Measuring disability: a systematic review of the validity and reliability of the Global Activity Limitations Indicator (GALI)

Measuring disability: a systematic review of the validity and reliability of the Global Activity... Background: GALI or Global Activity Limitation Indicator is a global survey instrument measuring participation restriction. GALI is the measure underlying the European indicator Healthy Life Years (HLY). Gali has a substantial policy use within the EU and its Member States. The objective of current paper is to bring together what is known from published manuscripts on the validity and the reliability of GALI. Methods: Following the PRISMA guidelines, two search strategies (PUBMED, Google Scholar) were combined to identify manuscripts published in English with publication date 2000 or beyond. Articles were classified as reliability studies, concurrent or predictive validity studies, in national or international populations. Results: Four cross-sectional studies (of which 2 international) studied how GALI relates to other health measures (concurrent validity). A dose-response effect by GALI severity level on the association with the other health status measures was observed in the national studies. The 2 international studies (SHARE, EHIS) concluded that the odds of reporting participation restriction was higher in subjects with self-reported or observed functional limitations. In SHARE, the size of the Odds Ratio’s (ORs) in the different countries was homogeneous, while in EHIS the size of the ORs varied more strongly. For the predictive validity, subjects were followed over time (4 studies of which one international). GALI proved, both in national and international data, to be a consistent predictor of future health outcomes both in terms of mortality and health care expenditure. As predictors of mortality, the two distinct health concepts, self-rated health and GALI, acted independently and complementary of each other. The one reliability study identified reported a sufficient reliability of GALI. Conclusion: GALI as inclusive one question instrument fits all conceptual characteristics specified for a global measure on participation restriction. In none of the studies, included in the review, there was evidence of a failing validity. The review shows that GALI has a good and sufficient concurrent and predictive validity, and reliability. Keywords: Disability, Participation restriction, Healthy life years, Validity, Reliability, Summary measure of population health, GALI Introduction reduction in the total lifetime days of disability, the Ageing of populations defies health and social policies. so-called compression of morbidity [1]. Population ill-health and especially disability are major The concept of disability is complex and multidimen- challenges as there is currently no consistent evidence sional. In initial medical models, disability was viewed as a that the lengthening of life expectancy goes with a problem residing solely in the persons affected. Disability referred to consequences of chronic or acute diseases or accidents on the functioning of specific body systems and on mental, physical and sensory functions in terms of (1) * Correspondence: herman.vanoyen@sciensano.be Department of Epidemiology and Public Health, Sciensano, J. Wytsmanstraat impairment or dysfunctions and structural abnormalities 14, 1050 Brussels, Belgium in specific body systems; (2) disability or restrictions in Department of Public Health, Ghent University, De Pintelaan 185, 9000 basic physical and mental actions and (3) handicaps or Ghent, Belgium Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Van Oyen et al. Archives of Public Health (2018) 76:25 Page 2 of 11 difficulties in doing activities of daily life [2–5]. More self-reported measure of participation restriction. Add- biophysical-social models introduce the person-environment itional conceptual criteria were the health relatedness of perspective of the disablement process: disability as the the cause of disability, the generic normative comparison outcome of the interaction of a person and his environ- in the level of participation, the long-standing duration of ment [6] and the dynamics of disability which is af- the disability (a duration of at least 6 months) and the fected by how a person’s capacity fits the environmental ability to measure levels of severity [15, 18]. To accommo- demand and results in participation [3]. Participation date the results of cognitive testing in relation to the se- restriction is defined as limitations in the performance verity options in the answer categories, the wording “to of roles and social involvement in different settings what extent” was added in the final version [20]: such as work and employment, school, leisure, parent- ing, housework, community, social and civic life [7]. Be- “For at least the past 6 months, to what extent have cause participation is influenced by environmental you been limited because of a health problem in factors and social norms, any measure of participation activities people usually do?” restriction cannot differentiate the impact of the im- pairment and functional limitations from the impact of Would you say you have been: severely limited, accommodations and enabling environments [8, 9]. Disability can occur in any human activity and set- limited but not severely, or tings. Adding to this complexity, instruments measuring disability differ in the domains of functioning included, not limited at all? in their goals to measure either capacity (without any personal or equipment assistance) or performance (with Being part of the MEHM, GALI is used in major Euro- assistance), or to measure also disability symptoms (pain, pean health and non-health surveys such as the Euro- weakness, endurance, …), levels of severity or the dur- pean Health Interview Survey (EHIS), Survey on Income ation of the disability. Traditional survey instruments and Living Conditions (SILC) and the Survey of Health, measure a limited number of tasks (5 to 7) in the do- Ageing and Retirement in Europe (SHARE). Since 2004, main of personal care (ADL (Activity of Daily Living)) or GALI is also the underlying measure of the European in- in the domain of household management (IADL (Instru- dicator “Healthy Life Years (HLY)”. HLY is a measure of mental ADL)). Other survey instruments have either in- disability free life expectancy and was presented in the creased the number of disability questions by adding set of structural indicators selected and defined to help more and more tasks [10] or have developed short set of measure progress in strategic European policies such as disability questions that have good coverage of activities the 2000 Lisbon strategy and the European 2020 strategy [11, 12]. At the same time there has been a quest to on Active and Healthy Ageing [21]. HLY is one of the measure disability with parsimony similar to the parsi- components of the Active Ageing Index [22]. At the na- mony in measuring health using the global one-item sur- tional level, countries such as France, have selected HLY vey instrument on self-rated health (SRH) [13, 14]. as one of their high level indicators for long term evalu- In response to the call for parsimony, a global survey in- ation of their economic, social and environmental pol- strument to measure disability, the Global Activity Limita- icies [23]. GALI also fits the requirement to follow-up tion Indicator or GALI, was proposed. The development European and United Nations disability policies, that of GALI occurred in the framework of the creation of a stress the importance of full and effective participation coherent set of indicators to monitor health across Europe as main policy outcome [24, 25]. More recently, GALI, [15]. GALI was part of set of 10 survey instruments in- as underlying health measure of HLY, contributes to the cluding three global one-item survey instruments were scoreboard indicators of the European Pillar of Social proposed covering distinct health concepts: perceived Rights [26] . Due to its high informational value, its rela- health, chronic morbidity and participation restriction tive simplicity and its compliance elaborated by the [16]. The 3 global questions define the Minimum Euro- European Union, HLY has been proposed to be the in- pean Health Module (MEHM) [17]. At the time of the de- strument in designing social security solutions [27]. velopment of the GALI, the beta version on the Given the use of GALI within the European Union, es- International classification of Functioning, Disability and pecially the fact that it is the measure underlying the Health (ICF) served as conceptual framework [6, 18, 19]. European indicator HLY, the objective of current paper Because of its implicit reference to the ability for societal is to bring together what is known from published man- participation in a variety of non-specified settings and uscripts on the validity and/or the reliability of GALI. non-specified domains of life (such as employment, Construct validity evaluation has been divided into school, housework, and leisure) using the wording “activ- translation validity, a more qualitative process and criter- ities people usually do”,GALI isintendedto beaglobal ion validity, a quantitative approach [28]. Current review Van Oyen et al. Archives of Public Health (2018) 76:25 Page 3 of 11 focus on the quantitative validity, as translation validity citations). The remaining 101 articles were screened to is linked to the conceptual criteria used for the GALI de- identify manuscripts published with the objective to esti- velopment [9, 15]. mate the validation or reliability of the GALI (N = 11). Double publications were excluded (N = 2). The double Methods publications were a result of the publication of institu- Two search strategies were combined by HVO and NB tional working papers or reports prior to the publication in January 2017 to identify peer reviewed manuscripts of a manuscript in a scientific journal. All 9 manuscripts published in English with publication date 2000 (the retained were found in both databases. Manuscripts period GALI was developed) or beyond. Following the were grouped as reliability studies, concurrent validity PRISMA guidelines, manuscripts were independently studies (cross-sectional studies measuring an associ- evaluated by HVO and NB first on the titles and ab- ation) or predictive validity studies (ability of GALI to stracts and in a second stage on the text. The result of predict an outcome) [28]. We further distinguished be- the search and manuscript selection is summarized in a tween national and international studies, as international PRISMA flow chart (Fig. 1)[29]. First, PUBMED data- studies may be more sensitive to total survey error due, base (https://www.ncbi.nlm.nih.gov/pubmed) was used for example, to the lack of international harmonization with {“global activity limitation”[All Fields] OR at different stages of the study [30]. (GALI[All Fields] NOT GALI[Author])} as search proto- col. Of the 69 publications 54 were excluded: content Results not related to the topic (e.g. Gali as part of a name of a The classification of the manuscripts by type of study is butterfly “Calisto franciscoi Gali”), language other than given in Table 1. English, Gali in name, email or contact address of one of the authors). Of the remaining 15 articles, 9 articles were Concurrent validity retained as manuscripts studying the validity or reliabil- Concurrent validation studies are cross-sectional studies ity of GALI. A second search used the Google Scholar with the objective to measure how GALI relates to other database (https://scholar.google.com/) using as search health measures. As there is no gold standard, the asso- protocol “global activity limitation”. Of the 208 refer- ciations are mainly measured using other health compo- ences 107 were excluded (content not related to the nents such as chronic (co)morbidity or other dimensions topic, language other than English, only an abstract, and of the disablement process, e.g. functional limitations in Fig. 1 PRISMA flow chart [29]: validation and reliability studies of the Global Activity Limitation Indicator (GALI) selection, 2000–2017 Van Oyen et al. Archives of Public Health (2018) 76:25 Page 4 of 11 � � � � � � � � � � � � Table 1 Studies estimating the validity or reliability of the Global Activity Limitation Indicator (GALI) by study design, period 2000–2017 Reference Study setting Health outcome used Severity Statistical measure Key findings and population in the comparison with level GALI the GALI Concurrent validity studies Van Oyen et al. National HIS* 2001 Self-reported: ADL*, SF-36* physical Yes Predicted probability distribution of The probability distribution of GALI by 2006 [32] Belgium Population domain score, number of self-reported GALI distribution, POR* from proportional severity level fits appropriately against aged 15+ N = 9168 chronic physical conditions out of a odds models, heterogeneity across indicators measuring mental and physical list of 29 conditions, number of mental demographic variables illness both in subjects with or without conditions (depression, anxiety, somatization, ADL limitations; sleep disorders) from the SCL-90R*, GHQ-12* 95% of subjects without ADL limitations scale for mental wellbeing and CMI* and no mental or physical health problems do not report participation restrictions; Subjects with ADL limitations report participation restrictions and the severity level of participation restriction is higher in function of the level of severity and the number ADL limitations and there is no evidence for heterogeneity across gender, age, education and language; The probability distribution of GALI by severity level is associated with the different physical and mental morbidity measures; A dose-response relationship is observed; The measures of associations are not as strong for mental health problems compared to physical health problems. Cabrero-Garcia National HIS 2006 Self-reported: physical and mental morbidity, Yes Spearman correlation, predicted probability GALI is primarily a measure of functional et al. 2014 [31] Spain Population functional disability (ADL, IADL and mobility) of GALI from fractional polynomial models, status and is secondarily associated with aged 65+ N = 7835 Concurrent comparison of the associations MOR* from multinomial logistical regression physical and mental morbidity whereas of GALI and the association of SRH* with the for SRH physical morbidity and to a lesser health outcomes: extend mental morbidity are the main FCI*, GHQ-12 scale for mental wellbeing, correlates; Functional disability* The odds of having (severe) participation restrictions increased with the level of functional limitations (number), the physical and mental morbidity, suggesting a dose- response relationship; Mental morbidity was as strong a correlate of GALI as of SRH, whereas physical morbidity waslessstrongacorrelateofGALI compared to SRH. Jagger et al. International SHARE* 2004 Measurement: maximum grip strength and No Predicted probability distribution of GALI, GALI effectively capture disability as measured 2010 [34] 11 EU countries Population walking speed (in subjects aged 75+) ORs from logistic regression models, by both the self-reported as objective measures aged 50+ N = 27340 Self-reported: ADL, IADL, and walking Random-effects meta-analysis to assess of functional limitations; limitations heterogeneity of associations between The likelihood of reporting participation countries restrictions increases as the severity of functional limitations increases in both the self-reported as objective measures of functional limitations; The likelihood of reporting no participation restriction in subjects with limitations is non-zero, though small and a minimum for the most severe measure, the ADL; Cross-country comparison did not provide any evidence for heterogeneity for the OR Van Oyen et al. Archives of Public Health (2018) 76:25 Page 5 of 11 � � � � � � � � � � � � � � � Table 1 Studies estimating the validity or reliability of the Global Activity Limitation Indicator (GALI) by study design, period 2000–2017 (Continued) Reference Study setting Health outcome used Severity Statistical measure Key findings and population in the comparison with level GALI the GALI of having participation restrictions in function of the self-reported ADL and the objective measures’; In all countries, the odds of having participation restrictions was higher in subjects with IADL limitations. The size of the effect was however more pronounced in some countries compared to others. Berger et al. International EHIS* Self-reported: ADL, IADL No Predicted probability distribution of GALI, GALI is significantly associated with ADL and 2015 [33] 2007–2010 14 EU and functional limitations ORs from logistic regression models, IADL limitations and functional limitations; countries Population Random-effects meta-analysis to assess The likelihood of reporting participation aged 15+ N = 152,796 heterogeneity of associations between restrictions increases as the number of ADL and countries IADL limitations and the severity of functional limitations increased; The likelihood of not reporting participation restrictions decreases as the number of ADL and IADL limitations and the severity of functional limitations increased; In all countries, the odds of having participation restrictions was higher in subjects with ADL, IADL and functional limitations. The size of the effect was more pronounced in some countries compared to others providing evidence for heterogeneity of the effect size. Predictive validity studies Berger et al. National HIS 2001 linked Mortality (follow-up to 10 years) Yes MRRs* from Poisson regression models; Compared to individuals without participation 2015 [36] with mortality and migration Comparison of relative predictive ability restrictions, subjects with moderate or severe database (National Register), of GALI compared to SRH* participation restriction have a 1.8 to 3.0 2001–2010 Belgium Population increased mortality rate over the 10 years aged 15+ N = 8583, 902 deaths of follow-up; The effect does not vary significantly by gender, education or age, except in subjects under age of 50 years; SRH and GALI are complementary predictors of mortality, with some indications of a stronger effect of SRH; The predictive effect of SRH and GALI slightly decrease over time. Van der Heyden National HIS 2008 linked with Mortality (follow-up to 2 years) No MRRs from Poisson regression models; Subjects with participation restriction have et al. 2015 [37] mortality within the Health Comparison of relative predictive ability a 2.4 increased mortality rate over the 2 Insurance database, 2008–2010 of GALI compared to SRH years of follow-up; Belgium Population aged The effect does not vary by gender; 65+ N = 1894, 178 deaths In men, SRH and GALI are complementary predictors of mortality, whereas in women this is only so for GALI. Van der Heyden National Health care expenditure Yes Linear regression after logistic transformation Moderate and severe participation restriction et al. 2015 [35] HIS 2008 linked with Health (Health insurance, out of costs; increases all health expenses by 3 to 6-times; Insurance database including of-pocket, supplement) Cost ratios were estimated to compare The increase is the more pronounced in the expenditure in 2008–2010 expenses to a reference; reimbursed health care expenditure; Belgium Population aged Decomposition of differences in expenses In absence of any chronic condition, moderate 15+ N = 7286 using the Blinder-Oaxaca method and severe participation restriction increases all health care expenditure by 2.5 to 4.5 times; Van Oyen et al. Archives of Public Health (2018) 76:25 Page 6 of 11 � � � � � � � � � Table 1 Studies estimating the validity or reliability of the Global Activity Limitation Indicator (GALI) by study design, period 2000–2017 (Continued) Reference Study setting Health outcome used Severity Statistical measure Key findings and population in the comparison with level GALI the GALI Chronic conditions explain only 22% of the differences in health care expenditure by level of participation restriction. Verropoulou International SHARE* 2004 Mortality (follow-up 2 to 3 years) Yes Hazard ratios from Cox proportional Both GALI and SRH are significant predictors et al. 2015 [38] with follow-up to re-interview hazard models; Comparison of relative of mortality in separate models; in wave 2006 /2007 11 EU predictive ability of GALI compared When adjusting for specific health indicators countries Population aged to SRH (asthma, cancer, depression, mobility, IADL, 50+ N = 17,941, 696 deaths orientation), GALI and SRH (only men) were significant but the magnitude diminished; GALI and SRH add information on top of specific health indicators; When GALI and SRH are included in one model, GALI was only significant in women, suggesting a partial conceptual overlap as there is a correlation between GALI and SRH; SRH and GALI represent different aspects of health. Reliability studies Cox et al. National Food Consumption Twice self-reported GALI Yes Pearson correlation coefficients, weighted Both Pearson (0.73) and Kappa coefficient (0.68) 2009 [39] Survey 1st and 2nd visit within time window Kappa coefficients indicate an acceptable reliability; Belgium, Population aged between 11 and 55 days Agreement is significantly higher for males 15+ N = 170 (Kappa = 0.82) compared to females (Kappa = 0.54); Agreements did not differ by education level, age, time span and language (French, Dutch). *HIS Health interview survey ADL Activities of Daily Living SF-36 Short Form Survey SCL-90R Symptoms Check List GHQ-12 General Health Questionnaire CMI Composite Morbidity Indicator: no illness, only mental illness, only physical illness and both mental and physical illness POR Proportional Odds Ratios MOR Multinomial Odds Ratios SRH Self-Rated Health FCI Functional Comorbidity Index based on a list of 16 chronic conditions including obesity, hearing and visual impairments Functional disability: based on a 27 items related to I/ADL and mobility Washington group instrument: ref. = 32,350 SHARE Survey of Health and Retirement in Europe 2004 survey was done in Austria, Belgium, Denmark, France, Germany, Greece, Italy, the Netherlands, Spain, Sweden, Switzerland EHIS European Health interview survey 2007–2010 surveys were done in Belgium, Bulgaria, Cyprus, Czech Republic, France, Greece, Hungary, Latvia, Malta, Poland, Romania, Slovakia, Slovenia, Spain MRR Mortality Rate Ratio Van Oyen et al. Archives of Public Health (2018) 76:25 Page 7 of 11 activities. Two studies were national [31, 32] and 2 were level. E.g., the predicted probability of no participation re- international [33, 34]. In one of the international studies, strictions in subjects without ADL limitations and without GALI could be evaluated against the results of objective any mental or physical illness was 0.95 and dropped to re- measures of functional limitations [34]. The age groups spectively 0.90, 0.80 and 0.57 in subjects reporting only included in the studies varied between subjects 15 years mental illness, only physical illness or both; in people with and older, 50 years and older or 65 years and older. One ADL limitations the predicted probabilities of no participa- of the international studies focused on the population tion restriction were respectively 0.46, 0.43, 0.37 and 0.12. 15 years and older but provided, for comparison pur- The second concurrent validation study used the 2006 poses, tables and graphs for the population 50 years and Spanish National Health Survey but included only sub- older as supplementary material [33]. The national stud- jects 65 years and older to test (1) if GALI is primarily ies considered GALI by severity level, while the inter- correlated with functional disability and secondarily with national studies ignored the severity level. morbidity, and (2) if Self Rated Health (SRH), in con- Van Oyen et al. used the 2001 Belgian Health Interview trast, is primarily correlated with morbidity and second- survey [32] to evaluate GALI against (1) Activities of Daily arily with functional disability [31]. Associations were Living (ADL) limitations (categorical by level of severity, sought with a functional comorbidity indicator (FCI) and by number of limitations); (2) the Short Form Survey based on a list of 16 chronic conditions including obes- (SF-36) physical domain score; (3) the number of ity, hearing and visual impairments, the GHQ-12 for self-reported chronic physical conditions out of a list of 29; mental ill-health and a functional disability measure (4) the number of mental conditions (depression, anxiety, (based on 27 items related to IADL/ADL and mobility). somatization, sleep disorders) based on 4 subscales of the The Spearman correlation coefficients of FCI, GHQ-12 Symptoms Check List (SCL-90R) and (5) a mental and functional disability were 0.35, 0.45 and 0.58 with well-being score using the General Health Questionnaire GALI compared to 0.46, 0.44 and 0.36 with SRH. The (GHQ-12). A composite morbidity indicator (CMI: catego- predicted probability of participation restriction indi- rized as no illness, only mental, only physical illness and cated a greater effect in function of the number of func- both) was used to measure the associations stratified by the tional disabilities compared to the comorbidity indicator ADL functional limitation status. The results indicated that while the inverse was observed for the predictive prob- all health indicators were positively associated with GALI. ability of not being in very good/good SRH. The pre- The participation restriction distribution by severity level dicted probabilities for GALI and SRH were similar in was positively associated with both the number and severity function of the GHQ-12. Compared to subjects with no of ADL limitations, the SF-36 physical domain score, the functional disability (ADL/IADL or mobility), the multi- number of chronic conditions and the mental health indi- nomial odds ratios (MORs) of participation restriction cators (mental health comorbidity score or the GHQ-12). and severe participation restriction were respectively E.g., without any ADL limitations the predicted probability 1.44 and 2.02 in subjects with one functional disability distribution of being without, with mild or with severe par- (ADL/IADL or mobility) and respectively 8.94 and 64.84 ticipation restriction was respectively 0.82, 0.15, 0.03 com- when limitations in 11 functions were reported. The pared to 0.20, 0.43, 0.37 and 0.13, 0.38, 0.49 in people with MORs for participation restriction and severe participa- at least one ADL limitation or with at least one severe ADL tion restriction were respectively 1.96 and 2.00 for sub- limitation. When people were limited in 6 ADLs, the GALI jects with a FCI score of 1 and respectively 7.49 and 7.96 probabilities of reporting no, mild and severe restrictions for people with a FCI score of 7. The MORs of having were respectively 0.03, 0.10 and 0.87. Using the GHQ-12 participation restriction and severe participation restric- mental well-being score, the predicted probabilities of no, tion in people with a GHQ-12 score of 1 and of 7 were mild and severe GALI restrictions changed from respect- respectively 1.32 and 1.61; and 3.42 and 8.05. The MORs ively 0.82, 0.14, 0.04 (best GHQ-12 score) to 0.36 0.36 0.28 of fair SRH and very poor/poor SRH indicated a similar (worst GHQ-12 score). The CMI was associated with par- pattern but were more extreme in function of the func- ticipation restriction both in people with and without ADL tional comorbidity score while less extreme in function limitations. In the population free of ADL limitations, and of the functional disability indicator. These results sug- compared to subjects reporting non illness, the participa- gest that GALI was primarily a measure of functional tion restriction prevalence and especially the prevalence of status and secondarily a measure of physical and mental being severely restricted increased gradually in people morbidity whereas for SRH, physical morbidity and to a reporting only mental illness, reporting only physical illness lesser extend mental morbidity were the main correlates. or reporting both mental and physical illness. A similar The two concurrent international studies [33, 34] trend was observed in subjects with ADL limitations but followed a similar statistical analysis plan using the data the prevalence of participation restriction and severe re- from SHARE and EHIS: the estimation of the predicted striction was substantially higher within each morbidity GALI probability distribution by fitting logistic regression Van Oyen et al. Archives of Public Health (2018) 76:25 Page 8 of 11 and random-effects meta-analysis models to evaluate het- (HIS-2001 [36], HIS-2008 [37]) while the other used the erogeneity of the association between countries (Table 1). international SHARE study, wave 2004 and wave The 2004 SHARE study covered 11 EU countries [34], 2006/2007 [38]. The duration of mortality follow-up ranged while the EHIS (2007–2010) used data from 14 EU coun- from 2 to 3 years [37, 38]to10years [36]. Two of the 3 tries [33]. Main differences are related to the age groups in- mortality studies used GALI by severity level [36, 38]. The cluded, the use of observed measures and the stronger age considered ranged from 15 years and older [36]to overall survey design homogeneity between countries in 50 years and older [38] and 65 years-plus [37]. In all 3 mor- the SHARE survey compared to the EHIS survey design. In tality studies, the predictive capacity of GALI was set off both studies, the severity level information of GALI was ig- against the predictive capacity of SRH. The 3 studies identi- nored in the analysis. The GALI was evaluated against the fied GALI and SRH as complementary predictors of mor- number of ADL and IADL limitations. The EHIS data also tality, indicating that GALI and SRH share some traits but provided an evaluation against a function limitation meas- add different dimensions: health and disability. In the two urebyseveritybased on theself-reported moderateor se- studies with a short follow-up period [37, 38], adjusting for vere problems in at least one of the following 6 functions: age, education and life style (SHARE only: physical activity, walking a certain distance, going up and down the stairs, smoking, BMI) both GALI and SRH were significant pre- carrying in the hands or arms, using hands and fingers to dictors of mortality: having participation restrictions dou- manipulate small objects, biting and chewing. The objective bled (mild) and tripled (severe) the mortality rate in instruments to evaluate the validity of the GALI in SHARE SHARE, while the mortality rate increased by a factor of measured the hand grip strength and the walking speed (in 2.4 when no severity level was accounted for in the Belgian 75 years or older only; reported walking limitations in those study [37]. In the Belgian study [37], when both GALI and 50–74 years). In the SHARE study, the probability of the re- SRH were included in the Poisson regression model, GALI ported participation restriction was lower when the grip remained a significant predictor of mortality next to SRH strength or walking speed were higher. Similarly, and in in both males and females, while SRH remained only sig- both studies, the probability of having participation restric- nificant in males. In the SHARE follow-up [38], the tion was higher as the number of ADL or IADL limitations fully-adjusted proportional hazard model, controlling for increasedorifthe levelofseverity ofthose limitationsin- specific morbidity indicators (asthma, cancer, depression) creased (EHIS only). In all countries and in both studies, and disability measures (mobility, IADL), GALI and SRH, the direction of the association, measured as ORs, between showed that SRH remained significantly associated with GALI and the other health measures was as expected. That mortality only in men while GALI remained significantly is: the odds of reporting participation restriction is higher associated with mortality only in women. Over a 10-year in subjects with poorer functioning and disability measures follow-up period [36], GALI as well as SRH were strong (either in function of the ADL, IADL, the physical func- predictors of mortality. Adjusting for age, gender and tional limitations or in function of low hand grip strength socio-economic position, people with mild and severe par- or walking speed). In the population 50 years and over, the ticipation restriction, compared to no participation restric- overall random effect meta-regression ORs were less ex- tion, had mortality rate ratio’s(MRR’s) of respectively 1.8 treme in the SHARE study compared to the EHIS: e.g. and 3.0. Compared to good/very good SRH the MRR in comparing subjects with at least one ADL-limitation vs. subjects with fair and bad/very bad SRH the MRR was re- those with none, the combined OR of having participation spectively 1.8 and 3.6. When including GALI and SRH in restriction were 8.3 in SHARE and 12.3 in EHIS; with re- the model, both remained significant predictors of mortal- spect to IADL limitations, the combined ORs were respect- ity: MRR for GALI were: 1.4 (mild) and 1.8 (severe); and ively 6.4 and 9.1. In the SHARE study, there was no MRR for SRH were: 1.5 (fair) and 2.5 (bad/very bad). The significant country variation in the ORs in function of predictive ability did not change with gender or ADL, hand grip strength and walking speed. For IADL in socio-economic position. However, in older subjects, the the SHARE and for all functions and disability measures in predictive ability of SRH was not as strong. The impact on the EHIS, the OR of having participation restriction was mortality of both GALI and SRH decreased over time but more pronounced in some countries compared to others, remained statistical significant in truncated follow-up pe- providing evidence of heterogeneity in the effect size. riods: 0–3 years, 3–6 years and 6–10 years. Van der Heyden et al. evaluated how GALI predicted Predictive validity health care expenditure using data linkage between each The predictive validity of GALI was reported by 4 studies participant to the 2008 Belgian Health Interview Survey (Table 1). One study evaluated the predictive effect on and the national health insurance data for the 12 months health care expenditure in Belgium [35]; the other 3 stud- following the date of the interview [35]. Participation re- ies focused on mortality. Two of the mortality follow-up striction was a strong determinant of the total health studies used Belgian Health Interview Surveys as baseline care expenditure: e.g. the population with participation Van Oyen et al. Archives of Public Health (2018) 76:25 Page 9 of 11 restriction (21% of the population) accounted for 49% of two distinct approaches. One option is to include more ac- the total health expenditure; for severe restriction (5% of tivities, more specific answer categories, more aspects such the population) this was 17% of the total expenditure. as disability symptoms and disability in more specific set- The association was stronger for the reimbursed health tings of life [10, 42, 43]. This option induces increased re- care cost compared to the out-of-pocket payments. In spondents burden, increased survey cost, more complex subjects with no chronic conditions compared to people analysis in order to provide condensed indicators for end without participation restriction, the cost ratio of the re- users. The other option seeks short sets and/or a one single imbursed cost in subjects with mild or severe participa- global instrument [11, 15]. GALI has been nominated the tion restriction was respectively 2.5 and 4.2. In people champion in parsimony [12], but the lack of simplicity and with one chronic condition or in people with ≥2 chronic the high density of concepts in one single question may conditions the cost ratio compared to no participation hamper its acceptability [9, 44]. Three different alternatives restriction were respectively 1.5 and 1.7 in subjects with (decomposing GALI using filtered and routed questions or mild participation restriction and 2.4 and 3.2 in people through omitting features such as the duration of the dis- with severe participation restrictions. The authors ability and/or the health relatedness) and the original GALI decomposed the health expenditure gap between people were evaluated against the short version (including 4 func- with and without participation restriction: differences in tional limitation questions) of the Washington Group on the age distribution (20%) and in the prevalence of Disability instrument [11]. The four variants were randomly chronic diseases (22%) between the two groups were the assigned to survey participants (N = 3009). The results, indi- main contributors to the explained differences (48%). cating a substantial higher sensitivity of GALI, no evidence Next to the confounding effect of age, the decompos- for a better understanding of the simplified alternatives but ition analysis also indicated that the impact of age on possibly, a small advantage in specificity when bringing the health expenditure differed by GALI severity level, sug- duration of the participation restriction to subsequent ques- gesting an interaction effect of age. However, in the un- tions, should be balanced against the cost of breaking an explained component, the coefficient of chronic established chronological series [45]. Currently, Eurostat fol- conditions did not differ significantly between GALI cat- lows for the upcoming EHIS wave III (2018–2019), the rec- egories, suggesting that chronic conditions in people ommendation of the EHLEIS working group on the with participation restrictions do not result in significant blueprint for an internationally harmonized Summary Meas- different health care expenditure compared to people ure of Population Health [9]. The EHLEIS working group, without participation restrictions. including experts from the EU, Japan, USA, OECD and WHO met 3 times (2012, 2013 and 2014) in Paris and pro- Reliability posed that of the different components of disability, partici- Only one study evaluated the reliability of GALI as part pation restriction in the first place and, in addition, of the evaluation of the MEHM [39]. The study used the functional limitations should be the main goals for inter- Belgian 2004 Food Consumption Survey in which people nationally harmonized global measures [9]. GALI fits the six were visited twice at home by the interviewers. The conceptual characteristics specified by the working group: 1. interquartile range between the 2 visits was 17–26 days comprehensive content of participation; 2. measure of with median time of 20 days. Both the Pearson correl- participation performance with current accommoda- ation (0.73) and the weighted Kappa coefficient (0.68) in- tion; 3. health relatedness of the cause of participation dicated an acceptable reliability. The stratification by restriction; 4. normative comparison in the level of par- gender showed a higher Kappa coefficient among males ticipation; 5. long-term duration of restriction; 6. meas- (0.82) compared to females (0.54). The agreement did ure severity of restriction in the response scale (at least not statistically differ by age (15–64 vs. 65+), by educa- three levels). As mentioned above, this comes at a cost tion (technical secondary or less vs. general secondary or of lack of conciseness and simplicity. higher), by language (Dutch vs. French) or time span be- Of the 3 global questions that constitutes the MEHM tween the interviews (≤20 days vs. > 20 days). [15], studies reporting on the concurrent and predictive validity and reliability of SRH have the longest history Discussion [46–48], while less evaluations have been done with re- Defining disability is not easy. Because it interweaves med- spect to the global question on chronic disease [49]. ical and social domains [40], the concept of disability has led In this manuscript, we summarize for the first time the to divergent interpretations and uses [41]. People may ex- current evidence of the validity of GALI including concur- perience disability due to health in any human activity; yet, rent and predictive validity studies and reliability studies. activities included in traditional instruments with focus on Current review has limitations. A first limitation of the ADL and/or IADL cover only a fraction of all activities [12]. review is that it only included peer-reviewed manuscripts The challenges of measuring disability have been tackled by published in English, identified using only one bibliographic Van Oyen et al. Archives of Public Health (2018) 76:25 Page 10 of 11 database in addition to Google Scholar. Google Scholar was trend toward a less healthy life style or less contacts with used to search also the grey literature. The authors were in social relations in function of more participation restric- close contact with the international research network on tions [51]. Using follow-up data, GALI proved both in health expectancies and the disablement process (REVES). national and international studies to be a consistent pre- Members of the REVES network were invited to participate dictor of future health outcomes both in terms of mor- in a survey [9] with the aim to identify additional manu- tality and health care expenditure. As predictors of scripts. Secondly, all studies included rely on self-reported mortality, the two distinct concepts - SRH and GALI - measures and the precision of the validity and reliability es- acted independently and complementary of each other. timates relies upon accurate reporting. Thirdly, no quality Only one reliability study was identified indicating a suf- related weighting was applied in describing the different ficient reliability of GALI. manuscripts. E.g in contrast to the mortality follow-up in Belgium [36], the mortality follow-up within SHARE [38] Conclusion did not use register data resulting in an under-numeration The strength of GALI as an inclusive one-question instru- of the number of deaths. A possible effect of a selection bias ment is that it fits all conceptual characteristics specified on the predictive power of GALI and SRH on mortality for a global measure on participation restriction. The re- cannot be excluded, but if any, the text of the manuscript view indicates that current version of GALI has a good and claims it should be limited [38]. Next, the review was ham- sufficient concurrent and predictive validity and reliability. pered by the methodological heterogeneity of the different Funding studies: e.g. the association of GALI with health indicators This paper is delivered in the context of the project ‘664691/BRIDGE Health’ which in the two national concurrent validations studies [31, 32] has received funding from the European Union’sHealthProgramme (2014–2020). was measured using different health indicators and different statistical modeling. Finally, the review was not registered. Authors’ contributions The literature searches and manuscript selection was done by HVO and NB. To summarize, in none of the 9 studies included, there The manuscript was written by HVO, and reviewed and discussed by PB, RY was evidence of a failing validity. The concurrent validity and NB. All authors read and approved the final manuscript. was evaluated in 4 studies. The two national concurrent validation studies indicated a dose-response effect by Ethics approval and consent to participate Not applicable. GALI severity level on the association with other health status measures with a somewhat weaker association re- Competing interests lated to the mental well-being score.The two inter- HVO is Editor-in-Chief at the Archives of Public Health. national studies did not consider the GALI severity level. They concluded that the odds of reporting participation Publisher’sNote restriction were higher in subjects with self-reported or Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. observed functional limitations. The strength of the as- sociation varied more strongly between the countries Author details using the EHIS compared to the SHARE survey. The dif- Department of Epidemiology and Public Health, Sciensano, J. Wytsmanstraat 14, 1050 Brussels, Belgium. Department of Public Health, Ghent University, ference in homogeneity can in part be explained by the De Pintelaan 185, 9000 Ghent, Belgium. Department of Sociology, Interface variation in the implementation of the EHIS, including Demography, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium. different wording across countries [33]. International Department of Social & Environmental Health Research, London School of Hygiene & Tropical Medicine, Keppel Street, London WC1E 7HT, UK. comparability of data submitted to Eurostat, including those on GALI is hampered as EU regulations does not Received: 8 December 2017 Accepted: 9 May 2018 include guidelines on the exact formulation of the ques- tions within one and between surveys [20] nor on the References data collection mode increasing the likelihood for differ- 1. Chatterji S, Byles J, Cutler D, Seeman T, Verdes E. Health, functioning, and ential total survey errors [30]. This under-valorizes the disability in older adults–present status and future implications. Lancet. substantial efforts done by Eurostat to harmonize instru- 2015;385:563–75. 2. Nagi SZ. The concept and measurement of disability. In: Berkowitz ED, ments in surveys [49]. Differences in accounting for the editor. Disability policies and government programs. New York: Praeger; institutionalized population may further have affected 1979. p. 1–15. the heterogeneity in the size of the association of the 3. Verbrugge LM, Jette AM. The disablement process. Soc Sci Med. 1994;38:1–14. 4. Wood PHN. The language of disablement: a glossary relating to disease and GALI with other health measures [50]. Although, with- its consequences. Int Rehab Med. 1980;2:86–92. out focus on validity, a recent study, using the 2013 Da- 5. World Health Organization. International Classification of Impairments, nish Health and Morbidity survey and 31 health-related Disabilities, and Handicaps. Geneva: World Health Organization; 1980. http:// apps.who.int/iris/bitstream/10665/41003/1/9241541261_eng.pdf indicators, corroborates the concurrent validity describ- 6. World Health Organization. Towards a common language for functioning, ing a trend towards poorer health and wellbeing in sub- disability and health: ICF. Geneva: WHO; 2002. p. 1–23. http://www.who.int/ jects with more participation restriction as well as a classifications/icf/training/icfbeginnersguide.pdf Van Oyen et al. Archives of Public Health (2018) 76:25 Page 11 of 11 7. Dijkers MP. Issues in the conceptualization and measurement of 31. Cabrero-Garcia J, Julia-Sanchis R. The global activity limitation index mainly participation: an overview. Arch Phys Med Rehabil. 2010;91:S5–16. measured functional disability, whereas self-rated health measured physical 8. Badley EM. Enhancing the conceptual clarity of the activity and morbidity. J Clin Epidemiol. 2014;67:468–76. participation components of the international classification of functioning, 32. Van Oyen H, Van der Heyden J, Perenboom R, Jagger C. Monitoring disability, and health. Soc Sci Med. 2008;66:2335–45. population disability: evaluation of a new global activity limitation Indicator (GALI). Soz Praventiv Med. 2006;51:153–61. 9. Berger N, Robine JM, Ojima T, Madans J, Van Oyen H. Harmonising summary measures of population health using global survey instruments. J 33. Berger N, Van Oyen H, Cambois E, Fouweather T, Jagger C, Nusselder W, et Epidemiol Community Health. 2016;70:1039–44. al. Assessing the validity of the global activity limitation Indicator in fourteen European countries. BMC Med Res Methodol. 2015;15:1. 10. Ustun TB, Chatterji S, Kostanjsek N, Rehm J, Kennedy C, Epping-Jordan J, et 34. Jagger C, Gillies C, Cambois E, Van Oyen H, Nusselder W, Robine JM. The al. Developing the World Health Organization disability assessment global activity limitation index measured function and disability similarly schedule 2.0. Bull World Health Organ. 2010;88:815–23. across European countries. J Clin Epidemiol. 2010;63:892–9. 11. Madans JH, Loeb ME, Altman BM. Measuring disability and monitoring the UN 35. Van der Heyden J, Van Oyen H, Berger N, De Bacquer D, Van Herck K. convention on the rights of persons with disabilities: the work of the Washington Activity limitations predict health care expenditures in the general group on disability statistics. BMC Public Health. 2011;11(Suppl 4):S4. population in Belgium. BMC Public Health. 2015;15:267. 12. Verbrugge LM. Disability experience and measurement. J Aging Health. 36. Berger N, Van der Heyden J, Van Oyen H. The global activity limitation 2016;28:1124–58. indicator and self-rated health: two complementary predictors of mortality. 13. Verbrugge LM. A global disability indicator: companion to self-rated health. Arch Public Health. 2015;73:25. In: Schechter S, editor. Cognitive Methods Staff Working Paper Series, No. 37. Van der Heyden J, Berger N, Van Oyen H. Comparison of self-rated health 10. Hyattsville: Office of Research and Methodology, NCHS; 1994. p. 60–88. and activity limitation as predictors of short term mortality in the older http://www.srl.uic.edu/links/CMS_WP10_Schechter_1994_Verbrugge.pdf. population. Public Health. 2015;129:283–5. 14. Verbrugge LM, Merrill SS, Liu X. Measuring disability with parsimony. Disabil 38. Verropoulou G. Specific versus general self-reported health indicators Rehabil. 1999;21:295–306. predicting mortality among older adults in Europe: disparities by gender 15. Robine JM, Jagger C, Egidi V, Gispert M, Nusselder W, Perenboom R, et al. employing SHARE longitudinal data. Int J Public Health. 2014;59:665–78. Creating a coherent set of indicators to monitor health across Europe: the 39. Cox B, Van Oyen H, Cambois E, Jagger C, Le Roy S, Robine JM, et al. The Euro-REVES 2 project. Eur J Publ Health. 2003;13:6–14. reliability of the minimum European health module. Int J Public Health. 16. Robine JM, Jagger C, Romieu I. Selection of a Coherent Set of Health 2009;54:55–60. Indicators for the European Union, Phase II. 1–129. Montpellier: Euro-Reves; 40. Leonardi M, Bickenbach J, Ustun TB, Kostanjsek N, Chatterji S. The definition 2002. https://ec.europa.eu/health/ph_projects/2000/monitoring/fp_ of disability: what is in a name? Lancet. 2006;368:1219–21. monitoring_2000_frep_03_en.pdf 41. Mont D. Measuring health and disability. Lancet. 2007;369:1658–63. 17. Eurostat. Minimum European Health Module. In: Glossary:Minimum 42. Freedman VA, Kasper JD, Spillman BC, Agree EM, Mor V, Wallace RB, et al. European Health Module (MEHM) - Statistics Explained; 2017. Behavioral adaptation and late-life disability: a new spectrum for assessing 18. Perenboom R, Van Oyen H, van Herten L: Limitations in usual activities, a public health impacts. Am J Public Health. 2014;104:e88–94. global approach. In Selection of a Coherent Set of Health Indicators for the 43. Leveille SG, Fried LP, McMullen W, Guralnik JM. Advancing the taxonomy of European Union, Phase II. Edited by Robine JM, Jagger C, Romieu I. Chapter disability in older adults. J Gerontol A Biol Sci Med Sci. 2004;59:86–93. 5. 68–80. 2002. Montpellier: Euro-Reves. https://ec.europa.eu/health/ph_ 44. Chan A, Saito Y, Robine JM. International perspectives on summary projects/2000/monitoring/fp_monitoring_2000_frep_03_en.pdf. measures of population health in an aging world. J Aging Health. 2016;28: 19. World Health Organization. International Classification of Functioning, 1119–23. Disability and Health: ICF. Geneva: World Health Organization; 2001. WHO | 45. Cambois E, Grobon S, Van Oyen H, Robine JM. Impact of question wording International Classification of Functioning, Disability and Health (ICF) on the measurement of activity limitation: evidence from a randomized test 20. Robine JM, Jagger C, Van Oyen H, Cambois E, Doblhammer G, Nusselder W, in France. J Aging Health. 2016;28:1315–38. et al. The minimum European health module. 2010_4.6. Montpellier: 46. Foraker RE, Rose KM, Chang PP, McNeill AM, Suchindran CM, Selvin E, et al. EHEMU. EHEMU Technical report; 2010. p. 1–29. http://www.eurohex.eu/pdf/ Socioeconomic status and the trajectory of self-rated health. Age Ageing. Reports_2010/2010TR4.6_Health%20Module.pdf 2011;40:706–11. 21. Lagiewka K. European innovation partnership on active and healthy ageing: 47. Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty- what have been the policy drivers and determinants to set a headline seven community studies. J Health Soc Behav. 1997;38:21–37. target of 2 additional healthy life years at birth at EU average by 2020? Arch 48. Mossey JM, Shapiro E. Self-rated health: a predictor of mortality among the Public Health. 2012;70:23. elderly. Am J Public Health. 1982;72:800–8. 22. Zaidi A, Gasior K, Hofmarcher M, Lelkes O, Marin B, Rodrigues R, et al. Active 49. Van der Heyden J, De Backer D, Tafforeau J, Van Herck K. Reliability and ageing index 2012. Concept, methodology, and final results. 1–76. Vienna: validity of a global question on self-reported chronic morbidity. J Public European Centre Vienna, UNECE; 2013. http://citeseerx.ist.psu.edu/viewdoc/ Health (Springer). 2014;22:371–80. download;jsessionid=D965DBF5E0AA4F6E8C3B2D173CD1DD07?doi=10.1.1. 50. Cambois E, Jagger C, Nusselder W, Van Oyen H, Robine JM. International 360.9488&rep=rep1&type=pdf comparisons of disability prevalence estimates: impact of accounting or not 23. Gouvernement Français. Esperance de vie en bonne santé. 2016. 17–2-2017. accounting for the institutionalized population. In: Altman BM, editor. Espérance de vie | Gouvernement.fr. International measurement of disability: purpose, method and application, 24. Leonardi M. Measuring health and disability: supporting policy the work of the Washington group, vol. 350. Hyattsville: National Center for development. The European MHADIE project. Disabil Rehabil. 2010;32:1–8. Health Statistics; 2016. p. 207–29. 25. United Nations. Convention on the Rights of Persons with Disabilities : 51. Johnsen NF, Davidsen M, Michelsen SI, Juel K. Health profile for Danish resolution adopted by the General Assembly. New York: UN General adults with activity limitation: a cross-sectional study. BMC Public Health. Assembly; 2007. Convention on the Rights of Persons with Disabilities [A/ 2017;18:46. RES/61/106] | United Nations Enable 26. European Commission. European Pillar of Social Rights. 2017. European Pillar of Social Rights | European Commission. 27. Romaniuk P, Brukalo K. Designing solutions for the retirement system - in search of balance between economy and health. Front Public Health. 2016;4:184. 28. Drost EA, editor. Validity and reliability in social science research. Educ Res Perspect. 2011;38:105–23. 29. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097. 30. Groves R, Lyberg L, editors. Total survey error. Past, present and future. Public Opin Q. 2010;74:849–79. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Archives of Public Health Springer Journals

Measuring disability: a systematic review of the validity and reliability of the Global Activity Limitations Indicator (GALI)

Free
11 pages

Loading next page...
 
/lp/springer_journal/measuring-disability-a-systematic-review-of-the-validity-and-mT5Qrrazgr
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Public Health; Medicine/Public Health, general
eISSN
2049-3258
D.O.I.
10.1186/s13690-018-0270-8
Publisher site
See Article on Publisher Site

Abstract

Background: GALI or Global Activity Limitation Indicator is a global survey instrument measuring participation restriction. GALI is the measure underlying the European indicator Healthy Life Years (HLY). Gali has a substantial policy use within the EU and its Member States. The objective of current paper is to bring together what is known from published manuscripts on the validity and the reliability of GALI. Methods: Following the PRISMA guidelines, two search strategies (PUBMED, Google Scholar) were combined to identify manuscripts published in English with publication date 2000 or beyond. Articles were classified as reliability studies, concurrent or predictive validity studies, in national or international populations. Results: Four cross-sectional studies (of which 2 international) studied how GALI relates to other health measures (concurrent validity). A dose-response effect by GALI severity level on the association with the other health status measures was observed in the national studies. The 2 international studies (SHARE, EHIS) concluded that the odds of reporting participation restriction was higher in subjects with self-reported or observed functional limitations. In SHARE, the size of the Odds Ratio’s (ORs) in the different countries was homogeneous, while in EHIS the size of the ORs varied more strongly. For the predictive validity, subjects were followed over time (4 studies of which one international). GALI proved, both in national and international data, to be a consistent predictor of future health outcomes both in terms of mortality and health care expenditure. As predictors of mortality, the two distinct health concepts, self-rated health and GALI, acted independently and complementary of each other. The one reliability study identified reported a sufficient reliability of GALI. Conclusion: GALI as inclusive one question instrument fits all conceptual characteristics specified for a global measure on participation restriction. In none of the studies, included in the review, there was evidence of a failing validity. The review shows that GALI has a good and sufficient concurrent and predictive validity, and reliability. Keywords: Disability, Participation restriction, Healthy life years, Validity, Reliability, Summary measure of population health, GALI Introduction reduction in the total lifetime days of disability, the Ageing of populations defies health and social policies. so-called compression of morbidity [1]. Population ill-health and especially disability are major The concept of disability is complex and multidimen- challenges as there is currently no consistent evidence sional. In initial medical models, disability was viewed as a that the lengthening of life expectancy goes with a problem residing solely in the persons affected. Disability referred to consequences of chronic or acute diseases or accidents on the functioning of specific body systems and on mental, physical and sensory functions in terms of (1) * Correspondence: herman.vanoyen@sciensano.be Department of Epidemiology and Public Health, Sciensano, J. Wytsmanstraat impairment or dysfunctions and structural abnormalities 14, 1050 Brussels, Belgium in specific body systems; (2) disability or restrictions in Department of Public Health, Ghent University, De Pintelaan 185, 9000 basic physical and mental actions and (3) handicaps or Ghent, Belgium Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Van Oyen et al. Archives of Public Health (2018) 76:25 Page 2 of 11 difficulties in doing activities of daily life [2–5]. More self-reported measure of participation restriction. Add- biophysical-social models introduce the person-environment itional conceptual criteria were the health relatedness of perspective of the disablement process: disability as the the cause of disability, the generic normative comparison outcome of the interaction of a person and his environ- in the level of participation, the long-standing duration of ment [6] and the dynamics of disability which is af- the disability (a duration of at least 6 months) and the fected by how a person’s capacity fits the environmental ability to measure levels of severity [15, 18]. To accommo- demand and results in participation [3]. Participation date the results of cognitive testing in relation to the se- restriction is defined as limitations in the performance verity options in the answer categories, the wording “to of roles and social involvement in different settings what extent” was added in the final version [20]: such as work and employment, school, leisure, parent- ing, housework, community, social and civic life [7]. Be- “For at least the past 6 months, to what extent have cause participation is influenced by environmental you been limited because of a health problem in factors and social norms, any measure of participation activities people usually do?” restriction cannot differentiate the impact of the im- pairment and functional limitations from the impact of Would you say you have been: severely limited, accommodations and enabling environments [8, 9]. Disability can occur in any human activity and set- limited but not severely, or tings. Adding to this complexity, instruments measuring disability differ in the domains of functioning included, not limited at all? in their goals to measure either capacity (without any personal or equipment assistance) or performance (with Being part of the MEHM, GALI is used in major Euro- assistance), or to measure also disability symptoms (pain, pean health and non-health surveys such as the Euro- weakness, endurance, …), levels of severity or the dur- pean Health Interview Survey (EHIS), Survey on Income ation of the disability. Traditional survey instruments and Living Conditions (SILC) and the Survey of Health, measure a limited number of tasks (5 to 7) in the do- Ageing and Retirement in Europe (SHARE). Since 2004, main of personal care (ADL (Activity of Daily Living)) or GALI is also the underlying measure of the European in- in the domain of household management (IADL (Instru- dicator “Healthy Life Years (HLY)”. HLY is a measure of mental ADL)). Other survey instruments have either in- disability free life expectancy and was presented in the creased the number of disability questions by adding set of structural indicators selected and defined to help more and more tasks [10] or have developed short set of measure progress in strategic European policies such as disability questions that have good coverage of activities the 2000 Lisbon strategy and the European 2020 strategy [11, 12]. At the same time there has been a quest to on Active and Healthy Ageing [21]. HLY is one of the measure disability with parsimony similar to the parsi- components of the Active Ageing Index [22]. At the na- mony in measuring health using the global one-item sur- tional level, countries such as France, have selected HLY vey instrument on self-rated health (SRH) [13, 14]. as one of their high level indicators for long term evalu- In response to the call for parsimony, a global survey in- ation of their economic, social and environmental pol- strument to measure disability, the Global Activity Limita- icies [23]. GALI also fits the requirement to follow-up tion Indicator or GALI, was proposed. The development European and United Nations disability policies, that of GALI occurred in the framework of the creation of a stress the importance of full and effective participation coherent set of indicators to monitor health across Europe as main policy outcome [24, 25]. More recently, GALI, [15]. GALI was part of set of 10 survey instruments in- as underlying health measure of HLY, contributes to the cluding three global one-item survey instruments were scoreboard indicators of the European Pillar of Social proposed covering distinct health concepts: perceived Rights [26] . Due to its high informational value, its rela- health, chronic morbidity and participation restriction tive simplicity and its compliance elaborated by the [16]. The 3 global questions define the Minimum Euro- European Union, HLY has been proposed to be the in- pean Health Module (MEHM) [17]. At the time of the de- strument in designing social security solutions [27]. velopment of the GALI, the beta version on the Given the use of GALI within the European Union, es- International classification of Functioning, Disability and pecially the fact that it is the measure underlying the Health (ICF) served as conceptual framework [6, 18, 19]. European indicator HLY, the objective of current paper Because of its implicit reference to the ability for societal is to bring together what is known from published man- participation in a variety of non-specified settings and uscripts on the validity and/or the reliability of GALI. non-specified domains of life (such as employment, Construct validity evaluation has been divided into school, housework, and leisure) using the wording “activ- translation validity, a more qualitative process and criter- ities people usually do”,GALI isintendedto beaglobal ion validity, a quantitative approach [28]. Current review Van Oyen et al. Archives of Public Health (2018) 76:25 Page 3 of 11 focus on the quantitative validity, as translation validity citations). The remaining 101 articles were screened to is linked to the conceptual criteria used for the GALI de- identify manuscripts published with the objective to esti- velopment [9, 15]. mate the validation or reliability of the GALI (N = 11). Double publications were excluded (N = 2). The double Methods publications were a result of the publication of institu- Two search strategies were combined by HVO and NB tional working papers or reports prior to the publication in January 2017 to identify peer reviewed manuscripts of a manuscript in a scientific journal. All 9 manuscripts published in English with publication date 2000 (the retained were found in both databases. Manuscripts period GALI was developed) or beyond. Following the were grouped as reliability studies, concurrent validity PRISMA guidelines, manuscripts were independently studies (cross-sectional studies measuring an associ- evaluated by HVO and NB first on the titles and ab- ation) or predictive validity studies (ability of GALI to stracts and in a second stage on the text. The result of predict an outcome) [28]. We further distinguished be- the search and manuscript selection is summarized in a tween national and international studies, as international PRISMA flow chart (Fig. 1)[29]. First, PUBMED data- studies may be more sensitive to total survey error due, base (https://www.ncbi.nlm.nih.gov/pubmed) was used for example, to the lack of international harmonization with {“global activity limitation”[All Fields] OR at different stages of the study [30]. (GALI[All Fields] NOT GALI[Author])} as search proto- col. Of the 69 publications 54 were excluded: content Results not related to the topic (e.g. Gali as part of a name of a The classification of the manuscripts by type of study is butterfly “Calisto franciscoi Gali”), language other than given in Table 1. English, Gali in name, email or contact address of one of the authors). Of the remaining 15 articles, 9 articles were Concurrent validity retained as manuscripts studying the validity or reliabil- Concurrent validation studies are cross-sectional studies ity of GALI. A second search used the Google Scholar with the objective to measure how GALI relates to other database (https://scholar.google.com/) using as search health measures. As there is no gold standard, the asso- protocol “global activity limitation”. Of the 208 refer- ciations are mainly measured using other health compo- ences 107 were excluded (content not related to the nents such as chronic (co)morbidity or other dimensions topic, language other than English, only an abstract, and of the disablement process, e.g. functional limitations in Fig. 1 PRISMA flow chart [29]: validation and reliability studies of the Global Activity Limitation Indicator (GALI) selection, 2000–2017 Van Oyen et al. Archives of Public Health (2018) 76:25 Page 4 of 11 � � � � � � � � � � � � Table 1 Studies estimating the validity or reliability of the Global Activity Limitation Indicator (GALI) by study design, period 2000–2017 Reference Study setting Health outcome used Severity Statistical measure Key findings and population in the comparison with level GALI the GALI Concurrent validity studies Van Oyen et al. National HIS* 2001 Self-reported: ADL*, SF-36* physical Yes Predicted probability distribution of The probability distribution of GALI by 2006 [32] Belgium Population domain score, number of self-reported GALI distribution, POR* from proportional severity level fits appropriately against aged 15+ N = 9168 chronic physical conditions out of a odds models, heterogeneity across indicators measuring mental and physical list of 29 conditions, number of mental demographic variables illness both in subjects with or without conditions (depression, anxiety, somatization, ADL limitations; sleep disorders) from the SCL-90R*, GHQ-12* 95% of subjects without ADL limitations scale for mental wellbeing and CMI* and no mental or physical health problems do not report participation restrictions; Subjects with ADL limitations report participation restrictions and the severity level of participation restriction is higher in function of the level of severity and the number ADL limitations and there is no evidence for heterogeneity across gender, age, education and language; The probability distribution of GALI by severity level is associated with the different physical and mental morbidity measures; A dose-response relationship is observed; The measures of associations are not as strong for mental health problems compared to physical health problems. Cabrero-Garcia National HIS 2006 Self-reported: physical and mental morbidity, Yes Spearman correlation, predicted probability GALI is primarily a measure of functional et al. 2014 [31] Spain Population functional disability (ADL, IADL and mobility) of GALI from fractional polynomial models, status and is secondarily associated with aged 65+ N = 7835 Concurrent comparison of the associations MOR* from multinomial logistical regression physical and mental morbidity whereas of GALI and the association of SRH* with the for SRH physical morbidity and to a lesser health outcomes: extend mental morbidity are the main FCI*, GHQ-12 scale for mental wellbeing, correlates; Functional disability* The odds of having (severe) participation restrictions increased with the level of functional limitations (number), the physical and mental morbidity, suggesting a dose- response relationship; Mental morbidity was as strong a correlate of GALI as of SRH, whereas physical morbidity waslessstrongacorrelateofGALI compared to SRH. Jagger et al. International SHARE* 2004 Measurement: maximum grip strength and No Predicted probability distribution of GALI, GALI effectively capture disability as measured 2010 [34] 11 EU countries Population walking speed (in subjects aged 75+) ORs from logistic regression models, by both the self-reported as objective measures aged 50+ N = 27340 Self-reported: ADL, IADL, and walking Random-effects meta-analysis to assess of functional limitations; limitations heterogeneity of associations between The likelihood of reporting participation countries restrictions increases as the severity of functional limitations increases in both the self-reported as objective measures of functional limitations; The likelihood of reporting no participation restriction in subjects with limitations is non-zero, though small and a minimum for the most severe measure, the ADL; Cross-country comparison did not provide any evidence for heterogeneity for the OR Van Oyen et al. Archives of Public Health (2018) 76:25 Page 5 of 11 � � � � � � � � � � � � � � � Table 1 Studies estimating the validity or reliability of the Global Activity Limitation Indicator (GALI) by study design, period 2000–2017 (Continued) Reference Study setting Health outcome used Severity Statistical measure Key findings and population in the comparison with level GALI the GALI of having participation restrictions in function of the self-reported ADL and the objective measures’; In all countries, the odds of having participation restrictions was higher in subjects with IADL limitations. The size of the effect was however more pronounced in some countries compared to others. Berger et al. International EHIS* Self-reported: ADL, IADL No Predicted probability distribution of GALI, GALI is significantly associated with ADL and 2015 [33] 2007–2010 14 EU and functional limitations ORs from logistic regression models, IADL limitations and functional limitations; countries Population Random-effects meta-analysis to assess The likelihood of reporting participation aged 15+ N = 152,796 heterogeneity of associations between restrictions increases as the number of ADL and countries IADL limitations and the severity of functional limitations increased; The likelihood of not reporting participation restrictions decreases as the number of ADL and IADL limitations and the severity of functional limitations increased; In all countries, the odds of having participation restrictions was higher in subjects with ADL, IADL and functional limitations. The size of the effect was more pronounced in some countries compared to others providing evidence for heterogeneity of the effect size. Predictive validity studies Berger et al. National HIS 2001 linked Mortality (follow-up to 10 years) Yes MRRs* from Poisson regression models; Compared to individuals without participation 2015 [36] with mortality and migration Comparison of relative predictive ability restrictions, subjects with moderate or severe database (National Register), of GALI compared to SRH* participation restriction have a 1.8 to 3.0 2001–2010 Belgium Population increased mortality rate over the 10 years aged 15+ N = 8583, 902 deaths of follow-up; The effect does not vary significantly by gender, education or age, except in subjects under age of 50 years; SRH and GALI are complementary predictors of mortality, with some indications of a stronger effect of SRH; The predictive effect of SRH and GALI slightly decrease over time. Van der Heyden National HIS 2008 linked with Mortality (follow-up to 2 years) No MRRs from Poisson regression models; Subjects with participation restriction have et al. 2015 [37] mortality within the Health Comparison of relative predictive ability a 2.4 increased mortality rate over the 2 Insurance database, 2008–2010 of GALI compared to SRH years of follow-up; Belgium Population aged The effect does not vary by gender; 65+ N = 1894, 178 deaths In men, SRH and GALI are complementary predictors of mortality, whereas in women this is only so for GALI. Van der Heyden National Health care expenditure Yes Linear regression after logistic transformation Moderate and severe participation restriction et al. 2015 [35] HIS 2008 linked with Health (Health insurance, out of costs; increases all health expenses by 3 to 6-times; Insurance database including of-pocket, supplement) Cost ratios were estimated to compare The increase is the more pronounced in the expenditure in 2008–2010 expenses to a reference; reimbursed health care expenditure; Belgium Population aged Decomposition of differences in expenses In absence of any chronic condition, moderate 15+ N = 7286 using the Blinder-Oaxaca method and severe participation restriction increases all health care expenditure by 2.5 to 4.5 times; Van Oyen et al. Archives of Public Health (2018) 76:25 Page 6 of 11 � � � � � � � � � Table 1 Studies estimating the validity or reliability of the Global Activity Limitation Indicator (GALI) by study design, period 2000–2017 (Continued) Reference Study setting Health outcome used Severity Statistical measure Key findings and population in the comparison with level GALI the GALI Chronic conditions explain only 22% of the differences in health care expenditure by level of participation restriction. Verropoulou International SHARE* 2004 Mortality (follow-up 2 to 3 years) Yes Hazard ratios from Cox proportional Both GALI and SRH are significant predictors et al. 2015 [38] with follow-up to re-interview hazard models; Comparison of relative of mortality in separate models; in wave 2006 /2007 11 EU predictive ability of GALI compared When adjusting for specific health indicators countries Population aged to SRH (asthma, cancer, depression, mobility, IADL, 50+ N = 17,941, 696 deaths orientation), GALI and SRH (only men) were significant but the magnitude diminished; GALI and SRH add information on top of specific health indicators; When GALI and SRH are included in one model, GALI was only significant in women, suggesting a partial conceptual overlap as there is a correlation between GALI and SRH; SRH and GALI represent different aspects of health. Reliability studies Cox et al. National Food Consumption Twice self-reported GALI Yes Pearson correlation coefficients, weighted Both Pearson (0.73) and Kappa coefficient (0.68) 2009 [39] Survey 1st and 2nd visit within time window Kappa coefficients indicate an acceptable reliability; Belgium, Population aged between 11 and 55 days Agreement is significantly higher for males 15+ N = 170 (Kappa = 0.82) compared to females (Kappa = 0.54); Agreements did not differ by education level, age, time span and language (French, Dutch). *HIS Health interview survey ADL Activities of Daily Living SF-36 Short Form Survey SCL-90R Symptoms Check List GHQ-12 General Health Questionnaire CMI Composite Morbidity Indicator: no illness, only mental illness, only physical illness and both mental and physical illness POR Proportional Odds Ratios MOR Multinomial Odds Ratios SRH Self-Rated Health FCI Functional Comorbidity Index based on a list of 16 chronic conditions including obesity, hearing and visual impairments Functional disability: based on a 27 items related to I/ADL and mobility Washington group instrument: ref. = 32,350 SHARE Survey of Health and Retirement in Europe 2004 survey was done in Austria, Belgium, Denmark, France, Germany, Greece, Italy, the Netherlands, Spain, Sweden, Switzerland EHIS European Health interview survey 2007–2010 surveys were done in Belgium, Bulgaria, Cyprus, Czech Republic, France, Greece, Hungary, Latvia, Malta, Poland, Romania, Slovakia, Slovenia, Spain MRR Mortality Rate Ratio Van Oyen et al. Archives of Public Health (2018) 76:25 Page 7 of 11 activities. Two studies were national [31, 32] and 2 were level. E.g., the predicted probability of no participation re- international [33, 34]. In one of the international studies, strictions in subjects without ADL limitations and without GALI could be evaluated against the results of objective any mental or physical illness was 0.95 and dropped to re- measures of functional limitations [34]. The age groups spectively 0.90, 0.80 and 0.57 in subjects reporting only included in the studies varied between subjects 15 years mental illness, only physical illness or both; in people with and older, 50 years and older or 65 years and older. One ADL limitations the predicted probabilities of no participa- of the international studies focused on the population tion restriction were respectively 0.46, 0.43, 0.37 and 0.12. 15 years and older but provided, for comparison pur- The second concurrent validation study used the 2006 poses, tables and graphs for the population 50 years and Spanish National Health Survey but included only sub- older as supplementary material [33]. The national stud- jects 65 years and older to test (1) if GALI is primarily ies considered GALI by severity level, while the inter- correlated with functional disability and secondarily with national studies ignored the severity level. morbidity, and (2) if Self Rated Health (SRH), in con- Van Oyen et al. used the 2001 Belgian Health Interview trast, is primarily correlated with morbidity and second- survey [32] to evaluate GALI against (1) Activities of Daily arily with functional disability [31]. Associations were Living (ADL) limitations (categorical by level of severity, sought with a functional comorbidity indicator (FCI) and by number of limitations); (2) the Short Form Survey based on a list of 16 chronic conditions including obes- (SF-36) physical domain score; (3) the number of ity, hearing and visual impairments, the GHQ-12 for self-reported chronic physical conditions out of a list of 29; mental ill-health and a functional disability measure (4) the number of mental conditions (depression, anxiety, (based on 27 items related to IADL/ADL and mobility). somatization, sleep disorders) based on 4 subscales of the The Spearman correlation coefficients of FCI, GHQ-12 Symptoms Check List (SCL-90R) and (5) a mental and functional disability were 0.35, 0.45 and 0.58 with well-being score using the General Health Questionnaire GALI compared to 0.46, 0.44 and 0.36 with SRH. The (GHQ-12). A composite morbidity indicator (CMI: catego- predicted probability of participation restriction indi- rized as no illness, only mental, only physical illness and cated a greater effect in function of the number of func- both) was used to measure the associations stratified by the tional disabilities compared to the comorbidity indicator ADL functional limitation status. The results indicated that while the inverse was observed for the predictive prob- all health indicators were positively associated with GALI. ability of not being in very good/good SRH. The pre- The participation restriction distribution by severity level dicted probabilities for GALI and SRH were similar in was positively associated with both the number and severity function of the GHQ-12. Compared to subjects with no of ADL limitations, the SF-36 physical domain score, the functional disability (ADL/IADL or mobility), the multi- number of chronic conditions and the mental health indi- nomial odds ratios (MORs) of participation restriction cators (mental health comorbidity score or the GHQ-12). and severe participation restriction were respectively E.g., without any ADL limitations the predicted probability 1.44 and 2.02 in subjects with one functional disability distribution of being without, with mild or with severe par- (ADL/IADL or mobility) and respectively 8.94 and 64.84 ticipation restriction was respectively 0.82, 0.15, 0.03 com- when limitations in 11 functions were reported. The pared to 0.20, 0.43, 0.37 and 0.13, 0.38, 0.49 in people with MORs for participation restriction and severe participa- at least one ADL limitation or with at least one severe ADL tion restriction were respectively 1.96 and 2.00 for sub- limitation. When people were limited in 6 ADLs, the GALI jects with a FCI score of 1 and respectively 7.49 and 7.96 probabilities of reporting no, mild and severe restrictions for people with a FCI score of 7. The MORs of having were respectively 0.03, 0.10 and 0.87. Using the GHQ-12 participation restriction and severe participation restric- mental well-being score, the predicted probabilities of no, tion in people with a GHQ-12 score of 1 and of 7 were mild and severe GALI restrictions changed from respect- respectively 1.32 and 1.61; and 3.42 and 8.05. The MORs ively 0.82, 0.14, 0.04 (best GHQ-12 score) to 0.36 0.36 0.28 of fair SRH and very poor/poor SRH indicated a similar (worst GHQ-12 score). The CMI was associated with par- pattern but were more extreme in function of the func- ticipation restriction both in people with and without ADL tional comorbidity score while less extreme in function limitations. In the population free of ADL limitations, and of the functional disability indicator. These results sug- compared to subjects reporting non illness, the participa- gest that GALI was primarily a measure of functional tion restriction prevalence and especially the prevalence of status and secondarily a measure of physical and mental being severely restricted increased gradually in people morbidity whereas for SRH, physical morbidity and to a reporting only mental illness, reporting only physical illness lesser extend mental morbidity were the main correlates. or reporting both mental and physical illness. A similar The two concurrent international studies [33, 34] trend was observed in subjects with ADL limitations but followed a similar statistical analysis plan using the data the prevalence of participation restriction and severe re- from SHARE and EHIS: the estimation of the predicted striction was substantially higher within each morbidity GALI probability distribution by fitting logistic regression Van Oyen et al. Archives of Public Health (2018) 76:25 Page 8 of 11 and random-effects meta-analysis models to evaluate het- (HIS-2001 [36], HIS-2008 [37]) while the other used the erogeneity of the association between countries (Table 1). international SHARE study, wave 2004 and wave The 2004 SHARE study covered 11 EU countries [34], 2006/2007 [38]. The duration of mortality follow-up ranged while the EHIS (2007–2010) used data from 14 EU coun- from 2 to 3 years [37, 38]to10years [36]. Two of the 3 tries [33]. Main differences are related to the age groups in- mortality studies used GALI by severity level [36, 38]. The cluded, the use of observed measures and the stronger age considered ranged from 15 years and older [36]to overall survey design homogeneity between countries in 50 years and older [38] and 65 years-plus [37]. In all 3 mor- the SHARE survey compared to the EHIS survey design. In tality studies, the predictive capacity of GALI was set off both studies, the severity level information of GALI was ig- against the predictive capacity of SRH. The 3 studies identi- nored in the analysis. The GALI was evaluated against the fied GALI and SRH as complementary predictors of mor- number of ADL and IADL limitations. The EHIS data also tality, indicating that GALI and SRH share some traits but provided an evaluation against a function limitation meas- add different dimensions: health and disability. In the two urebyseveritybased on theself-reported moderateor se- studies with a short follow-up period [37, 38], adjusting for vere problems in at least one of the following 6 functions: age, education and life style (SHARE only: physical activity, walking a certain distance, going up and down the stairs, smoking, BMI) both GALI and SRH were significant pre- carrying in the hands or arms, using hands and fingers to dictors of mortality: having participation restrictions dou- manipulate small objects, biting and chewing. The objective bled (mild) and tripled (severe) the mortality rate in instruments to evaluate the validity of the GALI in SHARE SHARE, while the mortality rate increased by a factor of measured the hand grip strength and the walking speed (in 2.4 when no severity level was accounted for in the Belgian 75 years or older only; reported walking limitations in those study [37]. In the Belgian study [37], when both GALI and 50–74 years). In the SHARE study, the probability of the re- SRH were included in the Poisson regression model, GALI ported participation restriction was lower when the grip remained a significant predictor of mortality next to SRH strength or walking speed were higher. Similarly, and in in both males and females, while SRH remained only sig- both studies, the probability of having participation restric- nificant in males. In the SHARE follow-up [38], the tion was higher as the number of ADL or IADL limitations fully-adjusted proportional hazard model, controlling for increasedorifthe levelofseverity ofthose limitationsin- specific morbidity indicators (asthma, cancer, depression) creased (EHIS only). In all countries and in both studies, and disability measures (mobility, IADL), GALI and SRH, the direction of the association, measured as ORs, between showed that SRH remained significantly associated with GALI and the other health measures was as expected. That mortality only in men while GALI remained significantly is: the odds of reporting participation restriction is higher associated with mortality only in women. Over a 10-year in subjects with poorer functioning and disability measures follow-up period [36], GALI as well as SRH were strong (either in function of the ADL, IADL, the physical func- predictors of mortality. Adjusting for age, gender and tional limitations or in function of low hand grip strength socio-economic position, people with mild and severe par- or walking speed). In the population 50 years and over, the ticipation restriction, compared to no participation restric- overall random effect meta-regression ORs were less ex- tion, had mortality rate ratio’s(MRR’s) of respectively 1.8 treme in the SHARE study compared to the EHIS: e.g. and 3.0. Compared to good/very good SRH the MRR in comparing subjects with at least one ADL-limitation vs. subjects with fair and bad/very bad SRH the MRR was re- those with none, the combined OR of having participation spectively 1.8 and 3.6. When including GALI and SRH in restriction were 8.3 in SHARE and 12.3 in EHIS; with re- the model, both remained significant predictors of mortal- spect to IADL limitations, the combined ORs were respect- ity: MRR for GALI were: 1.4 (mild) and 1.8 (severe); and ively 6.4 and 9.1. In the SHARE study, there was no MRR for SRH were: 1.5 (fair) and 2.5 (bad/very bad). The significant country variation in the ORs in function of predictive ability did not change with gender or ADL, hand grip strength and walking speed. For IADL in socio-economic position. However, in older subjects, the the SHARE and for all functions and disability measures in predictive ability of SRH was not as strong. The impact on the EHIS, the OR of having participation restriction was mortality of both GALI and SRH decreased over time but more pronounced in some countries compared to others, remained statistical significant in truncated follow-up pe- providing evidence of heterogeneity in the effect size. riods: 0–3 years, 3–6 years and 6–10 years. Van der Heyden et al. evaluated how GALI predicted Predictive validity health care expenditure using data linkage between each The predictive validity of GALI was reported by 4 studies participant to the 2008 Belgian Health Interview Survey (Table 1). One study evaluated the predictive effect on and the national health insurance data for the 12 months health care expenditure in Belgium [35]; the other 3 stud- following the date of the interview [35]. Participation re- ies focused on mortality. Two of the mortality follow-up striction was a strong determinant of the total health studies used Belgian Health Interview Surveys as baseline care expenditure: e.g. the population with participation Van Oyen et al. Archives of Public Health (2018) 76:25 Page 9 of 11 restriction (21% of the population) accounted for 49% of two distinct approaches. One option is to include more ac- the total health expenditure; for severe restriction (5% of tivities, more specific answer categories, more aspects such the population) this was 17% of the total expenditure. as disability symptoms and disability in more specific set- The association was stronger for the reimbursed health tings of life [10, 42, 43]. This option induces increased re- care cost compared to the out-of-pocket payments. In spondents burden, increased survey cost, more complex subjects with no chronic conditions compared to people analysis in order to provide condensed indicators for end without participation restriction, the cost ratio of the re- users. The other option seeks short sets and/or a one single imbursed cost in subjects with mild or severe participa- global instrument [11, 15]. GALI has been nominated the tion restriction was respectively 2.5 and 4.2. In people champion in parsimony [12], but the lack of simplicity and with one chronic condition or in people with ≥2 chronic the high density of concepts in one single question may conditions the cost ratio compared to no participation hamper its acceptability [9, 44]. Three different alternatives restriction were respectively 1.5 and 1.7 in subjects with (decomposing GALI using filtered and routed questions or mild participation restriction and 2.4 and 3.2 in people through omitting features such as the duration of the dis- with severe participation restrictions. The authors ability and/or the health relatedness) and the original GALI decomposed the health expenditure gap between people were evaluated against the short version (including 4 func- with and without participation restriction: differences in tional limitation questions) of the Washington Group on the age distribution (20%) and in the prevalence of Disability instrument [11]. The four variants were randomly chronic diseases (22%) between the two groups were the assigned to survey participants (N = 3009). The results, indi- main contributors to the explained differences (48%). cating a substantial higher sensitivity of GALI, no evidence Next to the confounding effect of age, the decompos- for a better understanding of the simplified alternatives but ition analysis also indicated that the impact of age on possibly, a small advantage in specificity when bringing the health expenditure differed by GALI severity level, sug- duration of the participation restriction to subsequent ques- gesting an interaction effect of age. However, in the un- tions, should be balanced against the cost of breaking an explained component, the coefficient of chronic established chronological series [45]. Currently, Eurostat fol- conditions did not differ significantly between GALI cat- lows for the upcoming EHIS wave III (2018–2019), the rec- egories, suggesting that chronic conditions in people ommendation of the EHLEIS working group on the with participation restrictions do not result in significant blueprint for an internationally harmonized Summary Meas- different health care expenditure compared to people ure of Population Health [9]. The EHLEIS working group, without participation restrictions. including experts from the EU, Japan, USA, OECD and WHO met 3 times (2012, 2013 and 2014) in Paris and pro- Reliability posed that of the different components of disability, partici- Only one study evaluated the reliability of GALI as part pation restriction in the first place and, in addition, of the evaluation of the MEHM [39]. The study used the functional limitations should be the main goals for inter- Belgian 2004 Food Consumption Survey in which people nationally harmonized global measures [9]. GALI fits the six were visited twice at home by the interviewers. The conceptual characteristics specified by the working group: 1. interquartile range between the 2 visits was 17–26 days comprehensive content of participation; 2. measure of with median time of 20 days. Both the Pearson correl- participation performance with current accommoda- ation (0.73) and the weighted Kappa coefficient (0.68) in- tion; 3. health relatedness of the cause of participation dicated an acceptable reliability. The stratification by restriction; 4. normative comparison in the level of par- gender showed a higher Kappa coefficient among males ticipation; 5. long-term duration of restriction; 6. meas- (0.82) compared to females (0.54). The agreement did ure severity of restriction in the response scale (at least not statistically differ by age (15–64 vs. 65+), by educa- three levels). As mentioned above, this comes at a cost tion (technical secondary or less vs. general secondary or of lack of conciseness and simplicity. higher), by language (Dutch vs. French) or time span be- Of the 3 global questions that constitutes the MEHM tween the interviews (≤20 days vs. > 20 days). [15], studies reporting on the concurrent and predictive validity and reliability of SRH have the longest history Discussion [46–48], while less evaluations have been done with re- Defining disability is not easy. Because it interweaves med- spect to the global question on chronic disease [49]. ical and social domains [40], the concept of disability has led In this manuscript, we summarize for the first time the to divergent interpretations and uses [41]. People may ex- current evidence of the validity of GALI including concur- perience disability due to health in any human activity; yet, rent and predictive validity studies and reliability studies. activities included in traditional instruments with focus on Current review has limitations. A first limitation of the ADL and/or IADL cover only a fraction of all activities [12]. review is that it only included peer-reviewed manuscripts The challenges of measuring disability have been tackled by published in English, identified using only one bibliographic Van Oyen et al. Archives of Public Health (2018) 76:25 Page 10 of 11 database in addition to Google Scholar. Google Scholar was trend toward a less healthy life style or less contacts with used to search also the grey literature. The authors were in social relations in function of more participation restric- close contact with the international research network on tions [51]. Using follow-up data, GALI proved both in health expectancies and the disablement process (REVES). national and international studies to be a consistent pre- Members of the REVES network were invited to participate dictor of future health outcomes both in terms of mor- in a survey [9] with the aim to identify additional manu- tality and health care expenditure. As predictors of scripts. Secondly, all studies included rely on self-reported mortality, the two distinct concepts - SRH and GALI - measures and the precision of the validity and reliability es- acted independently and complementary of each other. timates relies upon accurate reporting. Thirdly, no quality Only one reliability study was identified indicating a suf- related weighting was applied in describing the different ficient reliability of GALI. manuscripts. E.g in contrast to the mortality follow-up in Belgium [36], the mortality follow-up within SHARE [38] Conclusion did not use register data resulting in an under-numeration The strength of GALI as an inclusive one-question instru- of the number of deaths. A possible effect of a selection bias ment is that it fits all conceptual characteristics specified on the predictive power of GALI and SRH on mortality for a global measure on participation restriction. The re- cannot be excluded, but if any, the text of the manuscript view indicates that current version of GALI has a good and claims it should be limited [38]. Next, the review was ham- sufficient concurrent and predictive validity and reliability. pered by the methodological heterogeneity of the different Funding studies: e.g. the association of GALI with health indicators This paper is delivered in the context of the project ‘664691/BRIDGE Health’ which in the two national concurrent validations studies [31, 32] has received funding from the European Union’sHealthProgramme (2014–2020). was measured using different health indicators and different statistical modeling. Finally, the review was not registered. Authors’ contributions The literature searches and manuscript selection was done by HVO and NB. To summarize, in none of the 9 studies included, there The manuscript was written by HVO, and reviewed and discussed by PB, RY was evidence of a failing validity. The concurrent validity and NB. All authors read and approved the final manuscript. was evaluated in 4 studies. The two national concurrent validation studies indicated a dose-response effect by Ethics approval and consent to participate Not applicable. GALI severity level on the association with other health status measures with a somewhat weaker association re- Competing interests lated to the mental well-being score.The two inter- HVO is Editor-in-Chief at the Archives of Public Health. national studies did not consider the GALI severity level. They concluded that the odds of reporting participation Publisher’sNote restriction were higher in subjects with self-reported or Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. observed functional limitations. The strength of the as- sociation varied more strongly between the countries Author details using the EHIS compared to the SHARE survey. The dif- Department of Epidemiology and Public Health, Sciensano, J. Wytsmanstraat 14, 1050 Brussels, Belgium. Department of Public Health, Ghent University, ference in homogeneity can in part be explained by the De Pintelaan 185, 9000 Ghent, Belgium. Department of Sociology, Interface variation in the implementation of the EHIS, including Demography, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium. different wording across countries [33]. International Department of Social & Environmental Health Research, London School of Hygiene & Tropical Medicine, Keppel Street, London WC1E 7HT, UK. comparability of data submitted to Eurostat, including those on GALI is hampered as EU regulations does not Received: 8 December 2017 Accepted: 9 May 2018 include guidelines on the exact formulation of the ques- tions within one and between surveys [20] nor on the References data collection mode increasing the likelihood for differ- 1. Chatterji S, Byles J, Cutler D, Seeman T, Verdes E. Health, functioning, and ential total survey errors [30]. This under-valorizes the disability in older adults–present status and future implications. Lancet. substantial efforts done by Eurostat to harmonize instru- 2015;385:563–75. 2. Nagi SZ. The concept and measurement of disability. In: Berkowitz ED, ments in surveys [49]. Differences in accounting for the editor. Disability policies and government programs. New York: Praeger; institutionalized population may further have affected 1979. p. 1–15. the heterogeneity in the size of the association of the 3. Verbrugge LM, Jette AM. The disablement process. Soc Sci Med. 1994;38:1–14. 4. Wood PHN. The language of disablement: a glossary relating to disease and GALI with other health measures [50]. Although, with- its consequences. Int Rehab Med. 1980;2:86–92. out focus on validity, a recent study, using the 2013 Da- 5. World Health Organization. International Classification of Impairments, nish Health and Morbidity survey and 31 health-related Disabilities, and Handicaps. Geneva: World Health Organization; 1980. http:// apps.who.int/iris/bitstream/10665/41003/1/9241541261_eng.pdf indicators, corroborates the concurrent validity describ- 6. World Health Organization. Towards a common language for functioning, ing a trend towards poorer health and wellbeing in sub- disability and health: ICF. Geneva: WHO; 2002. p. 1–23. http://www.who.int/ jects with more participation restriction as well as a classifications/icf/training/icfbeginnersguide.pdf Van Oyen et al. Archives of Public Health (2018) 76:25 Page 11 of 11 7. Dijkers MP. Issues in the conceptualization and measurement of 31. Cabrero-Garcia J, Julia-Sanchis R. The global activity limitation index mainly participation: an overview. Arch Phys Med Rehabil. 2010;91:S5–16. measured functional disability, whereas self-rated health measured physical 8. Badley EM. Enhancing the conceptual clarity of the activity and morbidity. J Clin Epidemiol. 2014;67:468–76. participation components of the international classification of functioning, 32. Van Oyen H, Van der Heyden J, Perenboom R, Jagger C. Monitoring disability, and health. Soc Sci Med. 2008;66:2335–45. population disability: evaluation of a new global activity limitation Indicator (GALI). Soz Praventiv Med. 2006;51:153–61. 9. Berger N, Robine JM, Ojima T, Madans J, Van Oyen H. Harmonising summary measures of population health using global survey instruments. J 33. Berger N, Van Oyen H, Cambois E, Fouweather T, Jagger C, Nusselder W, et Epidemiol Community Health. 2016;70:1039–44. al. Assessing the validity of the global activity limitation Indicator in fourteen European countries. BMC Med Res Methodol. 2015;15:1. 10. Ustun TB, Chatterji S, Kostanjsek N, Rehm J, Kennedy C, Epping-Jordan J, et 34. Jagger C, Gillies C, Cambois E, Van Oyen H, Nusselder W, Robine JM. The al. Developing the World Health Organization disability assessment global activity limitation index measured function and disability similarly schedule 2.0. Bull World Health Organ. 2010;88:815–23. across European countries. J Clin Epidemiol. 2010;63:892–9. 11. Madans JH, Loeb ME, Altman BM. Measuring disability and monitoring the UN 35. Van der Heyden J, Van Oyen H, Berger N, De Bacquer D, Van Herck K. convention on the rights of persons with disabilities: the work of the Washington Activity limitations predict health care expenditures in the general group on disability statistics. BMC Public Health. 2011;11(Suppl 4):S4. population in Belgium. BMC Public Health. 2015;15:267. 12. Verbrugge LM. Disability experience and measurement. J Aging Health. 36. Berger N, Van der Heyden J, Van Oyen H. The global activity limitation 2016;28:1124–58. indicator and self-rated health: two complementary predictors of mortality. 13. Verbrugge LM. A global disability indicator: companion to self-rated health. Arch Public Health. 2015;73:25. In: Schechter S, editor. Cognitive Methods Staff Working Paper Series, No. 37. Van der Heyden J, Berger N, Van Oyen H. Comparison of self-rated health 10. Hyattsville: Office of Research and Methodology, NCHS; 1994. p. 60–88. and activity limitation as predictors of short term mortality in the older http://www.srl.uic.edu/links/CMS_WP10_Schechter_1994_Verbrugge.pdf. population. Public Health. 2015;129:283–5. 14. Verbrugge LM, Merrill SS, Liu X. Measuring disability with parsimony. Disabil 38. Verropoulou G. Specific versus general self-reported health indicators Rehabil. 1999;21:295–306. predicting mortality among older adults in Europe: disparities by gender 15. Robine JM, Jagger C, Egidi V, Gispert M, Nusselder W, Perenboom R, et al. employing SHARE longitudinal data. Int J Public Health. 2014;59:665–78. Creating a coherent set of indicators to monitor health across Europe: the 39. Cox B, Van Oyen H, Cambois E, Jagger C, Le Roy S, Robine JM, et al. The Euro-REVES 2 project. Eur J Publ Health. 2003;13:6–14. reliability of the minimum European health module. Int J Public Health. 16. Robine JM, Jagger C, Romieu I. Selection of a Coherent Set of Health 2009;54:55–60. Indicators for the European Union, Phase II. 1–129. Montpellier: Euro-Reves; 40. Leonardi M, Bickenbach J, Ustun TB, Kostanjsek N, Chatterji S. The definition 2002. https://ec.europa.eu/health/ph_projects/2000/monitoring/fp_ of disability: what is in a name? Lancet. 2006;368:1219–21. monitoring_2000_frep_03_en.pdf 41. Mont D. Measuring health and disability. Lancet. 2007;369:1658–63. 17. Eurostat. Minimum European Health Module. In: Glossary:Minimum 42. Freedman VA, Kasper JD, Spillman BC, Agree EM, Mor V, Wallace RB, et al. European Health Module (MEHM) - Statistics Explained; 2017. Behavioral adaptation and late-life disability: a new spectrum for assessing 18. Perenboom R, Van Oyen H, van Herten L: Limitations in usual activities, a public health impacts. Am J Public Health. 2014;104:e88–94. global approach. In Selection of a Coherent Set of Health Indicators for the 43. Leveille SG, Fried LP, McMullen W, Guralnik JM. Advancing the taxonomy of European Union, Phase II. Edited by Robine JM, Jagger C, Romieu I. Chapter disability in older adults. J Gerontol A Biol Sci Med Sci. 2004;59:86–93. 5. 68–80. 2002. Montpellier: Euro-Reves. https://ec.europa.eu/health/ph_ 44. Chan A, Saito Y, Robine JM. International perspectives on summary projects/2000/monitoring/fp_monitoring_2000_frep_03_en.pdf. measures of population health in an aging world. J Aging Health. 2016;28: 19. World Health Organization. International Classification of Functioning, 1119–23. Disability and Health: ICF. Geneva: World Health Organization; 2001. WHO | 45. Cambois E, Grobon S, Van Oyen H, Robine JM. Impact of question wording International Classification of Functioning, Disability and Health (ICF) on the measurement of activity limitation: evidence from a randomized test 20. Robine JM, Jagger C, Van Oyen H, Cambois E, Doblhammer G, Nusselder W, in France. J Aging Health. 2016;28:1315–38. et al. The minimum European health module. 2010_4.6. Montpellier: 46. Foraker RE, Rose KM, Chang PP, McNeill AM, Suchindran CM, Selvin E, et al. EHEMU. EHEMU Technical report; 2010. p. 1–29. http://www.eurohex.eu/pdf/ Socioeconomic status and the trajectory of self-rated health. Age Ageing. Reports_2010/2010TR4.6_Health%20Module.pdf 2011;40:706–11. 21. Lagiewka K. European innovation partnership on active and healthy ageing: 47. Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty- what have been the policy drivers and determinants to set a headline seven community studies. J Health Soc Behav. 1997;38:21–37. target of 2 additional healthy life years at birth at EU average by 2020? Arch 48. Mossey JM, Shapiro E. Self-rated health: a predictor of mortality among the Public Health. 2012;70:23. elderly. Am J Public Health. 1982;72:800–8. 22. Zaidi A, Gasior K, Hofmarcher M, Lelkes O, Marin B, Rodrigues R, et al. Active 49. Van der Heyden J, De Backer D, Tafforeau J, Van Herck K. Reliability and ageing index 2012. Concept, methodology, and final results. 1–76. Vienna: validity of a global question on self-reported chronic morbidity. J Public European Centre Vienna, UNECE; 2013. http://citeseerx.ist.psu.edu/viewdoc/ Health (Springer). 2014;22:371–80. download;jsessionid=D965DBF5E0AA4F6E8C3B2D173CD1DD07?doi=10.1.1. 50. Cambois E, Jagger C, Nusselder W, Van Oyen H, Robine JM. International 360.9488&rep=rep1&type=pdf comparisons of disability prevalence estimates: impact of accounting or not 23. Gouvernement Français. Esperance de vie en bonne santé. 2016. 17–2-2017. accounting for the institutionalized population. In: Altman BM, editor. Espérance de vie | Gouvernement.fr. International measurement of disability: purpose, method and application, 24. Leonardi M. Measuring health and disability: supporting policy the work of the Washington group, vol. 350. Hyattsville: National Center for development. The European MHADIE project. Disabil Rehabil. 2010;32:1–8. Health Statistics; 2016. p. 207–29. 25. United Nations. Convention on the Rights of Persons with Disabilities : 51. Johnsen NF, Davidsen M, Michelsen SI, Juel K. Health profile for Danish resolution adopted by the General Assembly. New York: UN General adults with activity limitation: a cross-sectional study. BMC Public Health. Assembly; 2007. Convention on the Rights of Persons with Disabilities [A/ 2017;18:46. RES/61/106] | United Nations Enable 26. European Commission. European Pillar of Social Rights. 2017. European Pillar of Social Rights | European Commission. 27. Romaniuk P, Brukalo K. Designing solutions for the retirement system - in search of balance between economy and health. Front Public Health. 2016;4:184. 28. Drost EA, editor. Validity and reliability in social science research. Educ Res Perspect. 2011;38:105–23. 29. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097. 30. Groves R, Lyberg L, editors. Total survey error. Past, present and future. Public Opin Q. 2010;74:849–79.

Journal

Archives of Public HealthSpringer Journals

Published: May 28, 2018

References

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