Health Interview Surveys (HIS) and combinations of

Evaluation of Health Interview Surveys and Health Examination Surveys in the European Union ARPO AROMAA, PAlVIKKI KOPONEN, JEAN TAFFOREAU, CLAUDINE VERMEIRE, AND THE HIS/HES CORE GROUP * Background: The project 'Health surveys in the EU: HIS and HIS/HES evaluations and models' aims to assess the coverage of specific health and health-related areas in national and international surveys by reviewing and evaluating surveys, their methods and comparability, and by recommending appropriate survey designs and methods. Methods: As basis for the evaluation, the project developed a health survey database. At present, Health Interview Surveys (HIS) and Health Examination Surveys (HES) from 8 Western European countries as well as from Canada, Australia and the USA are included. Results: National HISs have been carried out regularly in almost all Western European countries. The HIS may consist of short health sections or modules within multi-purpose surveys or lengthy health interviews with several questionnaires. National HESs with a comprehensive focus have been conducted at regular or irregular intervals in five countries. The HES may comprise an interview and a few measurements or a comprehensive health examination. Sampling frames, fieldwork, quality control procedures and response rates vary greatly. Differences between measurement instruments used, in the wording of questions and in examination protocols reduce the comparability of many findings. Conclusion: The Internet based HIS/HES database allows for a quick reference and comparison of methods and instruments used in national health surveys. It illustrates the need for improving comparability. Collaboration and co-ordination is needed to promote comprehensive health monitoring supporting the development of national and European-level health policy. Keywords: health examination, health interview, health surveys, survey methodology Health Interview Surveys (HIS) and combinations of Health Interview and Health Examination Surveys (HIS/HES) are central components of a comprehensive health monitoring system. HISs deliver valuable information on health status, illnesses, lifestyles, functional capacity and use of health care services. By definition, interviews and questionnaires are the only way to obtain data on perceived health, symptoms and health related behaviour. In addition, clinical measurement is needed to obtain valid information on many chronic conditions, functional limitations and disabilities, and on several key health determinants. Chronic diseases are often under-reported or overreported in HISs. ^Non-symptomatic persons may suffer from some conditions (e.g. hypertension) requiring treatment, the reported symptoms (e.g. visual or auditory acuity) may not be specific enough or there is a considerable recall bias (e.g. infections and immunisations). The results of physical examination and subjective reporting of pain and disability can differ. 5 A well-known example of differences in self-reports compared to actual * A. Aromaa, P. Koponen, J. Tafforeau, C. Vermeire, HIS/HES Core Group (see appendix) National Public Health Institute, Helsinki Finland Scientific Institute of Public Health, Brussels, Belgium Correspondence: Paivikki Koponen, PhD, National Public Health Institute, Department of Health and Functional Capacity, Mannerheimintie, FIN-OOOO Helsinki, Finland, tel. +58 9 4744 8795, fax +58 9 4744 894, e-mail: paivikki.koponen@ktl.fi measurements is under-reporting weight and over- or under-reporting height. To obtain comprehensive and comparable information on Community Health Indicators 7 in a valid and reliable manner, standardized health survey methods should be used. Comparability of information is needed to show differences and similarities between countries, and to point out problems and achievements. Valid comparisons serve to set goals to improve health. HES is more expensive and logistically more demanding than HIS, as it requires a variety of qualified personnel and careful training. In practice, an integral part of every HES is a HIS, and sometimes the HES is carried out on a sub-sample due to the special demands in the design and procedures. ' 8 Many measurements and examinations of a typical HES are closer to clinical medicine and some have better validity than HIS methods, which is not to say that physical examinations are not prone to error.^ One major advantage of the HIS/HES combination is in measurement of time trends and differences between population groups, since interpretation of the findings is facilitated by the availability of different types of partly independent measures. Data for health monitoring are regularly obtained also from other sources. Statistical sources and registers, e.g. hospital admission and discharge registers and general practice registers, can provide an overview of morbidity and suggest hypotheses for further investigation. Downloaded from https://academic.oup.com/eurpub/article-abstract//suppl_/7/4949 on September 8

EUROPEAN JOURNAL OF PUBLIC HEALTH VOL. NO. SUPPLEMENT Register data is valuable for the evaluation of health care services, but not sufficient for population health monitoring purposes. Comprehensive and valid register data is only available for some diseases and there are differences in the coverage and availability of data. Population surveys overcome much of the selection bias of health status data based on services, provided that participation is high. The response rates in health surveys vary greatly. ' Many of those who most likely have several chronic conditions and functional limitations, e.g. institutionalized persons, may be excluded from the sample or health problems are accumulated among the non-participants. ' Recruitment remains one of the most challenging and underestimated phases in the research process. 'The efforts made to enhance comparability of survey results through the use of common standardized instruments should be complemented by more attention paid to other methodological issues, such as sampling, recruitment and potential bias from non-participation. AIMS Within the framework of the EC Health Monitoring Programme (995-), the project 'Health Surveys in the EU: HIS and HIS/HES evaluations and modes' aims to produce an inventory of surveys in EU Member States and EFTA/EEA countries. A database on HIS, combinations of HIS/HES, and other national population surveys with a significant health component has been developed, including also health surveys from Canada, Australia and USA. The project will assess the coverage of health and health-related topics relevant for health monitoring by national and international surveys. The project's other focus is on evaluation and improvement of health survey instruments, methods and the implementation of surveys. The project has been carried out in two phases: the first phase in 999 and the second, ongoing phase in -. A third phase is planned. The specific aims of the current second phase of the project are to review, evaluate and recommend appropriate survey designs and methods for use in HIS and HIS/HES surveys, and to maintain, update and develop the health survey database created during the first phase of the project. Information has been gathered on existing recommendations, instruments and standards from several recent projects under the auspices of the EU Health Monitoring Programme. 7 "" 9 and international organizations. The database is a consultation forum on recommended questions and examination protocols. It enables comparisons of health survey methods, questions and examinations in different countries. It aims to support the development of questions and examinations in future health surveys leading to better validity and comparability of data and indicators. Institutions and contact persons for each health survey have been identified to maintain an operational European network. The evaluation will include the assessment of usefulness, feasibility and comparability of methods, instruments and excisting recommendations. Where several instruments and recommendations exist, attention is drawn to their differences and where these have not been tested in fieldwork with good results, attention will be drawn to this. Evaluation will be done in detail for a few health topics (mental health and physical functioning) and methods (sampling, inclusion of institutionalized persons, recruitment and participation), chosen for the subprojects under the HIS/HES project. HIS and HES survey results in three countries will also be compared on cardiovascular risk factors, subjective health and quality of life (measured by SF-). This serves as an example of merging, pooling and joint analysis of data from national health surveys. METHODS During the first phase of the project a core group of experts in the field of health survey research was established and comprises partners from eight countries (appendix). A European network of experts in HIS and HES methodologies was initiated. Inventories of national HIS and HIS/HES in EU Member States and EFTA countries were carried out and updated during the first and the second phase of the project. " Information was collected by literature reviews, personal communication, a systematic postal survey covering all EU/EFTA Member States, and methodological questionnaires sent to contact persons of identified surveys. Only one institute did not return the methodological questionnaire. Information on survey design, mode of data collection, target population, sampling frame, sampling procedures, sample size, non-response, survey personnel and their training, and quality assurance and control was collected by these questionnaires. Information on HISs and HIS/HESs was entered into the health survey database together with the recommended instruments and protocols of WHO/EURO and the EHRM-project. 9 The database covers methodological information, all interview questions (both in the original language and in English) and examination protocols, as well as the institutions in charge of the co-ordination of the surveys and contact persons responsible for each survey. The inclusion criteria for HIS and HIS/HES have been the following. The surveys must be based on nationally representative population samples, they must be repeated at more or less regular intervals, and they must be comprehensive (not disease/topic specific or restricted to a narrow age group, e.g. only children or the elderly). The January version of the database includes 9 HIS (with a total of 89 questions) and HES (with a total of test/examination protocols) in EU and EFTA Countries, and in Australia, Canada and USA. Seven international (WHO, European Community) HISs are also included. The inventory of health surveys will be complemented with information from new EU member states/ candidate countries. This work will be finalised in 4. Downloaded from https://academic.oup.com/eurpub/article-abstract//suppl_/7/4949 on September 8

SUPPLEMENT Evaluation of HIS and HES in the EU The HIS/HES database was first developed in Microsoft Access. A CD-Rom with this version has been distributed to more than 5 institutions, national governmental services (statistical and health) and some international organizations playing a key role in health monitoring (EC, WHO, OECD). The newly developed Internet compatible version (the HIS/HES Website) will be finalised during spring (https://www.iph.fgov.be/hishes). The database has been migrated to an SQL Server database with Coldfusion as the interface for Internet. Analysis of the content of health survey questionnaires and examination protocols led to the design of a classification of health and health related areas and health topics. This classification was adapted to the list of European Community Health Indicators.' Health survey questions and examination protocols have been classified using these topic codes, which can thus be used as search criteria in the database. The list of HIS topics has been divided into seven areas: demographic and socio-economic factors, health status, personal factors, life style factors, living and working conditions, prevention, health protection and health promotion, and health and social services. The list of HES topics is partly based on the ICF- classification, and is divided into 7 areas: e.g. risk factors, cardiovascular function and diseases, diabetes mellitus and other metabolic diseases, kidney, urinary tract and thyroid function Table National HIS and HIS/HES surveys in EU/EFTA countries (number of surveys included in the HIS/HES database, and carried out 998-) Austria Belgium Denmark Finland France Germany Greece Ireland Iceland Italy Luxembourg Netherlands Norway Portugal Spain Sweden Switzerland UK Total HIS a l c T 8 T 5 C 4 C HIS/HES b a: General HIS, disability surveys, health education/lifestyle or living conditions surveys etc. with a specific health module/section. b: Combination of HIS and HES, Surveys including a health examination to all/some participants of HIS. c: Including health modules within a 'Microcensus', General Census or a General socio-economic survey to all citizens. and disease, infections and inflammations, and sensory function, physical function and physical fitness. RESULTS: CURRENT AND PLANNED NATIONAL HIS AND HIS/HES IN EUROPE The inventory showed that national HISs have been regularly executed in most EU and EFTA countries (table ). In Finland, France, UK and the Netherlands many health surveys have been repeated regularly. In most countries one or two national HISs have been recorded, namely in Belgium, Denmark, Germany, Iceland, Ireland, Italy, Norway, Portugal, Spain, Sweden and Switzerland. The model of these national interview surveys varies from health modules or sections within living condition surveys to specific disability surveys, lifestyle and health education surveys, surveys on use of health services or comprehensive/general health surveys. In the near future about national HISs per year will be executed in EU and EFTA countries. There were similarities but also important differences in survey methodology (table ). The survey design was most frequently cross-sectional without follow-up. Samples were drawn from population registers or address files of households or individuals, most commonly by multistage probability sampling, stratified by geographic areas. Nearly half of the surveys had no age limit. When applied, the lower limit varied from to years, and the upper limit from 4 to 84 years. Institutionalized persons were included in 5 surveys from seven countries. Sample sizes varied from fewer than individuals (Iceland) to 79 households (Spain) and non-response rates from 5% to 48% for HIS and from 5% to 75% for HES. The high non-response rates can be explained, e.g. by the large number of people not contacted in some countries, linked to the level of accuracy in sampling frames. Further explanations for the low response rate in some HESs are the fact that only those who were first Table Methodological information on survey design and sampling in national HIS and HIS/HES surveys in EU/EFTA countries (number of HIS and HIS/HES carried out 998-) Survey design Cross-sectional without follow-up Cross-sectional with follow-up Panel Combination of panel and cross-sectional Unknown/missing Sample type Households Individuals Unknown/missing Sampling procedure Multistage probability Simple probability Other Unknown/missing HIS Total=4 9 HIS/HES Total- Downloaded from https://academic.oup.com/eurpub/article-abstract//suppl_/7/4949 on September 8

EUROPEAN JOURNAL OF PUBLIC HEALTH VOL. NO. SUPPLEMENT interviewed were invited to the examinations in two surveys, the restrictions in appointment times, and the poor feedback to the participants about their personal test results in one survey. There were also differences in the place of the examinations (home or clinic), in the data collection methods (table ), in the professional background of survey personnel, and in their training before and during fieldwork. More than one mode of data collection was applied in most HISs, e.g. both face-to-face interviews and selfadministered questionnaires. A summary of analyses of the HIS questionnaires included in the database is shown in table 4- Questions on the health status of respondents were most common. Diseasespecific morbidity, perceived health, activities of daily living, and chronic conditions were investigated in most surveys. Lifestyle factors, in particular smoking, were commonly present. Other frequent questions relate to the use of health and social services, like hospitalization, contacts with the GP, and the use of medicines. Rarely covered topics include consumption of illicit drugs, participation in cancer screening programmes, reproductive health, dental health and violence. The comparability of the questions in different national HISs has been studied. The comparability of most topics appears quite limited across all countries. An evolution to better comparability and to broadening the range of topics covered in health surveys is evident in the most recent period (surveys carried out in -). Several topics are already comparable between a number of European countries, e.g. self-reported morbidity (especially for hypertension and diabetes) and smoking. National population based HESs with a comprehensive focus have been conducted at regular or irregular intervals in five countries (Finland, Germany, Ireland, the Netherlands and UK) (table I). In these countries several local, regional and/or focused surveys have been carried out in addition to the national HESs. All of the national HESs included a HIS component preceding or parallel to the HES. National HESs in the planning or pilot stage were identified in two additional countries (France, Italy) and some preliminary plans for a national survey were reported in one more country (Portugal). In addition, Table Methodological information on fieldwork procedures for national/international HIS in Europe ( surveys carried out 998-) Mode of data-collection Face-to-face interview Telephone interview Self-administered questionnaires Use of proxies allowed in interviews For children For adults not able to respond For adults not at home No proxies allowed Computer aided interview Number of surveys 4/ 4/ 8/ / 8/ / / / surveys focused on specific age-groups (e.g. the elderly in Sweden) or based on regional rather than national samples (e.g. Norway) were identified. Surveys focused on specific age-groups, diseases or risk factors, and geographically limited HESs have been carried out in almost all EU/EFTA countries. A well-known example are the WHO-MONICA related surveys. 5 In replies to our questionnaire the major reason given for not carrying out national HESs until now was the high expense or the difficulties in implementing fieldwork. However, all respondents from countries without national HESs stated that such national examination surveys are necessary. Most respondents also felt that there is a need to develop a core module for HES in Europe. The HIS/HESs were carried out according to different models ranging from an interview with a few measurements and/or blood samples to a comprehensive health examination taking several hours to complete per person. Differences in the measurement protocols and other differences in the fieldwork phase limit comparability of results of the surveys. In the examinations there was a clear emphasis on cardiovascular diseases and their risk factors (table 5), but other topics were also covered, most often respiratory function, diabetes, liver function, haematological system function, infections and allergy. Topics covered in only one European examination survey, the Health survey in Finland, were examinations of musculoskeletal function and disorders (including bone density), sensory function (vision and hearing tests), and Table 4 Examples of health topics frequently covered in national/international HIS (topics covered in interviews) in Europe ( surveys carried out during 998-) Topic Demographic and socioeconomic factors Education and/or employment status Health status Self assessed/perceived health Long-standing illness/chronic conditions/disabilities Limitations of activities of daily living, personal General mental health Personal factors Body height and weight Life style factors Smoking Alcohol use and abuse Physical activity Working conditions Housing conditions Prevention Contraception Vaccinations Health and social services Use of services, GP Medication Hospitalization Number of surveys 4/ 49/ 49/ / / / 44/ / / 7/ 4/ / / / 7/ 7/ Downloaded from https://academic.oup.com/eurpub/article-abstract//suppl_/7/4949 on September 8

SUPPLEMENT Evaluation of HIS and HES in the EU dental health (clinical dental examination). In other countries mental and dental health issues were often studied in separate surveys. During the examinations diagnostic interviews for mental health were part of the general HIS/HESs in two countries, while specific national mental health surveys were carried out in two other European countries. DISCUSSION AND CONCLUSIONS The HIS/HES database allows for a quick reference and comparison of methods and instruments used in national health surveys. National HISs are a widely used tool of health policy and planning. Comprehensive national HESs are still relatively rare but interest in them is increasing in many countries. Comprehensive local or regional HESs have been carried out occasionally in several countries. Several regional and local/community level HIS/HESs have been carried out in most countries and most countries have participated in major international multicentre HIS/HESs carried out since the 95s. Table 5 Health topics covered in national HIS/HES (topics covered in clinical examinations) in Europe ( surveys carried out during 998-) Topic Risk factors Height and weight Waist and hip circumference Blood lipids Cardiovascular function Blood pressure Electrocardiogram Respiratory function Spirometry Diabetes mellitus and other metabolic function Blood glucose Kidney and urinary tract function Blood and/or urine tests (protein, albumin etc.) Liver function Gamma-GT and similar tests Haematological system function Blood count Infections Blood samples, general or specific markers Allergy Blood samples, immunoglobulins Physical function e.g. joint function Mental disorders Diagnostic measurement Cognitive function Memory testing Nutritional status Blood samples for vitamins, and/or minerals and trace elements Number of surveys / / / / / 4/ / / / / It has not always been easy to decide which surveys should be included in the database. In some countries there is only a short health module within a census while in other countries several comprehensive health surveys have been carried out. The difference between a regional and a national survey, and between a topic-specific and a comprehensive health survey is not always clear either. When we conclude that a certain topic has not been covered at all or that it is rarely covered, this may be explained by not including topic-specific or regional surveys in the database, or that this topic has not been studied at all in the country. However, we believe that by the questionnaires and consultations with contact persons from all EU and EFTA countries we have identified all major surveys in these countries. We welcome any information on remaining unidentified surveys. Updating and further development of the database will be a major challenge for the anticipated third phase of the HIS/HES project. Inclusion of major regional, topic- or age-group-specific surveys will be considered in collaboration with other topic-specific projects. The third phase will continue evaluation and development of recommendations on survey methodology, instruments and protocols, as well as making proposals for improving quality and comparability of national surveys. The evaluation of methods and instruments should be extended by taking into account other relevant ongoing projects. Also a systematic search for all existing recommendations needs to be carried out. Much more information is needed concerning field tests and validation. The national HIS and HIS/HESs have been carried out using different designs and models. There are also important differences in sampling frames, participation and fieldwork procedures. All this may cause serious bias in the results and invalidate comparisons. Recommendations for protocols and quality control are available, but for some measurements there are no common standards. Even when recommendations and standards exist, their application varies. One prerequisite of better comparability seems to be jointly organized training and quality control. Differences in measurement instruments used, in versions of these instruments, in the choice of wording of questions and in examination protocols restrict comparisons of many health survey topics. The health survey database can be used to facilitate future improvement of validity and comparability. Most recently developed and recommended instruments require further testing and evaluation before they are matured enough to be released for wider application. Continuous international collaboration of experts and co-ordination of the surveys is needed to promote comprehensive health monitoring at the European level. The project is financially supported by the European Commission. Heliovaara M, Aromaa A, Klaukka T, Knekt P, Joukamaa M, Impivaara O. Reliability and validity of interview data on chronic Downloaded from https://academic.oup.com/eurpub/article-abstract//suppl_/7/4949 on September 8

EUROPEAN JOURNAL OF PUBLIC HEALTH VOL. NO. SUPPLEMENT diseases: the Mini-Finland Health Survey. J Clin Epidemiol 99;4:8-9. Fisher G, Pappas G, Limb M. Prospects, problems, and prerequisites for national health examination surveys in developing countries. Soc Sci Med 99;4():9-5. Kasper J. Asking about access: challenges for surveys in a changing healthcare environment. Health Serv Res 998;():75-9. 4 ILSA (The Italian Longitudinal Study on Aging Working Group). Prevalence of chronic diseases in older Italians: comparing self-reported and clinical diagnoses. Int J Epidemiol 997;(5):995-. 5 Michel A, Kohlman T, Raspe H. The association between clinical findings on physical examination and self-reported severity in black pain. Spine 997;():9-. Bolton-Smith C, Woodward M, Tunstall-Pedoe H, Morrison C. Accuracy of the estimated prevalence of obesity from self reported height and weight in an adult Scottish Population. J Epidemiol Community Health ;54:4-8. 7 Kramers PG, the ECHI working group. Design for a set of European Community Health Indicators. Final report by the ECHI project. The Netherlands, RIVM,. 8 Armitage P. National Health Survey Systems in the European community. IntJ Epidemiol 97;5(4):-. 9 Koran LM. The reliability of clinical methods, data and judgements. N Eng J Med 975;9:4-, 95-7. Rose GA, Blackburn H, Gillum RF, Prineas RJ. Cardiovascular survey methods. Second edition. Geneva, WHO, 98. Aromaa A. Health observation and health reaporting in Europe. Rev Epidemiol et Sante Publ 998,4:48-9. O'Neill TW, Marsden D, Matthis C, Raspe H, Silman AJ. Survey response rates: national and regional differences in a European multicentre study of vertebral osteoporosis. J Epidemiol Community Health 995;49():87-9. de Marco R, Verlato G, Zanolin E, Bugiani M, Drane JW. Nonresponse bias in EC Respiratory Health Survey in Italy. Eur RespirJ 994;7():9-45. 4 Carter WB, Elward K, Malmgren J, Martin ML, Larson E. Participation of older adults in health programs and research: a critical review of the literature. Gerontologist 99;(5):584-9. 5 Riedel-Heller SG, Busse A, Angermeyer MC. Are cognitively impaired individuals adequately represented in community surveys? Recruitment challenges and strategies to facilitate participation in community surveys of older adults: a review. Eur J Epidemiol ; :87-5. Brown BA, Long HL, Gould H, Weitz T, Milliken N. A conceptual model for recruitment of diverse women into research studies. J Womens Health Gend Med ;9():5-. 7 Lehtinen V, Korkeila J, Morgan A, et al. Establishemnt of a set of mental health indicators for European Union, Final Report. Helsinki, Stakes,. 8 Robine J-M, Jagger C, Romieu I, editors. Selection of a coherent set of health indicators for the European Union. Phase II: Final Report. Montpellier, France, Euro-REVES,. 9 Tolonen H, Kuulasmaa K, Laatikainen T, Wolf H, and the European Health Risk Monitoring Project. Recommendation for indicators, international collaboration, protocol and manual of operations for chronic disease risk factor surveys. Helsinki, National Public Health Institute,. (http://www.ktl.fi/ehrm/documents/product) WHO. Health Interview Surveys. Towards international harmonization of methods and instruments. WHO Regional Publications, European Series, no 58. Geneva, WHO, 99. Hupkens C, van den Berg J, van der Zee J. National health interview surveys in Europe: an overview. Health Policy 999;47:45-8. Hupkens C, Swinkels H. Health interview surveys in the European Union: overview of methods and contents. The Netherlands, CBS,. Koponen P, Aromaa A. Health examination surveys (HES): review of literature and inventory of surveys in the EU/EFTA Member States. Helsinki, National Public Health Institute,. 4 WHO. International Classification of Functioning, Disability and Health. Geneva, WHO,. 5 Bothig S, WHO MONICA Project. WHO MONICA Project: objectives and design. IntJ Epidemiol 989;8(,Suppl.):S9-7. Appendix Participants and organization Phase J of the project was coordinated by Jaap van den Berg at Statistics Netherlands. The subcontractor was The National Public Health Institute (KTL, Professor Arpo Aromaa) in Finland. Phase of the project was coordinated by Professor Arpo Aromaa at KTL in Finland. The subcontractor was the Institute of Public Health (IPH) in Brussels (Dr. Jean Tafforeau). The core group members and HIS centres are: Jean Tafforeau/Claudine Vermeire, Institute of Public Health (IPH), Belgium Jaap van den Berg/Christianne Hupkens, Statistics Netherlands (CBS) Jean-Marie Robine/Emmanuelle Cambois, Equipe INSERM Demographie et Sante, REVES Network on Health Expectancy, France V. Egidi/V. Buratta/L. Gargiulo/L. Quattrociocchi, Institute Nazionale di Statistica (ISTAT), Rome, Italy Carlos Matias Dias, Observatorio Nacional de Saude Dr. Ricardo Jorge, Lisboa, Portugal The core group members and HS/HES centres are: Arpo Aromaa/Paivikki Koponen, KTL, Finland Jaap Seidell/Lucie Viet, RIVM, Bilthoven, The Netherlands Barbel-Maria Kurth, Robert Koch Institut, Berlin, Germany Michael Marmot/Paola Primatesta, University College of London, United Kingdom Gino Farchi/Susanna Conti, Instituto Superiore di Sanita, Rome, Italy Downloaded from https://academic.oup.com/eurpub/article-abstract//suppl_/7/4949 on September 8