06/05/2013

The European Health Interview Survey (EHIS) was developed between 2003 and 2006, during a process in which all the EU Member States (MS) were largely involved. It consists of four modules on health status, health care, health determinants, and background variables. Those modules may be implemented at the national level either as one specific survey or as elements of existing surveys (i.e. national health interview survey, labour force survey, other household surveys). The final version of the questionnaire was adopted by the MS at the Working Group on Public Health Statistics in November 2006.

The European Health Interview Survey (EHIS) aims at measuring on a harmonised basis and with a high degree of comparability among MS the health status, life style (health determinants) and health care services use of the EU citizens.

The survey contains around 130 questions split among the four modules covering the following topics:

Background variables on demography and socio economic status

Health status:

• Minimum European health module (MEHM): self-perceived health, limitation in daily activities and chronic conditions
• Disease specific morbidity
• Accidents / injuries
• Health related absenteeism from work
• Physical and sensory functional limitations
• Activities of daily living (ADL - feeding, bathing etc) and help received
• Instrumental activities of daily living (IADL - preparing meals, shopping, etc) and help received
• Pain
• Aspect of mental health

 Health care:

• Hospitalisation
• Consultations with doctors and dentists
• Unmet needs for consultation with a specialist
• Visits to specific health professionals
• Visits to specific categories of alternative medicine practitioners
• Use of home care services
• Satisfaction with services provided by health care providers
• Use of medicines
• Health care preventive actions (influenza vaccination, breast examination, cervical smear test etc)
• Out-of-pocket payments for medical care (self-completion form)

 Health determinants:

• Height and weight
• Physical activity
• Consumption of fruits, vegetables and juice
• Home external environment (noise, pollution, etc.)
• Workplace exposures
• Smoking behaviour (self-completion form)
• Alcohol consumption (self-completion form)
• Illicit drug use (self-completion form).

The first wave of the EHIS was implemented during the period 2006-2009 under a gentlemen's agreement. Nineteen countries have carried out it:

• 2006: AT, EE
• 2007: SI - CH
• 2008: BE, BG, CZ, CY, FR, LV, MT, RO, TR
• 2009: DE, EL, ES, HU, PL, SK

Germany provided aggregated data and for breakdowns with a strata size less than 20, the values were marked as confidental (flag ~c). No data are received for Switzerland an Turkey.

In total, 26 indicators (see document EHIS indicators guidelines) based on DG SANCO and DG EMPL needs and covering 3 domains (health status, health determinants and health care) are disseminated on Eurostat website.

Missing values, don't know and refusal answers (except the variable ISCED and the income) are not considered in the calculations. It means the indicators have been worked out on the respondents and validated answers only. Frequencies were not statistically adjusted for interaction effects between the background variables.

The indicators present distribution percentages and are calculated with different breakdown according to the indicator:

• sex, age group (10-years intervals from 15 years old and over) and educational level grouped (ISCED0-2, ISCED3-4, ISCED5-6);
• sex, age group (18-44, 45-54, 55-64, 65-74 and more then 75) and income quintiles.

For example: 4.5 % of Latvian women aged 25-34 are obese (BMI greater than 30).

All the indicators are worked out for the population aged 15 and over. Nevertheless, for some specific indicators, frequencies are calculated on different populations:

• the Body Mass Index (BMI) is calculated for adults only (18+);
• the self-reported prevalence of blood pressure (table hlth_ehis_st1) is computed for people aged 25+;
• the influenza vaccination rate (table hlth_ehis_hc1) is computed for people aged 65+;
• the breast cancer screening rate (table hlth_ehis_hc2) is computed on women aged 50-69;
• the cervical cancer screening rate (table hlth_ehis_hc3) is computed on women aged 20-69;
• the colon cancer screening rate (table hlth_ehis_hc4) is computed on people aged 50-74.

EHIS results are produced in accordance with the relevant international classification systems. The main classifications used is: ISCED 1997  for the level of education.

The EHIS is a general household survey in which by sample design no sectors are excluded.

The formula of the derivate indicators mentioned below can be found in the EHIS indicators guidelines. 

Body Mass Index (BMI): is defined as the weight in kilos divided by the square of the height in meters. For dissemination purpose, the following subdivision (according to the international obesity taskforce - IOTF) is used:

• less than 18.5: underweight
• between 18,5 and less than 25: normal weight
• between 25 and less than 30: overweight
• equal or greater than 30: obese 

Diseases - self-reported prevalence: proportion of individuals reporting to have been diagnosed with the disease which occurred during the past 12 months. Five self-reported prevalence have been computed with the EHIS: diabetes, chronic depression, asthma, chronic obstructive pulmonary disease - COPD and high blood pressure.

In France, variables measuring whether the condition was diagnosed by a general practitioner (GP) and whether it occurred during the past 12 months were not included in the questionnaire. Consequently, the 5 prevalence are worked out without taking on board this information.

In Austria and Estonia and also Germany, the chronic depression self-reported prevalence was not computed (no data).

Proportion of disease described as Chronic bronchitis, emphysema, other chronic obstructive pulmonary diseases was not computed in Germany.

Even though ICD10 codes were not explicitly asked in the survey, following scale is used for the dissemination:

• (any types of) diabetes E10-E14 (diabetes mellitus)
• chronic depression F32_F33 (depressive disorders)
• asthma J45 (asthma)
• J41 - J44 (Chronic bronchitis, emphysema, other chronic obstructive pulmonary diseases)
• High blood pressure I10-I15 (hypertensive diseases) 

Incidence of different types of accidents:

Indicator 1: proportion of individuals reporting to have had an accident during the past 12 months, which resulted in injury.

Indicator 2: proportion of individuals reporting to have had an accident during the past 12 months, which resulted in injury for which medical treatment was sought.

In Austria, Estonia and France and also Germany, both indicators were not computed (no data).

The following scale is used for the dissemination:

• Accident at home/school/ leisure accident
• Road traffic 

General musculoskeletal pain: proportion of persons with none / low / moderate / severe / very severe physical pain or physical discomfort during the past four weeks.

In Austria and France and also Germany, this indicator was not computed (no data). 

Psychological distress: the Mental Health Inventory (MHI-5) consists of three depression-related items and two anxiety-related items. It has a score of 0 to 100, where a score of 100 represents optimal mental health. The table contains average scores by sex, age and educational status. For instance in Bulgaria the average score of population aged 15-24 is equal to 85.4 whereas in 73.3 in Malta.

In Estonia and Germany, this indicator was not computed (no data). In France figures are not published, because data are not reliable. Not all question were asked.

Psychological well-being - Positive mental health: the Energy and Vitality Index (EVI) consists of two vitality-related items and two energy-related items. It has a score of 0 to 100, where a score of 100 represents optimal mental health. The table contains average scroe by sex, age and educational status. For instance in Austria population aged 15-24 has the average score of phychological well-being equal to 72 whereas population aged 75-84 in Austria has this average score 53.3.

In Estonia and Germany, this indicator was not computed (no data). In France figures are not published, because data are not reliable. Not all question were asked.

Daily cigarettes smokers: proportion of persons reporting to smoke cigarettes (manufactured and hand-rolled) daily. 

In Turkey, this indicator was not computed (no data). In France data are not reliable.

Smokers by number of cigarettes: proportion of persons who smoke less / more than 20 cigarettes per day. In France data are not reliable.

In Austria and Turkey, this indicator was not computed (no data). In France data are not reliable.

Hazardous alcohol consumption / binge drinking: proportion of persons who don't drink or seldom / who drink regularly but don't binge drink/ who drink regularly and binge drink regularly (less than monthly / monthly / weekly / daily or almost daily).

In Austria, Estonia and Turkey, this indicator was not computed (no data). In France, all data are flagged ' ~u '. In Poland different methodology was used.

Consumption of fruits (vegetables): proportion of  persons who eat fruits (respectively vegetables) twice or more a day / once a day / less than once a day but at least 4 times a week / less than 4 times a week but at least once a week /  less than once a week / who never eat fruits (respectively vegetables).

In Austria and also Germany, both indicators (consumption of fruits/ consumption of vegetables) were not computed (no data). 

Physical activity: percentage of the population practising at least 30 minutes of physical activity (moderate or intense) per day.

In Estonia, France and Belgium, Germany and Turkey this indicator was not computed (no data). In Romania, all data are flagged ' ~u '. 

Influenza vaccination rate in elderly population: percentage of persons (65+) who was vaccinated against flu during the past 12 months.

When respondents answered they were vaccinated against flu last year, a check was made to see whether it was really last year (months of the interview should be before the month of the vaccination otherwise the vaccination took place during the past 13, 14 months and not 12 months). For confidentiality reason, France didn't deliver the month of the interview but we know that the survey was carried out from April to July 2008 so only the people vaccinated between May and December 2007 are taken on board  when considering the last year period. In Romania, the questionnaire is a little bit different from the EHIS so special checking was also needed. Since the Romanian survey was carried out from May to June 2008, people vaccinated between June and December 2007 are selected when considering the last year period.

In Austria, Belgium and Spain this indicator was not computed (no data). 

Breast cancer screening - cervical cancer screening - colon cancer screening: percentage of persons reporting to have undergone a breast (respectively cervical / colon) cancer screening test within the past 12 months / more than 1 year, but not more than 2 years / more than 2 years, but not more than 3 years / not within the past 3 years / never.

In Austria and Estonia, the colorectal cancer screening rate was not computed (no data). 

Consultation of psychologist during the last year: proportion of individuals reporting to have had a contact with a psychologist or psychotherapist.

Self-reported visits to doctors:proportion of persons having consulted a doctor X times during the last 4 weeks. The scale used for the dissemination is 0 / 1 / 2 / 3 to 5/ more than 5 visits.

The following scale is used for the dissemination:

• physicians or doctors
• dentist or orthodontist
• medical or surgical specialist

In France, the number of consultations with a doctor (whatever is speciality) was not computed (no data).

In Estonia, the number of consultations with a medical or surgical specialist was not computed (no data).

Medicine use: percentage of persons reporting to take medication prescribed by a physician for a specific disease during the past 2 weeks. Five medicine use rates have been computed with the EHIS: for diabetes, chronic depression, asthma, chronic obstructive pulmonary disease - COPD and for high blood pressure.

In Estonia and France, these 5 measures were not computed (no data).

The following scale is used for the dissemination:

• (any types of) diabetes E10-E14 (diabetes mellitus)
• chronic depression F32_F33 (depressive disorders)
• asthma J45 (asthma)
• J41 - J44 (Chronic bronchitis, emphysema, other chronic obstructive pulmonary diseases)
• High blood pressure I10-I15 (hypertensive diseases) 

Online ICD codes.

For further details on the concepts and the calculation methods, please refer to EHIS indicators guidelines. 

Educational level: under EHIS, the attainment levels of individuals are classified according to the 'International Standard Classification of Education' version of 1997:

• Level 0: no formal education or below primary education
• Level 1: Primary education or first stage of basic education
• Level 2: Lower secondary or second stage of basic education
• Level 3: Upper secondary education
• Level 4: Post-secondary non-tertiary education
• Level 5: First stage of tertiary education
• Level 6: Second stage of tertiary education 

Income quintile: is computed by grouping the answer categories of the IN04 variable of the EHIS questionnaire. Perhaps you can provide the approximate range instead of your household's total net income per month. Would you (please look at this card and) tell me which group represents your household's total net monthly income from all these sources after deductions for income tax, National Insurance etc. Is it ...

• below 1st decile
• between 1st decile and 2nd decile
• between 2nd decile and 3rd decile
• etc
• between 8th decile and 9th decile
• above 9th decile

A code unknown has been created grouping missing and refusal answers.

Individuals aged 15 years old and over living in private households. For some countries there exists an upper age limit as well.

For specifications by country see EHIS wave I surveys - overview in the annex EHIS wave I surveys - overview.

All persons aged 15 years old and over living in private households and, for some countries, also in institutions like homes for elderly.

For specifications by country see EHIS wave I surveys - overview in the annex.

Only 19 countries participated in the first wave of EHIS, namely AT, BE, BG, CZ, CY, DE, EE, EL, ES, FR, HU, LV, MT, PL, RO, SI, SK, TR (CH).

Published data refer to EHIS wave I data collection round, which covers for most countries the period 2008-2009. However, the EHIS data are collected in different years depending on the country, going from 2006 to 2009. A tabular overview is given in the annex EHIS wave I surveys - overview.

Not applicable

Most indicators are reported as rates (distribution percentages).

The EHIS wave I data collection round refers to years 2006-2009 depending on the country (see the point 3.8 Time coverage).

Countries submit data to Eurostat on the basis of a gentlemen's agreement established in the framework of Eurostat Working Group on "Public Health Statistics".

A new Regulation on Community statistics on public health and health and safety at work (EC) No 1338/2008 was signed by the European Parliament and the Council on 16 December 2008. This Regulation is the framework of the EHIS data collection. Within the context of this framework Regulation, a specific Implementing Measure for wave II will be developed - within the ESS.

None

Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4)) of 11 March 2009 (OJ L 87, p. 164), stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.

See above on the flag setting

Access to EHIS microdata files by researchers should be available in the future. In order to ensure disclosure control and the confidentiality of microdata, a set of anonymisation rules (including  hiding variables or grouping values) is being discussed and tested.

There is no release calendar.

None

In line with the Community legal framework and the European Statistics Code of Practice Eurostat disseminates European statistics on Eurostat's website (see item 10 - 'Dissemination format') respecting professional independence and in an objective, professional and transparent manner in which all users are treated equitably. The detailed arrangements are governed by the Eurostat protocol on impartial access to Eurostat data for users.

First countries, which indicators were published were: AT, BG, CY, CZ, EE, FR, HU, LV, RO.

In March 2011, indicators were disseminated for: EL, MT, SI, SK.

By the end of 2011 were disseminated indicators also: BE, DE, ES, PL.

Finaly in April 2012 was disseminated TR. 

Next EHIS Wave II might start in 2014.

• Between 8% and 25% of adults are obese across Member States
• At the age of 65, both women and men are expected to live a further 9 years in a healthy condition

None

Please consult free data on-line.

Due to the confidential character of the EHIS microdata, direct access to the anonymised data will be only provided by means of research contracts. Access is in principle restricted to universities, research institutes, national statistical institutes, central banks inside the EU, as well as to the European Central Bank. Individuals cannot be granted direct access.

Contact point: estat-microdata access@ec.europa.eu

Internet address:http://ec.europa.eu/eurostat

Methodological guidelines as well as all relevant documents are available on CIRCABC: Eurostat - Public Health statistics: Health Interview survey - EHIS wave 1.

To be completed with reports available by July 2011, see point 21.3 "Annex".

Not available.

EHIS aims at achieving an input standardisation. The data collection methods for this survey have been prepared in detail in order to take into account the problems of comparability and of harmonisation between countries. A standard questionnaire (questions, answer categories, filters, etc.) was ready by 2006 as well as conceptual guidelines and rationales. Conceptual translation into all EU languages was requested. A standard translation protocol was used to translate the English questionnaire into national languages (see chapter 5 p27 onwards of the guidelines for the development and criteria for the adoption of Health Survey instruments).

This is complemented by Eurostat consistency and integrity checks on the microdata so that minimum output quality standard is reached. In addition, data are accompanied with quality reports stating the accuracy, coherence and comparability of the data. See also the forthcoming annex on the quality reports.

EHIS answers to DG SANCO and DG EMPL policy needs (incontrovertible data to launch new prevention plans, measure of key indicators). Data will also be used by researchers to make in-depth analyses on specific health issues. The EHIS aims at measuring on a harmonised basis and with a high degree of comparability among EU Member States the health status, life style (health determinants) and health care services use of the EU citizens. The topics included in the questionnaire both answer to policy driven needs and to scientific purposes. Within this framework, the EHIS concentrates on the main elements needed at EU level and does not intend to cover all detailed health aspects which can better be carried out via specific surveys or survey modules at national level, or at EU level when necessary.

The main users of the EHIS data are/will be:

• Institutional users like other Commission services, particularly DG SANCO and DG EMPL for their needs in relation to the OMC and ECHI indicators as well as national administrations (mainly those in charge of the monitoring of public health, or other international organisations);
• Statistical users in Eurostat or in Member States National Statistical Institutes to feed sectoral or transversal publications such as the Annual Progress Report on the Lisbon Strategy (structural indicators), the Eurostat yearbook and various pocketbooks, among other reports;
• Researchers having access to microdata;
• End users - including the media - interested in public health in the EU.

Not available.

EHIS covers only people living in private households (all persons aged 15 and over within the household are eligible for the operation), i.e. persons living in collective households and in institutions are generally excluded from the target population.

See also information on the non-delivery of some indicators in the paragraph 3.4 Statistical concepts and definitions.

Not available

Standard errors of key indicators are commonly used as a measure of the reliability of data collected through sample survey. In their national quality reports, Member states provided the standard error for 3 indicators based on the Minimum European Health Module (MEHM):

• Proportion of respondents in good or very good health
• Proportion of respondents with a longstanding illness
• Proportion of respondents severely limited in activities people usually do because of health problems for at least the past 6 months

A synthesis of the national quality reports will be publicly available by April 2011, see point 21.3 "Annex"

The term 'non-sampling error' is a generic one that encompasses any errors other than sampling errors. The non-sampling errors discussed in this section are: coverage errors, measurement and processing errors, and non-response errors.

Coverage errors

Coverage errors are caused by the imperfections of a sampling frame for the target population of the survey.

In EHIS two main groups can be defined in terms of the sampling frame used:

• Some countries have relied on individual information from population registers. In order to make the best coverage of the target population, registers have to be updated frequently. It means any modification in the population (both people moving in and people moving out) must be reported as quickly as possible.
• Other countries have used a sampling frame based on the Census. The databases also have to be updated to represent the units that have come into being after the Census and thus ensure the representativeness of the sample.

A systematic source of coverage problems is the time lag between the reference date for the selection of the sample and the fieldwork period, which should be made the shortest. 

Measurement and processing errors

Generally, measurement errors arise from the questionnaire, the interviewer, the interviewee and the data collection method used.

It is vital in a survey like EHIS, which collects a multitude of health components with different time period references, that the questionnaire is constructed so that the interviewee can provide all the correct information. In particular, experiences from pilot surveys were used in order to optimize the data collection process. The questionnaires were also tested (cognitive testing) in order to identify potential sources of problems. 

Due to the complexity and the sensitivity of the survey, the interviewees could not or did not want to give information about some modules (their alcohol consumption, their income etc).

Non-response errors

All surveys have to deal with non-response, i.e. information missing for some of the sample units. Unit non-response happens when no interview can be obtained, while item non-response does when only some of the items are missing. EHIS suffers from these two types of non-response:

• Unit non-response: when an individual refuses to cooperate or is away during the fieldwork period. Other reasons can explain unit non-response: the questionnaire is lost; the individual is unable to respond because of incapacity or illness...
• Item non-response: typically happens to questions the interviewee does not answer because he considers them personal or not easily understandable.

Non-response is a potential source of bias particularly if the non-responding units have specific survey patterns ('non-ignorable' non-response). For instance, persons with limitations (physical or sensorial) are less keen to give health information to an interviewer, thus some groups with particular features are under-represented in the sample and the estimates downwardly biased. 

A majority of the countries applied calibration methods (i.e. changes in the weighting factors) in order to correct non-response.

There is no fixed time for transmitting data to Eurostat.

Not applicable.

EHIS aims at achieving an input standardisation. The data collection methods for this survey have been prepared in detail in order to take into account the problems of comparability and of harmonisation between countries. A standard questionnaire (questions, answer categories, filters, etc.) was ready by 2006 as well as conceptual guidelines and rationales. Conceptual translation into all EU languages was requested. A standard translation protocol was used to translate the English questionnaire into national languages (see chapter 5 p27 onwards of the guidelines for the development and criteria for the adoption of Health Survey instruments.

None

Some EHIS wave I indicators can be compared with indicators computed with the EHIS 2004 round. See Public Health domain on Eurostat website under:

• Health care: indicators from surveys (SILC, HIS round 2004) à indicators from the national Health Interview Surveys (HIS round 2004: period 1999-2003)
• Health status: indicators from surveys (SILC, HIS, LFS) à indicators from the national Health Interview Surveys (HIS round 2004: period 1999-2003)

Any data considered as 'non consistent' are not published.

Not available

None

The published data are final, i.e. no revision is expected.

The data are collected via national surveys. EHIS modules may be either grouped in one separate national survey or be integrated into an existing national survey (i.e. national health interview survey, labour force survey, other household survey). In such a way Member states have had the maximum flexibility for implementation. However, across the EU the same data were collected according a common questionnaire. See also ESMS page.

Every 5 years

Data are collected via questionnaires and are obtained through face-to-face interviews, telephone interviews, self-administered questionnaires or by a combination of these means (depending on the country).

Testing data comparability with previous HIS round or with international database on health statistics was performed.

The strata sizes of the different samples were calculated as well as the missing percentages on the answers. Strata must be understood as the number of respondents belonging to the measured subgroup; for example the men in Bulgaria between 54 and 65 years old and having an ISCED level 6 for education. Flags for eliminating too small strata and for indicating unreliable strata were applied as follows:

• ":" no information available or the strata size is less than 20 sample observations or the item non-response exceeds 50%.
• published with a flag "u: unreliable" if the estimate is based on 20 to 49 sample observations, or if the item non-response exceeds 20 % and is lower than or equal to 50 %.
• no flag if the strata size is greater than 50 and the item non-response is below 20%.
• "c: confidental" is set for data of Germany where strata sizes are below 20 observations. (Germany delivered aggregated data and indicated those cells)

Due to different time periods and to incomplete coverage reasons, no EU aggregates are calculated.

For AT, BG, CZ, CY, GR, EE, FR, HU, LV, MT, RO, SI, SK and TR individual data were received and a weighted distribution with the weights specified for each person in the sample was applied (no weighted data file for MT). No other adjustments were performed on the national data.

See information in the paragraph 3.4 Statistical concepts and definitions.

Synthesis of the national quality reports

Comparisons of the EHIS questions

Comparability of EHIS results