Public health monitoring of hypertension, diabetes and elevated cholesterol: comparison of different data sources

Public health monitoring of hypertension, diabetes and elevated cholesterol: comparison of... Abstract Background Three data sources are generally used in monitoring health on the population level. Health interview surveys (HISs) are based on participants’ self-report. Health examination surveys (HESs) yield more objective data, and also persons who are unaware of their elevated risks can be detected. Medical records (MRs) and other administrative registers also provide objective data, but their availability, coverage and quality vary between countries. We summarized studies comparing self-reported data with (i) measured data from HESs or (ii) MRs. We aimed to describe differences in feasibility and comparability of different data sources for monitoring (i) elevated blood pressure or hypertension (ii) elevated blood glucose or diabetes and (iii) elevated total cholesterol. Methods We conducted a literature search to identify studies, which validated self-reported measures against objective measures. We found 30 studies published since the year 2000 fulfilling our inclusion criteria (targeted to adults and comparing prevalence among the same persons). Results Hypertension and elevated total cholesterol were prone to be under-estimated in HISs. The under-estimate was more pronounced, when the HIS data were compared with HES data, and lower when compared with MRs. For diabetes, the HISs and the objective methods resulted in fairly similar prevalence rates. Conclusion The three data sources measure different manifestations of the risk factors and cannot be expected to yield similar prevalence rates. Using HIS data only may lead to under-estimation of elevated risk factor levels or disease prevalence. Whenever possible, information from the three data sources should be evaluated and combined. Introduction In 2015, high blood pressure, high fasting blood glucose and high total cholesterol were among the 10 most common risk factors explaining disability-adjusted life years according to the Global Burden of Disease Study 2015.1 Their relevance as risk factors had increased since 2005. One of the six objectives in the WHO global action plan for the prevention and control of noncommunicable diseases 2013–20 is to monitor the trends and determinants of noncommunicable diseases and to evaluate progress in their prevention and control.2 The action plan calls for undertaking periodic data collection on cardiovascular and metabolic risk factors. Three indicators (i) prevalence of elevated blood pressure (defined as systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg) and mean systolic blood pressure, (ii) prevalence of elevated blood glucose or diabetes (defined as fasting plasma glucose concentration ≥7.0 mmol/l or on medication for elevated blood glucose) and (iii) prevalence of elevated total cholesterol (defined as total cholesterol ≥5.0 mmol/l) and mean total cholesterol concentration are among the 25 indicators that are recommended to be monitored to follow up the achievements in the voluntary global targets. Diabetes and blood pressure are also among the European Community Health Indicators (ECHI). In 2007, it was evaluated that data for these indicators were widely available in European countries, mostly from National Health Interview Surveys (HIS). In some cases, they were available from registers and health examination surveys (HESs).3,4 Based on information gathered during the European Community Health Indicators and Monitoring project (ECHIM), it was concluded that registers and statistical information systems are important sources of health data, but both national HISs and HESs are additionally needed to provide comprehensive information on morbidity, functioning and health determinants in the whole population. Several countries have conducted repeated national HESs in order to monitor key health indicators of the population. Collecting data with HESs including questionnaires, physical measurements and often also collection of biological samples, yields objective and up-to-date data, and also persons who are unaware of their elevated risk factor levels can be detected. However, HESs are costly and time-consuming and require proper standardization. For blood pressure measurement, as an example, thorough training of the survey personnel is essential.5,6 Furthermore, recruiting of participants is often challenging in HESs.7 Because of lower costs HISs are used more frequently than HESs. Self-administered questionnaires, face-to-face interviews and telephone interviews can be used. The disadvantages of self-report (SR) in HISs include that it rests on the subjects’ memory, awareness and ability to report on a health condition.5,8 The two objective data sources, HES data and medical records (MRs), describe different dimensions of the outcome. HES data mostly rely on measurements conducted on one occasion and clinical diagnoses cannot be made based on such data. However, people with elevated risk but without diagnosis can be identified from HES data. The availability and coverage of MRs vary between countries according to the national health care system.9 The data depend on access and use of health services. Undiagnosed persons that have not sought medical care but might fulfil diagnostic criteria cannot be identified. For example, routine registers reveal diabetes or cardiovascular disease only in those who have used services and have been diagnosed.10,11 Furthermore, many administrative registers only cover information from hospital care. There may be inconsistences in coding of the conditions by the health care personnel as well as in measurements, laboratory analysis and clinical practice.12 There are numerous studies that have examined the validity of self-reported data against more objective measures.8,13 To gather up-to date information on different data sources and to discuss their strengths and limitations as well as to evaluate their usability for public health monitoring purposes, we summarized studies that compared self-reported data with two more objective data types: (i) measured data from HESs or (ii) MRs or other administrative register-based data. Three key risk factors or disease diagnoses were chosen to this evaluation: (i) elevated blood pressure or hypertension, (ii) elevated blood glucose or diabetes and (iii) elevated total cholesterol. We aimed to describe differences in feasibility and comparability of different data sources for monitoring major cardiovascular risk factors. Methods Literature search The literature search was conducted in PubMed in October 2016 and again in the beginning of January 2017 to identify studies on elevated blood pressure or hypertension, elevated blood glucose or diabetes and elevated total cholesterol which validated self-reported measures against measures based on (i) HES data or (ii) MRs or other administrative register-based data. Henceforth, the latter is referred to as MRs. The search terms are listed in Supplementary file 1. Studies published since the year 2000 were included. Older studies were excluded based on three reasons. Firstly, we were mainly interested in clarifying the current situation in relation to health monitoring. Secondly, the methods to identify and treat the selected risk factors have developed, the treatment alternatives have increased, and the indications for treatment have changed. Furthermore, the awareness of one’s own risk factor status may have improved among the population. Thirdly, we wanted to focus on studies that were not included in earlier reviews.8,13,14 The following inclusion criteria were applied: at least one of the following conditions was included as an outcome: elevated blood pressure/hypertension, elevated blood glucose/diabetes or elevated total cholesterol (i.e. either a diagnosis or a measure that indicated an elevated risk of the outcome); self-reported data were compared with data from (i) HES and/or (ii) MRs or other administrative registers; prevalence of the outcome was reported; risk factor status was assessed among the same study subjects with the two methods; the study was performed among adults (18+); and the study was published in English. To compare population level estimates, we excluded studies in which persons with known hypertension, diabetes or elevated cholesterol were excluded. We also excluded studies where finger prick samples were used to analyze blood glucose or total cholesterol. Furthermore, studies using other lipid markers than total cholesterol such as low-density lipoprotein or triglycerides were not included. Altogether 17 studies that compared self-reported data with HES data and 13 studies that compared self-reported data with MRs were found (table 1). These studies were presented in 28 publications, as two publications included two studies: (i) in the publication by El Fakiri et al.15 SR was compared separately with both HES and MRs and is thus included in both categories and (ii) the two age groups in Navin Cristina et al.16 are treated as two separate studies because their results were presented separately in the publication. The publications comprised 8 studies from Europe,9,15,17–21 12 from the North America,22–33 1 from the South America,34 4 from Asia35–38 and 5 from Australia/New Zealand16,39–41 (references 41–49 in the Supplementary file 2). The study population covered general population in 23 studies and clinical population in 7 studies. Table 1 Description of studies and study populations First author and year  n (men/women)  Age (years)  Country  Population  Study  Hypertension/elevated blood pressure  Diabetes/elevated blood glucose  Elevated total cholesterol  SR vs. HES                  (a) Definition based on measurement and medication  Europe                  Molenaar et al. 200717  4950 (2221/2729)  18+  The Netherlands  General  Utrecht Health Project  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Huerta et al. 200918  1556 (719/837)  20+  Spain  General  Diabetes, Nutrition and Obesity study (DINO)  x  x  –  Tolonen et al. 20149  4127 (1816/2311)  25–64  12 European countries  General  EHES Pilot Project  x  x  x  North America                  Natarajan et al. 200222  8236 (3558/4678)  21+  USA  General  NHANES III (1988–94)  –  –  x  Ahluwalia et al. 200923  733 (0/733)  40–64  USA  Generala  WISEWOMAN  x  –  x  Dey et al. 201526  101 (49/52)  18–80  Canada  Patients with recent history of high-risk TIA and/or minor stroke  Cohort study at a Stroke Prevention Clinic and inpatient ward  –  x  –  Asia                  Goldman et al. 200335  1004 (585/419)  54+  Taiwan  General  Social Environment and Biomarkers of Aging Study  x  x  –  Chun et al. 201636  7270 (3096/4174)  50+  South Korea  General  KNHANES IV  x  x  x  South America                  Lima-Costa et al. 200434  970 (422/548)  18+  Brazil  General  Bambuí study  x  –  –  Australia/New Zealand                  Taylor et al. 201039  1525 (749/776)  18+  Australia  General  North West Adelaide Health Study (NWAHS)  x  –  x  (b) Definition based on measurement only  North America                  Cowie et al. 201024  13094 (NR)  20+  USA  General  NHANES (2003–2006)  –  x  –  Dave et al. 201325  16598 (5747/10087)  18+  USA  Generalb  Community Initiative to Eliminate Stroke  x  –  –  Fisher-Hoch et al. 201527  2838 (NR)  18+  USA  Generalc  Cameron County Hispanic Cohort  –  x  –  Asia                  Bao et al. 201538  7913 (2841/5072)  20–74  China  General  A population-based cross-sectional study in the city of Harbin  –  x  –  Ning et al. 201637  17708 (8479/9229)  45+  China  General  China Health and Retirement Longitudinal Study  x  x  –  Australia/New Zealand                  Peterson et al. 201640  7269 (3275/3994)  18+  Australia  General  Australian Health Survey  x  x  x  SR vs. MRs                  Europe                  Tormo et al. 200019  248 (44/204)  29–69  Spain  General  Subsample of Spanish EPIC cohort  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Englert et al. 201020  7640 (4271/3369)  18+  Germany  Hypercholesterolemia patients  Orbital study  x  x  –  Frost et al. 201221  600 (600/0)  60–74  Denmark  General  Cross-sectional study  x  x  –  North America                  Simpson et al. 200428  1002 (0/1002)  65+  USA  Disabled older women  Women’s Health and Aging Study I  –  x  –  Okura et al. 200429  2037 (981/1056)  45+  USA  General  Rochester Epidemiology Project  x  x  –  St Sauver et al. 200530  26162 (10192/15970)  20+  USA  General  Patients at Mayo Clinic  x  –  x  Muggah et al. 201331  85549 (38743/46806)  20+  Canada  General  Canadian community health survey  x  x  –  Leong et al. 201332  3322 (1555/1767)  20+  Canada  General  Quebec Statistical Institute (QSI) survey  –  x  –  Koller et al. 201433  3821 (NR)  18+  USA  Generald  Alaska EARTH study  x  x  x  Australia/New Zealand                  Teh et al. 201341  878 (395/483)  80–90  New Zealand  General  The Life and Living to Advanced Age: a Cohort Study in New Zealand (LiLACS NZ)  x  –  –  Navin Cristina et al. 201616  1002 (0/1002)  56–61  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Navin Cristina et al. 201616  1926 (0/1926)  82–87  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  First author and year  n (men/women)  Age (years)  Country  Population  Study  Hypertension/elevated blood pressure  Diabetes/elevated blood glucose  Elevated total cholesterol  SR vs. HES                  (a) Definition based on measurement and medication  Europe                  Molenaar et al. 200717  4950 (2221/2729)  18+  The Netherlands  General  Utrecht Health Project  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Huerta et al. 200918  1556 (719/837)  20+  Spain  General  Diabetes, Nutrition and Obesity study (DINO)  x  x  –  Tolonen et al. 20149  4127 (1816/2311)  25–64  12 European countries  General  EHES Pilot Project  x  x  x  North America                  Natarajan et al. 200222  8236 (3558/4678)  21+  USA  General  NHANES III (1988–94)  –  –  x  Ahluwalia et al. 200923  733 (0/733)  40–64  USA  Generala  WISEWOMAN  x  –  x  Dey et al. 201526  101 (49/52)  18–80  Canada  Patients with recent history of high-risk TIA and/or minor stroke  Cohort study at a Stroke Prevention Clinic and inpatient ward  –  x  –  Asia                  Goldman et al. 200335  1004 (585/419)  54+  Taiwan  General  Social Environment and Biomarkers of Aging Study  x  x  –  Chun et al. 201636  7270 (3096/4174)  50+  South Korea  General  KNHANES IV  x  x  x  South America                  Lima-Costa et al. 200434  970 (422/548)  18+  Brazil  General  Bambuí study  x  –  –  Australia/New Zealand                  Taylor et al. 201039  1525 (749/776)  18+  Australia  General  North West Adelaide Health Study (NWAHS)  x  –  x  (b) Definition based on measurement only  North America                  Cowie et al. 201024  13094 (NR)  20+  USA  General  NHANES (2003–2006)  –  x  –  Dave et al. 201325  16598 (5747/10087)  18+  USA  Generalb  Community Initiative to Eliminate Stroke  x  –  –  Fisher-Hoch et al. 201527  2838 (NR)  18+  USA  Generalc  Cameron County Hispanic Cohort  –  x  –  Asia                  Bao et al. 201538  7913 (2841/5072)  20–74  China  General  A population-based cross-sectional study in the city of Harbin  –  x  –  Ning et al. 201637  17708 (8479/9229)  45+  China  General  China Health and Retirement Longitudinal Study  x  x  –  Australia/New Zealand                  Peterson et al. 201640  7269 (3275/3994)  18+  Australia  General  Australian Health Survey  x  x  x  SR vs. MRs                  Europe                  Tormo et al. 200019  248 (44/204)  29–69  Spain  General  Subsample of Spanish EPIC cohort  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Englert et al. 201020  7640 (4271/3369)  18+  Germany  Hypercholesterolemia patients  Orbital study  x  x  –  Frost et al. 201221  600 (600/0)  60–74  Denmark  General  Cross-sectional study  x  x  –  North America                  Simpson et al. 200428  1002 (0/1002)  65+  USA  Disabled older women  Women’s Health and Aging Study I  –  x  –  Okura et al. 200429  2037 (981/1056)  45+  USA  General  Rochester Epidemiology Project  x  x  –  St Sauver et al. 200530  26162 (10192/15970)  20+  USA  General  Patients at Mayo Clinic  x  –  x  Muggah et al. 201331  85549 (38743/46806)  20+  Canada  General  Canadian community health survey  x  x  –  Leong et al. 201332  3322 (1555/1767)  20+  Canada  General  Quebec Statistical Institute (QSI) survey  –  x  –  Koller et al. 201433  3821 (NR)  18+  USA  Generald  Alaska EARTH study  x  x  x  Australia/New Zealand                  Teh et al. 201341  878 (395/483)  80–90  New Zealand  General  The Life and Living to Advanced Age: a Cohort Study in New Zealand (LiLACS NZ)  x  –  –  Navin Cristina et al. 201616  1002 (0/1002)  56–61  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Navin Cristina et al. 201616  1926 (0/1926)  82–87  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Note: NR, not reported; RCT, randomized clinical trial; CVD, cardiovascular disease. a Low-income women. b Persons of color, low-income, rural residency and persons for whom English was a second language. c Mexican American people. d American Indian and Alaska Native people. Table 1 Description of studies and study populations First author and year  n (men/women)  Age (years)  Country  Population  Study  Hypertension/elevated blood pressure  Diabetes/elevated blood glucose  Elevated total cholesterol  SR vs. HES                  (a) Definition based on measurement and medication  Europe                  Molenaar et al. 200717  4950 (2221/2729)  18+  The Netherlands  General  Utrecht Health Project  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Huerta et al. 200918  1556 (719/837)  20+  Spain  General  Diabetes, Nutrition and Obesity study (DINO)  x  x  –  Tolonen et al. 20149  4127 (1816/2311)  25–64  12 European countries  General  EHES Pilot Project  x  x  x  North America                  Natarajan et al. 200222  8236 (3558/4678)  21+  USA  General  NHANES III (1988–94)  –  –  x  Ahluwalia et al. 200923  733 (0/733)  40–64  USA  Generala  WISEWOMAN  x  –  x  Dey et al. 201526  101 (49/52)  18–80  Canada  Patients with recent history of high-risk TIA and/or minor stroke  Cohort study at a Stroke Prevention Clinic and inpatient ward  –  x  –  Asia                  Goldman et al. 200335  1004 (585/419)  54+  Taiwan  General  Social Environment and Biomarkers of Aging Study  x  x  –  Chun et al. 201636  7270 (3096/4174)  50+  South Korea  General  KNHANES IV  x  x  x  South America                  Lima-Costa et al. 200434  970 (422/548)  18+  Brazil  General  Bambuí study  x  –  –  Australia/New Zealand                  Taylor et al. 201039  1525 (749/776)  18+  Australia  General  North West Adelaide Health Study (NWAHS)  x  –  x  (b) Definition based on measurement only  North America                  Cowie et al. 201024  13094 (NR)  20+  USA  General  NHANES (2003–2006)  –  x  –  Dave et al. 201325  16598 (5747/10087)  18+  USA  Generalb  Community Initiative to Eliminate Stroke  x  –  –  Fisher-Hoch et al. 201527  2838 (NR)  18+  USA  Generalc  Cameron County Hispanic Cohort  –  x  –  Asia                  Bao et al. 201538  7913 (2841/5072)  20–74  China  General  A population-based cross-sectional study in the city of Harbin  –  x  –  Ning et al. 201637  17708 (8479/9229)  45+  China  General  China Health and Retirement Longitudinal Study  x  x  –  Australia/New Zealand                  Peterson et al. 201640  7269 (3275/3994)  18+  Australia  General  Australian Health Survey  x  x  x  SR vs. MRs                  Europe                  Tormo et al. 200019  248 (44/204)  29–69  Spain  General  Subsample of Spanish EPIC cohort  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Englert et al. 201020  7640 (4271/3369)  18+  Germany  Hypercholesterolemia patients  Orbital study  x  x  –  Frost et al. 201221  600 (600/0)  60–74  Denmark  General  Cross-sectional study  x  x  –  North America                  Simpson et al. 200428  1002 (0/1002)  65+  USA  Disabled older women  Women’s Health and Aging Study I  –  x  –  Okura et al. 200429  2037 (981/1056)  45+  USA  General  Rochester Epidemiology Project  x  x  –  St Sauver et al. 200530  26162 (10192/15970)  20+  USA  General  Patients at Mayo Clinic  x  –  x  Muggah et al. 201331  85549 (38743/46806)  20+  Canada  General  Canadian community health survey  x  x  –  Leong et al. 201332  3322 (1555/1767)  20+  Canada  General  Quebec Statistical Institute (QSI) survey  –  x  –  Koller et al. 201433  3821 (NR)  18+  USA  Generald  Alaska EARTH study  x  x  x  Australia/New Zealand                  Teh et al. 201341  878 (395/483)  80–90  New Zealand  General  The Life and Living to Advanced Age: a Cohort Study in New Zealand (LiLACS NZ)  x  –  –  Navin Cristina et al. 201616  1002 (0/1002)  56–61  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Navin Cristina et al. 201616  1926 (0/1926)  82–87  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  First author and year  n (men/women)  Age (years)  Country  Population  Study  Hypertension/elevated blood pressure  Diabetes/elevated blood glucose  Elevated total cholesterol  SR vs. HES                  (a) Definition based on measurement and medication  Europe                  Molenaar et al. 200717  4950 (2221/2729)  18+  The Netherlands  General  Utrecht Health Project  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Huerta et al. 200918  1556 (719/837)  20+  Spain  General  Diabetes, Nutrition and Obesity study (DINO)  x  x  –  Tolonen et al. 20149  4127 (1816/2311)  25–64  12 European countries  General  EHES Pilot Project  x  x  x  North America                  Natarajan et al. 200222  8236 (3558/4678)  21+  USA  General  NHANES III (1988–94)  –  –  x  Ahluwalia et al. 200923  733 (0/733)  40–64  USA  Generala  WISEWOMAN  x  –  x  Dey et al. 201526  101 (49/52)  18–80  Canada  Patients with recent history of high-risk TIA and/or minor stroke  Cohort study at a Stroke Prevention Clinic and inpatient ward  –  x  –  Asia                  Goldman et al. 200335  1004 (585/419)  54+  Taiwan  General  Social Environment and Biomarkers of Aging Study  x  x  –  Chun et al. 201636  7270 (3096/4174)  50+  South Korea  General  KNHANES IV  x  x  x  South America                  Lima-Costa et al. 200434  970 (422/548)  18+  Brazil  General  Bambuí study  x  –  –  Australia/New Zealand                  Taylor et al. 201039  1525 (749/776)  18+  Australia  General  North West Adelaide Health Study (NWAHS)  x  –  x  (b) Definition based on measurement only  North America                  Cowie et al. 201024  13094 (NR)  20+  USA  General  NHANES (2003–2006)  –  x  –  Dave et al. 201325  16598 (5747/10087)  18+  USA  Generalb  Community Initiative to Eliminate Stroke  x  –  –  Fisher-Hoch et al. 201527  2838 (NR)  18+  USA  Generalc  Cameron County Hispanic Cohort  –  x  –  Asia                  Bao et al. 201538  7913 (2841/5072)  20–74  China  General  A population-based cross-sectional study in the city of Harbin  –  x  –  Ning et al. 201637  17708 (8479/9229)  45+  China  General  China Health and Retirement Longitudinal Study  x  x  –  Australia/New Zealand                  Peterson et al. 201640  7269 (3275/3994)  18+  Australia  General  Australian Health Survey  x  x  x  SR vs. MRs                  Europe                  Tormo et al. 200019  248 (44/204)  29–69  Spain  General  Subsample of Spanish EPIC cohort  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Englert et al. 201020  7640 (4271/3369)  18+  Germany  Hypercholesterolemia patients  Orbital study  x  x  –  Frost et al. 201221  600 (600/0)  60–74  Denmark  General  Cross-sectional study  x  x  –  North America                  Simpson et al. 200428  1002 (0/1002)  65+  USA  Disabled older women  Women’s Health and Aging Study I  –  x  –  Okura et al. 200429  2037 (981/1056)  45+  USA  General  Rochester Epidemiology Project  x  x  –  St Sauver et al. 200530  26162 (10192/15970)  20+  USA  General  Patients at Mayo Clinic  x  –  x  Muggah et al. 201331  85549 (38743/46806)  20+  Canada  General  Canadian community health survey  x  x  –  Leong et al. 201332  3322 (1555/1767)  20+  Canada  General  Quebec Statistical Institute (QSI) survey  –  x  –  Koller et al. 201433  3821 (NR)  18+  USA  Generald  Alaska EARTH study  x  x  x  Australia/New Zealand                  Teh et al. 201341  878 (395/483)  80–90  New Zealand  General  The Life and Living to Advanced Age: a Cohort Study in New Zealand (LiLACS NZ)  x  –  –  Navin Cristina et al. 201616  1002 (0/1002)  56–61  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Navin Cristina et al. 201616  1926 (0/1926)  82–87  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Note: NR, not reported; RCT, randomized clinical trial; CVD, cardiovascular disease. a Low-income women. b Persons of color, low-income, rural residency and persons for whom English was a second language. c Mexican American people. d American Indian and Alaska Native people. Statistical methods Various statistical methods such as prevalence rates, κ coefficient, sensitivity and specificity and positive and negative predictive values were applied in describing the agreement between two compared methods. In this evaluation, we focus on prevalence rates. κ coefficients are also presented if available. Also 95% confidence intervals for prevalence rates and κ coefficients are reported whenever available. None of the original publications presented standard errors or standard deviations for prevalence rates. Results Comparability of data collection methods SR measures The self-reported data were collected by interviews or self-administered questionnaires. The format of questions varied between the studies (Supplementary file 1). Most studies reported the actual question whereas in some studies the questions were described less accurately. Typically, the self-reported data relied on questions like: Have you ever been told (by a doctor or other health professional) that you have high blood pressure/diabetes/high cholesterol? (tables 2–4, Supplementary file 1). In most studies, either current conditions or conditions that had ‘ever’ been observed were asked. In some studies, only diagnosed conditions or conditions that a health professional had ‘told’ about were enquired while in other studies diagnosis or health professionals were not specified. Table 2 The format of question used in SR, the definition of hypertension/elevated blood pressure used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and κ coefficients for agreement   SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of hypertension  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                Molenaar et al. 2007 (n = 4950)17  General practitioner or specialist  12 months  Systolic blood pressure ≥140 mmHg (<60y),  10.7 (9.8–11.6)  22.9 (21.7–24.1)  -12.2  NR  systolic blood pressure ≥160 mmHg (≥60y) or diastolic blood pressure≥90  or medication  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Systolic blood pressure ≥160 mmHg  52  59  -7  0.51 (0.43–0.59)  or diastolic blood pressure ≥95 mmHg  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Systolic blood pressure ≥140 mmHg  19.5 (17.6–21.6)  35.4 (33.0–37.8)  -15.9  0.51 (0.47–0.56)  or diastolic blood pressure ≥90 mmHg  or medication  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Systolic blood pressure ≥140 mmHg          or diastolic blood pressure ≥90 mmHg  or medication        Men  22.6  32.7  -10.1  NR        Women  25.4  21.9  +3.5  NR  North America                Ahluwalia et al. 2009 (n = 733)23  Doctor, nurse or other health professional  Ever  Systolic blood pressure ≥140 mmHg  49.6  56  -6.4  0.62 (0.56–0.67)  or diastolic blood pressure ≥90 mmHg  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Currently  Systolic blood pressure ≥140 mmHg  30.3 (27.5-33.2)  57.3 (54.3-60.4)  -27  0.41 (0.36–0.47)  or diastolic blood pressure ≥90 mmHg  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Systolic blood pressure ≥140 mmHgor diastolic blood pressure ≥90 mmHgor medication  36.4  48.8  -12.4  0.72 (NR)  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  Systolic blood pressure ≥140 mmHg  24  38.5  -14.5  0.57 (0.55–0.58)  or diastolic blood pressure ≥90 mmHg  or medication  South America                Lima-Costa et al. 2004 (n = 970)34  Doctor or other health professional  Ever  Systolic blood pressure ≥140 mmHg  27.2 (24.4-30.1)  23.3 (20.7-26.1)  +3.9  NR  or diastolic blood pressure ≥90 mmHg  or medication  Australia/New Zealand                Taylor et al. 2010 (n = 1525)39  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  15.8  30.6  -14.8  0.55 (NR)  or diastolic blood pressure ≥90 mmHg  or medication  (b) Definition based on measurement only  North America                Dave et al. 2013 (n = 16 598)25  Physician, doctor or nursea  Evera  Systolic blood pressure ≥140 mmHg  16.15  24.81  -8.66  0.25 (NR)  or diastolic blood pressure ≥90 mmHgb  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  17.4 (16.5-18.3)  23.9 (22.9-24.9)  -6.5  0.21 (0.18–0.23)  or diastolic blood pressure ≥90 mmHg  SR vs.MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                Tormo et al. 2000 (n = 248)19  Physician  Ever  Diagnosis of high blood pressure  27.4  26.6  +0.8  0.65 (0.53–0.76)  or medication  or attending a hypertension control programme run only for hypertensive people  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  ICPC (International Classification of Primary Care) code K86/K87  52  44  +8  0.63 (0.55–0.70)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  56  62  -6  0.69 (0.66–0.71)  Frost et al. 2012 (n = 600)21  Not defined  NR  Diagnosis of hypertension          or medication        Men  22.2  36.7  -14.5  NR  North America                Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of hypertension  35.8  37.7  -1.9  0.75 (0.72–0.78)  and medication  or the words ‘borderline’ or ‘labile’ used in reference to blood pressure with documentation of two blood pressure measurements (consecutively but may be subsequent visits) systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg within a 12-month period  St Sauver et al. 2005 (n = 26 162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with hypertension  23.5  26.8  -3.3  NR  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  One hospital admission with a hypertension diagnosis code or one OHIP record with a hypertension diagnosis code, followed within 2 years by another OHIP record or a hospital admission with a hypertension diagnosis code.  20.8  27.6  -6.8  0.66 (0.65–0.66)  Codes used: ICD-9: 401x, 402x, 403x, 404x, 405x (any type), ICD-10-CA: I10, I11, I12, I13, I15 (any type), OHIP diagnosis code: 401, 402, 403, 404, 405 (any type)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for hypertension  25.1  25.8  -0.7  0.62 (NR)  Australia/New Zealand                Teh et al. 2013 (n = 878)41  Doctor  Ever  Primary health care record: Diagnoses were either ascertained from READ codes, a standardized primary care coding system, from hospital discharge letter, or from reading through the medical records.  54.8  68.4  -13.6  0.44 (0.38–0.50)  Administrative hospitalization discharge diagnosis records: Diagnosis codes for hypertension ICD-10: I10, ICD-9: 401.0, 401.1, 401.9 were identified.  Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  32.5 (29.7-35.3)  12.8 (10.8-14.8)  +19.7  0.35 (0.29–0.41)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  57.8 (55.0-60.7)  38.2 (35.3-41.0)  +19.6  0.21 (0.15–0.26)    SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of hypertension  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                Molenaar et al. 2007 (n = 4950)17  General practitioner or specialist  12 months  Systolic blood pressure ≥140 mmHg (<60y),  10.7 (9.8–11.6)  22.9 (21.7–24.1)  -12.2  NR  systolic blood pressure ≥160 mmHg (≥60y) or diastolic blood pressure≥90  or medication  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Systolic blood pressure ≥160 mmHg  52  59  -7  0.51 (0.43–0.59)  or diastolic blood pressure ≥95 mmHg  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Systolic blood pressure ≥140 mmHg  19.5 (17.6–21.6)  35.4 (33.0–37.8)  -15.9  0.51 (0.47–0.56)  or diastolic blood pressure ≥90 mmHg  or medication  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Systolic blood pressure ≥140 mmHg          or diastolic blood pressure ≥90 mmHg  or medication        Men  22.6  32.7  -10.1  NR        Women  25.4  21.9  +3.5  NR  North America                Ahluwalia et al. 2009 (n = 733)23  Doctor, nurse or other health professional  Ever  Systolic blood pressure ≥140 mmHg  49.6  56  -6.4  0.62 (0.56–0.67)  or diastolic blood pressure ≥90 mmHg  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Currently  Systolic blood pressure ≥140 mmHg  30.3 (27.5-33.2)  57.3 (54.3-60.4)  -27  0.41 (0.36–0.47)  or diastolic blood pressure ≥90 mmHg  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Systolic blood pressure ≥140 mmHgor diastolic blood pressure ≥90 mmHgor medication  36.4  48.8  -12.4  0.72 (NR)  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  Systolic blood pressure ≥140 mmHg  24  38.5  -14.5  0.57 (0.55–0.58)  or diastolic blood pressure ≥90 mmHg  or medication  South America                Lima-Costa et al. 2004 (n = 970)34  Doctor or other health professional  Ever  Systolic blood pressure ≥140 mmHg  27.2 (24.4-30.1)  23.3 (20.7-26.1)  +3.9  NR  or diastolic blood pressure ≥90 mmHg  or medication  Australia/New Zealand                Taylor et al. 2010 (n = 1525)39  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  15.8  30.6  -14.8  0.55 (NR)  or diastolic blood pressure ≥90 mmHg  or medication  (b) Definition based on measurement only  North America                Dave et al. 2013 (n = 16 598)25  Physician, doctor or nursea  Evera  Systolic blood pressure ≥140 mmHg  16.15  24.81  -8.66  0.25 (NR)  or diastolic blood pressure ≥90 mmHgb  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  17.4 (16.5-18.3)  23.9 (22.9-24.9)  -6.5  0.21 (0.18–0.23)  or diastolic blood pressure ≥90 mmHg  SR vs.MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                Tormo et al. 2000 (n = 248)19  Physician  Ever  Diagnosis of high blood pressure  27.4  26.6  +0.8  0.65 (0.53–0.76)  or medication  or attending a hypertension control programme run only for hypertensive people  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  ICPC (International Classification of Primary Care) code K86/K87  52  44  +8  0.63 (0.55–0.70)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  56  62  -6  0.69 (0.66–0.71)  Frost et al. 2012 (n = 600)21  Not defined  NR  Diagnosis of hypertension          or medication        Men  22.2  36.7  -14.5  NR  North America                Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of hypertension  35.8  37.7  -1.9  0.75 (0.72–0.78)  and medication  or the words ‘borderline’ or ‘labile’ used in reference to blood pressure with documentation of two blood pressure measurements (consecutively but may be subsequent visits) systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg within a 12-month period  St Sauver et al. 2005 (n = 26 162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with hypertension  23.5  26.8  -3.3  NR  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  One hospital admission with a hypertension diagnosis code or one OHIP record with a hypertension diagnosis code, followed within 2 years by another OHIP record or a hospital admission with a hypertension diagnosis code.  20.8  27.6  -6.8  0.66 (0.65–0.66)  Codes used: ICD-9: 401x, 402x, 403x, 404x, 405x (any type), ICD-10-CA: I10, I11, I12, I13, I15 (any type), OHIP diagnosis code: 401, 402, 403, 404, 405 (any type)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for hypertension  25.1  25.8  -0.7  0.62 (NR)  Australia/New Zealand                Teh et al. 2013 (n = 878)41  Doctor  Ever  Primary health care record: Diagnoses were either ascertained from READ codes, a standardized primary care coding system, from hospital discharge letter, or from reading through the medical records.  54.8  68.4  -13.6  0.44 (0.38–0.50)  Administrative hospitalization discharge diagnosis records: Diagnosis codes for hypertension ICD-10: I10, ICD-9: 401.0, 401.1, 401.9 were identified.  Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  32.5 (29.7-35.3)  12.8 (10.8-14.8)  +19.7  0.35 (0.29–0.41)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  57.8 (55.0-60.7)  38.2 (35.3-41.0)  +19.6  0.21 (0.15–0.26)  Note: NR, not reported. a The question format was as follows: ‘Do you suffer from high blood pressure and/or has a physician/doctor/nurse diagnosed you as a hypertensive?’ b Persons who were taking blood pressure lowering medications were excluded. c Includes survey responses from the last survey (survey 5) only (‘case 1’ in the publication). Table 2 The format of question used in SR, the definition of hypertension/elevated blood pressure used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and κ coefficients for agreement   SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of hypertension  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                Molenaar et al. 2007 (n = 4950)17  General practitioner or specialist  12 months  Systolic blood pressure ≥140 mmHg (<60y),  10.7 (9.8–11.6)  22.9 (21.7–24.1)  -12.2  NR  systolic blood pressure ≥160 mmHg (≥60y) or diastolic blood pressure≥90  or medication  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Systolic blood pressure ≥160 mmHg  52  59  -7  0.51 (0.43–0.59)  or diastolic blood pressure ≥95 mmHg  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Systolic blood pressure ≥140 mmHg  19.5 (17.6–21.6)  35.4 (33.0–37.8)  -15.9  0.51 (0.47–0.56)  or diastolic blood pressure ≥90 mmHg  or medication  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Systolic blood pressure ≥140 mmHg          or diastolic blood pressure ≥90 mmHg  or medication        Men  22.6  32.7  -10.1  NR        Women  25.4  21.9  +3.5  NR  North America                Ahluwalia et al. 2009 (n = 733)23  Doctor, nurse or other health professional  Ever  Systolic blood pressure ≥140 mmHg  49.6  56  -6.4  0.62 (0.56–0.67)  or diastolic blood pressure ≥90 mmHg  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Currently  Systolic blood pressure ≥140 mmHg  30.3 (27.5-33.2)  57.3 (54.3-60.4)  -27  0.41 (0.36–0.47)  or diastolic blood pressure ≥90 mmHg  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Systolic blood pressure ≥140 mmHgor diastolic blood pressure ≥90 mmHgor medication  36.4  48.8  -12.4  0.72 (NR)  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  Systolic blood pressure ≥140 mmHg  24  38.5  -14.5  0.57 (0.55–0.58)  or diastolic blood pressure ≥90 mmHg  or medication  South America                Lima-Costa et al. 2004 (n = 970)34  Doctor or other health professional  Ever  Systolic blood pressure ≥140 mmHg  27.2 (24.4-30.1)  23.3 (20.7-26.1)  +3.9  NR  or diastolic blood pressure ≥90 mmHg  or medication  Australia/New Zealand                Taylor et al. 2010 (n = 1525)39  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  15.8  30.6  -14.8  0.55 (NR)  or diastolic blood pressure ≥90 mmHg  or medication  (b) Definition based on measurement only  North America                Dave et al. 2013 (n = 16 598)25  Physician, doctor or nursea  Evera  Systolic blood pressure ≥140 mmHg  16.15  24.81  -8.66  0.25 (NR)  or diastolic blood pressure ≥90 mmHgb  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  17.4 (16.5-18.3)  23.9 (22.9-24.9)  -6.5  0.21 (0.18–0.23)  or diastolic blood pressure ≥90 mmHg  SR vs.MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                Tormo et al. 2000 (n = 248)19  Physician  Ever  Diagnosis of high blood pressure  27.4  26.6  +0.8  0.65 (0.53–0.76)  or medication  or attending a hypertension control programme run only for hypertensive people  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  ICPC (International Classification of Primary Care) code K86/K87  52  44  +8  0.63 (0.55–0.70)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  56  62  -6  0.69 (0.66–0.71)  Frost et al. 2012 (n = 600)21  Not defined  NR  Diagnosis of hypertension          or medication        Men  22.2  36.7  -14.5  NR  North America                Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of hypertension  35.8  37.7  -1.9  0.75 (0.72–0.78)  and medication  or the words ‘borderline’ or ‘labile’ used in reference to blood pressure with documentation of two blood pressure measurements (consecutively but may be subsequent visits) systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg within a 12-month period  St Sauver et al. 2005 (n = 26 162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with hypertension  23.5  26.8  -3.3  NR  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  One hospital admission with a hypertension diagnosis code or one OHIP record with a hypertension diagnosis code, followed within 2 years by another OHIP record or a hospital admission with a hypertension diagnosis code.  20.8  27.6  -6.8  0.66 (0.65–0.66)  Codes used: ICD-9: 401x, 402x, 403x, 404x, 405x (any type), ICD-10-CA: I10, I11, I12, I13, I15 (any type), OHIP diagnosis code: 401, 402, 403, 404, 405 (any type)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for hypertension  25.1  25.8  -0.7  0.62 (NR)  Australia/New Zealand                Teh et al. 2013 (n = 878)41  Doctor  Ever  Primary health care record: Diagnoses were either ascertained from READ codes, a standardized primary care coding system, from hospital discharge letter, or from reading through the medical records.  54.8  68.4  -13.6  0.44 (0.38–0.50)  Administrative hospitalization discharge diagnosis records: Diagnosis codes for hypertension ICD-10: I10, ICD-9: 401.0, 401.1, 401.9 were identified.  Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  32.5 (29.7-35.3)  12.8 (10.8-14.8)  +19.7  0.35 (0.29–0.41)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  57.8 (55.0-60.7)  38.2 (35.3-41.0)  +19.6  0.21 (0.15–0.26)    SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of hypertension  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                Molenaar et al. 2007 (n = 4950)17  General practitioner or specialist  12 months  Systolic blood pressure ≥140 mmHg (<60y),  10.7 (9.8–11.6)  22.9 (21.7–24.1)  -12.2  NR  systolic blood pressure ≥160 mmHg (≥60y) or diastolic blood pressure≥90  or medication  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Systolic blood pressure ≥160 mmHg  52  59  -7  0.51 (0.43–0.59)  or diastolic blood pressure ≥95 mmHg  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Systolic blood pressure ≥140 mmHg  19.5 (17.6–21.6)  35.4 (33.0–37.8)  -15.9  0.51 (0.47–0.56)  or diastolic blood pressure ≥90 mmHg  or medication  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Systolic blood pressure ≥140 mmHg          or diastolic blood pressure ≥90 mmHg  or medication        Men  22.6  32.7  -10.1  NR        Women  25.4  21.9  +3.5  NR  North America                Ahluwalia et al. 2009 (n = 733)23  Doctor, nurse or other health professional  Ever  Systolic blood pressure ≥140 mmHg  49.6  56  -6.4  0.62 (0.56–0.67)  or diastolic blood pressure ≥90 mmHg  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Currently  Systolic blood pressure ≥140 mmHg  30.3 (27.5-33.2)  57.3 (54.3-60.4)  -27  0.41 (0.36–0.47)  or diastolic blood pressure ≥90 mmHg  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Systolic blood pressure ≥140 mmHgor diastolic blood pressure ≥90 mmHgor medication  36.4  48.8  -12.4  0.72 (NR)  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  Systolic blood pressure ≥140 mmHg  24  38.5  -14.5  0.57 (0.55–0.58)  or diastolic blood pressure ≥90 mmHg  or medication  South America                Lima-Costa et al. 2004 (n = 970)34  Doctor or other health professional  Ever  Systolic blood pressure ≥140 mmHg  27.2 (24.4-30.1)  23.3 (20.7-26.1)  +3.9  NR  or diastolic blood pressure ≥90 mmHg  or medication  Australia/New Zealand                Taylor et al. 2010 (n = 1525)39  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  15.8  30.6  -14.8  0.55 (NR)  or diastolic blood pressure ≥90 mmHg  or medication  (b) Definition based on measurement only  North America                Dave et al. 2013 (n = 16 598)25  Physician, doctor or nursea  Evera  Systolic blood pressure ≥140 mmHg  16.15  24.81  -8.66  0.25 (NR)  or diastolic blood pressure ≥90 mmHgb  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  17.4 (16.5-18.3)  23.9 (22.9-24.9)  -6.5  0.21 (0.18–0.23)  or diastolic blood pressure ≥90 mmHg  SR vs.MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                Tormo et al. 2000 (n = 248)19  Physician  Ever  Diagnosis of high blood pressure  27.4  26.6  +0.8  0.65 (0.53–0.76)  or medication  or attending a hypertension control programme run only for hypertensive people  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  ICPC (International Classification of Primary Care) code K86/K87  52  44  +8  0.63 (0.55–0.70)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  56  62  -6  0.69 (0.66–0.71)  Frost et al. 2012 (n = 600)21  Not defined  NR  Diagnosis of hypertension          or medication        Men  22.2  36.7  -14.5  NR  North America                Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of hypertension  35.8  37.7  -1.9  0.75 (0.72–0.78)  and medication  or the words ‘borderline’ or ‘labile’ used in reference to blood pressure with documentation of two blood pressure measurements (consecutively but may be subsequent visits) systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg within a 12-month period  St Sauver et al. 2005 (n = 26 162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with hypertension  23.5  26.8  -3.3  NR  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  One hospital admission with a hypertension diagnosis code or one OHIP record with a hypertension diagnosis code, followed within 2 years by another OHIP record or a hospital admission with a hypertension diagnosis code.  20.8  27.6  -6.8  0.66 (0.65–0.66)  Codes used: ICD-9: 401x, 402x, 403x, 404x, 405x (any type), ICD-10-CA: I10, I11, I12, I13, I15 (any type), OHIP diagnosis code: 401, 402, 403, 404, 405 (any type)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for hypertension  25.1  25.8  -0.7  0.62 (NR)  Australia/New Zealand                Teh et al. 2013 (n = 878)41  Doctor  Ever  Primary health care record: Diagnoses were either ascertained from READ codes, a standardized primary care coding system, from hospital discharge letter, or from reading through the medical records.  54.8  68.4  -13.6  0.44 (0.38–0.50)  Administrative hospitalization discharge diagnosis records: Diagnosis codes for hypertension ICD-10: I10, ICD-9: 401.0, 401.1, 401.9 were identified.  Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  32.5 (29.7-35.3)  12.8 (10.8-14.8)  +19.7  0.35 (0.29–0.41)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  57.8 (55.0-60.7)  38.2 (35.3-41.0)  +19.6  0.21 (0.15–0.26)  Note: NR, not reported. a The question format was as follows: ‘Do you suffer from high blood pressure and/or has a physician/doctor/nurse diagnosed you as a hypertensive?’ b Persons who were taking blood pressure lowering medications were excluded. c Includes survey responses from the last survey (survey 5) only (‘case 1’ in the publication). Table 3 The format of question used in SR, the definition of diabetes/elevated blood glucose used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and kappa coefficients for agreement   SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of diabetes or elevated blood glucose  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Fasting glucose ≥7.0 mmol/l  29  31  −2  0.76 (0.69–0.83)  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Fasting blood glucose ≥7.0 mmol/l  7.8 (6.5–9.3)  10.6 (9.1–12.3)  −2.8  0.78 (0.73–0.84)  or treatment (insulin, hypoglycemic drugs or diet)  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l          or HbA1C≥6.5%  or medication        Men  5.8  6.6  −0.8  NR        Women  4.8  3.9  +0.9  NR  North America                Dey et al. 2015 (n = 101)26  Physician  Ever  Fasting plasma glucose ≥6.9 mmol/l  23.8  26.7  −2.9  0.76 (0.61–0.91)  or HbA1C≥6.5%  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Current  HbA1c≥7.0%  14.6 (12.4–16.7)  15.5 (13.2–17.7)  −0.9  0.86 (0.79–0.92)  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l  13.6  15.4  −1.8  0.82 (NR)  or treatment  (b) Definition based on measurement only  North America                Cowie et al. 2010 (n = 13 094)24  Doctor or health care provider  Ever (other than during pregnancy)  HbA1c≥6.5%  7.8 (7.0–8.6)  9.6 (8.7–10.5)  −1.8  NR  Fisher-Hoch et al. 2015 (n = 2838)27  Health care provider  Ever  Fasting plasma glucose ≥7.0 mmol/l (or other criteria of the 2010 American Diabetes Association definition)  16.4  27.6  −11.2  NR  Asia                Bao et al. 2015 (n = 7913)38  Doctor  Ever (other than during pregnancy for women)  Fasting plasma glucose ≥7.0 mmol/l and/or 2-h post-load plasma glucose ≥11.1 mmol/l  4.4  12.7  −8.3  NR  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  HbA1c≥6.5%  5.8  6.9  −1.1  0.65 (0.62–0.68)  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Current  Fasting plasma glucose ≥7.0 mmol/l  6.1 (5.6–6.7)  4.5 (4.0–4.9)  +1.6  0.58 (0.54–0.62)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  International Classification of Primary Care code T90  29  29  0  0.84 (0.78–0.89)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  20  23  −3  0.89 (0.86–0.92)  Frost et al. 2012 (n = 600)21  Not defined  Not defined  Diagnosis of type II diabetes          or medication        Men  6.5  7.2  −0.7  NR  North America                Simpson et al. 2004 (n = 1002)28  Physician  Ever  ‘Based on standardized specific criteria using data from medical history, standardized research physical examination (e.g. electrocardiograms, ankle brachial index, spirometric testing), review of all medications, review of hospital records, x-rays, and a physician questionnaire.’  17  17  0  0.92 (0.86–0.98)  Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of diabetes mellitus  5.2  7.4  −2.2  0.76 (0.70–0.82)  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  Based on Ontario Health Insurance Plan diagnosis codes and DAD admissions. Different criteria for different age groups.  6.8  8.4  −1.6  0.80 (0.80–0.81)  Codes used: ICD-9: 250 (any type), ICD-10-CA: E10, E11, E13, E14 (any type), OHIP diagnosis code: 250, OHIP fee code: Q040, K029, K030  Leong et al. 2013 (n = 3322)32  Doctor or another health professional  Ever  Two or more physical billings for diabetes and/or one or more hospitalizations for diabetes  7.9  8.5  −0.6  0.79 (0.76–0.83)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for diabetes  5.1  6.5  −1.4  0.68 (NR)  Australia/New Zealand                Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  8.6 (6.9–10.3)  7.7 (6.1–9.3)  +0.9  0.75 (0.68–0.83)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  12.8 (10.8–14.7)  12.7 (10.7–14.6)  +0.1  0.77 (0.72–0.83)    SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of diabetes or elevated blood glucose  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Fasting glucose ≥7.0 mmol/l  29  31  −2  0.76 (0.69–0.83)  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Fasting blood glucose ≥7.0 mmol/l  7.8 (6.5–9.3)  10.6 (9.1–12.3)  −2.8  0.78 (0.73–0.84)  or treatment (insulin, hypoglycemic drugs or diet)  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l          or HbA1C≥6.5%  or medication        Men  5.8  6.6  −0.8  NR        Women  4.8  3.9  +0.9  NR  North America                Dey et al. 2015 (n = 101)26  Physician  Ever  Fasting plasma glucose ≥6.9 mmol/l  23.8  26.7  −2.9  0.76 (0.61–0.91)  or HbA1C≥6.5%  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Current  HbA1c≥7.0%  14.6 (12.4–16.7)  15.5 (13.2–17.7)  −0.9  0.86 (0.79–0.92)  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l  13.6  15.4  −1.8  0.82 (NR)  or treatment  (b) Definition based on measurement only  North America                Cowie et al. 2010 (n = 13 094)24  Doctor or health care provider  Ever (other than during pregnancy)  HbA1c≥6.5%  7.8 (7.0–8.6)  9.6 (8.7–10.5)  −1.8  NR  Fisher-Hoch et al. 2015 (n = 2838)27  Health care provider  Ever  Fasting plasma glucose ≥7.0 mmol/l (or other criteria of the 2010 American Diabetes Association definition)  16.4  27.6  −11.2  NR  Asia                Bao et al. 2015 (n = 7913)38  Doctor  Ever (other than during pregnancy for women)  Fasting plasma glucose ≥7.0 mmol/l and/or 2-h post-load plasma glucose ≥11.1 mmol/l  4.4  12.7  −8.3  NR  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  HbA1c≥6.5%  5.8  6.9  −1.1  0.65 (0.62–0.68)  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Current  Fasting plasma glucose ≥7.0 mmol/l  6.1 (5.6–6.7)  4.5 (4.0–4.9)  +1.6  0.58 (0.54–0.62)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  International Classification of Primary Care code T90  29  29  0  0.84 (0.78–0.89)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  20  23  −3  0.89 (0.86–0.92)  Frost et al. 2012 (n = 600)21  Not defined  Not defined  Diagnosis of type II diabetes          or medication        Men  6.5  7.2  −0.7  NR  North America                Simpson et al. 2004 (n = 1002)28  Physician  Ever  ‘Based on standardized specific criteria using data from medical history, standardized research physical examination (e.g. electrocardiograms, ankle brachial index, spirometric testing), review of all medications, review of hospital records, x-rays, and a physician questionnaire.’  17  17  0  0.92 (0.86–0.98)  Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of diabetes mellitus  5.2  7.4  −2.2  0.76 (0.70–0.82)  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  Based on Ontario Health Insurance Plan diagnosis codes and DAD admissions. Different criteria for different age groups.  6.8  8.4  −1.6  0.80 (0.80–0.81)  Codes used: ICD-9: 250 (any type), ICD-10-CA: E10, E11, E13, E14 (any type), OHIP diagnosis code: 250, OHIP fee code: Q040, K029, K030  Leong et al. 2013 (n = 3322)32  Doctor or another health professional  Ever  Two or more physical billings for diabetes and/or one or more hospitalizations for diabetes  7.9  8.5  −0.6  0.79 (0.76–0.83)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for diabetes  5.1  6.5  −1.4  0.68 (NR)  Australia/New Zealand                Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  8.6 (6.9–10.3)  7.7 (6.1–9.3)  +0.9  0.75 (0.68–0.83)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  12.8 (10.8–14.7)  12.7 (10.7–14.6)  +0.1  0.77 (0.72–0.83)  Note: NR, not reported. a Includes survey responses from the last survey (survey 5) only (‘case 1’ in the publication). Table 3 The format of question used in SR, the definition of diabetes/elevated blood glucose used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and kappa coefficients for agreement   SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of diabetes or elevated blood glucose  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Fasting glucose ≥7.0 mmol/l  29  31  −2  0.76 (0.69–0.83)  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Fasting blood glucose ≥7.0 mmol/l  7.8 (6.5–9.3)  10.6 (9.1–12.3)  −2.8  0.78 (0.73–0.84)  or treatment (insulin, hypoglycemic drugs or diet)  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l          or HbA1C≥6.5%  or medication        Men  5.8  6.6  −0.8  NR        Women  4.8  3.9  +0.9  NR  North America                Dey et al. 2015 (n = 101)26  Physician  Ever  Fasting plasma glucose ≥6.9 mmol/l  23.8  26.7  −2.9  0.76 (0.61–0.91)  or HbA1C≥6.5%  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Current  HbA1c≥7.0%  14.6 (12.4–16.7)  15.5 (13.2–17.7)  −0.9  0.86 (0.79–0.92)  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l  13.6  15.4  −1.8  0.82 (NR)  or treatment  (b) Definition based on measurement only  North America                Cowie et al. 2010 (n = 13 094)24  Doctor or health care provider  Ever (other than during pregnancy)  HbA1c≥6.5%  7.8 (7.0–8.6)  9.6 (8.7–10.5)  −1.8  NR  Fisher-Hoch et al. 2015 (n = 2838)27  Health care provider  Ever  Fasting plasma glucose ≥7.0 mmol/l (or other criteria of the 2010 American Diabetes Association definition)  16.4  27.6  −11.2  NR  Asia                Bao et al. 2015 (n = 7913)38  Doctor  Ever (other than during pregnancy for women)  Fasting plasma glucose ≥7.0 mmol/l and/or 2-h post-load plasma glucose ≥11.1 mmol/l  4.4  12.7  −8.3  NR  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  HbA1c≥6.5%  5.8  6.9  −1.1  0.65 (0.62–0.68)  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Current  Fasting plasma glucose ≥7.0 mmol/l  6.1 (5.6–6.7)  4.5 (4.0–4.9)  +1.6  0.58 (0.54–0.62)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  International Classification of Primary Care code T90  29  29  0  0.84 (0.78–0.89)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  20  23  −3  0.89 (0.86–0.92)  Frost et al. 2012 (n = 600)21  Not defined  Not defined  Diagnosis of type II diabetes          or medication        Men  6.5  7.2  −0.7  NR  North America                Simpson et al. 2004 (n = 1002)28  Physician  Ever  ‘Based on standardized specific criteria using data from medical history, standardized research physical examination (e.g. electrocardiograms, ankle brachial index, spirometric testing), review of all medications, review of hospital records, x-rays, and a physician questionnaire.’  17  17  0  0.92 (0.86–0.98)  Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of diabetes mellitus  5.2  7.4  −2.2  0.76 (0.70–0.82)  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  Based on Ontario Health Insurance Plan diagnosis codes and DAD admissions. Different criteria for different age groups.  6.8  8.4  −1.6  0.80 (0.80–0.81)  Codes used: ICD-9: 250 (any type), ICD-10-CA: E10, E11, E13, E14 (any type), OHIP diagnosis code: 250, OHIP fee code: Q040, K029, K030  Leong et al. 2013 (n = 3322)32  Doctor or another health professional  Ever  Two or more physical billings for diabetes and/or one or more hospitalizations for diabetes  7.9  8.5  −0.6  0.79 (0.76–0.83)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for diabetes  5.1  6.5  −1.4  0.68 (NR)  Australia/New Zealand                Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  8.6 (6.9–10.3)  7.7 (6.1–9.3)  +0.9  0.75 (0.68–0.83)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  12.8 (10.8–14.7)  12.7 (10.7–14.6)  +0.1  0.77 (0.72–0.83)    SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of diabetes or elevated blood glucose  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Fasting glucose ≥7.0 mmol/l  29  31  −2  0.76 (0.69–0.83)  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Fasting blood glucose ≥7.0 mmol/l  7.8 (6.5–9.3)  10.6 (9.1–12.3)  −2.8  0.78 (0.73–0.84)  or treatment (insulin, hypoglycemic drugs or diet)  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l          or HbA1C≥6.5%  or medication        Men  5.8  6.6  −0.8  NR        Women  4.8  3.9  +0.9  NR  North America                Dey et al. 2015 (n = 101)26  Physician  Ever  Fasting plasma glucose ≥6.9 mmol/l  23.8  26.7  −2.9  0.76 (0.61–0.91)  or HbA1C≥6.5%  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Current  HbA1c≥7.0%  14.6 (12.4–16.7)  15.5 (13.2–17.7)  −0.9  0.86 (0.79–0.92)  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l  13.6  15.4  −1.8  0.82 (NR)  or treatment  (b) Definition based on measurement only  North America                Cowie et al. 2010 (n = 13 094)24  Doctor or health care provider  Ever (other than during pregnancy)  HbA1c≥6.5%  7.8 (7.0–8.6)  9.6 (8.7–10.5)  −1.8  NR  Fisher-Hoch et al. 2015 (n = 2838)27  Health care provider  Ever  Fasting plasma glucose ≥7.0 mmol/l (or other criteria of the 2010 American Diabetes Association definition)  16.4  27.6  −11.2  NR  Asia                Bao et al. 2015 (n = 7913)38  Doctor  Ever (other than during pregnancy for women)  Fasting plasma glucose ≥7.0 mmol/l and/or 2-h post-load plasma glucose ≥11.1 mmol/l  4.4  12.7  −8.3  NR  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  HbA1c≥6.5%  5.8  6.9  −1.1  0.65 (0.62–0.68)  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Current  Fasting plasma glucose ≥7.0 mmol/l  6.1 (5.6–6.7)  4.5 (4.0–4.9)  +1.6  0.58 (0.54–0.62)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  International Classification of Primary Care code T90  29  29  0  0.84 (0.78–0.89)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  20  23  −3  0.89 (0.86–0.92)  Frost et al. 2012 (n = 600)21  Not defined  Not defined  Diagnosis of type II diabetes          or medication        Men  6.5  7.2  −0.7  NR  North America                Simpson et al. 2004 (n = 1002)28  Physician  Ever  ‘Based on standardized specific criteria using data from medical history, standardized research physical examination (e.g. electrocardiograms, ankle brachial index, spirometric testing), review of all medications, review of hospital records, x-rays, and a physician questionnaire.’  17  17  0  0.92 (0.86–0.98)  Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of diabetes mellitus  5.2  7.4  −2.2  0.76 (0.70–0.82)  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  Based on Ontario Health Insurance Plan diagnosis codes and DAD admissions. Different criteria for different age groups.  6.8  8.4  −1.6  0.80 (0.80–0.81)  Codes used: ICD-9: 250 (any type), ICD-10-CA: E10, E11, E13, E14 (any type), OHIP diagnosis code: 250, OHIP fee code: Q040, K029, K030  Leong et al. 2013 (n = 3322)32  Doctor or another health professional  Ever  Two or more physical billings for diabetes and/or one or more hospitalizations for diabetes  7.9  8.5  −0.6  0.79 (0.76–0.83)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for diabetes  5.1  6.5  −1.4  0.68 (NR)  Australia/New Zealand                Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  8.6 (6.9–10.3)  7.7 (6.1–9.3)  +0.9  0.75 (0.68–0.83)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  12.8 (10.8–14.7)  12.7 (10.7–14.6)  +0.1  0.77 (0.72–0.83)  Note: NR, not reported. a Includes survey responses from the last survey (survey 5) only (‘case 1’ in the publication). Table 4 The format of question used in SR, the definition of elevated total cholesterol used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and κ coefficients for agreement   SR, the format of question   Objective data (HES or MR)               Diagnosed/told by  Time frame covered  Definition of elevated total cholesterol  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  Total cholesterol ≥6.5 mmol/l  34  36  -2  0.55 (0.42–0.59)  or medication  Tolonen et al. 2014 (n =4127)9  Medical doctor  Ever  Total cholesterol ≥5.0 mmol/l          or medication        Men  21.5  71.1  -49.6  NR        Women  19  62.6  -43.6  NR  North America                Natarajan et al. 2002 (n =8236)22  Doctor or other health professional  Ever  Total cholesterol ≥5.17 mmol/l  32.1  59.4  -27.3  NR  or medication  Ahluwalia et al. 2009 (n =733)23  Doctor, nurse or other health professional  Ever  Total cholesterol ≥6.2 mmol/l  56  44.3  +11.7  0.51 (0.44–0.57)  or medication  Asia                Chun et al. 2016 (n =7270)36  Doctor  Ever  Total cholesterol ≥6.2 mmol/l  11.7  16.7  -5  0.48 (NR)  or medication  Australia/New Zealand  Taylor et al. 2010 (n =1525)39  Doctor or nurse  Currently (‘still’)  Total cholesterol ≥5.5 mmol/l  12.3  42.8  -30.5  0.30 (NR)  or medication  (b) Definition based on measurement only  Australia/New Zealand  Peterson et al. 2016 (n =7269)40  Doctor or nurse  Current  Total cholesterol ≥5.5 mmol/l  12.2 (11.5–13.0)  37.3 (36.2–38.4)  -25.1  -0.02 (-0.04–0.01)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  ICPC (International Classification of Primary Care) code T93  34  23  +11  0.48 (0.39–0.57)  or medication  and/or search terms or registration codes specific for the primary health care center  North America                St Sauver et al. 2005 (n =26162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with high cholesterol  22.8  27.2  -4.4  NR  Koller et a. 2014 (n =3821)33  Doctor or health care provider  Ever  ICD-9 codes for elevated cholesterol  17.3  18.5  -1.2  0.57 (NR)    SR, the format of question   Objective data (HES or MR)               Diagnosed/told by  Time frame covered  Definition of elevated total cholesterol  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  Total cholesterol ≥6.5 mmol/l  34  36  -2  0.55 (0.42–0.59)  or medication  Tolonen et al. 2014 (n =4127)9  Medical doctor  Ever  Total cholesterol ≥5.0 mmol/l          or medication        Men  21.5  71.1  -49.6  NR        Women  19  62.6  -43.6  NR  North America                Natarajan et al. 2002 (n =8236)22  Doctor or other health professional  Ever  Total cholesterol ≥5.17 mmol/l  32.1  59.4  -27.3  NR  or medication  Ahluwalia et al. 2009 (n =733)23  Doctor, nurse or other health professional  Ever  Total cholesterol ≥6.2 mmol/l  56  44.3  +11.7  0.51 (0.44–0.57)  or medication  Asia                Chun et al. 2016 (n =7270)36  Doctor  Ever  Total cholesterol ≥6.2 mmol/l  11.7  16.7  -5  0.48 (NR)  or medication  Australia/New Zealand  Taylor et al. 2010 (n =1525)39  Doctor or nurse  Currently (‘still’)  Total cholesterol ≥5.5 mmol/l  12.3  42.8  -30.5  0.30 (NR)  or medication  (b) Definition based on measurement only  Australia/New Zealand  Peterson et al. 2016 (n =7269)40  Doctor or nurse  Current  Total cholesterol ≥5.5 mmol/l  12.2 (11.5–13.0)  37.3 (36.2–38.4)  -25.1  -0.02 (-0.04–0.01)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  ICPC (International Classification of Primary Care) code T93  34  23  +11  0.48 (0.39–0.57)  or medication  and/or search terms or registration codes specific for the primary health care center  North America                St Sauver et al. 2005 (n =26162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with high cholesterol  22.8  27.2  -4.4  NR  Koller et a. 2014 (n =3821)33  Doctor or health care provider  Ever  ICD-9 codes for elevated cholesterol  17.3  18.5  -1.2  0.57 (NR)  Note: NR, not reported. Table 4 The format of question used in SR, the definition of elevated total cholesterol used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and κ coefficients for agreement   SR, the format of question   Objective data (HES or MR)               Diagnosed/told by  Time frame covered  Definition of elevated total cholesterol  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  Total cholesterol ≥6.5 mmol/l  34  36  -2  0.55 (0.42–0.59)  or medication  Tolonen et al. 2014 (n =4127)9  Medical doctor  Ever  Total cholesterol ≥5.0 mmol/l          or medication        Men  21.5  71.1  -49.6  NR        Women  19  62.6  -43.6  NR  North America                Natarajan et al. 2002 (n =8236)22  Doctor or other health professional  Ever  Total cholesterol ≥5.17 mmol/l  32.1  59.4  -27.3  NR  or medication  Ahluwalia et al. 2009 (n =733)23  Doctor, nurse or other health professional  Ever  Total cholesterol ≥6.2 mmol/l  56  44.3  +11.7  0.51 (0.44–0.57)  or medication  Asia                Chun et al. 2016 (n =7270)36  Doctor  Ever  Total cholesterol ≥6.2 mmol/l  11.7  16.7  -5  0.48 (NR)  or medication  Australia/New Zealand  Taylor et al. 2010 (n =1525)39  Doctor or nurse  Currently (‘still’)  Total cholesterol ≥5.5 mmol/l  12.3  42.8  -30.5  0.30 (NR)  or medication  (b) Definition based on measurement only  Australia/New Zealand  Peterson et al. 2016 (n =7269)40  Doctor or nurse  Current  Total cholesterol ≥5.5 mmol/l  12.2 (11.5–13.0)  37.3 (36.2–38.4)  -25.1  -0.02 (-0.04–0.01)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  ICPC (International Classification of Primary Care) code T93  34  23  +11  0.48 (0.39–0.57)  or medication  and/or search terms or registration codes specific for the primary health care center  North America                St Sauver et al. 2005 (n =26162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with high cholesterol  22.8  27.2  -4.4  NR  Koller et a. 2014 (n =3821)33  Doctor or health care provider  Ever  ICD-9 codes for elevated cholesterol  17.3  18.5  -1.2  0.57 (NR)    SR, the format of question   Objective data (HES or MR)               Diagnosed/told by  Time frame covered  Definition of elevated total cholesterol  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  Total cholesterol ≥6.5 mmol/l  34  36  -2  0.55 (0.42–0.59)  or medication  Tolonen et al. 2014 (n =4127)9  Medical doctor  Ever  Total cholesterol ≥5.0 mmol/l          or medication        Men  21.5  71.1  -49.6  NR        Women  19  62.6  -43.6  NR  North America                Natarajan et al. 2002 (n =8236)22  Doctor or other health professional  Ever  Total cholesterol ≥5.17 mmol/l  32.1  59.4  -27.3  NR  or medication  Ahluwalia et al. 2009 (n =733)23  Doctor, nurse or other health professional  Ever  Total cholesterol ≥6.2 mmol/l  56  44.3  +11.7  0.51 (0.44–0.57)  or medication  Asia                Chun et al. 2016 (n =7270)36  Doctor  Ever  Total cholesterol ≥6.2 mmol/l  11.7  16.7  -5  0.48 (NR)  or medication  Australia/New Zealand  Taylor et al. 2010 (n =1525)39  Doctor or nurse  Currently (‘still’)  Total cholesterol ≥5.5 mmol/l  12.3  42.8  -30.5  0.30 (NR)  or medication  (b) Definition based on measurement only  Australia/New Zealand  Peterson et al. 2016 (n =7269)40  Doctor or nurse  Current  Total cholesterol ≥5.5 mmol/l  12.2 (11.5–13.0)  37.3 (36.2–38.4)  -25.1  -0.02 (-0.04–0.01)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  ICPC (International Classification of Primary Care) code T93  34  23  +11  0.48 (0.39–0.57)  or medication  and/or search terms or registration codes specific for the primary health care center  North America                St Sauver et al. 2005 (n =26162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with high cholesterol  22.8  27.2  -4.4  NR  Koller et a. 2014 (n =3821)33  Doctor or health care provider  Ever  ICD-9 codes for elevated cholesterol  17.3  18.5  -1.2  0.57 (NR)  Note: NR, not reported. Objective data HES measures In HES’s, blood pressure was measured twice or three times. In some studies, the average of two or three measurements was used in the analyses, whereas in others the result of only one measurement was used (Supplementary file 1). Diabetes or elevated blood glucose and elevated total cholesterol status were analyzed from blood samples. For elevated blood glucose, fasting plasma glucose (FPG) or HbA1c from non-fasting samples were used (table 3, Supplementary file 1). In one study, also glucose tolerance test was used.38 Total cholesterol was analyzed from fasting samples in four studies,15,36,39,40 both fasting and non-fasting samples in two studies9,22 and non-fasting samples in one study23 (Supplementary file 1). The definitions and cut-off values for the selected risk factors or diseases varied between the studies (see tables 2–4). In most studies, the use of medication for the condition was taken into account in the definition (measured value exceeded the cut-off value and/or medication was used for the condition). We focus on these studies. Studies based on measurements only are listed in the tables but not used in comparison. For studies that compared self-reported data with HES data the terms ‘hypertension’ and ‘diabetes’ are used when the definition in HES data was based on both measurements (blood pressure or fasting blood glucose/HbA1C) and medication for the condition (tables 2 and 3). We are aware that all respondents in these groups may not fulfil the clinical diagnostic criteria. When the definition was based on measurements only, the terms ‘elevated blood pressure’ and ‘elevated blood glucose’ are used. For cholesterol, the term ‘elevated total cholesterol’ is used in both cases (table 4). Data from MRs The data on MRs or other administrative registers typically relied on diagnostic codes and also the use of medication was taken into account in several studies (tables 2–4). For studies that compared self-reported data with MRs the terms ‘hypertension’, ‘diabetes’ and ‘elevated total cholesterol’ are used. Comparability of prevalence rates obtained from different data sources Self-reported data under-estimated the prevalence of hypertension and elevated total cholesterol in the majority of studies compared with HES data (tables 2 and 4). The difference between SR and MRs was less pronounced. For diabetes, under-estimate was negligible (table 3). Hypertension Self-reported data under-estimated hypertension prevalence in eight out of nine studies where the data for men and women were combined, and the reference data came from HESs (table 2). In a study that showed the prevalence rates for men and women separately, the self-reported hypertension prevalence rate was 10.1 percentage points lower among men and 3.5 percentage points higher among women compared with HES data.9 Among four studies where CI’s were presented, statistically significant under-estimation was observed in three studies. The difference between the prevalence rates by self-reported data and HES ranged from −15.9 to +3.9 percentage points including figures both for men and women combined and separately. The differences between self-reported data and MR data were less consistent (table 2). Hypertension was under-estimated by self-reported data in six out of eight studies that only showed the results for men and women combined. Under-estimation of hypertension was also seen in a study including men only.21 On the contrary, substantial and significant over-estimation of hypertension by SRs was observed in the two age groups in a study for older women.16 The difference between the hypertension prevalence rates by self-reporting and MRs ranged from −14.5 to +19.7 percentage points. The κ coefficients assessing the agreement between two methods ranged from 0.41 to 0.72 for self-reported data vs. HES and from 0.21 to 0.75 for self-reported data vs. MRs (table 2). Diabetes The self-reported data provided fairly similar prevalence rates as did HES and MRs (table 3). The difference in prevalence rates between self-reporting and HES ranged from −2.9 to +0.9 percentage points. When self-reported data were compared with MRs, the difference between the methods ranged from −3.0 to +0.9 percentage points. The κ coefficients ranged from 0.76 to 0.86 for self-reported data vs. HES and from 0.68 to 0.92 for self-reported data vs. MRs (table 3). Elevated total cholesterol Under-reporting of elevated cholesterol was observed in four out of five studies combining data for men and women and comparing self-reporting with HES data (table 4). Substantial under-estimation was observed both among men and women (−49.6 and −43.6 percentage points, respectively) in a study where the data for men and women were analyzed separately.9 The difference between the prevalence rates by self-reporting and HES ranged from −49.6 to +11.7 percentage points. Under-reporting was observed in two out of three studies, when data from self-reporting were compared with data from MRs. The difference between the prevalence rates by self-reporting and MRs ranged from −4.4 to +11 percentage points. The κ coefficients ranged from 0.30 to 0.55 for self-reported data vs. HES. The two κ coefficients reported in the selected publications for self-reported data vs. MRs were 0.48 and 0.57 (table 4). Discussion We compared self-reported data on hypertension, diabetes and elevated total cholesterol with two more objective data sources, HES and MRs or other register based data, to increase knowledge on which methods are most feasible, and produce reliable, accurate and comparable information for health monitoring purposes. Studies published in 2000–16 that compared self-reported data with either HES data or MRs were evaluated. Self-reported data tended to under-estimate prevalence rates especially for hypertension and elevated total cholesterol, which is in line with earlier reviews.8,13 The under-estimate was more pronounced, when the self-reported data were compared with HES data instead of MRs. Hypertension and elevated total cholesterol are typically asymptomatic for a long time and remain therefore easily unmeasured and undetected. Therefore, they may be under-estimated in MR data. In a study from the Netherlands, both HES and MRs were included as reference material.15 Higher hypertension prevalence was reported with HES than with MRs (59% and 44%, respectively), which also suggests that MRs do not cover all hypertensive subjects. Diabetes was more accurately self-reported both compared with HES data and MRs, which is also in line with earlier reviews.13 As the screening of blood glucose and the use risk tests for diabetes have become more common the awareness of the condition has improved. The κ coefficients were consistent with the comparisons based on prevalence rates in that the overall agreement between SR and HES or MRs was higher for diabetes than for hypertension or elevated cholesterol. In an earlier review, the following factors, among others, were reported to explain the inaccuracy of SR: respondents may lack the knowledge to accurately answer the questions posed, and poorly designed survey instruments could result in respondents not fully comprehending the questions.13 Participants’ awareness of the disease or risk factor levels is crucial to obtain accurate reporting in HIS studies.9 Unless a diagnosis is set or risk factors examined and thoroughly explained to the patient by medical professionals, the participants are not aware of them, as many conditions are asymptomatic. In general, women have been more often aware of their hypertension than men,42–44 which can be partly explained by regular follow-up during pregnancies. This applied also to the evaluated studies reporting hypertension separately for men and women.9,36,39 If SR is based on having ever been told to have elevated blood pressure or elevated blood glucose, pregnancy complications may lead to higher prevalence rates for women. This might explain a part of the slight overestimation of hypertension among women.9 In contrast, the substantial overestimation of hypertension among older women cannot be explained with conditions related to pregnancies as the question concerned the past 3 years only.16 Instead, the authors’ presumed hypertension to be underestimated in the hospital data. The awareness of one’s own conditions or risk factor levels may have increased over time in many populations. Measuring blood pressure with digital devices at home has become more common and cholesterol and blood glucose measurements may be available more widely than earlier and also outside medical services, e.g. at pharmacies and health clubs. Hypertension awareness has increased significantly among hypertensives43,44 and this development can be presumed to have continued. The comparability of studies included in this evaluation was hindered by many discrepancies in their methods. Self-reported data on risk factors or diseases were collected heterogeneously as varying questions and definitions were used in different studies and the time frame covered with the question varied. Both population level coverage of the service and coverage in recording diagnoses and measures limit comparability of data from MRs or other administrative registers. For HES data, the use of medication was not taken into account in the definition of the conditions in all studies. We decided to focus on studies where the definition was based on both measured values and the use of medication, although the issue of medication is not straightforward. Antihypertensive drugs and statins are also used for other indications than elevated blood pressure or cholesterol. Considering medication based on recorded product names may hence have led to overestimation of hypertension or elevated cholesterol if the indication was not specified. Furthermore, different cut-off values were used, especially for elevated total cholesterol. The cut-offs used for elevated total cholesterol ranged from 5.0 to 6.5 mmol/l. The lowest cut-off was applied in a study in which also the largest underestimate of elevated total cholesterol by SR was seen.9 In addition, results for two cut-offs were shown in one study.23 Using the lower cut-off (5.17 mmol/l) resulted in a conclusion that SR underestimated, whereas using the higher cut-off (6.2 mmol/l) resulted in a conclusion that SR overestimated elevated total cholesterol. This might be related to practicing clinicians using different thresholds for communicating elevated cholesterol levels.22 Medication was included in the definition only in the case of the higher cut-off, which is why we only included the results by the higher cut-off from the study in question in this evaluation. Masked or white coat hypertension may also affect the prevalence rates for elevated blood pressure, as in HESs blood pressure is measured on one occasion only and most commonly in an office at the examination site.45,46 Methodological differences, such as the format of questions, response rates and survey modes, among national surveys have hampered valid mutual comparisons,4,47,48 but recent efforts for standardization in European countries will improve the situation.6 However, rising non-participation rates cause problems and attention needs to be given to recruitment methods.7 While electronic patient records have improved the availability of data and provide information not available from survey data such as comorbidity and diagnostic and treatment details, they cannot fully substitute survey data. MRs may not always be up-to-date and there may be socio-economic differences in access to medical care.49 In clinical practice, screening for cardiovascular risk factors and control of diagnosed diseases may not be systematic. The measures rely on health professionals’ practice patterns and accuracy of recording.12 MRs may also lack information on diagnoses performed in different health care sectors (public vs. private sector), and background and health behaviours. There may also be limitations in availability of or access to MRs for research purposes. Conclusions All the data sources have their strengths and limitations, and none of them alone should be regarded as a gold standard. However, using only self-reported data in monitoring vascular diseases and their risk at the population level may lead to underestimation of the true prevalence especially in regards to hypertension and elevated total cholesterol. Whenever feasible, combined information from standardized interviews and measurements supplemented with register data to exploit the advantages of all of them might be used for public health monitoring. Supplementary data Supplementary data are available at EURPUB online. Funding This work was supported by the European Commission/DG SANTÉ (BRIDGE Health Project, grant number 664691). The views expressed here are those of the authors and they do not represent the Commission’s official position. Conflicts of interest: None declared Key points The prevalence rates of hypertension and elevated total cholesterol were under-estimated with self-reported data compared with data from HESs and to a smaller extent compared with medical records. Instead, all the three data sources resulted in quite similar diabetes prevalence rates. The methods and the wordings used in questions varied between the studies which hindered the comparison. Common standardized protocols and questions would facilitate the comparison between surveys. Combination of different data sources should be used in risk factor and disease monitoring whenever possible to increase the reliability and coverage of prevalence estimates. References 1 GBD 2015 Risk Factors Collaborators. Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet  2016; 388: 1659– 724. CrossRef Search ADS PubMed  2 World Health Organization. Global Action Plan for the Prevention and Control of Noncommunicable Diseases . Geneva: World Health Organization, 2013. 3 Kilpeläinen K, Tuomi-Nikula A, Thelen J. Health indicators in Europe: availability and data needs. Eur J Public Health  2012; 22: 716– 21. Google Scholar CrossRef Search ADS PubMed  4 Verschuuren M, Gissler M, Kilpeläinen K. Public health indicators for the EU: the joint action for ECHIM (European Community Health Indicators & Monitoring). Arch Public Health  2013; 71: 12. Google Scholar CrossRef Search ADS PubMed  5 Tolonen H, Koponen P, Mindell J, et al.   European Health Examination Survey—towards a sustainable monitoring system. Eur J Public Health  2014; 24: 338– 44. Google Scholar CrossRef Search ADS PubMed  6 Tolonen H, Koponen P, Naska A, et al.   Challenges in standardization of blood pressure measurement at the population level. BMC Med Res Methodol  2015; 15: 3. Google Scholar CrossRef Search ADS PubMed  7 Mindell JS, Giampaoli S, Goesswald A, et al.   Sample selection, recruitment and participation rates in health examination surveys in Europe—experience from seven national surveys. BMC Med Res Methodol  2015; 15: 4. Google Scholar CrossRef Search ADS PubMed  8 Gorber S, Tremblay M, Campbell N, Hardt J. The accuracy of self-reported hypertension: a systematic review and meta-analysis. Curr Hypertension Rev  2008; 4: 36– 62. Google Scholar CrossRef Search ADS   9 Tolonen H, Koponen P, Mindell JS, et al.   Under-estimation of obesity, hypertension and high cholesterol by self-reported data: comparison of self-reported information and objective measures from health examination surveys. Eur J Public Health  2014; 24: 941– 8. Google Scholar CrossRef Search ADS PubMed  10 Elo SL, Karlberg IH. Validity and utilization of epidemiological data: a study of ischaemic heart disease and coronary risk factors in a local population. Public Health  2009; 123: 52– 7. Google Scholar CrossRef Search ADS PubMed  11 Wild S, Fischbacher C, McKnight J. (on Behalf of the Scottish Diabetes Research Network Epidemiology Group). Using Large Diabetes Databases for Research. J Diabetes Sci Technol  2016; 10: 1073– 8. Google Scholar CrossRef Search ADS PubMed  12 Baus A, Hendryx M, Pollard C. Identifying patients with hypertension: a case for auditing electronic health record data. Perspect Health Inf Manag  2012; 9: 1e. Google Scholar PubMed  13 Newell SA, Girgis A, Sanson-Fisher RW, Savolainen NJ. The accuracy of self-reported health behaviors and risk factors relating to cancer and cardiovascular disease in the general population: a critical review. Am J Prev Med  1999; 17: 211– 29. Google Scholar CrossRef Search ADS PubMed  14 Saydah SH, Geiss LS, Tierney E, et al.   Review of the performance of methods to identify diabetes cases among vital statistics, administrative, and survey data. Ann Epidemiol  2004; 14: 507– 16. Google Scholar CrossRef Search ADS PubMed  15 El Fakiri F, Bruijnzeels MA, Hoes AW. No evidence for marked ethnic differences in accuracy of self-reported diabetes, hypertension, and hypercholesterolemia. J Clin Epidemiol  2007; 60: 1271– 9. Google Scholar CrossRef Search ADS PubMed  16 Navin Cristina TJ, Stewart Williams JA, Parkinson L, et al.   Identification of diabetes, heart disease, hypertension and stroke in mid- and older-aged women: comparing self-report and administrative hospital data records. Geriatr Gerontol Int  2016; 16: 95– 102. Google Scholar CrossRef Search ADS PubMed  17 Molenaar EA, Ameijden EJCV, Grobbee DE, Numans ME. Comparison of routine care self-reported and biometrical data on hypertension and diabetes: results of the Utrecht Health Project. Eur J Public Health  2007; 17: 199– 205. Google Scholar CrossRef Search ADS PubMed  18 Huerta JM, Tormo MJ, Egea-Caparros JM, et al.   Accuracy of self-reported diabetes, hypertension and hyperlipidemia in the adult Spanish population. DINO study findings. Rev Esp Cardiol  2009; 62: 143– 52. Google Scholar CrossRef Search ADS PubMed  19 Tormo MJ, Navarro C, Chirlaque MD, Barber X. Validation of self diagnosis of high blood pressure in a sample of the Spanish EPIC cohort: overall agreement and predictive values. EPIC Group of Spain. J Epidemiol Commun Health  2000; 54: 221– 6. Google Scholar CrossRef Search ADS   20 Englert H, Muller-Nordhorn J, Seewald S, et al.   Is patient self-report an adequate tool for monitoring cardiovascular conditions in patients with hypercholesterolemia? J Public Health (Oxford)  2010; 32: 387– 94. Google Scholar CrossRef Search ADS   21 Frost M, Wraae K, Gudex C, et al.   Chronic diseases in elderly men: underreporting and underdiagnosis. Age Ageing  2012; 41: 177– 83. Google Scholar CrossRef Search ADS PubMed  22 Natarajan S, Lipsitz SR, Nietert PJ. Self-report of high cholesterol: determinants of validity in U.S. adults. Am J Prev Med  2002; 23: 13– 21. Google Scholar CrossRef Search ADS PubMed  23 Ahluwalia IB, Tessaro I, Rye S, Parker L. Self-reported and clinical measurement of three chronic disease risks among low-income women in West Virginia. J Womens Health (Larchmt)  2009; 18: 1857– 62. Google Scholar CrossRef Search ADS PubMed  24 Cowie CC, Rust KF, Byrd-Holt DD, et al.   Prevalence of diabetes and high risk for diabetes using A1C criteria in the U.S. population in 1988-2006. Diabetes Care  2010; 33: 562– 8. Google Scholar CrossRef Search ADS PubMed  25 Dave GJ, Bibeau DL, Schulz MR, et al.   Predictors of congruency between self-reported hypertension status and measured blood pressure in the stroke belt. J Am Soc Hypertens  2013; 7: 370– 8. Google Scholar CrossRef Search ADS PubMed  26 Dey AK, Alyass A, Muir RT, et al.   Validity of self-report of cardiovascular risk factors in a population at high risk for stroke. J Stroke Cerebrovasc Dis  2015; 24: 2860– 5. Google Scholar CrossRef Search ADS PubMed  27 Fisher-Hoch SP, Vatcheva KP, Rahbar MH, McCormick JB. Undiagnosed diabetes and pre-diabetes in health disparities. PLoS One  2015; 10: e0133135. Google Scholar CrossRef Search ADS PubMed  28 Simpson CF, Boyd CM, Carlson MC, et al.   Agreement between self-report of disease diagnoses and medical record validation in disabled older women: factors that modify agreement. J Am Geriatr Soc  2004; 52: 123– 7. Google Scholar CrossRef Search ADS PubMed  29 Okura Y, Urban LH, Mahoney DW, et al.   Agreement between self-report questionnaires and medical record data was substantial for diabetes, hypertension, myocardial infarction and stroke but not for heart failure. J Clin Epidemiol  2004; 57: 1096– 103. Google Scholar CrossRef Search ADS PubMed  30 St Sauver JL, Hagen PT, Cha SS, et al.   Agreement between patient reports of cardiovascular disease and patient medical records. Mayo Clin Proc  2005; 80: 203– 10. Google Scholar CrossRef Search ADS PubMed  31 Muggah E, Graves E, Bennett C, Manuel DG. Ascertainment of chronic diseases using population health data: a comparison of health administrative data and patient self-report. BMC Public Health  2013; 13: 16. Google Scholar CrossRef Search ADS PubMed  32 Leong A, Dasgupta K, Chiasson JL, Rahme E. Estimating the population prevalence of diagnosed and undiagnosed diabetes. Diabetes Care  2013; 36: 3002– 8. Google Scholar CrossRef Search ADS PubMed  33 Koller KR, Wilson AS, Asay ED, et al.   Agreement between self-report and medical record prevalence of 16 chronic conditions in the Alaska EARTH Study. J Prim Care Community Health  2014; 5: 160– 5. Google Scholar CrossRef Search ADS PubMed  34 Lima-Costa MF, Peixoto SV, Firmo JO. Validity of self-reported hypertension and its determinants (the Bambui study). Rev Saude Publica  2004; 38: 637– 42. Google Scholar CrossRef Search ADS PubMed  35 Goldman N, Lin IF, Weinstein M, Lin YH. Evaluating the quality of self-reports of hypertension and diabetes. J Clin Epidemiol  2003; 56: 148– 54. Google Scholar CrossRef Search ADS PubMed  36 Chun H, Kim IH, Min KD. Accuracy of self-reported hypertension, diabetes, and hypercholesterolemia: analysis of a representative sample of Korean older adults. Osong Public Health Res Perspect  2016; 7: 108– 15. Google Scholar CrossRef Search ADS PubMed  37 Ning M, Zhang Q, Yang M. Comparison of self-reported and biomedical data on hypertension and diabetes: findings from the China Health and Retirement Longitudinal Study (CHARLS). BMJ Open  2016; 6: e009836. Google Scholar CrossRef Search ADS PubMed  38 Bao C, Zhang D, Sun B, et al.   Optimal cut-off points of fasting plasma glucose for two-step strategy in estimating prevalence and screening undiagnosed diabetes and pre-diabetes in Harbin, China. PLoS One  2015; 10: e0119510. Google Scholar CrossRef Search ADS PubMed  39 Taylor A, Dal Grande E, Gill T, et al.   Comparing self-reported and measured high blood pressure and high cholesterol status using data from a large representative cohort study. Aust N Z J Public Health  2010; 34: 394– 400. Google Scholar CrossRef Search ADS PubMed  40 Peterson KL, Jacobs JP, Allender S, et al.   Characterising the extent of misreporting of high blood pressure, high cholesterol, and diabetes using the Australian Health Survey. BMC Public Health  2016; 16: 695. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The European Journal of Public Health Oxford University Press

Public health monitoring of hypertension, diabetes and elevated cholesterol: comparison of different data sources

Loading next page...
 
/lp/ou_press/public-health-monitoring-of-hypertension-diabetes-and-elevated-MVbxs7rfqa
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved.
ISSN
1101-1262
eISSN
1464-360X
D.O.I.
10.1093/eurpub/cky020
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Three data sources are generally used in monitoring health on the population level. Health interview surveys (HISs) are based on participants’ self-report. Health examination surveys (HESs) yield more objective data, and also persons who are unaware of their elevated risks can be detected. Medical records (MRs) and other administrative registers also provide objective data, but their availability, coverage and quality vary between countries. We summarized studies comparing self-reported data with (i) measured data from HESs or (ii) MRs. We aimed to describe differences in feasibility and comparability of different data sources for monitoring (i) elevated blood pressure or hypertension (ii) elevated blood glucose or diabetes and (iii) elevated total cholesterol. Methods We conducted a literature search to identify studies, which validated self-reported measures against objective measures. We found 30 studies published since the year 2000 fulfilling our inclusion criteria (targeted to adults and comparing prevalence among the same persons). Results Hypertension and elevated total cholesterol were prone to be under-estimated in HISs. The under-estimate was more pronounced, when the HIS data were compared with HES data, and lower when compared with MRs. For diabetes, the HISs and the objective methods resulted in fairly similar prevalence rates. Conclusion The three data sources measure different manifestations of the risk factors and cannot be expected to yield similar prevalence rates. Using HIS data only may lead to under-estimation of elevated risk factor levels or disease prevalence. Whenever possible, information from the three data sources should be evaluated and combined. Introduction In 2015, high blood pressure, high fasting blood glucose and high total cholesterol were among the 10 most common risk factors explaining disability-adjusted life years according to the Global Burden of Disease Study 2015.1 Their relevance as risk factors had increased since 2005. One of the six objectives in the WHO global action plan for the prevention and control of noncommunicable diseases 2013–20 is to monitor the trends and determinants of noncommunicable diseases and to evaluate progress in their prevention and control.2 The action plan calls for undertaking periodic data collection on cardiovascular and metabolic risk factors. Three indicators (i) prevalence of elevated blood pressure (defined as systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg) and mean systolic blood pressure, (ii) prevalence of elevated blood glucose or diabetes (defined as fasting plasma glucose concentration ≥7.0 mmol/l or on medication for elevated blood glucose) and (iii) prevalence of elevated total cholesterol (defined as total cholesterol ≥5.0 mmol/l) and mean total cholesterol concentration are among the 25 indicators that are recommended to be monitored to follow up the achievements in the voluntary global targets. Diabetes and blood pressure are also among the European Community Health Indicators (ECHI). In 2007, it was evaluated that data for these indicators were widely available in European countries, mostly from National Health Interview Surveys (HIS). In some cases, they were available from registers and health examination surveys (HESs).3,4 Based on information gathered during the European Community Health Indicators and Monitoring project (ECHIM), it was concluded that registers and statistical information systems are important sources of health data, but both national HISs and HESs are additionally needed to provide comprehensive information on morbidity, functioning and health determinants in the whole population. Several countries have conducted repeated national HESs in order to monitor key health indicators of the population. Collecting data with HESs including questionnaires, physical measurements and often also collection of biological samples, yields objective and up-to-date data, and also persons who are unaware of their elevated risk factor levels can be detected. However, HESs are costly and time-consuming and require proper standardization. For blood pressure measurement, as an example, thorough training of the survey personnel is essential.5,6 Furthermore, recruiting of participants is often challenging in HESs.7 Because of lower costs HISs are used more frequently than HESs. Self-administered questionnaires, face-to-face interviews and telephone interviews can be used. The disadvantages of self-report (SR) in HISs include that it rests on the subjects’ memory, awareness and ability to report on a health condition.5,8 The two objective data sources, HES data and medical records (MRs), describe different dimensions of the outcome. HES data mostly rely on measurements conducted on one occasion and clinical diagnoses cannot be made based on such data. However, people with elevated risk but without diagnosis can be identified from HES data. The availability and coverage of MRs vary between countries according to the national health care system.9 The data depend on access and use of health services. Undiagnosed persons that have not sought medical care but might fulfil diagnostic criteria cannot be identified. For example, routine registers reveal diabetes or cardiovascular disease only in those who have used services and have been diagnosed.10,11 Furthermore, many administrative registers only cover information from hospital care. There may be inconsistences in coding of the conditions by the health care personnel as well as in measurements, laboratory analysis and clinical practice.12 There are numerous studies that have examined the validity of self-reported data against more objective measures.8,13 To gather up-to date information on different data sources and to discuss their strengths and limitations as well as to evaluate their usability for public health monitoring purposes, we summarized studies that compared self-reported data with two more objective data types: (i) measured data from HESs or (ii) MRs or other administrative register-based data. Three key risk factors or disease diagnoses were chosen to this evaluation: (i) elevated blood pressure or hypertension, (ii) elevated blood glucose or diabetes and (iii) elevated total cholesterol. We aimed to describe differences in feasibility and comparability of different data sources for monitoring major cardiovascular risk factors. Methods Literature search The literature search was conducted in PubMed in October 2016 and again in the beginning of January 2017 to identify studies on elevated blood pressure or hypertension, elevated blood glucose or diabetes and elevated total cholesterol which validated self-reported measures against measures based on (i) HES data or (ii) MRs or other administrative register-based data. Henceforth, the latter is referred to as MRs. The search terms are listed in Supplementary file 1. Studies published since the year 2000 were included. Older studies were excluded based on three reasons. Firstly, we were mainly interested in clarifying the current situation in relation to health monitoring. Secondly, the methods to identify and treat the selected risk factors have developed, the treatment alternatives have increased, and the indications for treatment have changed. Furthermore, the awareness of one’s own risk factor status may have improved among the population. Thirdly, we wanted to focus on studies that were not included in earlier reviews.8,13,14 The following inclusion criteria were applied: at least one of the following conditions was included as an outcome: elevated blood pressure/hypertension, elevated blood glucose/diabetes or elevated total cholesterol (i.e. either a diagnosis or a measure that indicated an elevated risk of the outcome); self-reported data were compared with data from (i) HES and/or (ii) MRs or other administrative registers; prevalence of the outcome was reported; risk factor status was assessed among the same study subjects with the two methods; the study was performed among adults (18+); and the study was published in English. To compare population level estimates, we excluded studies in which persons with known hypertension, diabetes or elevated cholesterol were excluded. We also excluded studies where finger prick samples were used to analyze blood glucose or total cholesterol. Furthermore, studies using other lipid markers than total cholesterol such as low-density lipoprotein or triglycerides were not included. Altogether 17 studies that compared self-reported data with HES data and 13 studies that compared self-reported data with MRs were found (table 1). These studies were presented in 28 publications, as two publications included two studies: (i) in the publication by El Fakiri et al.15 SR was compared separately with both HES and MRs and is thus included in both categories and (ii) the two age groups in Navin Cristina et al.16 are treated as two separate studies because their results were presented separately in the publication. The publications comprised 8 studies from Europe,9,15,17–21 12 from the North America,22–33 1 from the South America,34 4 from Asia35–38 and 5 from Australia/New Zealand16,39–41 (references 41–49 in the Supplementary file 2). The study population covered general population in 23 studies and clinical population in 7 studies. Table 1 Description of studies and study populations First author and year  n (men/women)  Age (years)  Country  Population  Study  Hypertension/elevated blood pressure  Diabetes/elevated blood glucose  Elevated total cholesterol  SR vs. HES                  (a) Definition based on measurement and medication  Europe                  Molenaar et al. 200717  4950 (2221/2729)  18+  The Netherlands  General  Utrecht Health Project  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Huerta et al. 200918  1556 (719/837)  20+  Spain  General  Diabetes, Nutrition and Obesity study (DINO)  x  x  –  Tolonen et al. 20149  4127 (1816/2311)  25–64  12 European countries  General  EHES Pilot Project  x  x  x  North America                  Natarajan et al. 200222  8236 (3558/4678)  21+  USA  General  NHANES III (1988–94)  –  –  x  Ahluwalia et al. 200923  733 (0/733)  40–64  USA  Generala  WISEWOMAN  x  –  x  Dey et al. 201526  101 (49/52)  18–80  Canada  Patients with recent history of high-risk TIA and/or minor stroke  Cohort study at a Stroke Prevention Clinic and inpatient ward  –  x  –  Asia                  Goldman et al. 200335  1004 (585/419)  54+  Taiwan  General  Social Environment and Biomarkers of Aging Study  x  x  –  Chun et al. 201636  7270 (3096/4174)  50+  South Korea  General  KNHANES IV  x  x  x  South America                  Lima-Costa et al. 200434  970 (422/548)  18+  Brazil  General  Bambuí study  x  –  –  Australia/New Zealand                  Taylor et al. 201039  1525 (749/776)  18+  Australia  General  North West Adelaide Health Study (NWAHS)  x  –  x  (b) Definition based on measurement only  North America                  Cowie et al. 201024  13094 (NR)  20+  USA  General  NHANES (2003–2006)  –  x  –  Dave et al. 201325  16598 (5747/10087)  18+  USA  Generalb  Community Initiative to Eliminate Stroke  x  –  –  Fisher-Hoch et al. 201527  2838 (NR)  18+  USA  Generalc  Cameron County Hispanic Cohort  –  x  –  Asia                  Bao et al. 201538  7913 (2841/5072)  20–74  China  General  A population-based cross-sectional study in the city of Harbin  –  x  –  Ning et al. 201637  17708 (8479/9229)  45+  China  General  China Health and Retirement Longitudinal Study  x  x  –  Australia/New Zealand                  Peterson et al. 201640  7269 (3275/3994)  18+  Australia  General  Australian Health Survey  x  x  x  SR vs. MRs                  Europe                  Tormo et al. 200019  248 (44/204)  29–69  Spain  General  Subsample of Spanish EPIC cohort  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Englert et al. 201020  7640 (4271/3369)  18+  Germany  Hypercholesterolemia patients  Orbital study  x  x  –  Frost et al. 201221  600 (600/0)  60–74  Denmark  General  Cross-sectional study  x  x  –  North America                  Simpson et al. 200428  1002 (0/1002)  65+  USA  Disabled older women  Women’s Health and Aging Study I  –  x  –  Okura et al. 200429  2037 (981/1056)  45+  USA  General  Rochester Epidemiology Project  x  x  –  St Sauver et al. 200530  26162 (10192/15970)  20+  USA  General  Patients at Mayo Clinic  x  –  x  Muggah et al. 201331  85549 (38743/46806)  20+  Canada  General  Canadian community health survey  x  x  –  Leong et al. 201332  3322 (1555/1767)  20+  Canada  General  Quebec Statistical Institute (QSI) survey  –  x  –  Koller et al. 201433  3821 (NR)  18+  USA  Generald  Alaska EARTH study  x  x  x  Australia/New Zealand                  Teh et al. 201341  878 (395/483)  80–90  New Zealand  General  The Life and Living to Advanced Age: a Cohort Study in New Zealand (LiLACS NZ)  x  –  –  Navin Cristina et al. 201616  1002 (0/1002)  56–61  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Navin Cristina et al. 201616  1926 (0/1926)  82–87  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  First author and year  n (men/women)  Age (years)  Country  Population  Study  Hypertension/elevated blood pressure  Diabetes/elevated blood glucose  Elevated total cholesterol  SR vs. HES                  (a) Definition based on measurement and medication  Europe                  Molenaar et al. 200717  4950 (2221/2729)  18+  The Netherlands  General  Utrecht Health Project  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Huerta et al. 200918  1556 (719/837)  20+  Spain  General  Diabetes, Nutrition and Obesity study (DINO)  x  x  –  Tolonen et al. 20149  4127 (1816/2311)  25–64  12 European countries  General  EHES Pilot Project  x  x  x  North America                  Natarajan et al. 200222  8236 (3558/4678)  21+  USA  General  NHANES III (1988–94)  –  –  x  Ahluwalia et al. 200923  733 (0/733)  40–64  USA  Generala  WISEWOMAN  x  –  x  Dey et al. 201526  101 (49/52)  18–80  Canada  Patients with recent history of high-risk TIA and/or minor stroke  Cohort study at a Stroke Prevention Clinic and inpatient ward  –  x  –  Asia                  Goldman et al. 200335  1004 (585/419)  54+  Taiwan  General  Social Environment and Biomarkers of Aging Study  x  x  –  Chun et al. 201636  7270 (3096/4174)  50+  South Korea  General  KNHANES IV  x  x  x  South America                  Lima-Costa et al. 200434  970 (422/548)  18+  Brazil  General  Bambuí study  x  –  –  Australia/New Zealand                  Taylor et al. 201039  1525 (749/776)  18+  Australia  General  North West Adelaide Health Study (NWAHS)  x  –  x  (b) Definition based on measurement only  North America                  Cowie et al. 201024  13094 (NR)  20+  USA  General  NHANES (2003–2006)  –  x  –  Dave et al. 201325  16598 (5747/10087)  18+  USA  Generalb  Community Initiative to Eliminate Stroke  x  –  –  Fisher-Hoch et al. 201527  2838 (NR)  18+  USA  Generalc  Cameron County Hispanic Cohort  –  x  –  Asia                  Bao et al. 201538  7913 (2841/5072)  20–74  China  General  A population-based cross-sectional study in the city of Harbin  –  x  –  Ning et al. 201637  17708 (8479/9229)  45+  China  General  China Health and Retirement Longitudinal Study  x  x  –  Australia/New Zealand                  Peterson et al. 201640  7269 (3275/3994)  18+  Australia  General  Australian Health Survey  x  x  x  SR vs. MRs                  Europe                  Tormo et al. 200019  248 (44/204)  29–69  Spain  General  Subsample of Spanish EPIC cohort  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Englert et al. 201020  7640 (4271/3369)  18+  Germany  Hypercholesterolemia patients  Orbital study  x  x  –  Frost et al. 201221  600 (600/0)  60–74  Denmark  General  Cross-sectional study  x  x  –  North America                  Simpson et al. 200428  1002 (0/1002)  65+  USA  Disabled older women  Women’s Health and Aging Study I  –  x  –  Okura et al. 200429  2037 (981/1056)  45+  USA  General  Rochester Epidemiology Project  x  x  –  St Sauver et al. 200530  26162 (10192/15970)  20+  USA  General  Patients at Mayo Clinic  x  –  x  Muggah et al. 201331  85549 (38743/46806)  20+  Canada  General  Canadian community health survey  x  x  –  Leong et al. 201332  3322 (1555/1767)  20+  Canada  General  Quebec Statistical Institute (QSI) survey  –  x  –  Koller et al. 201433  3821 (NR)  18+  USA  Generald  Alaska EARTH study  x  x  x  Australia/New Zealand                  Teh et al. 201341  878 (395/483)  80–90  New Zealand  General  The Life and Living to Advanced Age: a Cohort Study in New Zealand (LiLACS NZ)  x  –  –  Navin Cristina et al. 201616  1002 (0/1002)  56–61  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Navin Cristina et al. 201616  1926 (0/1926)  82–87  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Note: NR, not reported; RCT, randomized clinical trial; CVD, cardiovascular disease. a Low-income women. b Persons of color, low-income, rural residency and persons for whom English was a second language. c Mexican American people. d American Indian and Alaska Native people. Table 1 Description of studies and study populations First author and year  n (men/women)  Age (years)  Country  Population  Study  Hypertension/elevated blood pressure  Diabetes/elevated blood glucose  Elevated total cholesterol  SR vs. HES                  (a) Definition based on measurement and medication  Europe                  Molenaar et al. 200717  4950 (2221/2729)  18+  The Netherlands  General  Utrecht Health Project  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Huerta et al. 200918  1556 (719/837)  20+  Spain  General  Diabetes, Nutrition and Obesity study (DINO)  x  x  –  Tolonen et al. 20149  4127 (1816/2311)  25–64  12 European countries  General  EHES Pilot Project  x  x  x  North America                  Natarajan et al. 200222  8236 (3558/4678)  21+  USA  General  NHANES III (1988–94)  –  –  x  Ahluwalia et al. 200923  733 (0/733)  40–64  USA  Generala  WISEWOMAN  x  –  x  Dey et al. 201526  101 (49/52)  18–80  Canada  Patients with recent history of high-risk TIA and/or minor stroke  Cohort study at a Stroke Prevention Clinic and inpatient ward  –  x  –  Asia                  Goldman et al. 200335  1004 (585/419)  54+  Taiwan  General  Social Environment and Biomarkers of Aging Study  x  x  –  Chun et al. 201636  7270 (3096/4174)  50+  South Korea  General  KNHANES IV  x  x  x  South America                  Lima-Costa et al. 200434  970 (422/548)  18+  Brazil  General  Bambuí study  x  –  –  Australia/New Zealand                  Taylor et al. 201039  1525 (749/776)  18+  Australia  General  North West Adelaide Health Study (NWAHS)  x  –  x  (b) Definition based on measurement only  North America                  Cowie et al. 201024  13094 (NR)  20+  USA  General  NHANES (2003–2006)  –  x  –  Dave et al. 201325  16598 (5747/10087)  18+  USA  Generalb  Community Initiative to Eliminate Stroke  x  –  –  Fisher-Hoch et al. 201527  2838 (NR)  18+  USA  Generalc  Cameron County Hispanic Cohort  –  x  –  Asia                  Bao et al. 201538  7913 (2841/5072)  20–74  China  General  A population-based cross-sectional study in the city of Harbin  –  x  –  Ning et al. 201637  17708 (8479/9229)  45+  China  General  China Health and Retirement Longitudinal Study  x  x  –  Australia/New Zealand                  Peterson et al. 201640  7269 (3275/3994)  18+  Australia  General  Australian Health Survey  x  x  x  SR vs. MRs                  Europe                  Tormo et al. 200019  248 (44/204)  29–69  Spain  General  Subsample of Spanish EPIC cohort  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Englert et al. 201020  7640 (4271/3369)  18+  Germany  Hypercholesterolemia patients  Orbital study  x  x  –  Frost et al. 201221  600 (600/0)  60–74  Denmark  General  Cross-sectional study  x  x  –  North America                  Simpson et al. 200428  1002 (0/1002)  65+  USA  Disabled older women  Women’s Health and Aging Study I  –  x  –  Okura et al. 200429  2037 (981/1056)  45+  USA  General  Rochester Epidemiology Project  x  x  –  St Sauver et al. 200530  26162 (10192/15970)  20+  USA  General  Patients at Mayo Clinic  x  –  x  Muggah et al. 201331  85549 (38743/46806)  20+  Canada  General  Canadian community health survey  x  x  –  Leong et al. 201332  3322 (1555/1767)  20+  Canada  General  Quebec Statistical Institute (QSI) survey  –  x  –  Koller et al. 201433  3821 (NR)  18+  USA  Generald  Alaska EARTH study  x  x  x  Australia/New Zealand                  Teh et al. 201341  878 (395/483)  80–90  New Zealand  General  The Life and Living to Advanced Age: a Cohort Study in New Zealand (LiLACS NZ)  x  –  –  Navin Cristina et al. 201616  1002 (0/1002)  56–61  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Navin Cristina et al. 201616  1926 (0/1926)  82–87  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  First author and year  n (men/women)  Age (years)  Country  Population  Study  Hypertension/elevated blood pressure  Diabetes/elevated blood glucose  Elevated total cholesterol  SR vs. HES                  (a) Definition based on measurement and medication  Europe                  Molenaar et al. 200717  4950 (2221/2729)  18+  The Netherlands  General  Utrecht Health Project  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Huerta et al. 200918  1556 (719/837)  20+  Spain  General  Diabetes, Nutrition and Obesity study (DINO)  x  x  –  Tolonen et al. 20149  4127 (1816/2311)  25–64  12 European countries  General  EHES Pilot Project  x  x  x  North America                  Natarajan et al. 200222  8236 (3558/4678)  21+  USA  General  NHANES III (1988–94)  –  –  x  Ahluwalia et al. 200923  733 (0/733)  40–64  USA  Generala  WISEWOMAN  x  –  x  Dey et al. 201526  101 (49/52)  18–80  Canada  Patients with recent history of high-risk TIA and/or minor stroke  Cohort study at a Stroke Prevention Clinic and inpatient ward  –  x  –  Asia                  Goldman et al. 200335  1004 (585/419)  54+  Taiwan  General  Social Environment and Biomarkers of Aging Study  x  x  –  Chun et al. 201636  7270 (3096/4174)  50+  South Korea  General  KNHANES IV  x  x  x  South America                  Lima-Costa et al. 200434  970 (422/548)  18+  Brazil  General  Bambuí study  x  –  –  Australia/New Zealand                  Taylor et al. 201039  1525 (749/776)  18+  Australia  General  North West Adelaide Health Study (NWAHS)  x  –  x  (b) Definition based on measurement only  North America                  Cowie et al. 201024  13094 (NR)  20+  USA  General  NHANES (2003–2006)  –  x  –  Dave et al. 201325  16598 (5747/10087)  18+  USA  Generalb  Community Initiative to Eliminate Stroke  x  –  –  Fisher-Hoch et al. 201527  2838 (NR)  18+  USA  Generalc  Cameron County Hispanic Cohort  –  x  –  Asia                  Bao et al. 201538  7913 (2841/5072)  20–74  China  General  A population-based cross-sectional study in the city of Harbin  –  x  –  Ning et al. 201637  17708 (8479/9229)  45+  China  General  China Health and Retirement Longitudinal Study  x  x  –  Australia/New Zealand                  Peterson et al. 201640  7269 (3275/3994)  18+  Australia  General  Australian Health Survey  x  x  x  SR vs. MRs                  Europe                  Tormo et al. 200019  248 (44/204)  29–69  Spain  General  Subsample of Spanish EPIC cohort  x  –  –  El Fakiri et al. 200715  430 (198/232)  30–70  The Netherlands  High risk for CVD  RCT on CVD  x  x  x  Englert et al. 201020  7640 (4271/3369)  18+  Germany  Hypercholesterolemia patients  Orbital study  x  x  –  Frost et al. 201221  600 (600/0)  60–74  Denmark  General  Cross-sectional study  x  x  –  North America                  Simpson et al. 200428  1002 (0/1002)  65+  USA  Disabled older women  Women’s Health and Aging Study I  –  x  –  Okura et al. 200429  2037 (981/1056)  45+  USA  General  Rochester Epidemiology Project  x  x  –  St Sauver et al. 200530  26162 (10192/15970)  20+  USA  General  Patients at Mayo Clinic  x  –  x  Muggah et al. 201331  85549 (38743/46806)  20+  Canada  General  Canadian community health survey  x  x  –  Leong et al. 201332  3322 (1555/1767)  20+  Canada  General  Quebec Statistical Institute (QSI) survey  –  x  –  Koller et al. 201433  3821 (NR)  18+  USA  Generald  Alaska EARTH study  x  x  x  Australia/New Zealand                  Teh et al. 201341  878 (395/483)  80–90  New Zealand  General  The Life and Living to Advanced Age: a Cohort Study in New Zealand (LiLACS NZ)  x  –  –  Navin Cristina et al. 201616  1002 (0/1002)  56–61  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Navin Cristina et al. 201616  1926 (0/1926)  82–87  Australia  Recent acute care episode  Australian Longitudinal Study on Women's Health (ALSWH)  x  x  –  Note: NR, not reported; RCT, randomized clinical trial; CVD, cardiovascular disease. a Low-income women. b Persons of color, low-income, rural residency and persons for whom English was a second language. c Mexican American people. d American Indian and Alaska Native people. Statistical methods Various statistical methods such as prevalence rates, κ coefficient, sensitivity and specificity and positive and negative predictive values were applied in describing the agreement between two compared methods. In this evaluation, we focus on prevalence rates. κ coefficients are also presented if available. Also 95% confidence intervals for prevalence rates and κ coefficients are reported whenever available. None of the original publications presented standard errors or standard deviations for prevalence rates. Results Comparability of data collection methods SR measures The self-reported data were collected by interviews or self-administered questionnaires. The format of questions varied between the studies (Supplementary file 1). Most studies reported the actual question whereas in some studies the questions were described less accurately. Typically, the self-reported data relied on questions like: Have you ever been told (by a doctor or other health professional) that you have high blood pressure/diabetes/high cholesterol? (tables 2–4, Supplementary file 1). In most studies, either current conditions or conditions that had ‘ever’ been observed were asked. In some studies, only diagnosed conditions or conditions that a health professional had ‘told’ about were enquired while in other studies diagnosis or health professionals were not specified. Table 2 The format of question used in SR, the definition of hypertension/elevated blood pressure used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and κ coefficients for agreement   SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of hypertension  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                Molenaar et al. 2007 (n = 4950)17  General practitioner or specialist  12 months  Systolic blood pressure ≥140 mmHg (<60y),  10.7 (9.8–11.6)  22.9 (21.7–24.1)  -12.2  NR  systolic blood pressure ≥160 mmHg (≥60y) or diastolic blood pressure≥90  or medication  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Systolic blood pressure ≥160 mmHg  52  59  -7  0.51 (0.43–0.59)  or diastolic blood pressure ≥95 mmHg  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Systolic blood pressure ≥140 mmHg  19.5 (17.6–21.6)  35.4 (33.0–37.8)  -15.9  0.51 (0.47–0.56)  or diastolic blood pressure ≥90 mmHg  or medication  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Systolic blood pressure ≥140 mmHg          or diastolic blood pressure ≥90 mmHg  or medication        Men  22.6  32.7  -10.1  NR        Women  25.4  21.9  +3.5  NR  North America                Ahluwalia et al. 2009 (n = 733)23  Doctor, nurse or other health professional  Ever  Systolic blood pressure ≥140 mmHg  49.6  56  -6.4  0.62 (0.56–0.67)  or diastolic blood pressure ≥90 mmHg  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Currently  Systolic blood pressure ≥140 mmHg  30.3 (27.5-33.2)  57.3 (54.3-60.4)  -27  0.41 (0.36–0.47)  or diastolic blood pressure ≥90 mmHg  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Systolic blood pressure ≥140 mmHgor diastolic blood pressure ≥90 mmHgor medication  36.4  48.8  -12.4  0.72 (NR)  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  Systolic blood pressure ≥140 mmHg  24  38.5  -14.5  0.57 (0.55–0.58)  or diastolic blood pressure ≥90 mmHg  or medication  South America                Lima-Costa et al. 2004 (n = 970)34  Doctor or other health professional  Ever  Systolic blood pressure ≥140 mmHg  27.2 (24.4-30.1)  23.3 (20.7-26.1)  +3.9  NR  or diastolic blood pressure ≥90 mmHg  or medication  Australia/New Zealand                Taylor et al. 2010 (n = 1525)39  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  15.8  30.6  -14.8  0.55 (NR)  or diastolic blood pressure ≥90 mmHg  or medication  (b) Definition based on measurement only  North America                Dave et al. 2013 (n = 16 598)25  Physician, doctor or nursea  Evera  Systolic blood pressure ≥140 mmHg  16.15  24.81  -8.66  0.25 (NR)  or diastolic blood pressure ≥90 mmHgb  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  17.4 (16.5-18.3)  23.9 (22.9-24.9)  -6.5  0.21 (0.18–0.23)  or diastolic blood pressure ≥90 mmHg  SR vs.MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                Tormo et al. 2000 (n = 248)19  Physician  Ever  Diagnosis of high blood pressure  27.4  26.6  +0.8  0.65 (0.53–0.76)  or medication  or attending a hypertension control programme run only for hypertensive people  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  ICPC (International Classification of Primary Care) code K86/K87  52  44  +8  0.63 (0.55–0.70)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  56  62  -6  0.69 (0.66–0.71)  Frost et al. 2012 (n = 600)21  Not defined  NR  Diagnosis of hypertension          or medication        Men  22.2  36.7  -14.5  NR  North America                Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of hypertension  35.8  37.7  -1.9  0.75 (0.72–0.78)  and medication  or the words ‘borderline’ or ‘labile’ used in reference to blood pressure with documentation of two blood pressure measurements (consecutively but may be subsequent visits) systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg within a 12-month period  St Sauver et al. 2005 (n = 26 162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with hypertension  23.5  26.8  -3.3  NR  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  One hospital admission with a hypertension diagnosis code or one OHIP record with a hypertension diagnosis code, followed within 2 years by another OHIP record or a hospital admission with a hypertension diagnosis code.  20.8  27.6  -6.8  0.66 (0.65–0.66)  Codes used: ICD-9: 401x, 402x, 403x, 404x, 405x (any type), ICD-10-CA: I10, I11, I12, I13, I15 (any type), OHIP diagnosis code: 401, 402, 403, 404, 405 (any type)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for hypertension  25.1  25.8  -0.7  0.62 (NR)  Australia/New Zealand                Teh et al. 2013 (n = 878)41  Doctor  Ever  Primary health care record: Diagnoses were either ascertained from READ codes, a standardized primary care coding system, from hospital discharge letter, or from reading through the medical records.  54.8  68.4  -13.6  0.44 (0.38–0.50)  Administrative hospitalization discharge diagnosis records: Diagnosis codes for hypertension ICD-10: I10, ICD-9: 401.0, 401.1, 401.9 were identified.  Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  32.5 (29.7-35.3)  12.8 (10.8-14.8)  +19.7  0.35 (0.29–0.41)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  57.8 (55.0-60.7)  38.2 (35.3-41.0)  +19.6  0.21 (0.15–0.26)    SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of hypertension  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                Molenaar et al. 2007 (n = 4950)17  General practitioner or specialist  12 months  Systolic blood pressure ≥140 mmHg (<60y),  10.7 (9.8–11.6)  22.9 (21.7–24.1)  -12.2  NR  systolic blood pressure ≥160 mmHg (≥60y) or diastolic blood pressure≥90  or medication  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Systolic blood pressure ≥160 mmHg  52  59  -7  0.51 (0.43–0.59)  or diastolic blood pressure ≥95 mmHg  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Systolic blood pressure ≥140 mmHg  19.5 (17.6–21.6)  35.4 (33.0–37.8)  -15.9  0.51 (0.47–0.56)  or diastolic blood pressure ≥90 mmHg  or medication  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Systolic blood pressure ≥140 mmHg          or diastolic blood pressure ≥90 mmHg  or medication        Men  22.6  32.7  -10.1  NR        Women  25.4  21.9  +3.5  NR  North America                Ahluwalia et al. 2009 (n = 733)23  Doctor, nurse or other health professional  Ever  Systolic blood pressure ≥140 mmHg  49.6  56  -6.4  0.62 (0.56–0.67)  or diastolic blood pressure ≥90 mmHg  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Currently  Systolic blood pressure ≥140 mmHg  30.3 (27.5-33.2)  57.3 (54.3-60.4)  -27  0.41 (0.36–0.47)  or diastolic blood pressure ≥90 mmHg  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Systolic blood pressure ≥140 mmHgor diastolic blood pressure ≥90 mmHgor medication  36.4  48.8  -12.4  0.72 (NR)  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  Systolic blood pressure ≥140 mmHg  24  38.5  -14.5  0.57 (0.55–0.58)  or diastolic blood pressure ≥90 mmHg  or medication  South America                Lima-Costa et al. 2004 (n = 970)34  Doctor or other health professional  Ever  Systolic blood pressure ≥140 mmHg  27.2 (24.4-30.1)  23.3 (20.7-26.1)  +3.9  NR  or diastolic blood pressure ≥90 mmHg  or medication  Australia/New Zealand                Taylor et al. 2010 (n = 1525)39  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  15.8  30.6  -14.8  0.55 (NR)  or diastolic blood pressure ≥90 mmHg  or medication  (b) Definition based on measurement only  North America                Dave et al. 2013 (n = 16 598)25  Physician, doctor or nursea  Evera  Systolic blood pressure ≥140 mmHg  16.15  24.81  -8.66  0.25 (NR)  or diastolic blood pressure ≥90 mmHgb  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  17.4 (16.5-18.3)  23.9 (22.9-24.9)  -6.5  0.21 (0.18–0.23)  or diastolic blood pressure ≥90 mmHg  SR vs.MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                Tormo et al. 2000 (n = 248)19  Physician  Ever  Diagnosis of high blood pressure  27.4  26.6  +0.8  0.65 (0.53–0.76)  or medication  or attending a hypertension control programme run only for hypertensive people  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  ICPC (International Classification of Primary Care) code K86/K87  52  44  +8  0.63 (0.55–0.70)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  56  62  -6  0.69 (0.66–0.71)  Frost et al. 2012 (n = 600)21  Not defined  NR  Diagnosis of hypertension          or medication        Men  22.2  36.7  -14.5  NR  North America                Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of hypertension  35.8  37.7  -1.9  0.75 (0.72–0.78)  and medication  or the words ‘borderline’ or ‘labile’ used in reference to blood pressure with documentation of two blood pressure measurements (consecutively but may be subsequent visits) systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg within a 12-month period  St Sauver et al. 2005 (n = 26 162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with hypertension  23.5  26.8  -3.3  NR  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  One hospital admission with a hypertension diagnosis code or one OHIP record with a hypertension diagnosis code, followed within 2 years by another OHIP record or a hospital admission with a hypertension diagnosis code.  20.8  27.6  -6.8  0.66 (0.65–0.66)  Codes used: ICD-9: 401x, 402x, 403x, 404x, 405x (any type), ICD-10-CA: I10, I11, I12, I13, I15 (any type), OHIP diagnosis code: 401, 402, 403, 404, 405 (any type)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for hypertension  25.1  25.8  -0.7  0.62 (NR)  Australia/New Zealand                Teh et al. 2013 (n = 878)41  Doctor  Ever  Primary health care record: Diagnoses were either ascertained from READ codes, a standardized primary care coding system, from hospital discharge letter, or from reading through the medical records.  54.8  68.4  -13.6  0.44 (0.38–0.50)  Administrative hospitalization discharge diagnosis records: Diagnosis codes for hypertension ICD-10: I10, ICD-9: 401.0, 401.1, 401.9 were identified.  Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  32.5 (29.7-35.3)  12.8 (10.8-14.8)  +19.7  0.35 (0.29–0.41)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  57.8 (55.0-60.7)  38.2 (35.3-41.0)  +19.6  0.21 (0.15–0.26)  Note: NR, not reported. a The question format was as follows: ‘Do you suffer from high blood pressure and/or has a physician/doctor/nurse diagnosed you as a hypertensive?’ b Persons who were taking blood pressure lowering medications were excluded. c Includes survey responses from the last survey (survey 5) only (‘case 1’ in the publication). Table 2 The format of question used in SR, the definition of hypertension/elevated blood pressure used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and κ coefficients for agreement   SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of hypertension  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                Molenaar et al. 2007 (n = 4950)17  General practitioner or specialist  12 months  Systolic blood pressure ≥140 mmHg (<60y),  10.7 (9.8–11.6)  22.9 (21.7–24.1)  -12.2  NR  systolic blood pressure ≥160 mmHg (≥60y) or diastolic blood pressure≥90  or medication  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Systolic blood pressure ≥160 mmHg  52  59  -7  0.51 (0.43–0.59)  or diastolic blood pressure ≥95 mmHg  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Systolic blood pressure ≥140 mmHg  19.5 (17.6–21.6)  35.4 (33.0–37.8)  -15.9  0.51 (0.47–0.56)  or diastolic blood pressure ≥90 mmHg  or medication  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Systolic blood pressure ≥140 mmHg          or diastolic blood pressure ≥90 mmHg  or medication        Men  22.6  32.7  -10.1  NR        Women  25.4  21.9  +3.5  NR  North America                Ahluwalia et al. 2009 (n = 733)23  Doctor, nurse or other health professional  Ever  Systolic blood pressure ≥140 mmHg  49.6  56  -6.4  0.62 (0.56–0.67)  or diastolic blood pressure ≥90 mmHg  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Currently  Systolic blood pressure ≥140 mmHg  30.3 (27.5-33.2)  57.3 (54.3-60.4)  -27  0.41 (0.36–0.47)  or diastolic blood pressure ≥90 mmHg  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Systolic blood pressure ≥140 mmHgor diastolic blood pressure ≥90 mmHgor medication  36.4  48.8  -12.4  0.72 (NR)  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  Systolic blood pressure ≥140 mmHg  24  38.5  -14.5  0.57 (0.55–0.58)  or diastolic blood pressure ≥90 mmHg  or medication  South America                Lima-Costa et al. 2004 (n = 970)34  Doctor or other health professional  Ever  Systolic blood pressure ≥140 mmHg  27.2 (24.4-30.1)  23.3 (20.7-26.1)  +3.9  NR  or diastolic blood pressure ≥90 mmHg  or medication  Australia/New Zealand                Taylor et al. 2010 (n = 1525)39  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  15.8  30.6  -14.8  0.55 (NR)  or diastolic blood pressure ≥90 mmHg  or medication  (b) Definition based on measurement only  North America                Dave et al. 2013 (n = 16 598)25  Physician, doctor or nursea  Evera  Systolic blood pressure ≥140 mmHg  16.15  24.81  -8.66  0.25 (NR)  or diastolic blood pressure ≥90 mmHgb  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  17.4 (16.5-18.3)  23.9 (22.9-24.9)  -6.5  0.21 (0.18–0.23)  or diastolic blood pressure ≥90 mmHg  SR vs.MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                Tormo et al. 2000 (n = 248)19  Physician  Ever  Diagnosis of high blood pressure  27.4  26.6  +0.8  0.65 (0.53–0.76)  or medication  or attending a hypertension control programme run only for hypertensive people  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  ICPC (International Classification of Primary Care) code K86/K87  52  44  +8  0.63 (0.55–0.70)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  56  62  -6  0.69 (0.66–0.71)  Frost et al. 2012 (n = 600)21  Not defined  NR  Diagnosis of hypertension          or medication        Men  22.2  36.7  -14.5  NR  North America                Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of hypertension  35.8  37.7  -1.9  0.75 (0.72–0.78)  and medication  or the words ‘borderline’ or ‘labile’ used in reference to blood pressure with documentation of two blood pressure measurements (consecutively but may be subsequent visits) systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg within a 12-month period  St Sauver et al. 2005 (n = 26 162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with hypertension  23.5  26.8  -3.3  NR  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  One hospital admission with a hypertension diagnosis code or one OHIP record with a hypertension diagnosis code, followed within 2 years by another OHIP record or a hospital admission with a hypertension diagnosis code.  20.8  27.6  -6.8  0.66 (0.65–0.66)  Codes used: ICD-9: 401x, 402x, 403x, 404x, 405x (any type), ICD-10-CA: I10, I11, I12, I13, I15 (any type), OHIP diagnosis code: 401, 402, 403, 404, 405 (any type)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for hypertension  25.1  25.8  -0.7  0.62 (NR)  Australia/New Zealand                Teh et al. 2013 (n = 878)41  Doctor  Ever  Primary health care record: Diagnoses were either ascertained from READ codes, a standardized primary care coding system, from hospital discharge letter, or from reading through the medical records.  54.8  68.4  -13.6  0.44 (0.38–0.50)  Administrative hospitalization discharge diagnosis records: Diagnosis codes for hypertension ICD-10: I10, ICD-9: 401.0, 401.1, 401.9 were identified.  Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  32.5 (29.7-35.3)  12.8 (10.8-14.8)  +19.7  0.35 (0.29–0.41)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  57.8 (55.0-60.7)  38.2 (35.3-41.0)  +19.6  0.21 (0.15–0.26)    SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of hypertension  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                Molenaar et al. 2007 (n = 4950)17  General practitioner or specialist  12 months  Systolic blood pressure ≥140 mmHg (<60y),  10.7 (9.8–11.6)  22.9 (21.7–24.1)  -12.2  NR  systolic blood pressure ≥160 mmHg (≥60y) or diastolic blood pressure≥90  or medication  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Systolic blood pressure ≥160 mmHg  52  59  -7  0.51 (0.43–0.59)  or diastolic blood pressure ≥95 mmHg  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Systolic blood pressure ≥140 mmHg  19.5 (17.6–21.6)  35.4 (33.0–37.8)  -15.9  0.51 (0.47–0.56)  or diastolic blood pressure ≥90 mmHg  or medication  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Systolic blood pressure ≥140 mmHg          or diastolic blood pressure ≥90 mmHg  or medication        Men  22.6  32.7  -10.1  NR        Women  25.4  21.9  +3.5  NR  North America                Ahluwalia et al. 2009 (n = 733)23  Doctor, nurse or other health professional  Ever  Systolic blood pressure ≥140 mmHg  49.6  56  -6.4  0.62 (0.56–0.67)  or diastolic blood pressure ≥90 mmHg  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Currently  Systolic blood pressure ≥140 mmHg  30.3 (27.5-33.2)  57.3 (54.3-60.4)  -27  0.41 (0.36–0.47)  or diastolic blood pressure ≥90 mmHg  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Systolic blood pressure ≥140 mmHgor diastolic blood pressure ≥90 mmHgor medication  36.4  48.8  -12.4  0.72 (NR)  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  Systolic blood pressure ≥140 mmHg  24  38.5  -14.5  0.57 (0.55–0.58)  or diastolic blood pressure ≥90 mmHg  or medication  South America                Lima-Costa et al. 2004 (n = 970)34  Doctor or other health professional  Ever  Systolic blood pressure ≥140 mmHg  27.2 (24.4-30.1)  23.3 (20.7-26.1)  +3.9  NR  or diastolic blood pressure ≥90 mmHg  or medication  Australia/New Zealand                Taylor et al. 2010 (n = 1525)39  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  15.8  30.6  -14.8  0.55 (NR)  or diastolic blood pressure ≥90 mmHg  or medication  (b) Definition based on measurement only  North America                Dave et al. 2013 (n = 16 598)25  Physician, doctor or nursea  Evera  Systolic blood pressure ≥140 mmHg  16.15  24.81  -8.66  0.25 (NR)  or diastolic blood pressure ≥90 mmHgb  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Currently  Systolic blood pressure ≥140 mmHg  17.4 (16.5-18.3)  23.9 (22.9-24.9)  -6.5  0.21 (0.18–0.23)  or diastolic blood pressure ≥90 mmHg  SR vs.MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                Tormo et al. 2000 (n = 248)19  Physician  Ever  Diagnosis of high blood pressure  27.4  26.6  +0.8  0.65 (0.53–0.76)  or medication  or attending a hypertension control programme run only for hypertensive people  El Fakiri et al. 2007 (n = 430)15  Not defined  Current  ICPC (International Classification of Primary Care) code K86/K87  52  44  +8  0.63 (0.55–0.70)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  56  62  -6  0.69 (0.66–0.71)  Frost et al. 2012 (n = 600)21  Not defined  NR  Diagnosis of hypertension          or medication        Men  22.2  36.7  -14.5  NR  North America                Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of hypertension  35.8  37.7  -1.9  0.75 (0.72–0.78)  and medication  or the words ‘borderline’ or ‘labile’ used in reference to blood pressure with documentation of two blood pressure measurements (consecutively but may be subsequent visits) systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg within a 12-month period  St Sauver et al. 2005 (n = 26 162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with hypertension  23.5  26.8  -3.3  NR  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  One hospital admission with a hypertension diagnosis code or one OHIP record with a hypertension diagnosis code, followed within 2 years by another OHIP record or a hospital admission with a hypertension diagnosis code.  20.8  27.6  -6.8  0.66 (0.65–0.66)  Codes used: ICD-9: 401x, 402x, 403x, 404x, 405x (any type), ICD-10-CA: I10, I11, I12, I13, I15 (any type), OHIP diagnosis code: 401, 402, 403, 404, 405 (any type)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for hypertension  25.1  25.8  -0.7  0.62 (NR)  Australia/New Zealand                Teh et al. 2013 (n = 878)41  Doctor  Ever  Primary health care record: Diagnoses were either ascertained from READ codes, a standardized primary care coding system, from hospital discharge letter, or from reading through the medical records.  54.8  68.4  -13.6  0.44 (0.38–0.50)  Administrative hospitalization discharge diagnosis records: Diagnosis codes for hypertension ICD-10: I10, ICD-9: 401.0, 401.1, 401.9 were identified.  Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  32.5 (29.7-35.3)  12.8 (10.8-14.8)  +19.7  0.35 (0.29–0.41)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) code I10                Womenc  57.8 (55.0-60.7)  38.2 (35.3-41.0)  +19.6  0.21 (0.15–0.26)  Note: NR, not reported. a The question format was as follows: ‘Do you suffer from high blood pressure and/or has a physician/doctor/nurse diagnosed you as a hypertensive?’ b Persons who were taking blood pressure lowering medications were excluded. c Includes survey responses from the last survey (survey 5) only (‘case 1’ in the publication). Table 3 The format of question used in SR, the definition of diabetes/elevated blood glucose used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and kappa coefficients for agreement   SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of diabetes or elevated blood glucose  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Fasting glucose ≥7.0 mmol/l  29  31  −2  0.76 (0.69–0.83)  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Fasting blood glucose ≥7.0 mmol/l  7.8 (6.5–9.3)  10.6 (9.1–12.3)  −2.8  0.78 (0.73–0.84)  or treatment (insulin, hypoglycemic drugs or diet)  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l          or HbA1C≥6.5%  or medication        Men  5.8  6.6  −0.8  NR        Women  4.8  3.9  +0.9  NR  North America                Dey et al. 2015 (n = 101)26  Physician  Ever  Fasting plasma glucose ≥6.9 mmol/l  23.8  26.7  −2.9  0.76 (0.61–0.91)  or HbA1C≥6.5%  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Current  HbA1c≥7.0%  14.6 (12.4–16.7)  15.5 (13.2–17.7)  −0.9  0.86 (0.79–0.92)  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l  13.6  15.4  −1.8  0.82 (NR)  or treatment  (b) Definition based on measurement only  North America                Cowie et al. 2010 (n = 13 094)24  Doctor or health care provider  Ever (other than during pregnancy)  HbA1c≥6.5%  7.8 (7.0–8.6)  9.6 (8.7–10.5)  −1.8  NR  Fisher-Hoch et al. 2015 (n = 2838)27  Health care provider  Ever  Fasting plasma glucose ≥7.0 mmol/l (or other criteria of the 2010 American Diabetes Association definition)  16.4  27.6  −11.2  NR  Asia                Bao et al. 2015 (n = 7913)38  Doctor  Ever (other than during pregnancy for women)  Fasting plasma glucose ≥7.0 mmol/l and/or 2-h post-load plasma glucose ≥11.1 mmol/l  4.4  12.7  −8.3  NR  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  HbA1c≥6.5%  5.8  6.9  −1.1  0.65 (0.62–0.68)  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Current  Fasting plasma glucose ≥7.0 mmol/l  6.1 (5.6–6.7)  4.5 (4.0–4.9)  +1.6  0.58 (0.54–0.62)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  International Classification of Primary Care code T90  29  29  0  0.84 (0.78–0.89)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  20  23  −3  0.89 (0.86–0.92)  Frost et al. 2012 (n = 600)21  Not defined  Not defined  Diagnosis of type II diabetes          or medication        Men  6.5  7.2  −0.7  NR  North America                Simpson et al. 2004 (n = 1002)28  Physician  Ever  ‘Based on standardized specific criteria using data from medical history, standardized research physical examination (e.g. electrocardiograms, ankle brachial index, spirometric testing), review of all medications, review of hospital records, x-rays, and a physician questionnaire.’  17  17  0  0.92 (0.86–0.98)  Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of diabetes mellitus  5.2  7.4  −2.2  0.76 (0.70–0.82)  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  Based on Ontario Health Insurance Plan diagnosis codes and DAD admissions. Different criteria for different age groups.  6.8  8.4  −1.6  0.80 (0.80–0.81)  Codes used: ICD-9: 250 (any type), ICD-10-CA: E10, E11, E13, E14 (any type), OHIP diagnosis code: 250, OHIP fee code: Q040, K029, K030  Leong et al. 2013 (n = 3322)32  Doctor or another health professional  Ever  Two or more physical billings for diabetes and/or one or more hospitalizations for diabetes  7.9  8.5  −0.6  0.79 (0.76–0.83)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for diabetes  5.1  6.5  −1.4  0.68 (NR)  Australia/New Zealand                Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  8.6 (6.9–10.3)  7.7 (6.1–9.3)  +0.9  0.75 (0.68–0.83)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  12.8 (10.8–14.7)  12.7 (10.7–14.6)  +0.1  0.77 (0.72–0.83)    SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of diabetes or elevated blood glucose  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Fasting glucose ≥7.0 mmol/l  29  31  −2  0.76 (0.69–0.83)  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Fasting blood glucose ≥7.0 mmol/l  7.8 (6.5–9.3)  10.6 (9.1–12.3)  −2.8  0.78 (0.73–0.84)  or treatment (insulin, hypoglycemic drugs or diet)  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l          or HbA1C≥6.5%  or medication        Men  5.8  6.6  −0.8  NR        Women  4.8  3.9  +0.9  NR  North America                Dey et al. 2015 (n = 101)26  Physician  Ever  Fasting plasma glucose ≥6.9 mmol/l  23.8  26.7  −2.9  0.76 (0.61–0.91)  or HbA1C≥6.5%  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Current  HbA1c≥7.0%  14.6 (12.4–16.7)  15.5 (13.2–17.7)  −0.9  0.86 (0.79–0.92)  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l  13.6  15.4  −1.8  0.82 (NR)  or treatment  (b) Definition based on measurement only  North America                Cowie et al. 2010 (n = 13 094)24  Doctor or health care provider  Ever (other than during pregnancy)  HbA1c≥6.5%  7.8 (7.0–8.6)  9.6 (8.7–10.5)  −1.8  NR  Fisher-Hoch et al. 2015 (n = 2838)27  Health care provider  Ever  Fasting plasma glucose ≥7.0 mmol/l (or other criteria of the 2010 American Diabetes Association definition)  16.4  27.6  −11.2  NR  Asia                Bao et al. 2015 (n = 7913)38  Doctor  Ever (other than during pregnancy for women)  Fasting plasma glucose ≥7.0 mmol/l and/or 2-h post-load plasma glucose ≥11.1 mmol/l  4.4  12.7  −8.3  NR  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  HbA1c≥6.5%  5.8  6.9  −1.1  0.65 (0.62–0.68)  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Current  Fasting plasma glucose ≥7.0 mmol/l  6.1 (5.6–6.7)  4.5 (4.0–4.9)  +1.6  0.58 (0.54–0.62)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  International Classification of Primary Care code T90  29  29  0  0.84 (0.78–0.89)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  20  23  −3  0.89 (0.86–0.92)  Frost et al. 2012 (n = 600)21  Not defined  Not defined  Diagnosis of type II diabetes          or medication        Men  6.5  7.2  −0.7  NR  North America                Simpson et al. 2004 (n = 1002)28  Physician  Ever  ‘Based on standardized specific criteria using data from medical history, standardized research physical examination (e.g. electrocardiograms, ankle brachial index, spirometric testing), review of all medications, review of hospital records, x-rays, and a physician questionnaire.’  17  17  0  0.92 (0.86–0.98)  Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of diabetes mellitus  5.2  7.4  −2.2  0.76 (0.70–0.82)  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  Based on Ontario Health Insurance Plan diagnosis codes and DAD admissions. Different criteria for different age groups.  6.8  8.4  −1.6  0.80 (0.80–0.81)  Codes used: ICD-9: 250 (any type), ICD-10-CA: E10, E11, E13, E14 (any type), OHIP diagnosis code: 250, OHIP fee code: Q040, K029, K030  Leong et al. 2013 (n = 3322)32  Doctor or another health professional  Ever  Two or more physical billings for diabetes and/or one or more hospitalizations for diabetes  7.9  8.5  −0.6  0.79 (0.76–0.83)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for diabetes  5.1  6.5  −1.4  0.68 (NR)  Australia/New Zealand                Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  8.6 (6.9–10.3)  7.7 (6.1–9.3)  +0.9  0.75 (0.68–0.83)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  12.8 (10.8–14.7)  12.7 (10.7–14.6)  +0.1  0.77 (0.72–0.83)  Note: NR, not reported. a Includes survey responses from the last survey (survey 5) only (‘case 1’ in the publication). Table 3 The format of question used in SR, the definition of diabetes/elevated blood glucose used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and kappa coefficients for agreement   SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of diabetes or elevated blood glucose  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Fasting glucose ≥7.0 mmol/l  29  31  −2  0.76 (0.69–0.83)  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Fasting blood glucose ≥7.0 mmol/l  7.8 (6.5–9.3)  10.6 (9.1–12.3)  −2.8  0.78 (0.73–0.84)  or treatment (insulin, hypoglycemic drugs or diet)  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l          or HbA1C≥6.5%  or medication        Men  5.8  6.6  −0.8  NR        Women  4.8  3.9  +0.9  NR  North America                Dey et al. 2015 (n = 101)26  Physician  Ever  Fasting plasma glucose ≥6.9 mmol/l  23.8  26.7  −2.9  0.76 (0.61–0.91)  or HbA1C≥6.5%  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Current  HbA1c≥7.0%  14.6 (12.4–16.7)  15.5 (13.2–17.7)  −0.9  0.86 (0.79–0.92)  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l  13.6  15.4  −1.8  0.82 (NR)  or treatment  (b) Definition based on measurement only  North America                Cowie et al. 2010 (n = 13 094)24  Doctor or health care provider  Ever (other than during pregnancy)  HbA1c≥6.5%  7.8 (7.0–8.6)  9.6 (8.7–10.5)  −1.8  NR  Fisher-Hoch et al. 2015 (n = 2838)27  Health care provider  Ever  Fasting plasma glucose ≥7.0 mmol/l (or other criteria of the 2010 American Diabetes Association definition)  16.4  27.6  −11.2  NR  Asia                Bao et al. 2015 (n = 7913)38  Doctor  Ever (other than during pregnancy for women)  Fasting plasma glucose ≥7.0 mmol/l and/or 2-h post-load plasma glucose ≥11.1 mmol/l  4.4  12.7  −8.3  NR  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  HbA1c≥6.5%  5.8  6.9  −1.1  0.65 (0.62–0.68)  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Current  Fasting plasma glucose ≥7.0 mmol/l  6.1 (5.6–6.7)  4.5 (4.0–4.9)  +1.6  0.58 (0.54–0.62)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  International Classification of Primary Care code T90  29  29  0  0.84 (0.78–0.89)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  20  23  −3  0.89 (0.86–0.92)  Frost et al. 2012 (n = 600)21  Not defined  Not defined  Diagnosis of type II diabetes          or medication        Men  6.5  7.2  −0.7  NR  North America                Simpson et al. 2004 (n = 1002)28  Physician  Ever  ‘Based on standardized specific criteria using data from medical history, standardized research physical examination (e.g. electrocardiograms, ankle brachial index, spirometric testing), review of all medications, review of hospital records, x-rays, and a physician questionnaire.’  17  17  0  0.92 (0.86–0.98)  Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of diabetes mellitus  5.2  7.4  −2.2  0.76 (0.70–0.82)  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  Based on Ontario Health Insurance Plan diagnosis codes and DAD admissions. Different criteria for different age groups.  6.8  8.4  −1.6  0.80 (0.80–0.81)  Codes used: ICD-9: 250 (any type), ICD-10-CA: E10, E11, E13, E14 (any type), OHIP diagnosis code: 250, OHIP fee code: Q040, K029, K030  Leong et al. 2013 (n = 3322)32  Doctor or another health professional  Ever  Two or more physical billings for diabetes and/or one or more hospitalizations for diabetes  7.9  8.5  −0.6  0.79 (0.76–0.83)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for diabetes  5.1  6.5  −1.4  0.68 (NR)  Australia/New Zealand                Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  8.6 (6.9–10.3)  7.7 (6.1–9.3)  +0.9  0.75 (0.68–0.83)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  12.8 (10.8–14.7)  12.7 (10.7–14.6)  +0.1  0.77 (0.72–0.83)    SR, the format of question   Objective data (HES or MR)             Diagnosed/told by  Time frame covered  Definition of diabetes or elevated blood glucose  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  Fasting glucose ≥7.0 mmol/l  29  31  −2  0.76 (0.69–0.83)  or medication  Huerta et al. 2009 (n = 1556)18  Not defined  Ever  Fasting blood glucose ≥7.0 mmol/l  7.8 (6.5–9.3)  10.6 (9.1–12.3)  −2.8  0.78 (0.73–0.84)  or treatment (insulin, hypoglycemic drugs or diet)  Tolonen et al. 2014 (n = 4127)9  Medical doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l          or HbA1C≥6.5%  or medication        Men  5.8  6.6  −0.8  NR        Women  4.8  3.9  +0.9  NR  North America                Dey et al. 2015 (n = 101)26  Physician  Ever  Fasting plasma glucose ≥6.9 mmol/l  23.8  26.7  −2.9  0.76 (0.61–0.91)  or HbA1C≥6.5%  or medication  Asia                Goldman et al. 2003 (n = 1004)35  Not defined  Current  HbA1c≥7.0%  14.6 (12.4–16.7)  15.5 (13.2–17.7)  −0.9  0.86 (0.79–0.92)  or medication  Chun et al. 2016 (n = 7270)36  Doctor  Ever  Fasting plasma glucose ≥7.0 mmol/l  13.6  15.4  −1.8  0.82 (NR)  or treatment  (b) Definition based on measurement only  North America                Cowie et al. 2010 (n = 13 094)24  Doctor or health care provider  Ever (other than during pregnancy)  HbA1c≥6.5%  7.8 (7.0–8.6)  9.6 (8.7–10.5)  −1.8  NR  Fisher-Hoch et al. 2015 (n = 2838)27  Health care provider  Ever  Fasting plasma glucose ≥7.0 mmol/l (or other criteria of the 2010 American Diabetes Association definition)  16.4  27.6  −11.2  NR  Asia                Bao et al. 2015 (n = 7913)38  Doctor  Ever (other than during pregnancy for women)  Fasting plasma glucose ≥7.0 mmol/l and/or 2-h post-load plasma glucose ≥11.1 mmol/l  4.4  12.7  −8.3  NR  Ning et al. 2016 (n = 17 708)37  Doctor  Ever  HbA1c≥6.5%  5.8  6.9  −1.1  0.65 (0.62–0.68)  Australia/New Zealand                Peterson et al. 2016 (n = 7269)40  Doctor or nurse  Current  Fasting plasma glucose ≥7.0 mmol/l  6.1 (5.6–6.7)  4.5 (4.0–4.9)  +1.6  0.58 (0.54–0.62)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n = 430)15  Not defined  Current  International Classification of Primary Care code T90  29  29  0  0.84 (0.78–0.89)  or medication  and/or search terms or registration codes specific for the primary health care center  Englert et al. 2010 (n = 7640)20  Not defined  Ever  ‘…coded according to the internationally agreed medical dictionary for regulatory activities (MedDRA 5.0) and converted into a yes/no format….’  20  23  −3  0.89 (0.86–0.92)  Frost et al. 2012 (n = 600)21  Not defined  Not defined  Diagnosis of type II diabetes          or medication        Men  6.5  7.2  −0.7  NR  North America                Simpson et al. 2004 (n = 1002)28  Physician  Ever  ‘Based on standardized specific criteria using data from medical history, standardized research physical examination (e.g. electrocardiograms, ankle brachial index, spirometric testing), review of all medications, review of hospital records, x-rays, and a physician questionnaire.’  17  17  0  0.92 (0.86–0.98)  Okura et al. 2004 (n = 2037)29  Medical provider  Ever  Diagnosis of diabetes mellitus  5.2  7.4  −2.2  0.76 (0.70–0.82)  Muggah et al. 2013 (n = 85 549)31  Not defined  Current  Based on Ontario Health Insurance Plan diagnosis codes and DAD admissions. Different criteria for different age groups.  6.8  8.4  −1.6  0.80 (0.80–0.81)  Codes used: ICD-9: 250 (any type), ICD-10-CA: E10, E11, E13, E14 (any type), OHIP diagnosis code: 250, OHIP fee code: Q040, K029, K030  Leong et al. 2013 (n = 3322)32  Doctor or another health professional  Ever  Two or more physical billings for diabetes and/or one or more hospitalizations for diabetes  7.9  8.5  −0.6  0.79 (0.76–0.83)  Koller et al. 2014 (n = 3821)33  Doctor or health care provider  Ever  ICD-9 codes for diabetes  5.1  6.5  −1.4  0.68 (NR)  Australia/New Zealand                Navin Cristina et al. 2016 (n = 1002)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  8.6 (6.9–10.3)  7.7 (6.1–9.3)  +0.9  0.75 (0.68–0.83)  Navin Cristina et al. 2016 (n = 1926)16  Not defined  Past 3 years  ICD-10-AM (Australian Modification) codes E10, E11, E13, E14                Womena  12.8 (10.8–14.7)  12.7 (10.7–14.6)  +0.1  0.77 (0.72–0.83)  Note: NR, not reported. a Includes survey responses from the last survey (survey 5) only (‘case 1’ in the publication). Table 4 The format of question used in SR, the definition of elevated total cholesterol used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and κ coefficients for agreement   SR, the format of question   Objective data (HES or MR)               Diagnosed/told by  Time frame covered  Definition of elevated total cholesterol  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  Total cholesterol ≥6.5 mmol/l  34  36  -2  0.55 (0.42–0.59)  or medication  Tolonen et al. 2014 (n =4127)9  Medical doctor  Ever  Total cholesterol ≥5.0 mmol/l          or medication        Men  21.5  71.1  -49.6  NR        Women  19  62.6  -43.6  NR  North America                Natarajan et al. 2002 (n =8236)22  Doctor or other health professional  Ever  Total cholesterol ≥5.17 mmol/l  32.1  59.4  -27.3  NR  or medication  Ahluwalia et al. 2009 (n =733)23  Doctor, nurse or other health professional  Ever  Total cholesterol ≥6.2 mmol/l  56  44.3  +11.7  0.51 (0.44–0.57)  or medication  Asia                Chun et al. 2016 (n =7270)36  Doctor  Ever  Total cholesterol ≥6.2 mmol/l  11.7  16.7  -5  0.48 (NR)  or medication  Australia/New Zealand  Taylor et al. 2010 (n =1525)39  Doctor or nurse  Currently (‘still’)  Total cholesterol ≥5.5 mmol/l  12.3  42.8  -30.5  0.30 (NR)  or medication  (b) Definition based on measurement only  Australia/New Zealand  Peterson et al. 2016 (n =7269)40  Doctor or nurse  Current  Total cholesterol ≥5.5 mmol/l  12.2 (11.5–13.0)  37.3 (36.2–38.4)  -25.1  -0.02 (-0.04–0.01)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  ICPC (International Classification of Primary Care) code T93  34  23  +11  0.48 (0.39–0.57)  or medication  and/or search terms or registration codes specific for the primary health care center  North America                St Sauver et al. 2005 (n =26162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with high cholesterol  22.8  27.2  -4.4  NR  Koller et a. 2014 (n =3821)33  Doctor or health care provider  Ever  ICD-9 codes for elevated cholesterol  17.3  18.5  -1.2  0.57 (NR)    SR, the format of question   Objective data (HES or MR)               Diagnosed/told by  Time frame covered  Definition of elevated total cholesterol  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  Total cholesterol ≥6.5 mmol/l  34  36  -2  0.55 (0.42–0.59)  or medication  Tolonen et al. 2014 (n =4127)9  Medical doctor  Ever  Total cholesterol ≥5.0 mmol/l          or medication        Men  21.5  71.1  -49.6  NR        Women  19  62.6  -43.6  NR  North America                Natarajan et al. 2002 (n =8236)22  Doctor or other health professional  Ever  Total cholesterol ≥5.17 mmol/l  32.1  59.4  -27.3  NR  or medication  Ahluwalia et al. 2009 (n =733)23  Doctor, nurse or other health professional  Ever  Total cholesterol ≥6.2 mmol/l  56  44.3  +11.7  0.51 (0.44–0.57)  or medication  Asia                Chun et al. 2016 (n =7270)36  Doctor  Ever  Total cholesterol ≥6.2 mmol/l  11.7  16.7  -5  0.48 (NR)  or medication  Australia/New Zealand  Taylor et al. 2010 (n =1525)39  Doctor or nurse  Currently (‘still’)  Total cholesterol ≥5.5 mmol/l  12.3  42.8  -30.5  0.30 (NR)  or medication  (b) Definition based on measurement only  Australia/New Zealand  Peterson et al. 2016 (n =7269)40  Doctor or nurse  Current  Total cholesterol ≥5.5 mmol/l  12.2 (11.5–13.0)  37.3 (36.2–38.4)  -25.1  -0.02 (-0.04–0.01)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  ICPC (International Classification of Primary Care) code T93  34  23  +11  0.48 (0.39–0.57)  or medication  and/or search terms or registration codes specific for the primary health care center  North America                St Sauver et al. 2005 (n =26162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with high cholesterol  22.8  27.2  -4.4  NR  Koller et a. 2014 (n =3821)33  Doctor or health care provider  Ever  ICD-9 codes for elevated cholesterol  17.3  18.5  -1.2  0.57 (NR)  Note: NR, not reported. Table 4 The format of question used in SR, the definition of elevated total cholesterol used for the HES or MRs data, prevalence rates from SR and HES or MR (%), the difference between the prevalence rates from SR and HES or MR (%) and κ coefficients for agreement   SR, the format of question   Objective data (HES or MR)               Diagnosed/told by  Time frame covered  Definition of elevated total cholesterol  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  Total cholesterol ≥6.5 mmol/l  34  36  -2  0.55 (0.42–0.59)  or medication  Tolonen et al. 2014 (n =4127)9  Medical doctor  Ever  Total cholesterol ≥5.0 mmol/l          or medication        Men  21.5  71.1  -49.6  NR        Women  19  62.6  -43.6  NR  North America                Natarajan et al. 2002 (n =8236)22  Doctor or other health professional  Ever  Total cholesterol ≥5.17 mmol/l  32.1  59.4  -27.3  NR  or medication  Ahluwalia et al. 2009 (n =733)23  Doctor, nurse or other health professional  Ever  Total cholesterol ≥6.2 mmol/l  56  44.3  +11.7  0.51 (0.44–0.57)  or medication  Asia                Chun et al. 2016 (n =7270)36  Doctor  Ever  Total cholesterol ≥6.2 mmol/l  11.7  16.7  -5  0.48 (NR)  or medication  Australia/New Zealand  Taylor et al. 2010 (n =1525)39  Doctor or nurse  Currently (‘still’)  Total cholesterol ≥5.5 mmol/l  12.3  42.8  -30.5  0.30 (NR)  or medication  (b) Definition based on measurement only  Australia/New Zealand  Peterson et al. 2016 (n =7269)40  Doctor or nurse  Current  Total cholesterol ≥5.5 mmol/l  12.2 (11.5–13.0)  37.3 (36.2–38.4)  -25.1  -0.02 (-0.04–0.01)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  ICPC (International Classification of Primary Care) code T93  34  23  +11  0.48 (0.39–0.57)  or medication  and/or search terms or registration codes specific for the primary health care center  North America                St Sauver et al. 2005 (n =26162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with high cholesterol  22.8  27.2  -4.4  NR  Koller et a. 2014 (n =3821)33  Doctor or health care provider  Ever  ICD-9 codes for elevated cholesterol  17.3  18.5  -1.2  0.57 (NR)    SR, the format of question   Objective data (HES or MR)               Diagnosed/told by  Time frame covered  Definition of elevated total cholesterol  SR (%) (95% CI)  HES (%) (95% CI)  SR-HES (%)  κ (95% CI)  SR vs. HES                (a) Definition based on measurement and medication  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  Total cholesterol ≥6.5 mmol/l  34  36  -2  0.55 (0.42–0.59)  or medication  Tolonen et al. 2014 (n =4127)9  Medical doctor  Ever  Total cholesterol ≥5.0 mmol/l          or medication        Men  21.5  71.1  -49.6  NR        Women  19  62.6  -43.6  NR  North America                Natarajan et al. 2002 (n =8236)22  Doctor or other health professional  Ever  Total cholesterol ≥5.17 mmol/l  32.1  59.4  -27.3  NR  or medication  Ahluwalia et al. 2009 (n =733)23  Doctor, nurse or other health professional  Ever  Total cholesterol ≥6.2 mmol/l  56  44.3  +11.7  0.51 (0.44–0.57)  or medication  Asia                Chun et al. 2016 (n =7270)36  Doctor  Ever  Total cholesterol ≥6.2 mmol/l  11.7  16.7  -5  0.48 (NR)  or medication  Australia/New Zealand  Taylor et al. 2010 (n =1525)39  Doctor or nurse  Currently (‘still’)  Total cholesterol ≥5.5 mmol/l  12.3  42.8  -30.5  0.30 (NR)  or medication  (b) Definition based on measurement only  Australia/New Zealand  Peterson et al. 2016 (n =7269)40  Doctor or nurse  Current  Total cholesterol ≥5.5 mmol/l  12.2 (11.5–13.0)  37.3 (36.2–38.4)  -25.1  -0.02 (-0.04–0.01)  SR vs. MR        SR (%) (95% CI)  MR (%) (95% CI)  SR-MR (%)  κ (95% CI)  Europe                El Fakiri et al. 2007 (n =430)15  Not defined  Current  ICPC (International Classification of Primary Care) code T93  34  23  +11  0.48 (0.39–0.57)  or medication  and/or search terms or registration codes specific for the primary health care center  North America                St Sauver et al. 2005 (n =26162)30  Not defined  Ever  HICDA-8 (Hospital Adaptation of the International Classification of Diseases, Eight Revision) codes associated with high cholesterol  22.8  27.2  -4.4  NR  Koller et a. 2014 (n =3821)33  Doctor or health care provider  Ever  ICD-9 codes for elevated cholesterol  17.3  18.5  -1.2  0.57 (NR)  Note: NR, not reported. Objective data HES measures In HES’s, blood pressure was measured twice or three times. In some studies, the average of two or three measurements was used in the analyses, whereas in others the result of only one measurement was used (Supplementary file 1). Diabetes or elevated blood glucose and elevated total cholesterol status were analyzed from blood samples. For elevated blood glucose, fasting plasma glucose (FPG) or HbA1c from non-fasting samples were used (table 3, Supplementary file 1). In one study, also glucose tolerance test was used.38 Total cholesterol was analyzed from fasting samples in four studies,15,36,39,40 both fasting and non-fasting samples in two studies9,22 and non-fasting samples in one study23 (Supplementary file 1). The definitions and cut-off values for the selected risk factors or diseases varied between the studies (see tables 2–4). In most studies, the use of medication for the condition was taken into account in the definition (measured value exceeded the cut-off value and/or medication was used for the condition). We focus on these studies. Studies based on measurements only are listed in the tables but not used in comparison. For studies that compared self-reported data with HES data the terms ‘hypertension’ and ‘diabetes’ are used when the definition in HES data was based on both measurements (blood pressure or fasting blood glucose/HbA1C) and medication for the condition (tables 2 and 3). We are aware that all respondents in these groups may not fulfil the clinical diagnostic criteria. When the definition was based on measurements only, the terms ‘elevated blood pressure’ and ‘elevated blood glucose’ are used. For cholesterol, the term ‘elevated total cholesterol’ is used in both cases (table 4). Data from MRs The data on MRs or other administrative registers typically relied on diagnostic codes and also the use of medication was taken into account in several studies (tables 2–4). For studies that compared self-reported data with MRs the terms ‘hypertension’, ‘diabetes’ and ‘elevated total cholesterol’ are used. Comparability of prevalence rates obtained from different data sources Self-reported data under-estimated the prevalence of hypertension and elevated total cholesterol in the majority of studies compared with HES data (tables 2 and 4). The difference between SR and MRs was less pronounced. For diabetes, under-estimate was negligible (table 3). Hypertension Self-reported data under-estimated hypertension prevalence in eight out of nine studies where the data for men and women were combined, and the reference data came from HESs (table 2). In a study that showed the prevalence rates for men and women separately, the self-reported hypertension prevalence rate was 10.1 percentage points lower among men and 3.5 percentage points higher among women compared with HES data.9 Among four studies where CI’s were presented, statistically significant under-estimation was observed in three studies. The difference between the prevalence rates by self-reported data and HES ranged from −15.9 to +3.9 percentage points including figures both for men and women combined and separately. The differences between self-reported data and MR data were less consistent (table 2). Hypertension was under-estimated by self-reported data in six out of eight studies that only showed the results for men and women combined. Under-estimation of hypertension was also seen in a study including men only.21 On the contrary, substantial and significant over-estimation of hypertension by SRs was observed in the two age groups in a study for older women.16 The difference between the hypertension prevalence rates by self-reporting and MRs ranged from −14.5 to +19.7 percentage points. The κ coefficients assessing the agreement between two methods ranged from 0.41 to 0.72 for self-reported data vs. HES and from 0.21 to 0.75 for self-reported data vs. MRs (table 2). Diabetes The self-reported data provided fairly similar prevalence rates as did HES and MRs (table 3). The difference in prevalence rates between self-reporting and HES ranged from −2.9 to +0.9 percentage points. When self-reported data were compared with MRs, the difference between the methods ranged from −3.0 to +0.9 percentage points. The κ coefficients ranged from 0.76 to 0.86 for self-reported data vs. HES and from 0.68 to 0.92 for self-reported data vs. MRs (table 3). Elevated total cholesterol Under-reporting of elevated cholesterol was observed in four out of five studies combining data for men and women and comparing self-reporting with HES data (table 4). Substantial under-estimation was observed both among men and women (−49.6 and −43.6 percentage points, respectively) in a study where the data for men and women were analyzed separately.9 The difference between the prevalence rates by self-reporting and HES ranged from −49.6 to +11.7 percentage points. Under-reporting was observed in two out of three studies, when data from self-reporting were compared with data from MRs. The difference between the prevalence rates by self-reporting and MRs ranged from −4.4 to +11 percentage points. The κ coefficients ranged from 0.30 to 0.55 for self-reported data vs. HES. The two κ coefficients reported in the selected publications for self-reported data vs. MRs were 0.48 and 0.57 (table 4). Discussion We compared self-reported data on hypertension, diabetes and elevated total cholesterol with two more objective data sources, HES and MRs or other register based data, to increase knowledge on which methods are most feasible, and produce reliable, accurate and comparable information for health monitoring purposes. Studies published in 2000–16 that compared self-reported data with either HES data or MRs were evaluated. Self-reported data tended to under-estimate prevalence rates especially for hypertension and elevated total cholesterol, which is in line with earlier reviews.8,13 The under-estimate was more pronounced, when the self-reported data were compared with HES data instead of MRs. Hypertension and elevated total cholesterol are typically asymptomatic for a long time and remain therefore easily unmeasured and undetected. Therefore, they may be under-estimated in MR data. In a study from the Netherlands, both HES and MRs were included as reference material.15 Higher hypertension prevalence was reported with HES than with MRs (59% and 44%, respectively), which also suggests that MRs do not cover all hypertensive subjects. Diabetes was more accurately self-reported both compared with HES data and MRs, which is also in line with earlier reviews.13 As the screening of blood glucose and the use risk tests for diabetes have become more common the awareness of the condition has improved. The κ coefficients were consistent with the comparisons based on prevalence rates in that the overall agreement between SR and HES or MRs was higher for diabetes than for hypertension or elevated cholesterol. In an earlier review, the following factors, among others, were reported to explain the inaccuracy of SR: respondents may lack the knowledge to accurately answer the questions posed, and poorly designed survey instruments could result in respondents not fully comprehending the questions.13 Participants’ awareness of the disease or risk factor levels is crucial to obtain accurate reporting in HIS studies.9 Unless a diagnosis is set or risk factors examined and thoroughly explained to the patient by medical professionals, the participants are not aware of them, as many conditions are asymptomatic. In general, women have been more often aware of their hypertension than men,42–44 which can be partly explained by regular follow-up during pregnancies. This applied also to the evaluated studies reporting hypertension separately for men and women.9,36,39 If SR is based on having ever been told to have elevated blood pressure or elevated blood glucose, pregnancy complications may lead to higher prevalence rates for women. This might explain a part of the slight overestimation of hypertension among women.9 In contrast, the substantial overestimation of hypertension among older women cannot be explained with conditions related to pregnancies as the question concerned the past 3 years only.16 Instead, the authors’ presumed hypertension to be underestimated in the hospital data. The awareness of one’s own conditions or risk factor levels may have increased over time in many populations. Measuring blood pressure with digital devices at home has become more common and cholesterol and blood glucose measurements may be available more widely than earlier and also outside medical services, e.g. at pharmacies and health clubs. Hypertension awareness has increased significantly among hypertensives43,44 and this development can be presumed to have continued. The comparability of studies included in this evaluation was hindered by many discrepancies in their methods. Self-reported data on risk factors or diseases were collected heterogeneously as varying questions and definitions were used in different studies and the time frame covered with the question varied. Both population level coverage of the service and coverage in recording diagnoses and measures limit comparability of data from MRs or other administrative registers. For HES data, the use of medication was not taken into account in the definition of the conditions in all studies. We decided to focus on studies where the definition was based on both measured values and the use of medication, although the issue of medication is not straightforward. Antihypertensive drugs and statins are also used for other indications than elevated blood pressure or cholesterol. Considering medication based on recorded product names may hence have led to overestimation of hypertension or elevated cholesterol if the indication was not specified. Furthermore, different cut-off values were used, especially for elevated total cholesterol. The cut-offs used for elevated total cholesterol ranged from 5.0 to 6.5 mmol/l. The lowest cut-off was applied in a study in which also the largest underestimate of elevated total cholesterol by SR was seen.9 In addition, results for two cut-offs were shown in one study.23 Using the lower cut-off (5.17 mmol/l) resulted in a conclusion that SR underestimated, whereas using the higher cut-off (6.2 mmol/l) resulted in a conclusion that SR overestimated elevated total cholesterol. This might be related to practicing clinicians using different thresholds for communicating elevated cholesterol levels.22 Medication was included in the definition only in the case of the higher cut-off, which is why we only included the results by the higher cut-off from the study in question in this evaluation. Masked or white coat hypertension may also affect the prevalence rates for elevated blood pressure, as in HESs blood pressure is measured on one occasion only and most commonly in an office at the examination site.45,46 Methodological differences, such as the format of questions, response rates and survey modes, among national surveys have hampered valid mutual comparisons,4,47,48 but recent efforts for standardization in European countries will improve the situation.6 However, rising non-participation rates cause problems and attention needs to be given to recruitment methods.7 While electronic patient records have improved the availability of data and provide information not available from survey data such as comorbidity and diagnostic and treatment details, they cannot fully substitute survey data. MRs may not always be up-to-date and there may be socio-economic differences in access to medical care.49 In clinical practice, screening for cardiovascular risk factors and control of diagnosed diseases may not be systematic. The measures rely on health professionals’ practice patterns and accuracy of recording.12 MRs may also lack information on diagnoses performed in different health care sectors (public vs. private sector), and background and health behaviours. There may also be limitations in availability of or access to MRs for research purposes. Conclusions All the data sources have their strengths and limitations, and none of them alone should be regarded as a gold standard. However, using only self-reported data in monitoring vascular diseases and their risk at the population level may lead to underestimation of the true prevalence especially in regards to hypertension and elevated total cholesterol. Whenever feasible, combined information from standardized interviews and measurements supplemented with register data to exploit the advantages of all of them might be used for public health monitoring. Supplementary data Supplementary data are available at EURPUB online. Funding This work was supported by the European Commission/DG SANTÉ (BRIDGE Health Project, grant number 664691). The views expressed here are those of the authors and they do not represent the Commission’s official position. Conflicts of interest: None declared Key points The prevalence rates of hypertension and elevated total cholesterol were under-estimated with self-reported data compared with data from HESs and to a smaller extent compared with medical records. Instead, all the three data sources resulted in quite similar diabetes prevalence rates. The methods and the wordings used in questions varied between the studies which hindered the comparison. Common standardized protocols and questions would facilitate the comparison between surveys. Combination of different data sources should be used in risk factor and disease monitoring whenever possible to increase the reliability and coverage of prevalence estimates. References 1 GBD 2015 Risk Factors Collaborators. Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet  2016; 388: 1659– 724. CrossRef Search ADS PubMed  2 World Health Organization. Global Action Plan for the Prevention and Control of Noncommunicable Diseases . Geneva: World Health Organization, 2013. 3 Kilpeläinen K, Tuomi-Nikula A, Thelen J. Health indicators in Europe: availability and data needs. Eur J Public Health  2012; 22: 716– 21. Google Scholar CrossRef Search ADS PubMed  4 Verschuuren M, Gissler M, Kilpeläinen K. Public health indicators for the EU: the joint action for ECHIM (European Community Health Indicators & Monitoring). Arch Public Health  2013; 71: 12. Google Scholar CrossRef Search ADS PubMed  5 Tolonen H, Koponen P, Mindell J, et al.   European Health Examination Survey—towards a sustainable monitoring system. Eur J Public Health  2014; 24: 338– 44. Google Scholar CrossRef Search ADS PubMed  6 Tolonen H, Koponen P, Naska A, et al.   Challenges in standardization of blood pressure measurement at the population level. BMC Med Res Methodol  2015; 15: 3. Google Scholar CrossRef Search ADS PubMed  7 Mindell JS, Giampaoli S, Goesswald A, et al.   Sample selection, recruitment and participation rates in health examination surveys in Europe—experience from seven national surveys. BMC Med Res Methodol  2015; 15: 4. Google Scholar CrossRef Search ADS PubMed  8 Gorber S, Tremblay M, Campbell N, Hardt J. The accuracy of self-reported hypertension: a systematic review and meta-analysis. Curr Hypertension Rev  2008; 4: 36– 62. Google Scholar CrossRef Search ADS   9 Tolonen H, Koponen P, Mindell JS, et al.   Under-estimation of obesity, hypertension and high cholesterol by self-reported data: comparison of self-reported information and objective measures from health examination surveys. Eur J Public Health  2014; 24: 941– 8. Google Scholar CrossRef Search ADS PubMed  10 Elo SL, Karlberg IH. Validity and utilization of epidemiological data: a study of ischaemic heart disease and coronary risk factors in a local population. Public Health  2009; 123: 52– 7. Google Scholar CrossRef Search ADS PubMed  11 Wild S, Fischbacher C, McKnight J. (on Behalf of the Scottish Diabetes Research Network Epidemiology Group). Using Large Diabetes Databases for Research. J Diabetes Sci Technol  2016; 10: 1073– 8. Google Scholar CrossRef Search ADS PubMed  12 Baus A, Hendryx M, Pollard C. Identifying patients with hypertension: a case for auditing electronic health record data. Perspect Health Inf Manag  2012; 9: 1e. Google Scholar PubMed  13 Newell SA, Girgis A, Sanson-Fisher RW, Savolainen NJ. The accuracy of self-reported health behaviors and risk factors relating to cancer and cardiovascular disease in the general population: a critical review. Am J Prev Med  1999; 17: 211– 29. Google Scholar CrossRef Search ADS PubMed  14 Saydah SH, Geiss LS, Tierney E, et al.   Review of the performance of methods to identify diabetes cases among vital statistics, administrative, and survey data. Ann Epidemiol  2004; 14: 507– 16. Google Scholar CrossRef Search ADS PubMed  15 El Fakiri F, Bruijnzeels MA, Hoes AW. No evidence for marked ethnic differences in accuracy of self-reported diabetes, hypertension, and hypercholesterolemia. J Clin Epidemiol  2007; 60: 1271– 9. Google Scholar CrossRef Search ADS PubMed  16 Navin Cristina TJ, Stewart Williams JA, Parkinson L, et al.   Identification of diabetes, heart disease, hypertension and stroke in mid- and older-aged women: comparing self-report and administrative hospital data records. Geriatr Gerontol Int  2016; 16: 95– 102. Google Scholar CrossRef Search ADS PubMed  17 Molenaar EA, Ameijden EJCV, Grobbee DE, Numans ME. Comparison of routine care self-reported and biometrical data on hypertension and diabetes: results of the Utrecht Health Project. Eur J Public Health  2007; 17: 199– 205. Google Scholar CrossRef Search ADS PubMed  18 Huerta JM, Tormo MJ, Egea-Caparros JM, et al.   Accuracy of self-reported diabetes, hypertension and hyperlipidemia in the adult Spanish population. DINO study findings. Rev Esp Cardiol  2009; 62: 143– 52. Google Scholar CrossRef Search ADS PubMed  19 Tormo MJ, Navarro C, Chirlaque MD, Barber X. Validation of self diagnosis of high blood pressure in a sample of the Spanish EPIC cohort: overall agreement and predictive values. EPIC Group of Spain. J Epidemiol Commun Health  2000; 54: 221– 6. Google Scholar CrossRef Search ADS   20 Englert H, Muller-Nordhorn J, Seewald S, et al.   Is patient self-report an adequate tool for monitoring cardiovascular conditions in patients with hypercholesterolemia? J Public Health (Oxford)  2010; 32: 387– 94. Google Scholar CrossRef Search ADS   21 Frost M, Wraae K, Gudex C, et al.   Chronic diseases in elderly men: underreporting and underdiagnosis. Age Ageing  2012; 41: 177– 83. Google Scholar CrossRef Search ADS PubMed  22 Natarajan S, Lipsitz SR, Nietert PJ. Self-report of high cholesterol: determinants of validity in U.S. adults. Am J Prev Med  2002; 23: 13– 21. Google Scholar CrossRef Search ADS PubMed  23 Ahluwalia IB, Tessaro I, Rye S, Parker L. Self-reported and clinical measurement of three chronic disease risks among low-income women in West Virginia. J Womens Health (Larchmt)  2009; 18: 1857– 62. Google Scholar CrossRef Search ADS PubMed  24 Cowie CC, Rust KF, Byrd-Holt DD, et al.   Prevalence of diabetes and high risk for diabetes using A1C criteria in the U.S. population in 1988-2006. Diabetes Care  2010; 33: 562– 8. Google Scholar CrossRef Search ADS PubMed  25 Dave GJ, Bibeau DL, Schulz MR, et al.   Predictors of congruency between self-reported hypertension status and measured blood pressure in the stroke belt. J Am Soc Hypertens  2013; 7: 370– 8. Google Scholar CrossRef Search ADS PubMed  26 Dey AK, Alyass A, Muir RT, et al.   Validity of self-report of cardiovascular risk factors in a population at high risk for stroke. J Stroke Cerebrovasc Dis  2015; 24: 2860– 5. Google Scholar CrossRef Search ADS PubMed  27 Fisher-Hoch SP, Vatcheva KP, Rahbar MH, McCormick JB. Undiagnosed diabetes and pre-diabetes in health disparities. PLoS One  2015; 10: e0133135. Google Scholar CrossRef Search ADS PubMed  28 Simpson CF, Boyd CM, Carlson MC, et al.   Agreement between self-report of disease diagnoses and medical record validation in disabled older women: factors that modify agreement. J Am Geriatr Soc  2004; 52: 123– 7. Google Scholar CrossRef Search ADS PubMed  29 Okura Y, Urban LH, Mahoney DW, et al.   Agreement between self-report questionnaires and medical record data was substantial for diabetes, hypertension, myocardial infarction and stroke but not for heart failure. J Clin Epidemiol  2004; 57: 1096– 103. Google Scholar CrossRef Search ADS PubMed  30 St Sauver JL, Hagen PT, Cha SS, et al.   Agreement between patient reports of cardiovascular disease and patient medical records. Mayo Clin Proc  2005; 80: 203– 10. Google Scholar CrossRef Search ADS PubMed  31 Muggah E, Graves E, Bennett C, Manuel DG. Ascertainment of chronic diseases using population health data: a comparison of health administrative data and patient self-report. BMC Public Health  2013; 13: 16. Google Scholar CrossRef Search ADS PubMed  32 Leong A, Dasgupta K, Chiasson JL, Rahme E. Estimating the population prevalence of diagnosed and undiagnosed diabetes. Diabetes Care  2013; 36: 3002– 8. Google Scholar CrossRef Search ADS PubMed  33 Koller KR, Wilson AS, Asay ED, et al.   Agreement between self-report and medical record prevalence of 16 chronic conditions in the Alaska EARTH Study. J Prim Care Community Health  2014; 5: 160– 5. Google Scholar CrossRef Search ADS PubMed  34 Lima-Costa MF, Peixoto SV, Firmo JO. Validity of self-reported hypertension and its determinants (the Bambui study). Rev Saude Publica  2004; 38: 637– 42. Google Scholar CrossRef Search ADS PubMed  35 Goldman N, Lin IF, Weinstein M, Lin YH. Evaluating the quality of self-reports of hypertension and diabetes. J Clin Epidemiol  2003; 56: 148– 54. Google Scholar CrossRef Search ADS PubMed  36 Chun H, Kim IH, Min KD. Accuracy of self-reported hypertension, diabetes, and hypercholesterolemia: analysis of a representative sample of Korean older adults. Osong Public Health Res Perspect  2016; 7: 108– 15. Google Scholar CrossRef Search ADS PubMed  37 Ning M, Zhang Q, Yang M. Comparison of self-reported and biomedical data on hypertension and diabetes: findings from the China Health and Retirement Longitudinal Study (CHARLS). BMJ Open  2016; 6: e009836. Google Scholar CrossRef Search ADS PubMed  38 Bao C, Zhang D, Sun B, et al.   Optimal cut-off points of fasting plasma glucose for two-step strategy in estimating prevalence and screening undiagnosed diabetes and pre-diabetes in Harbin, China. PLoS One  2015; 10: e0119510. Google Scholar CrossRef Search ADS PubMed  39 Taylor A, Dal Grande E, Gill T, et al.   Comparing self-reported and measured high blood pressure and high cholesterol status using data from a large representative cohort study. Aust N Z J Public Health  2010; 34: 394– 400. Google Scholar CrossRef Search ADS PubMed  40 Peterson KL, Jacobs JP, Allender S, et al.   Characterising the extent of misreporting of high blood pressure, high cholesterol, and diabetes using the Australian Health Survey. BMC Public Health  2016; 16: 695. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved.

Journal

The European Journal of Public HealthOxford University Press

Published: Feb 16, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off