conditions, accidents, dental

health• Consent to SMR and mortality record linkage sought• Originally based on 1994 Health Survey for England, still has lots of topics

in common (not always identical measures)

• Contractors:• Scottish Centre for Social Research (95-15) with UCL (95-11), MRC/CSO

Social and Public Health Sciences Unit (03-15), U of Edinburgh (2012-15), U of Aberdeen (2012-15)

History of the Scottish Health Survey (1995-2003)

• 1995• 7932 adults aged 16-64• 6958 also had a nurse visit and 6183 gave a blood sample

• 1998• 9047 adults aged 16-74• 3892 children aged 2-15• 7455 adult and 3211 child nurse visits• 6178 adults and 466 children (11-15) gave a blood sample

• 2003• 8148 adults aged 16+ • 3324 children aged 0-15• 5444 adult and 2224 child nurse visits• 4264 adults and 300 children (11-15) gave a blood sample

2005/6 Review recommendations

• Continuous survey from 2008• Core and modular structure

• From 2008, topics are either in every year, or alternating years• 1 adult in sub-sample of households completes Health Scotland’s

Knowledge, Attitudes and Motivations module

• Scotland level results annually• NHS Board results by amalgamating years together• Unclustered sample• NHS Boards able to purchase boosts to enhance their

samples• Fife, Borders and Grampian have boosted samples

• Adult nurse visits in a sub-sample of households

History & future of the Survey (2008-2015)

• 2008-2011• ~7000 adults aged 16+ per year• ~2000 children aged 0-15 per year• Sub-sample of adult nurse visits ~ 1000 per year • No nurse visits for children

• 2012-15• Major redesign as part of SG review of population surveys• Clustered sample drawn by SG• Harmonised core questionnaire common to all 3 Scottish population surveys• KAM module discontinued• SHeS still continuous, but smaller sample:

• ~4000 adults aged 16+ per year• ~1800 children aged 0-15 per year

• No nurse visits• Interviewer biomeasures instead

2008-11 progress update1995 1998 2003 2008-

20112012 2013-15

Blood pressure Int Int

Lung function

Waist/hip (16+) Int Int

Cotinine blood saliva

saliva saliva saliva saliva

HDL cholesterol ??

Total cholesterol ??

Fibrinogen

C Reactive protein ??

Glycated haemoglobin ??

Urinary sodium/ potassium/creatinine

Int Int

ECG

Biomeasures - adults

Datasets available in the Data Archive

• 1995, 1998 and 2003• 2008• 2009• 2008 and 2009 combined data• Jan 2012: 2010 and 2008/2009/2010 combined data• Jan 2013: 2011 and 2008/2009/2010/2011 combined data• Jan 2013: 1995-2011 trend analysis mini-dataset

Note – the combined datasets are not a substitute for the annual datasets, their purpose is to increase n for questions included in multiple years. Analysis of a single year should be run on the relevant single year dataset.

Scottish Health Survey-SMR linked data

• File maintained by ISD• Over 90% of participants consent to linkage• 1981-2009 linkage now available for each survey up to 2009• Consent wording from 2012 will increase scope of linkage

• Two versions of file• Minimum SHeS-SMR free to research community:

SMR01 (general/acute inpatient/day case inc. reasons for admission derived from SMR records)SMR04 (mental health inpatient/day case)SMR06 (cancer register)Causes of death

• Full dataset with individual SMR data, available by seeking permission from Privacy Advisory Committee

• Further info:• Article in International Journal of Epidemiology (2010, 39: 345-350)• Email: Leanne Hopkins at ISD ([email protected])

Negative implications of 2012 changes

• Wider confidence intervals for health board data• Current minimum adult sample per HB=784, from 2012=500

• More reliance on combined datasets for detailed sub-group analysis / fewer analyses stratified by gender

• CVD / diabetes risk analysis based on blood samples might not be possible

• Potential for interrupting time series (waist/BP)• Continued need to use combined datasets for

biomeasure analysis• Additional uncertainty about sample sizes in early years until

response rates are established

Positive implications of 2012 changes

• Increased ability to analyse trends• Better estimates of future CVD & diabetes risk

based on family history• Potentially higher response to biomeasures• Greater focus on respiratory health• Ability to analyse smoking prevalence, long-term

conditions and self-reported health using pooled sample from Health, Household and Crime surveys• First time that Scottish data will exist for smaller populations

such as minority ethnic groups

Scottish Health Survey: key contacts

• At Scottish Government:• Queries: [email protected]• Main website: www.scotland.gov.uk/ScottishHealthSurvey• Includes all Scottish Govt survey publications and links to other

work based on the survey• Useful analysis site:

www.scotland.gov.uk/Topics/Statistics/Browse/Health/scottish-health-survey/SHeSAnalystHelp

• Includes advice on analysis, variables in each report table, full SPSS syntax files and more

• At ScotCen:• Queries: [email protected]