Publication - Statistics publication

Scottish Health Survey 2013: Volume 2: Technical Report

Published: 2 Dec 2014
Part of:
Statistics
ISBN:
9781784129439

Presents information on the methodology and fieldwork for the Scottish Health Survey 2013.

419 page PDF

16.2 MB

419 page PDF

16.2 MB

Contents
Scottish Health Survey 2013: Volume 2: Technical Report
CHAPTER 1: METHODOLOGY AND RESPONSE

419 page PDF

16.2 MB

CHAPTER 1: METHODOLOGY AND RESPONSE

Joan Corbett, Sarah Martin, Alex Stannard, Shanna Dowling, Stephen Hinchliffe & Lisa Rutherford

1.1 INTRODUCTION

1.1.1 The Scottish Health Survey series

The Scottish Health Survey (SHeS) series was established, in 1995, by the then Scottish Office to provide data about the health of the population living in private households in Scotland. It was repeated in 1998 and 2003 and has been carried out annually since 2008.

The 1995 and 1998 surveys were carried out by the Joint Health Surveys Unit of the National Centre for Social Research (NatCen Social Research) and the Department of Epidemiology and Public Health, University College London Medical School (UCL). The MRC Social and Public Health Sciences Unit at the University of Glasgow (MRC SPHSU) joined the consortium in 2003.

After a Scottish Executive[1] review of the series in 2005/2006, the survey switched to a continuous format and has been continuous since 2008. A consortium made up of ScotCen Social Research (a branch of NatCen Social Research), UCL and MRC SPHSU carried out the 2008-2011 surveys.

The 2012-2015 surveys are being conducted by ScotCen Social Research and MRC SPHSU in collaboration with the Centre for Population Health Sciences, Edinburgh University and the Public Health Nutrition Research Group, Aberdeen University.

1.1.2 Aims of the Scottish Health Survey

The purpose of SHeS is to provide information at the national level about the health of the population and the ways in which lifestyle factors are associated with health. This level of information is not available from administrative or operational databases, as hospitals and GPs are not able to collect detailed information about lifestyles. In addition, it is crucial that the Scottish Government has information about the health of the population, including those who do not access health services regularly.

The specific aims of SHeS are:

  • 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 England,
  • to monitor trends in the population's health over time, and
  • to make a major contribution to monitoring progress towards health targets.

Each year, the survey consists of a set of core questions and measurements (for example, anthropometric and, if applicable, blood pressure measurements and analysis of urine and saliva samples), plus modules of questions on specific health conditions. As with the earlier surveys in the series, the principal focus of the 2012-2015 surveys remains to be cardiovascular disease (CVD) and related risk factors. CVD is one of the leading contributors to the global disease burden. Its main components are ischaemic heart disease (IHD) and stroke. IHD is the second most common cause of death in Scotland after cancer.[2] The SHeS series now has trend data going back 19 years; providing the time series is an important function of the survey.

1.1.3 Key changes to the survey methodology in 2012-2015

A number of changes to the survey methodology were proposed during the 2011 Scottish Government review of Scotland's major household surveys. The key changes to SHeS introduced in 2012 for the 2012-2015 surveys were:

  • Sample of addresses drawn by the Scottish Government
  • Inclusion of a set of harmonised core questions asked across all major Scottish Government household surveys.
  • Reduction in the achieved sample size
  • Discontinuation of a module of questions on Knowledge, Attitudes and Motivations (KAM) to health
  • Introduction of interviewer administered biological samples and measurements to replace the nurse interview.

These changes are discussed in greater detail in Volume 2 of the 2012 technical report.[3]

1.1.4 The 2013 survey

The 2013 SHeS was designed to provide data at national level about the population living in private households in Scotland. The eligible age range for the survey was all aged 0+.

An initial sample of 9,658 addresses was drawn from the Postcode Address File (PAF). These addresses were comprised of three sample types: Main (core) sample, child boost screening sample and Health Board boost sample. Ayrshire and Arran, Fife, Grampian and Western Isles NHS Health Boards opted to boost the number of adults (16+) interviewed in their area in 2013. The table below shows the number of addresses drawn for each sample type and the number of people eligible for interview within each sample type.

The 9,658 addresses were grouped into 504 interviewer assignments, with around 42 assignments being issued to interviewers each month between January 2013 and December 2013.

Sample type

Number of addresses issued in 2013

Eligible for interview

Main

4,456

Max of 10 adults (age 16+) and 2 children (age 0-15)

Child Boost

4,147

Only households containing children aged 0-15 were eligible to participate

(up to two children at these households were eligible to be interviewed)

Health Board Boost

1,055

Adults only (age 16+) (Max of 10)

Total

9,658

Data collection involved a main computer assisted interview (CAI) interview, paper self-completion questionnaire, height and weight measurements and, if applicable, adults also completed the biological module. Of the 4,456 main addresses issued, 1,719 were flagged as eligible for the 'biological module sample.' At these addresses all adults (16+) that participated in the main interview were eligible to take part in the module. Only interviewers that were specially trained in administering biological measures and samples were allocated to work on these addresses.

1.1.5 The 2013 annual report

The 2013 SHeS report consists of two volumes, published as a set as 'The Scottish Health Survey 2013.' Volume 1 presents results for adults and children on a variety of health topics. This report (Volume 2) provides methodological information and survey documentation. Both volumes are available on the Scottish Government's SHeS website along with a short summary report of the key findings from the 2013 survey (www.scotland.gov.uk/scottishhealthsurvey). Supplementary web tables are also available on the website.

1.1.6 Comparisons with previous surveys in the SHeS series

In the 2013 report comparisons are made with data collected earlier in the series (1995-2012). Having such an extensive trend period makes it possible to comment on whether any changes in health behaviours identified between years were real or an instance of sample fluctuation.

In addition, this report includes analysis of the first pooled dataset of the 2012 to 2015 series (the 2012 and 2013 combined dataset). Combining data across years in this way allows for analysis of questions with small sample sizes in one survey year - for example children's eating habits or blood pressure measurements collected in the biological module. Tables in the report indicate whether the figures presented are based on a single year's data or combined data from across survey years.

1.1.7 Health Board level analysis

Since 2008, the SHeS sample has been designed to be representative at the Health Board level (for all Boards) after four years of data collection have taken place. Analysis of the 2008 to 2011 data by NHS Health Board was published in 2012 and is available on the SHeS website (www.scotland.gov.uk/scottishhealthsurvey). The 2013 report does not include any analysis by NHS Health Board and the next Board level results will be published in 2016 (for the survey years 2012-2015 combined). Health Boards with larger samples may be able to analyse data at their Board level before this and users should consult the SHeS website for further guidance on sub-geographies analysis. Changes in the sample design between 2008-2011 and 2012-2015 mean that users are not advised to combine data outwith the four-year cycles from which they have been drawn.

1.1.8 Access to data

Data from the 2013 survey will be deposited at the UK Data Service along with the combined 2012/2013 dataset. Datasets from earlier years in the series are also deposited here (www.data-archive.ac.uk).

1.2 SAMPLE DESIGN

1.2.1 Requirements

The sample specification for the 2012-2015 Scottish Health Survey (SHeS) was designed by the Scottish Government. The design was coordinated with the designs for the Scottish Household Survey and the Scottish Crime and Justice Survey as part of a survey efficiency project and to allow the samples of the three surveys to be pooled for further analysis.[4]

There were three elements to the SHeS sample in 2012-2015:

1) Main adult sample - to allow annual reporting of Scotland level results and results at Health Board level at the end of the 2012-2015 four year cycle. This required an annual interview target of 4,000 adults for Scotland as a whole and a minimum of 125 for each Health Board. There was an additional requirement for 1,000 adults to complete a biological module each year.

2) Child sample boost - overall there was a requirement for 1,780 child interviews for Scotland. As the main sample was only expected to yield 780 child interviews, a further 1,000 interviews were required from a separate boost sample.

3) Health Board boosts - in 2013, Ayrshire and Arran, Fife, Grampian and Western Isles Health Boards commissioned boosts to increase the number of adult interviews in their Board area. Each of the Health Boards specified the target number of interviews for their boost.

1.2.2 Sample design and assumptions

For all three elements a two-stage clustered sample design with intermediate geographies randomly selected at the first stage and address points at the second stage, was used. With the exception of Orkney, Shetland and Eilean Siar, the sample was clustered by intermediate geographies (IG) with one quarter of IGs selected for each year of fieldwork. This means that over four years of fieldwork all IGs are included in the sample and the combined 2012-2015 sample is unclustered.

1.2.3 Main sample

As stated above, the annual sample size for Scotland was 4,000 adults with a minimum Health Board sample size of 125 adults. These sample sizes were the minimum required to allow effective reporting of Scotland-level results annually and Health Board results at the end of the four year cycle. An iterative approach was taken to efficiently allocate the sample across all Health Boards. For the first iteration 4,000 adult interviews were allocated across Health Boards in proportion to the adult population. Any Health Boards allocated fewer than 125 adult interviews had their allocation increased to 125.

The remaining sample was then allocated over the remaining Health Boards. Where allocations were not whole numbers the number was rounded up. This resulted in a total target of 4,006 adult interviews. The results of the allocation are shown in Figure 1A.

Figure 1A: SHeS target annual adult interviews, 2012-2015, by Health Board

Health Board

Target Annual Adult Interviews

Ayrshire and Arran

256

Borders

125

Dumfries and Galloway

125

Fife

252

Forth Valley

201

Grampian

379

Greater Glasgow and Clyde

836

Highland

217

Lanarkshire

383

Lothian

578

Orkney

125

Shetland

125

Tayside

279

Western Isles

125

Total

4,006

While the required sample sizes were set at Health Board level, to allow for coordination with the sample selection of the SHS and SCJS, the sample design was implemented using Local Authorities as strata. This was done by allocating the target Health Board samples to Local Authorities proportionate to population.

There was a slight complication in the design due to Local Authority boundaries not being concurrent with Health Board boundaries at the time the sample was drawn. Where there were fewer than 2% of the Local Authority population outwith its main Health Board, it was assumed that the Local Authority was entirely located within the Health Board. This gave the following assumptions:

  • Dumfries and Galloway LA lies completely in Dumfries and Galloway HB
  • Perth and Kinross LA lies completely in Tayside HB
  • Scottish Borders LA lies completely in Borders HB
  • Falkirk LA lies completely in Forth Valley HB
  • East Renfrewshire LA lies completely in Greater Glasgow and Clyde HB
  • Glasgow City LA lies completely in the Greater Glasgow and Clyde HB
  • Stirling LA lies completely in the Forth Valley HB
  • West Lothian LA lies completely Lothian HB.

For North Lanarkshire and South Lanarkshire 5.6% and 22.3% of the respective populations were located in Greater Glasgow and Clyde Health Board rather than Lanarkshire Health Board. Therefore, the sample stratification for the North Lanarkshire and South Lanarkshire Local Authority areas were each split into two with the west elements lying in Greater Glasgow and Clyde and the east elements lying in Lanarkshire.

The number of addresses selected in order to provide the target number of interviews was calculated by:

1) Estimating the number of productive adult interviews per co-operating household. Based on response data to the 2008 and 2009 surveys, it was estimated that for Greater Glasgow and Clyde there would be 1.5 interviews per co-operating household, with 1.55 interviews in all other Health Boards.

2) Allocation of the target interviews and associated estimate of co-operating households to Local Authority strata proportionate to population.

3) The sample was stratified by Local Authorities, however it was found that sample sizes for individual authorities were too small to base response rate assumptions on. It was also found that it would be inappropriate to base assumptions on Health Boards as there was a high degree of variability for Local Authorities within the same Health Board. Therefore, Local Authorities were placed in 9 groups which had common attributes and comparable response rates in 2009. The response rate assumptions for the Local Authority groups for the 2013 survey were based on the weighted average of Local Authority response rates 2009-2011. Figure 1B below shows the Local Authority groupings.

4) The final step was to estimate the level of ineligible addresses. The estimates were calculated at Local Authority level and based on the average level of ineligible addresses from the Scottish Health Survey, Scottish Household Survey, Scottish Crime and Justice Survey, and Scottish House Condition Survey from 2007 to 2009/2010.

Figure 1B: Local Authority groupings for response rate assumptions

Local authority group

Constituent local authorities

Ayrshire & Arran and Dumfries & Galloway

East Ayrshire, North Ayrshire, South Ayrshire, Dumfries and Galloway

Highlands

Aberdeenshire, Argyll and Bute, Highland, Moray

Islands

Eilean Siar, Orkney Islands, Shetland Islands

West

East Renfrewshire, East Dunbartonshire, West Dunbartonshire, Renfrewshire, Inverclyde

Large Cities

Aberdeen City, City of Edinburgh, Glasgow City

Lothian and Borders

West Lothian, East Lothian, Midlothian, Scottish Borders

Tayside

Perth and Kinross, Angus, Dundee City

Forth Valley and Fife

Clackmannanshire, Fife, Stirling, Falkirk

Lanarkshire

North Lanarkshire, South Lanarkshire

Figure 1E shows the number of selected addresses for the main sample in 2013.

1.2.4 Child boost sample

For the 2012-2015 surveys, 1,780 child interviews were required each year. Based on the 2009 survey, it was estimated that the main sample would provide 780 child interviews, therefore, to reach the target number of child interviews, a child boost sample was required to yield a further 1,000 interviews.

While the target number of child interviews was specified at Health Board level, as with the main sample, the child boost sample was also stratified by Local Authority. The process for calculating the number of addresses to select for the child boost sample was as follows:

1) The overall target sample of 1,780 child interviews was allocated proportionally to Health Boards based on the child (under 16) population. The expected number of child interviews from the main sample was then subtracted from the overall target sample to obtain the child boost target sample for Health Boards. If the number expected from the main sample was greater than the overall required number of child interviews for a Health Board then the boost target was set to zero and the remaining sample was redistributed so the overall target remained 1,780. The following table shows the target sample sizes for the main sample and child boost sample by Health Board.

Figure 1C: Target annual child interviews by Health Board

Health Board

Expected child interviews from main sample

Child interviews from boost

Total child interviews

Ayrshire and Arran

50

73

123

Borders

24

14

38

Dumfries and Galloway

24

24

48

Fife

49

75

124

Forth Valley

39

64

103

Grampian

73

108

181

Greater Glasgow and Clyde

167

229

396

Highland

42

60

102

Lanarkshire

74

127

201

Lothian

112

155

267

Orkney

24

0

24

Shetland

24

0

24

Tayside

54

76

130

Western Isles

24

0

24

Total

780

1005

1785

2) The number of co-operating households with children required in each Health Board for the child boost sample was estimated using the performance of the child boost sample in the 2008 and 2009 surveys.

3) To estimate the proportion of child-less households data from the 2007-2008 Scottish Household Survey was used. As there was little variation across different areas a Scotland level estimate of households without children (74.6%) was used.

4) Analysis of survey response to the child boost samples in the 2008 and 2009 surveys found that the response rate was consistently higher for the child boost than the main sample. Therefore, for each Local Authority area, the estimated response rates for the child boost sample were set at 6% higher than the main sample response rate.

5) The assumptions made on ineligible addresses for the main sample were applied to the address calculations for the child boost sample.

The total number of addresses selected for the child boost sample are shown in Figure 1E below.

1.2.5 Health Board boost samples

Each year individual Health Boards are given the opportunity to fund a boost sample to enable them to boost the number of adult interviews in their Board area. For the 2013 survey, Ayrshire and Arran, Fife, Grampian and Western Isles opted to boost the main sample in their areas. The following table shows the target sample size for each of the boosts.

Figure 1D: Target sample for Health Board boosts in 2013

Health Board

Target interviews for boost

Ayrshire and Arran

200

Fife

300

Grampian

300

Western Isles

200

Total

1000

As the main sample was selected before boost areas were confirmed, boost samples were supplementary to the main sample. For Ayrshire and Arran, Grampian and Western Isles the sample design followed the process outline above for the main sample. Fife Health Board requested that the combined main sample and boost sample was equally distributed across its three Community Health Partnership areas. This required an extra level of stratification for Fife before the process above could be followed. For the boost samples the same active PSUs as the main sample were used.

Figure 1E: Selected addresses by strata in 2013

Sample strata

Main sample selected addresses

Health Board boost samples

Child boost sample selected addresses

Total sampled

Aberdeen City

193

153

177

523

Aberdeenshire

169

134

200

503

Angus

78

-

83

161

Argyll & Bute

73

-

69

142

Clackmannanshire

35

-

43

78

Dumfries & Galloway

128

-

92

220

Dundee City

105

-

103

208

East Ayrshire

82

65

90

237

East Dunbartonshire

84

-

87

171

East Lothian

68

-

80

148

East Renfrewshire

71

-

83

154

Edinburgh, City of

442

-

373

815

Eilean Siar

128

205

0

333

Falkirk

106

-

127

233

Fife

259

311

288

858

Glasgow City

551

-

503

1054

Highland

165

-

177

342

Inverclyde

70

-

69

139

Midlothian

56

-

65

121

Moray

64

51

69

184

North Ayrshire

95

74

106

275

North Lanarkshire East

226

-

287

513

North Lanarkshire West

15

-

18

33

Orkney Islands

121

-

0

121

Perth & Kinross

108

-

110

218

Renfrewshire

143

-

145

288

Scottish Borders

134

-

56

190

Shetland Islands

118

-

0

118

South Ayrshire

79

62

78

219

South Lanarkshire East

189

-

225

414

South Lanarkshire West

45

-

46

91

Stirling

64

-

75

139

West Dunbartonshire

76

-

78

154

West Lothian

116

-

145

261

Total

4,456

1,055

4,147

9,658

1.2.6 Sample Selection

The Royal Mail's small user Postcode Address File (PAF) was used as the sample frame for the address selection. The advantages of using the PAF are as follows:

  • It has previously been used as the sample frame for Scottish Government surveys so previously recorded levels of ineligible addresses can be used to inform assumptions for the 2012-2015 sample design
  • It has excellent coverage of addresses in Scotland
  • The small user version excludes the majority of businesses.

The Assessor's Portal, the council tax list of all dwellings in Scotland, was considered as an alternative sample frame but since it had not previously been used as a sample frame for large scale surveys in Scotland there would have been a greater risk attached to assumptions for response rates and ineligible addresses.

The PAF does still include a number of ineligible addresses, such as small businesses, second homes, holiday rental accommodation and vacant properties. A review of the previous performance of individual surveys found that they each recorded fairly consistent levels of ineligible address for each Local Authority. This meant that robust assumptions could be made for the expected levels of ineligible addresses in the sample size calculations.

As the samples for the SHS, SHeS and SCJS have all been selected by the Scottish Government since 2012, addresses selected for any of the surveys are removed from the sample frame so that they cannot be re-sampled for another survey. This will help to reduce respondent burden. The addresses are removed from the sample frame for a minimum of 4 years.

The sample design specified in Section 1.2 was implemented in three stages:

1) All primary sampling units (datazones on the islands, intermediate geographies elsewhere) were randomly allocated to one of the four years of fieldwork. This meant that the sample was drawn from one quarter of PSUs each year and ensured that over four years (2012 to 2015) of fieldwork all addresses had a non-zero probability of selection. One quarter of target adult sample was required to complete the biological module. To make fieldwork more efficient, rather than randomly allocating addresses from the entire survey to the module, each year PSUs were allocated to the biological module and all selected addresses within those PSUs were eligible for the biological interview. To guard against a lower response rate, a proportion higher than the required one quarter of PSUs (38.6% in 2013) were allocated to the biological module.

Figure 1F: Primary sampling units selected in 2013

Health Board

PSUs in 2013 Sample

Total PSUs

Angus

22

90

Ayrshire and Arran

23

92

Borders

7

29

Dumfries and Galloway

9

35

Fife

26

103

Forth Valley

19

74

Grampian

32

128

Greater Glasgow and Clyde

68

273

Highland

19

76

Lanarkshire

34

137

Lothian

44

177

Orkney

7

27

Shetland

8

30

Western Isles

9

36

Total

327

1,307

2) The required number of addresses for the main sample and child boost sample were combined to give an overall total of addresses to sample for each stratum (local authorities plus Lanarkshires split). The overall number of addresses for each stratum was then sampled from the sample frame of addresses in active PSUs. Systematic random sampling was used with addresses within PSUs ordered by urban-rural classification, SIMD rank and postcode.

3) Once the overall sample was selected, each address was randomly allocated to the main sample or the child boost sample.

1.2.7 Selecting households at addresses with multiple dwellings

A small number of addresses have only one entry in the Postcode Address File (PAF) but contain multiple dwelling units. Such addresses are identified in the PAF by the Multiple Occupancy Indicator (MOI). To ensure that households within MOI addresses had the same probability of selection as other households, the likelihood of selecting the addresses was increased in proportion to the MOI. At addresses with more than one dwelling unit fieldworkers have a programme to randomly select the household at which interviews should be sought. There are generally a few cases where the MOI on the PAF is inconsistent with the actual number of dwelling units. When this occurred the fieldworkers recorded the information and a correction was made through the survey weighting.

1.2.8 Selecting individuals within households

For both the main and Health Board boost samples all adults aged 16 and over in responding households were selected for interview. To ease respondent burden, for child interviews for both the main sample and the child boost sample a maximum of two children were interviewed at each household. If a household contained more than two children then two were randomly selected for interview.

1.3 Topic coverage

1.3.1 Introduction

The most recent consultation on the content of the SHeS questionnaire was carried out in 2011, ahead of the 2012-2015 surveys.[5] Many of the topics included in previous years are being included again in the 2012 to 2015 surveys. As with previous years, the survey had a focus on cardio-vascular disease (CVD) and its associated risk factors.

1.3.2 Documentation

Copies of all the documents used in data collection are included in Appendix A. Full copies of the questionnaire documentation used in the main interview and biological module are also included in Appendix A. Protocols for taking measurements (height, weight, waist and blood pressure) and collecting biological samples (saliva and urine) are included in Appendix B. A summary of the main interview content and the content of the biological module is provided below.

1.3.3 Main interview

Information was collected at both the household and individual level. The table that follows summarises the content of the individual level interviews for all participants. The topics a participant was asked depended both on their age and the sample type to which their address had been allocated to. The age criteria for each topic is included in brackets after the topic name.

Version A households accounted for approximately 60-65% of the main (core) sample. At these households the questionnaire included the core questions and the questions in the Version A rotating module. In 2013, the topics included in the Version A rotating module were: accidents, stress at work, experience of discrimination and harassment, social capital, and dental health services.

Version B households accounted for the remaining 35-40% of the main (core) sample. At these addresses participants were only asked the core questions during the main interview with participating adults (aged 16+) also eligible to complete the biological module.

Figure 1G: Content of the 2013 Interview

CORE SAMPLE - Main interview outline

Version A

Version B

Household questionnaire including household composition

General health (0+) including caring (4+)

General CVD (16+)

Use of health services (0+)

Respiratory (16+)

-

Asthma (0+)

Accidents (0+)

-

Physical activity adults (16+) and children (2-15)

Sedentary activity adults (16+) and children (2-15)

Barriers and motivations to exercise (16+)

Eating habits children (2-15)

Fruit and veg consumption (2+)

Vitamins and supplements (16+)

Smoking and Drinking (16+) [16-19 in a self completion]

Dental health (16+)

Dental services (16+)

Social capital (16+)

Discrimination and harassment (16+)

Economic activity (16+)

Stress at work (16+)

Education (16+)

Ethnic background, religion and country of birth(0+)

Family health background and parental job details (16+)

Self-completions (13+ & parents of 4-12 yr olds)

Height (2+) and Weight (2+)

Data linkage and follow-up research consents (0+)

-

Biological module (16+)

A significant number of changes were made to the questionnaire content in advance of the 2012 to 2015 surveys based on the 2011 consultation. These changes are discussed in the 2012 Technical Report[3] and the SHeS Questionnaire Review Report 2012-2015.[5] The only change to the questionnaire in 2013 (excluding the rotating module) was the addition of a new question in the paper self-completion questionnaires on knowledge of the physical activity recommendations.

1.3.4 Self-completion questionnaire

Participants aged 13 and over were asked to fill in a self-completion booklet during the interview. In all, four different booklets were used in the survey. The version administered was dependent on a participant's age.

The booklet for young adults aged 16-17 asked about smoking and drinking behaviour (instead of being asked in the CAPI interview). Interviewers also had the option of using this young adults booklet for those aged 18-19 if they felt that it would be more appropriate for them to answer the questions in this format rather can face to face (e.g. might be more likely to give more honest answers than in the face to face interview when other household members including parents may be present).

Paper questionnaire booklets were administered for the following groups in the 2013 survey:

Adults

AUDIT questionnaire (designed to identify signs of hazardous or harmful drinking or possible alcohol dependence), General Health Questionnaire 12 (GHQ12), Warwick Edinburgh Mental Well-being scale (WEMWBS), gambling, use of contraception, knowledge of the physical activity recommendations, and sexual orientation

Young adults

Smoking, drinking, AUDIT questionnaire (designed to identify signs of hazardous or harmful drinking or possible alcohol dependence), GHQ12, WEMWBS, gambling, use of contraception, knowledge of the physical activity recommendations, and sexual orientation

13-15 year olds

GHQ12, WEMWBS and knowledge of the physical activity recommendations

Parents of 4-12 year olds

Strengths and Difficulties Questionnaire (SDQ) designed to detect behavioural, emotional and relationship difficulties in children, and knowledge of the physical activity recommendations.

1.3.5 Height and weight

Interviewers measured the height and weight of all participants aged 2 and over if they agreed. Protocols for taking height and weight measures are included in Appendix B.

1.3.6 Biological module

As highlighted previously, at a sub-sample (of around 35-40%) of main core sample addresses, adults (aged 16 and over) were eligible to complete the biological module. From 2012 to 2015, specially trained interviewers are collecting some of the measurements and samples which were collected by nurses prior to 2012.

Since the same interviewer administered the main interview and the biological module, the latter could either be completed immediately after the main interview or on a separate occasion.

As part of the module, participants were asked whether they used any medicines, pills, syrups, ointments, puffers or injections prescribed to them by a doctor or nurse. If participants answered yes to questions in the main interview about taking medication for high blood pressure, a heart condition or stroke then they would be asked to give the names of the drugs to the interviewer. This information is used to interpret blood pressure readings.

In addition to height and weight measurements, interviewers also took the following measurements from participants taking part in the biological module: blood pressure and waist circumference. Written agreement was also sought to take samples of saliva (for the analysis of cotinine, a derivative of nicotine) and spot urine samples (for the analysis of dietary sodium).

Participants were also asked a set of questions about depression, anxiety, suicide attempts and self-harm (taken from the Adult Psychiatric Morbidity Survey) in computer assisted self-interviewing (CASI) format.

Figure 1H: Content of the 2013 Biological Module

Outline of the Biological Module (age 16+)

Prescribed medicines (if has heart condition, high blood pressure or has had stroke)

Blood pressure

Waist measurement

Use of Nicotine Replacement therapy

Saliva sample

Urine sample

Depression, anxiety, suicidal attempts and self-harm

1.4 fieldwork procedures

1.4.1 Advance letters

Each sampled address was sent an advance letter that introduced the survey and stated that an interviewer would be calling to seek permission to interview. Three versions of the advance letter were used in 2013; one for the core version A and Health Board boost addresses, one for the core version B (with the biological module) addresses, and a separate version was sent to child boost addresses. A copy of the survey leaflet was included with every advance letter. The survey leaflet introduced the survey, described its purpose in more detail and included some summary findings from previous surveys.

1.4.2 Making contact

At initial contact, the interviewer established the number of dwelling units (DUs) and/or households (HHs) at an address and made any necessary selections (see Section 1.2).

The interviewer then attempted to make contact with each household. In the main sample they attempted to interview all adults (up to a maximum of ten) and up to two children aged 0-15 (see Section 1.2). At child boost sample households, interviewers first screened for children aged 0-15. In those households where children were present up to two children were selected for interview. Interviewers sought the consent of the parent and the child before the interview. At Health Board boost sample households interviewers attempted to interview a maximum of ten adults at selected households.

1.4.3 Collecting data

Interviewers used computer assisted interviewing.

At each co-operating eligible household in all sample types, the interviewer first completed a household questionnaire, information being obtained from the household reference person[6] or their partner wherever possible. This questionnaire obtained basic information (including date of birth and relationship to other household members) about all members of the household, regardless of age. The CAPI program then created individual questionnaires for each eligible participant in the household.

An individual interview was carried out with all selected adults and children in a household. In order to reduce the amount of time spent in the home, interviews could be carried out concurrently, with the program allowing up to four participants to be interviewed in a session.

Height and weight measurements were obtained towards the end of the interview.

In addition to an advance letter and general survey leaflet, participants were also given a more detailed leaflet describing the contents and purpose of the interview. Adults in households eligible for the biological module were given a longer version of this leaflet, providing information on the measurements and samples being taken.

A separate version of this leaflet was used for children in the main and child boost households. Parents at child boost addresses were also given a leaflet containing background information on the survey. Participant leaflets are included in Appendix A.

1.4.4 Introducing the biological module

Only a sub-sample of adults in the main sample was eligible to take part in the biological module. At the end of the main interview, adult participants in Version B addresses were given a Measurement Record Card which had additional information about the measurements and samples and asked for their agreement to take part in the biological module stage of the survey. Wherever possible, interviewers would carry out the measurements and collect the samples directly after the main interview. If this was not possible the interviewer would arrange to go back at a convenient time to complete this part of the survey. During the module the interviewer carried out the measurements described in Section 1.3.6 and obtained the saliva and urine samples from those adults eligible and willing to provide them. Written consent was obtained from participants before saliva and urine samples were taken. The consent statements are included in Appendix A.

1.4.5 Interviewing and measuring children

Children aged 13-15 were interviewed directly by interviewers, permission having first been obtained from the child's parent or guardian. Interviewers were instructed to ensure that the child's parent or guardian was present in the home throughout the interview. Information about younger children was collected directly from a parent/guardian. Whenever possible, younger children were present while their parent/guardian answered questions about their health. This was partly because the interviewer had to measure their height and weight and it also ensured that the child could contribute information where appropriate.

1.4.6 Feedback to participants

If participants wished, interviewers recorded height and weight measurements on their measurement record card.

Participants eligible for the biological module were given the Measurement Record Card for reference. If participants had their waist measurement and blood pressure taken then interviewers recorded their results on this card (if the participant wished).

Interviewers were issued with a set of guidelines to follow when commenting on participants' blood pressure readings (see Appendix B for details). If an adult's blood pressure reading was severely raised, interviewers were instructed to contact the Survey Doctor at the earliest opportunity. The Survey Doctor would then phone the participant and advise them to contact their GP as soon as possible.

1.5 Fieldwork quality control and ethical clearance

1.5.1 Training interviewers

Interviewers were fully briefed on the administration of the survey, including screening for households with children for the child boost sample. They were also trained and accredited in measuring height and weight.

Interviewers who had not previously worked on SHeS were accompanied by an interviewer supervisor during the early stages of their work to ensure that interviews and protocols were being correctly administered.

Interviewers interested in administering the biological module were initially screened for suitability. Minimum competency levels and experience in relation to SHeS were set and only interviewers that met the criteria were invited to the training and accreditation sessions.

Training to administer the biological module took place over three days. At the end of the training session interviewers were accredited on administering each of the measurements and samples and were only able to work on the module if they passed their accreditation.

Interviewers were also accompanied by a nurse supervisor (with previous experience of working on the survey) on their first visit. They are also supervised in the field annually by an experienced survey nurse to ensure they are administering the measurements and samples in line with SHeS protocols. They are also reaccredited annually by the research team.

Full sets of written instructions, covering both survey procedures and measurement protocols, were provided for interviewers (Appendix B contains a copy of the measurement protocols).

1.5.2 Checking interviewer and measurement quality

A large number of quality control measures were built into the survey to check on the quality of interviewer performance at both the data collection stage and subsequently. Recalls to check on the work of interviewers were carried out at 10% of productive households.

The computer program used by interviewers had in-built soft checks (which can be suppressed) and hard checks (which cannot be suppressed) which included messages querying uncommon or unlikely answers as well as answers outside an acceptable range. For example, if someone aged 16 or over had a height entered in excess of 1.93 metres, a message asked the interviewer to confirm that this was a correct entry (a soft check), and if someone said they had carried out an activity on more than 28 days in the last four weeks the interviewer would not be able to enter this (a hard check). For children, the checks were age specific. Some infants were weighed by having an adult hold them; the weight of the adult on their own was entered into the computer followed by the combined weight of the infant and adult. A hard check was used to ensure that the weight entered for the adult alone did not exceed the weight of the infant and adult combined.

1.5.3 Ethical clearance

Ethical approval for the 2013-2015 surveys was obtained from the Multi-Centre Research Ethics Committee for Wales (REC reference number: 12/WA/0261).

1.6 SURVEY RESPONSE

1.6.1 Introduction

This section presents the fieldwork outcomes for the sampled addresses. Survey response is an important indicator of survey quality as non-response can introduce bias into survey estimates. Standardised outcome codes (based on an updated version of those published in Lynn et al (2001)[7] for survey fieldwork were applied across the SHeS, SHS and SCJS. This enables consistent reporting of fieldwork performance and effective comparison between the performance of the surveys.

1.6.2 Household response

Table 1.1 shows a detailed breakdown of the SHeS response for all sampled addresses for Scotland in 2013. Addresses with unknown eligibility have been allocated as eligible and ineligible proportional to the levels of eligibility for the remainder of the sample. This approach provides a conservative estimate of the response rate as it estimates a high proportion of eligible cases amongst addresses with unknown eligibility.

At each selected household in the main sample all adults and a maximum of two children were eligible for interview. When considering the household response rate, households classed as "responding" were those where at least one eligible person was interviewed. The table shows that for the combined main and Health Board boost sample 66.3% of eligible households were classed as responding with all individual interviews complete at 51.4% of households.

For the child boost sample over 70% of households were ineligible as they did not contain any children under the age of 16. For eligible households 72.4% were classed as responding, with almost all interviews being completed in the responding households.

Table 1.2 shows that across Heath Boards the household response rate ranges from 59% (Greater Glasgow & Clyde) to 86% (Eilean Siar). Fully cooperating households were those where all eligible individuals were interviewed, all height and weight measured and, if eligible, completed the biological module. The definition of a fully cooperating household changed in 2012 and is therefore not comparable with fully cooperating figures prior to this.

Table 1.3 shows that the household response rate for eligible addresses in the child boost sample varied from 44% (Borders) to 88% (Dumfries and Galloway). Table 1.1 - Table 1.3

1.6.3 Individual response for adults

Overall there were 4,894 adult responses to the 2013 SHeS with 1,254 responses to the biological module, detailed in Table 1.4.

In order to calculate the adult response rate, since all adults in households were eligible for interview, the number of adults in non-responding households had to be estimated to calculate the total number of adults in all households. This was undertaken by calculating the average number of men and women per household for responding households and non-responding households (where information on the composition is known) and applying this to the households where nothing is known. The total estimated number of adults from sampled addresses eligible for interview is referred to as the "set" sample. For 2013 the set sample of men was 4,163 and for women 4,564.

Table 1.4 shows the adult response rate broken down by gender. The adult response rate was 51% for men, 60% for women and 56% overall.

In responding households (those households where at least one interview was completed) additional information on respondents allowed the consideration of response to stages of the survey by gender and age group. This is shown in Tables 1.5 and 1.6. For both men and women the younger age groups were found to have a lower response rate (49% for men aged 16 to 24 and 53% for women aged 16 to 24) than older age groups (over 90% response rate for men and women over 65).

As part of the biological module, respondents were asked to have their waist and blood pressure measured and to provide saliva and urine samples. Almost all individuals completing the biological module interview allowed the waist and blood pressure measurements to be taken but there was a drop off in providing the samples. Of those eligible for the biological module, a urine sample was not obtained or refused by 9% of men and 13% of women.

Table 1.9 shows that men are under-represented in the SHeS sample compared to NRS population estimates as they made up 44% of the sample but 48% of the population. Younger age groups were also under-represented in the SHeS sample when compared to NRS population estimates. In particular, men and women under 25 were under-represented. Conversely, men and women over 65 were over represented in the sample. Tables 1.4 - 1.6, Table 1.9

1.6.4 Individual response for children (0-15)

Interviews were undertaken with 1,839 children aged 0 to 15, with 910 interviews taking place as part of the main sample and 929 as part of the child boost.

As was the case with the adult sample, in order to calculate the response rate for children, the number of eligible children in selected households (the "set" sample) had to be estimated. This was done by assuming that, for both the main sample and the child boost sample, the non-responding and responding households contained the same average number of children.

Table 1.7 shows that overall response rates for both the main sample and the child boost sample were the same for boys and girls with a response rate of 56% for the main sample and 72% from the child boost sample.

Child response rates have also been calculated for children in responding households. Table 1.8 shows that for age groups under 11 years old the response rate for boys and girls was fairly consistent at over 96% for every group, however, the response rate for children aged 11 to 15 was slightly lower at 88% for boys and 90% for girls.

Table 1.7, Table 1.8

1.7 WEIGHTING THE DATA

1.7.1 Introduction

This section presents information on the weighting procedures applied to the survey data. Since 2012 the weighting for SHeS has been undertaken by the Scottish Government rather than the survey contractor (as had previously been the case), but the methodology applied was largely consistent with that of the 2008 to 2011 sweeps of the survey. The procedures for the implementation of the weighting methodology were developed by the Scottish Government working with the Methodology Advisory Service at the Office for National Statistics.[8]

To undertake the calibration weighting the ReGenesees Package for R was used and within this to execute the calibration a raking function was implemented.

1.7.2 Main adult weights

The main adult weight is applicable for all adults interviewed as part of the main sample and the health board boosts. There were six steps to calculating the overall adult weights. These were as follows:

1) Address selection weights (w1)

The address selection weights were calculated to compensate for unequal probabilities of selection of addresses in different survey strata. For the main sample with the health board boost there were 36 strata overall (one for each local authority, an extra strata for the Lanarkshires and two extra strata in Fife as a result of the boost). The address selection weight for each stratum was calculated as:

Mathematical Equation

2) Dwelling unit selection weights (w2)

As stated in Section 1.2.7, the MOI for the PAF was used to ensure that if there were multiple dwelling units at a single address point then they would have the same selection probability as individual addresses. However, there were some cases where the MOI was incorrect. The following correction was applied where this was the case:

Mathematical Equation

With w2 trimmed to a maximum of 3.

3) Household selection weights (w3)

Similarly, within a very small number of dwelling units fieldworkers found multiple households, of which only one was selected for participation in the SHeS. The following correction was applied for multiple households:

Mathematical Equation

With w3 trimmed to a maximum of 3.

4) Calibrated household weights (w4)

The three selection weights were combined (w1*w2*w3) before the household calibration stage. This combined weight was applied to the survey data to act as entry weights for the calibration. The execution of the calibration step then modified the entry weights so that the weighted total of all members of responding households matched the population totals for Health Boards, Scotland-level population totals for age/sex breakdown, and the population within SIMD15 areas. The population totals that were used were the National Records of Scotland's (NRS) mid-2012 estimates for private households. Figures 1I - 1K show the target populations.

5) Adult non-response weights (w5)

All adults within selected households were eligible for interview, but within responding households not all individuals completed an interview. The profiles of household members that did not complete the interview were different from those that do. Information on all individuals within responding households was available through information gathered as part of the household interview. This allowed the differential response rates for individuals within households to be modelled using logistic regression to calculate a probability of responding based on their profiles. The logistic regression was only applicable for households containing more than one adult since households consisting of only one adult either responded to the household and individual interviews or did not respond at all.

The following variables were considered for inclusion in the model:

  • Health Board
  • Age/sex
  • Number of adults in the household
  • Employment status of household reference person
  • Presence of a smoker in the household
  • Frequency of eating meals together
  • Marital status
  • Tenure
  • Urban/rural classification
  • Access to a car
  • Located within SIMD15 area

Through running backwards and forwards selection procedures for the logistic regression the following variables were included in the final model:

  • Health Board
  • Age/sex
  • Number of adults in the household
  • Frequency of eating meals together
  • Marital status
  • Employment status of household reference person
  • Located within SIMD15 area

The final logistic regression model was then used to calculate the probability of response for all individuals that did respond. The adult non-response weight (w5) was then calculated as the reciprocal of this probability:

Mathematical Equation

For households of only one adult the non-response weight was one.

6) Individual calibration and final adult weight (int13wt)

The household (w4) and non-response (w5) were combined (w4*w5) and applied to the survey data prior to the final stage of calibration weighting which matched weighted totals for the survey data to the NRS 2012 mid-year population estimates for Health Boards, age/sex distribution at Scotland level and age/sex distribution for the Glasgow and Greater Clyde Health Board.

Figure 1I: 2012 Mid-year population estimates for private households in Scotland by Health Board

Health Board

Children

Adults

Total

Ayrshire & Arran

63,841

304,827

368,668

Borders

19,121

93,368

112,489

Dumfries & Galloway

24,560

124,357

148,917

Fife

64,218

294,164

358,382

Forth Valley

53,170

238,664

291,834

Grampian

96,917

462,970

559,887

Greater Glasgow & Clyde

206,161

985,431

1,191,592

Highland

54,534

258,781

313,315

Lanarkshire

105,704

464,663

570,367

Lothian

142,343

679,596

821,939

Orkney

3,522

17,758

21,280

Shetland

4,333

18,616

22,949

Tayside

68,033

332,162

400,195

Western Isles

4,565

22,652

27,217

Total

911,022

4,298,009

5,209,031

Total figures might not be exact due to rounding

Figure 1J: 2012 Mid-year population estimates for private households in Scotland by SIMD15 indicator

SIMD15

Total population

15% most deprived datazones

762,340

All other datazones

4,446,691

Total

5,209,031

Total figures might not be exact due to rounding

Figure 1K: 2012 Mid-year population estimates for private households in Scotland by age group

Age group

Male

Female

Total

0-4

150,320

144,371

294,691

5-9

140,433

134,009

274,442

0-15

174,934

166,955

341,889

16-24

297,681

293,045

590,726

25-34

324,980

341,837

666,817

35-44

339,837

361,542

701,379

45-54

384,547

405,572

790,119

55-64

321,546

336,264

657,810

65-74

236,421

265,757

502,178

75+

156,563

232,417

388,980

Total

2,527,262

2,681,769

5,209,031

Total figures might not be exact due to rounding

1.7.3 Biological module weights

A similar process was applied to derive the weights for the biological module. This is outlined below.

1) Address selection weight (bw1)

New address selection weights were calculated using the same process as described for w1 but with the Health Board boost addresses excluded.

2) Dwelling unit (w2) and household selection weights (w3)

The dwelling unit and household selection weights from the main adult weight were applied as above.

3) Calibrated household weight (bw4)

The three selection weights were combined (bw1*w2*w3) and applied to the survey data before the household calibration was run so that survey data matched the population totals for Health Boards, Scotland-level age/sex breakdown, and the population within SIMD15 areas.

4) Adjustment for biological module selection (bw5)

Approximately one third (38.6%) of the main sample was allocated to the biological module. To incorporate this probability of selection a correction was applied to the calibrated household weight (bw4). The correction was:

Mathematical Equation

5) Application of adult non-response (w5)

For within household non-response, the non-response weight (w5) calculated for all households was also applicable for the biological module.

6) Non-response weight for biological module interview

Not all of the adults that responded to the main section of the interview responded to the biological module. Using the information collected for the respondent in the main interview and household interview the likelihood of responding to the biological module was modelled with logistic regression. The following variables were considered for inclusion in the model:

  • Health Board
  • Age/sex
  • Number of adults in the household
  • Employment status of Household reference person
  • Presence of a smoker in the household
  • Frequency of eating meals together
  • Self-assessed general health
  • Gardening/DIY/building work in past 4 weeks
  • Any physical activity in past 4 weeks
  • Economic activity (working/retired/sick)
  • Any housework in past 4 weeks
  • Marital status
  • Tenure
  • Urban/rural classification
  • Access to a car
  • Located within SIMD15 area
  • Long-term illness or disability
  • Highest qualification held
  • Ever had high blood pressure
  • Current smoker
  • Currently drink alcohol
  • Number of natural teeth
  • Any barriers to entry to the property

Through running backwards and forwards selection procedures for the logistic regression the following variables were included in the final model for response to the biological module:

  • Health Board
  • Age/sex
  • Number of adults in the household
  • Employment status of Household reference person
  • Frequency of eating meals together
  • Economic activity (working/retired/sick)
  • Marital status
  • Urban/rural classification
  • Located within SIMD15 area
  • Long-term illness or disability
  • Ever had high blood pressure

The final logistic regression model was then used to estimate the probability of response for all individuals that did respond to the biological module. The biological module non-response weight (bw6) was then calculated as the reciprocal of this probability:

Mathematical Equation

7) Final calibration for biological module (bio13_wt)

The household (bw4), biological sample correction (bw5), adult non-response (w5), and biological non-response (bw6) weights were combined (bw4*bw5*w5*bw6) and applied to the survey data.

For the final stage of biological module weighting the weighted totals for the survey data were calibrated to match the NRS 2012 mid-year population estimates for private households for Health Boards, age/sex distribution at Scotland level. However, due to the low sample size for the module a number of the categories had to be collapsed. In terms of Health Boards, all areas except for Grampian, Greater Glasgow and Clyde, Highland, and Lanarkshire were grouped together. For the age groups, the lowest two age groups were combined as were the highest two age groups.

1.7.4 Urine sample weights

A similar method was used to generate the adult urine sample weights. A urine sample was not obtained from every adult that responded to the biological module so a weight was calculated to correct for non-response. The method used was to start with the biological module weight and use logistic regression to model the probability that a respondent from the biological module would give a urine sample.

The following variables were considered for inclusion in the model:

  • Health Board
  • Age/sex
  • Number of adults in the household
  • Employment status of Household reference person
  • Presence of a smoker in the household
  • Frequency of eating meals together
  • Self-assessed general health
  • Gardening/DIY/building work in past 4 weeks
  • Any physical activity in past 4 weeks
  • Economic activity (working/retired/sick)
  • Any housework in past 4 weeks Marital status
  • Tenure
  • Urban/rural classification
  • Access to a car
  • Located within SIMD15 area
  • Long-term illness or disability
  • Highest qualification held
  • Ever had high blood pressure
  • Current smoker
  • Currently drink alcohol
  • Number of natural teeth
  • Any barriers to entry to the property

Through running backwards and forwards selection procedures for the logistic regression the following variables were included in the final model for response to the urine sample:

  • Health Board
  • Age/sex
  • Located within SIMD15 area
  • Gardening/DIY/building work in past 4 weeks
  • Any physical activity in past 4 weeks
  • Economic activity (working)
  • Any housework in past 4 weeks

The non-response weight for the urine sample was calculated as the reciprocal of the probability of providing a sample given by the logistic regression model. This non-response weight was combined with the pre-calibration biological module weight and applied to the survey data to give interim weighted totals. These totals were calibrated to match health board and age/sex population totals to arrive at the final urine weight (urine13_wt).

1.7.5 Non-biological module weights (Version A)

A weight titled "Version A" was calculated for the individual respondents in the main sample that were not selected for the biological module. This consisted of the main sample without the Health Board boost sample. The following steps were followed to derive the weight:

1) Address selection weight (bw1)

As derived in the first step of the biological module weight.

2) Dwelling unit (w2) and household selection weights (w3)

The dwelling unit and household selection weights from the main adult weight were applied as above.

3) Calibrated household weight (bw4)

As derived for the biological module.

4) Adult non-response weight (w5)

For within household non-response, the non-response weight (w5) calculated for all households was also applicable for the biological module.

5) Final calibration for Version A weight (verA13wt)

The household (bw4) and adult non-response (w5) weights were combined (bw4 *w5) and applied to the survey data. As was the case with the main adult weight and biological module weight, the weighted totals for the survey data were calibrated to match the NRS 2012 mid-year population estimates for private households for Health Boards, age/sex distribution at Scotland level.

1.7.6 Overall child weights

An overall child weight was derived for child responses from the main sample and from the child boost combined. Separate logistic regression non-response weights were not required for the child samples as the response rate for children within cooperating households was sufficiently high at 94%. The weighting steps are shown below. Steps (1) and (2) followed the same process as described in 3.2 above.

1) Address selection weight for main sample and child boost combined (cw1)

2) Dwelling unit (cw2) and household (cw3) selection weights

3) Selection of children within each household (cw4)

A maximum of two children were eligible for interview in each household. To ensure that children in larger households were not under-represented in the final sample the following child selection weight was calculated for households with more than two children to compensate for the probability of selection:

Mathematical Equation

For households with two or less children cw4=1.

4) Calibration for child interview weight (cint13wt)

The address selection (cw1), dwelling unit (cw2), household (cw3) and child selection weights (cw4) were combined (cw1*cw2*cw3*cw4) and applied to the survey data. The weighted totals for the survey data were calibrated to match the NRS 2012 mid-year population estimates private households for Health Boards, age/sex distribution at Scotland level.

1.7.7 Combined weights

Several weights have also been calculated to allow for analysis of combined data from the 2012 and 2013 surveys.

The weights provided for combined years of data are:

Weight name

Purpose of combined weight

int1213_wt

For analysis of 2012 and 2013 combined adult data

cint1213_wt

For analysis of 2012 and 2013 combined child data

Bio1213_wt

For analysis of 2012 and 2013 combined biological data

urine1213_wt

For analysis of 2012 and 2013 combined urine data

In each case, the calculation of the weights followed the same procedure. The pre-calibration weights which had already been calculated for the individual years (which take into account selection weighting and (except for the child weights) non-response weighting) were combined and calibrated to Health Board and age/sex 2012 population totals for private households.

1.8 data analysis and REPORTING

SHeS is a cross-sectional survey of the population. It examines associations between health states, personal characteristics and behaviour. However, such associations do not necessarily imply causality. In particular, associations between current health states and current behaviour need careful interpretation, as current health may reflect past, rather than present, behaviour. Similarly, current behaviour may be influenced by advice or treatment for particular health conditions.

1.8.1 Reporting age variables

Defining age for data collection

A considerable part of the data collected in the 2013 SHeS is age specific, with different questions directed to different age groups. During the interview the participant's date of birth was ascertained. For data collection purposes, a participant's age was defined as their age on their last birthday before the interview.

Age as an analysis variable

Age is a continuous variable, and an exact age variable on the data file expresses it as such (so that, for example, someone whose 24th birthday was on January 1 2013 and was interviewed on October 1 2013 would be classified as being aged 24.75 (24¾)).

The presentation of tabular data involves classifying the sample into year bands. This can be done in two ways, age at last birthday and 'rounded age', that is, rounded to the nearest integer. In this report all references to age are age at last birthday.

Age standardisation

Some of the adult data included in the 2013 report have been age-standardised to allow comparisons between groups after adjusting for the effects of any differences in their age distributions. If data reported has been age-standardised this is highlighted in the title to the table or chart. When different sub-groups are compared in respect of a variable on which age has an important influence, any differences in age distributions between these sub-groups are likely to affect the observed differences in the proportions of interest.

It should be noted that all analyses in the report are presented separately for men and women and on some occasions data for all adults are also presented. All age standardisation has been undertaken separately within each sex, expressing male data to the overall male population and female data to the overall female population. When comparing data for the two sexes, it should be remembered that no age standardisation has been introduced to remove the effects of the sexes' different age distributions.

Age standardisation was carried out using the direct standardisation method. The standard population to which the age distribution of sub-groups was adjusted was the mid-year 2012 private household population estimates for Scotland. The age-standardised proportion p' was calculated as follows, where pi is the age specific proportion in age group i and Ni is the standard population size in age group i:

Mathematical Equation

Therefore p' can be viewed as a weighted mean of pi using the weights Ni. Age standardisation was carried out using the age groups: 16-24, 25-34, 35-44, 45-54, 55-64, 65-74 and 75 and over. The variance of the standardised proportion can be estimated by:

Mathematical Equation

1.8.2 Standard analysis breakdowns

Household income

The 2013 survey included questions designed to measure participants' household income. While household income alone can be used as an analysis variable, the analysis conducted for this report used an adjusted measure which took account of the number of persons within the household. The McClements method was used to equivalise incomes; this is detailed in the Glossary at the end of this report. The equivalised income measure was divided into quintiles for the presentation of analysis within the report, but the full continuous data is available on the dataset.

Scottish Index of Multiple Deprivation (SIMD)

The analysis was based on the 2012 version of the Scottish Index of Multiple Deprivation.[9] It is based on 38 indicators in seven individual domains of current income, employment, housing, health, education, skills and training, geographic access to services and crime. SIMD is calculated at data zone level, enabling small pockets of deprivation to be identified. The data zones are ranked from most deprived (1) to least deprived (6505) on the overall SIMD index. The result is a comprehensive picture of relative area deprivation across Scotland. The index was divided into quintiles for the presentation of analysis within this report, a version divided into deciles is also available on the dataset. The full index is not available on the archived dataset due to concerns about its potential for identifying individual respondents or households.

1.8.3 Reporting biological data

As highlighted in Section 1.3.6 interviewers specially trained in administering biological measurements and sample collections have been conducting the biological module since 2012. This has enabled the continued collection of some of the measurements and samples that were previously carried out by survey nurses (between 1995 and 2011). Analysis of the interviewer administered data is included in the 2013 report for the first time, combining data collected by interviewers in the 2012 and 2013 surveys.

Urine and saliva data

The protocol for collecting urine and saliva samples did not change when interviewer administered measurements and sample collection was introduced to SHeS in 2012. Therefore it was not considered necessary to adjust the data, presented in Chapters 4 and 5 in Volume 1, in any way to take account of the change from nurses to interviewers taking the sample.

Waist circumference data

For waist circumference, both the person administering the measurement and the protocol for taking it changed in 2012 (See Annex B for protocol used to collect waist circumference in 2012). To assess the impact of the switch in design and to help inform time-series analysis, a small validation study was carried out in 2012.[10] As part of the study, calibration equations allowing for survey statistics to be 'adjusted' to account for the change from nurses to interviewers taking the measurements as well as the change in protocol were produced. A report on the methodology and results of the validation study is available on the Scottish Government survey website.[10]

Waist measurement data is presented in Chapter 7 of Volume 1 to this report. Two sets of data for the 2012/ 2013 years combined are presented in Chapter 7:

  • Unadjusted 'interviewer collected' data which was obtained by interviewers during the biological module
  • Adjusted 'nurse equivalent' data, that is, the data a nurse would have collected had they conducted the measure using the nurse protocol

The 'nurse equivalent' data was created by applying calibration equations to the data collected by interviewers to produce equivalent 2012/2013 nurse data. Separate calibration equations were created for men and women for this report:

  • Nurse equivalent waist measurement (men) =1.30+0.98(interviewer waist)
  • Nurse equivalent waist measurement (women) =6.68+0.89(interviewer waist)

The calibration equations were produced by a regression model on data from 302 paired nurse and interviewer measurements obtained during the validation study. The value of R[2] calculated for the association between the nurse measurements (using the nurse protocol) and the interviewers' measurements (using the interviewer protocol) was 0.93 for men, and 0.88 for women. This indicates that the equations produced for both men and women can be considered to accurately convert interviewer measurements to nurse measurements, and vice-versa, with very little unexplained variation. Hence no problems are foreseen in their use for converting interviewer measurements to nurse measurements on the Scottish Health Survey. Further information on the calculation of the calibration equations as well as additional calibration equations can be found in the validation study report.[10]

In Table 7.7 the trend in adult waist circumference in Scotland is updated. The adjusted 'nurse equivalent' data should be used when monitoring the trend in adult waist circumference across survey years. The unadjusted 'interviewer collected' data has also been included in Table 7.3.

Both sets of data (unadjusted and adjusted) will be included in the datasets deposited with the UK Data Archive.

Blood pressure data

The protocol and the equipment used did not change when interviewers started taking blood pressure readings in 2012 (see Annex B for details on the protocol used by both interviewers and nurses prior to this). However it was necessary to assess the impact of switching from nurses to interviewers in terms of analysing trends in blood pressure measurements among the population over time. As part of the validation study discussed with reference to waist measurement above, calibration equations were also produced for blood pressure readings in order to enable the readings to be converted into equivalent nurse measurements.

Blood pressure data are presented in Chapter 8 of Volume 1. As with waist measurement, two sets of data are presented for the years 2012/2013 combined:

  • Unadjusted 'interviewer collected' blood pressure data which was obtained by interviewers during the biological module
  • Adjusted 'nurse equivalent' data, that is, the data a nurse would have collected had they collected blood pressure readings from the same participants

The 'nurse adjusted' data was produced by applying calibration equations to the data collected by interviewers in order to produce 2012/2013 nurse equivalent data. Separate calibration equations were created for systolic and diastolic measurements for use in this report:

  • Nurse adjusted systolic blood measurement =17.23+0.88(interviewer systolic)
  • Nurse adjusted diastolic blood measurement =10.84+0.86(interviewer diastolic)

Similarly to the calibration equations created for the interviewer collected waist measurements, the calibration equations for the blood pressure measurements were produced using data collected during the validation study. A regression model was run on the data from 306 paired nurse and interviewer measurements taken during the validation study. The value of R2 calculated for the association between the nurse measurements and the interviewers' measurements was 0.77 for systolic blood pressure measurements and 0.73 for diastolic measurements. This suggests that there was a good fit between the predicted nurse measurement based on the interviewer measurement and can therefore be applied to 2012/2013 combined data to monitor change over time. Further information on the validation study and calculation of the calibration equations can be found in the validation study report.[10]

In Table 8.7 the trend in adult blood pressure in Scotland is updated. The adjusted 'nurse equivalent' data should be used when monitoring the trend in blood pressure across survey years. The unadjusted 'interviewer collected' data has been included in Table 8.6 - Table 8.9 which examine 2012/2013 data only. Both sets of data (unadjusted and adjusted) will be included in the datasets deposited with the UK Data Archive.

1.8.4 Latent Class Analysis

Latent Class analysis (LCA) has been used in the Chapter on Multiple Risks and Vulnerabilities (Volume 1, Chapter 9) to categorise individuals into different groups or 'latent classes' of risk profiles depending on their response to a set of questions.

LCA operates by identifying the number of classes or groups that best fit the data and generating probabilities of membership of each group for every eligible participant. Once this is complete, a participant is assigned to the class for which they have the highest probability of membership. The first step is to identify how many different classes or groups best fit the data. To test this, a number of models, each containing a pre-specified number of classes, were produced. Results from each model were compared and the most appropriate solution selected. Once the latent classes were identified,[11] information about the age and deprivation profile of each class was used to help complete their description.

A six class model was selected to use in the analysis for Chapter 9; the classes identified were as follows (and % of all adults in each group):

Risk Group

% of all adults in group

Younger, low risks, no morbidity

48

Older, average risks, multimorbid

18

Older, high risks, low morbidity

13

Younger, average / high risks, no morbidity

9

Older, multiple high risks, multimorbid

8

Younger, multiple high risks, multimorbid

5

1.8.5 Design effects and true standard errors

SHeS 2013 used a clustered, stratified multi-stage sample design. In addition, weights were applied when obtaining survey estimates. One of the effects of using the complex design and weighting is that standard errors for survey estimates are generally higher than the standard errors that would be derived from an unweighted simple random sample of the same size. The calculations of standard errors shown in tables, and comments on statistical significance throughout the report, have taken the clustering, stratification and weighting into account. The ratio of the standard error of the complex sample to that of a simple random sample of the same size is known as the design factor. Put another way, the design factor (or 'deft') is the factor by which the standard error of an estimate from a simple random sample has to be multiplied to give the true standard error of the complex design. The true standard errors and defts for SHeS 2013 have been calculated using a Taylor Series expansion method. The deft values and true standard errors (which are themselves estimates subject to random sampling error) are shown in Tables 1.10 to 1.17 for selected survey estimates presented in the main report. Tables 1.10 to 1.17


Contact

Email: Julie Landsberg