ARCHIVED - Chronic Diseases in Canada

Volume 29, no. 4, October 2009

A deprivation index for health planning in Canada

R. Pampalon, PhD (1); D. Hamel, MSc (1); P. Gamache, BSc (1); G. Raymond, BSc (2)

https://doi.org/10.24095/hpcdp.29.4.05

Author References

1. Institut national de santé publique du Québec, Québec, QC
2. Ministère de la santé et des services sociaux du Québec, Québec, QC

Correspondence: Robert Pampalon, PhD, Institut national de santé publique du Québec, 945 Wolfe Street, Québec, QC,  G1V 5B3, Tel.: (418) 650-5115 ext. 5719, Fax: (418) 643-5099, Email: robert.pampalon@inspq.qc.ca

Abstract

Administrative databases in the Canadian health sector do not contain socio-economic information. To facilitate the monitoring of social inequalities for health planning, this study proposes a material and social deprivation index for Canada. After explaining the concept of deprivation, we describe the methodological aspects of the index and apply it to the example of premature mortality (i.e. death before the age of 75). We illustrate variations in deprivation and the links between deprivation and mortality nationwide and in different geographic areas including the census metropolitan areas (CMAs) of Toronto, Montréal and Vancouver; other CMAs; average-size cities, referred to as census agglomerations (CAs); small towns and rural communities; and five regions of Canada, namely Atlantic, Quebec, Ontario, the Prairies and British Columbia. Material and social deprivation and their links to mortality vary considerably by geographic area. We comment on the results as well as the limitations of the index and its advantages for health planning.

Key words: Social inequalities, deprivation, health, health planning, premature mortality, Canada, geographical areas, metropolitan areas, urban areas, regions

Introduction

At a recent Canadian conference on health indicators, the participants proposed a list of 150 indicators as a means of giving the public, care providers and health authorities reliable and comparable data on health and the health system.^1a The participants also pointed out the need to report on inequalities in health, especially those resulting from socio-economic status and urban or rural location of residence.

Since the late 1970s, the production of health surveys such as the Canada Health Survey,2 the National Population Health Survey (NPHS)3 and the Canadian Community Health Survey (CCHS)4 have addressed this need. They contain general measures of health and health service use, as well as information on income, education, family structure and other socio-economic characteristics of respondents which can easily be cross-tabulated. The same cannot be said of the administrative databases created by provincial authorities to track the progression of vital statistics, such as mortality, or the use of health services, such as hospital admissions and primary care; these databases contain no socio-economic data on the individuals concerned.

To make up for this shortcoming, researchers generally use geographic proxies. These pieces of socio-economic information relate to small areas that can be introduced into administrative databases by linking these areas to the data available in the databases. This approach was initiated in Great Britain^5a and then introduced to other countries,^6-8 including Canada.^9a-13

All the Canadian studies that have used geographic proxies tracked social inequalities in health, generally using mortality as a health indicator, although some also considered measures of morbidity and use of health services. These analyses have also largely focused on urban areas and have tended to use only one indicator of social disparity—neighbourhood income.

The contribution made by these studies is undeniable. However, while income is a powerful indicator of health and has ramifications for other determinants of health, it cannot take the place of all those other determinants.^{14a, 15a} This is why more complex measures, namely deprivation indexes, have been developed in Great Britain^16-23 and elsewhere in Europe (Sweden,²⁴ Italy,^25a Spain,²⁶ France^27a), as well as in the United States,^{27b, 28a, 29} Japan³⁰ and New Zealand.^31a Such indexes cover a wide range of domains, from material deprivation alone^{17, 20a, 23} to seven separate domains, including income, employment, health, education, crime, housing and living environment.²² Such indexes have already been proposed in Canada, namely in Manitoba and Quebec, and in the metropolitan area of Vancouver.^32-35 They vary substantially in content and design and none covers Canada as a whole.

The deprivation index developed in Quebec has been widely used in the health sector. Since 2000, it has been introduced into a dozen administrative databases covering mortality, births, hospital admissions, medical services, nursing homes, youth protection services, clients of Centres Locaux de Services Communautaires (CLSCs) and community organizations.^34,36a-44 The index is also being used for regional resource allocation in Quebec^40a and is associated with various products (for example, SAS program used to assign the index, population tables based on the index, interactive index mapping, etc.) that are available free of charge on the Internet.⁴⁵

We propose a national version of the deprivation index developed in Quebec. Our goal is to describe the conceptual and methodological bases of the index, to explore its validity and variations according to geographic areas that reflect Canada’s diversity, and to illustrate its use in health planning through a single example—premature mortality.

We begin by defining the concept of deprivation and go on to describe the data and methods that are the foundation of the index. We present the results in relation to premature mortality and discuss the advantages of using the index in health planning.

The concept of deprivation

In the mid-1980s, Peter Townsend proposed a definition of deprivation as “a state of observable and demonstrable disadvantage relative to the local community or the wider society or nation to which an individual, family or group belongs.”^46a This disadvantage may occur at various levels, for example, with regard to food, housing, education, work or social ties. A person is considered deprived if he or she falls below the level attained by the majority of the population or below what is considered socially acceptable. Townsend distinguishes two forms of deprivation: material and social. Material deprivation involves deprivation of the goods and conveniences that are part of modern life, such as adequate housing, a car, a television set, or a neighbourhood with recreational areas. Social deprivation refers to relations within the family and in the workplace and community.

This idea of deprivation is related to a number of other concepts. Material deprivation evokes the concept of poverty, ^47a as in a lack of financial resources. For Townsend, however, poverty leads to deprivation in that it stands in the way of the acquisition of the goods and conveniences that are part of modern life. Social deprivation is related to the concept of social capital^47b and associated concepts, such as social fragmentation⁴⁸ and social isolation.^49a In all cases, it is a question of the type of social interactions (mutual trust and help, for example), as well as the intensity and quality of such interactions.

In summary, what we need to retain from Townsend’s definition is that deprivation cannot be reduced to a single material or economic dimension; it must also take into account social interactions.

Data and methodology

Basic spatial unit

The deprivation index is based on a spatial unit. Since the index is intended as a substitute for measures of individuals, the selected unit must be as small as possible in order to ensure a high degree of homogeneity in the socio-economic conditions attributed to each resident in this unit.⁸ The selected unit is the dissemination area (DA),⁵⁰ which comprises one or more neighbouring blocks of houses, with a population of 400 to 700 persons.

We constructed the index in two stages. In the first stage, we set aside DAs comprising no population, DAs with a high proportion of collective households or institutionalized persons (more than 15% of the total population or over 80 people living in collective households), DAs that had no B profile (socio-economic) or income data (sparsely populated DAs), and DAs in Nunavut Territory or located on a First Nations reserve. This left 42 430 DAs covering slightly more than 93% of the Canadian population. In the second stage, we projected the obtained deprivation values onto an additional number of DAs, including the DAs for which it was possible to adequately impute* an income value (3572 DAs); the DAs located in Nunavut and on First Nations reserves with a complete B profile or imputed income value (857 DAs); and DAs that had been excluded due to their high proportion of collective households or institutionalized persons but whose population with a B profile (or imputed income) accounted for more than 85% of the total population (605 DAs). As a result, a deprivation index was established for 47 464 DAs, or close to 98% of the population of Canada.

Socio-economic indicators

The indicators used to construct the index were selected on the basis of a literature review. To be selected, indicators needed to meet four criteria: have known links with health, previous use as geographic proxies, affinity with the material or social dimensions of deprivation, and availability by DA.05b, 14b, 15b, 25b, 35a, 49b, 51a, 52a This approach made it possible to identify the six indicators that were taken into account to construct the index: the proportion of people aged 15 years and older with no high school diploma (SCOLAR); the employment/population ratio of people aged 15 years and older (EMPLOI); the average income of people aged 15 years and older (REVENU); the proportion of individuals aged 15 years and older living alone (SEULES); the proportion of individuals aged 15 years and older who are separated, divorced or widowed (S_D_V); and the proportion of single-parent families (F_MONO).†

In some instances, the selected indicators varied significantly with the age and sex of the population. This was true of education, for example, since many young people less than 20 years old have not completed their schooling while many older people have a low level of education. Since the variations being tracked are socio-economic rather than demographic, these indicators, with the exception of F_MONO, were adjusted according to the age-sex structure of the Canadian population^31b using direct standardization.⁵³ Moreover, certain indicators were transformed in order to normalize their distribution.⁵⁴ For example, the REVENU variable was transformed into its log values and the SEULES variable into its arcsine values.

Integration of indicators

The integration of indicators in the form of a deprivation index was carried out using principal component analysis (PCA), the preferred approach for developing such indexes.25c-29, 31c, 49c, 55 This analysis yields fewer dimensions, reflecting the spatial organization of socio-economic indicators. A varimax rotation was applied to these dimensions to increase readability and to make them independent (or orthogonal). To validate the relevance of this factor structure across Canada, the PCA was repeated for the three largest census metropolitan areas (major CMAs), Toronto, Montréal and Vancouver; various other geographic areas, namely other CMAs, census agglomerations (CAs), small towns and rural communities; and each of five regions, Atlantic, Quebec, Ontario, the Prairies and British Columbia. The literature shows that measures of deprivation perform differently in urban and rural settings.52b, 56-60

For each component identified, the PCA produces a factor score which represents the value of the component in each DA. To ensure statistical accuracy in analyzing social inequalities in health, the DAs were grouped together. The DAs were first ranked according to their factor score from the most to the least privileged. Then, the distribution of DAs was broken down into quintiles, with each quintile representing 20% of the population. Quintile 1 (Q1) represents the most privileged population and quintile 5 (Q5), the least. These operations were carried out separately for each component of the analysis. Since deprivation is seen as a relative disadvantage compared with the community to which people belong, different versions of the index were produced by modifying the reference territory. Accordingly, there is a national version, a version by major CMA, a version by geographic area and a version by region. These versions are based on the PCA conducted in each setting and on the distribution of factor scores, ensuring an equal distribution of the population (20%) per material and social quintile.

Any of these versions can be used to reflect the discrepancies in deprivation that exist in each setting and also to compare populations of the same proportion. In the following analysis, the version of the index varies according to the reference territory considered. The values presented for Canada as a whole stem from the national version. Those presented for the geographic areas, major CMAs and regions of Canada stem from the geographic area, major CMA, and region of Canada versions, respectively.

Premature mortality

To illustrate how the index can be used to study socio-economic indicators of health for the purpose of health planning, we use the example of premature mortality, or death before 75 years of age. This is a general measure of population health^1b whose relationships with socio-economic conditions have been extensively documented on an international scale.^61-66

Taking into account deaths in 2001 and using the reference population from the census of the same year, we estimated the mortality rates using the negative binomial regression model, a generalization of the Poisson regression model that takes into account the problem of over-dispersion.⁶⁷ We estimated models in each geographic area for all deaths (all causes combined) and the entire population (both sexes combined). In these models, mortality rates were estimated for each quintile of material and social deprivation, from the most privileged (Q1) to the most deprived (Q5), and for the extreme quintiles on both dimensions (Q1-Q1 and Q5-Q5), adjusting for age, sex and, where applicable, geographic area and the other form of deprivation (material or social). Thus, when the mortality rate varies with both forms of deprivation simultaneously, this signifies that each form of deprivation is contributing to mortality independently. An interaction term between the two forms of deprivation was introduced into the models when a significance threshold of 5% was reached. The variability of adjusted rates was estimated using the Delta method.⁶⁸

To obtain a satisfactory portrait of the inequalities in mortality according to deprivation, we considered three measures: the adjusted mortality rate, the ratio, and the difference in adjusted mortality rates.^69-71 The mortality rate (per 100 000 inhabitants) expresses the level of mortality in each group. The ratio and the difference in the mortality rates illustrate, respectively, the relative and absolute discrepancies in mortality rates between groups at the extreme ends of the deprivation spectrum. Taking both forms of deprivation into consideration simultaneously, the mortality ratio is obtained by dividing the rate for the most deprived group (Q5-Q5) by that of the most privileged group (Q1-Q1). The difference in mortality is obtained by subtracting the rate for the most privileged group (Q1-Q1) from that of the most deprived group (Q5-Q5).

Results

The deprivation index

The deprivation index covers almost 98% of the Canadian population, and this percentage varies little from one geographic area to the next (Table 1). The index covers close to 90% of DAs in Canada, with a higher proportion of DAs in CMAs than in smaller towns and rural communities. This discrepancy is due to the greater number of DAs with no population in smaller towns and rural communities.

The Canada-wide PCA reveals the presence of a two-component factor structure (Table 2). Each of these components summarizes approximately one-third of the variations associated with the six indicators considered, for a total of 67% of these variations. The meaning of the components differs considerably. Whereas the first component primarily portrays variations associated with education, employment and income, the second indicates the state of being separated, divorced or widowed, living alone, or being a member of a single-parent family. This configuration echoes Townsend’s material and social dimensions of deprivation.^46b For this reason, and to facilitate the ensuing analysis and discussion, these two components will be referred to as material and social. The PCAs carried out in the different geographic areas and regions of Canada show that these two components were present everywhere, with the exception that the proportion of single-parent families in CMAs is associated with both material and social components equally (Table 3 and Table 4). The explained variance for both components is only slightly lower in small towns and rural communities and decreases from east to west across the country.

Material and social deprivation in Canada

The deprivation index reveals appreciable discrepancies in socio-economic conditions in Canada (Table 5). Material deprivation is accompanied by well-recognized variations in education, employment and income and, to a lesser degree, single-parent families. Social deprivation is more prevalent with single-parent families, with people living alone, and with those who are separated, divorced or widowed. This form of deprivation is also not totally independent of employment and income, as well as a certain degree of population aging, even though the indicators considered have been adjusted for age. By combining the two forms of deprivation and comparing the most privileged population (Q1 and Q1) with the least privileged population (Q5 and Q5) on both the material and social dimensions simultaneously, we note startling contrasts for all indicators that make up the deprivation index.

Such contrasts are observed across Canada, regardless of the geographic area or region (Table 6). However, the magnitude of socio-economic disparities can vary by area or region. The discrepancies in material and social deprivation are generally greater in the major CMAs than in census agglomerations (CAs), small towns and rural communities, and the Atlantic region. Conversely, however, the average level of material deprivation is lower in CMAs than in small towns and rural communities, and the Atlantic region.

Deprivation and premature mortality in Canada

Approximately 94% of premature deaths in 2001 were given a deprivation index, for a total of 85 614 deaths (Table 7). Of the deaths that were not given a deprivation index (n = 5 625), 14% were the result of erroneous postal codes and 86% were the result of DAs with no index corresponding to institutionalized populations.

The adjusted premature mortality rate in 2001, 310 deaths per 100 000, progresses in line with both material and social deprivation (Figure 1). The mortality ratio between material and social deprivation groups at extreme ends of the spectrum is 2.41 and the difference in mortality is 302 deaths per 100 000, a value equivalent to that observed for Canada as a whole.

While such discrepancies can be seen everywhere in Canada, their magnitude varies enormously by geographic area and region. Accordingly, among the most deprived individuals in Canada, we find that those who live in CAs as well as in small towns and rural communities have the highest rates of premature death (Figure 2). Conversely, in small towns and rural communities, the relative and absolute discrepancies in the mortality rate (ratio and difference) according to deprivation are relatively low (Figure 3). In terms of the regions of Canada, the greatest disparities in mortality according to deprivation are found in the Prairies and in British Columbia, whereas at the CMA level, they are seen in Vancouver and in the “other CMAs” group. Of the three major Canadian CMAs, Toronto has the smallest disparities.

Figure 1
Premature mortality rate by quintile of material and social deprivation Canada, 2001

Figure 2
Premature mortality rate in the most and least deprived persons (material and social) by geographic area and region of Canada, 2001

Figure 3
Ratio and difference in premature mortality between the most and least deprived persons (material and social) by geographic area and region of Canada, 2001

Discussion

The deprivation index comprises six indicators grouped into two components, material deprivation and social deprivation. These two components occur nationwide, in rural settings and in all the various urban settings (large CMAs, other CMAs and CAs). They point to major socio-economic inequalities in income, education, employment and family structure everywhere, demonstrating the relevance and applicability of the index beyond the urban settings that are usually preferred for the production of geographic proxies.^{09b, 11-13, 33, 35b}

Variations in the deprivation index are closely linked to variations in premature mortality. Material and social deprivation contribute independently to mortality, and this contribution increases gradually with the level of deprivation (Figure 1). Such gradients can be observed everywhere in Canada, including in large CMAs and other geographic areas, and in all regions. (Data not presented, available upon request.) Thus, deprivation not only affects groups that are extremely deprived: it is a matter of concern for the entire population.

The combined effect of the two forms of deprivation can be observed by comparing the mortality of groups at the extreme ends of social and material deprivation - Q5Q5 vs. Q1Q1 (Figure 2 and Figure 3). The combined effect is also observable—although in a less marked fashion—in populations whose size is similar to that of populations considered for each dimension separately, that is, on a quintile basis. Hence, in Canada, the mortality rate ratio and rate difference between extreme quintiles (Q5 vs. Q1) were, respectively, 1.82 (95% CI, 1.73-1.92) and 192 deaths per 100 000 (95% CI, 174-210) when both dimensions of deprivation are considered simultaneously, as opposed to 1.50 (95% CI, 1.45-1.55) and 125 deaths (95% CI, 115-136) for the material dimension and 1.65 (95% CI, 1.60-1.70) and 161 deaths (95% CI, 151-172) for the social dimension, treated separately. Similar differences can be seen in the various geographic settings. (Data available upon request.)

Other studies have already identified social disparities in mortality in Canadian CMAs.^09c-12 This study shows that these inequalities extend to all geographic areas that reflect Canada’s diversity. Due to different study methodologies and due to the absence of research on the geography of social disparities in health—in Canada and internationally—it is difficult to compare these results with those obtained elsewhere. In fact, the deprivation index sheds new light on the social disparities in health in Canada by expressing their variability by geographic area.

These initial results on premature mortality require further study, either to identify the exact causes of death, to determine if there is a difference in effect on sex, or to decipher the underlying factors. For instance, it would be interesting to explore factors such as relative deprivation,⁷² the presence of Aboriginal people,⁷³ recent immigration,⁷⁴ and the risks associated with the use of geographic proxies.⁷⁵ The use of such proxies may explain, at least in part, the presence of weak ratios and differences in mortality in small towns and rural communities. These initial results could also be compared to those generated with respect to other socio-economic indicators, such as low income. In order to be useful and correctly carried out, however, such a comparison should consider several socio-economic and health indicators simultaneously, with due attention paid to their conceptual foundations and their performance in relation to technical and political criteria,^{5c, 7b, 20b, 21, 28b, 35c, 51b, 76-78} an exercise that is well beyond the framework of this study.

The deprivation index has its limitations. It is not an individual measure of socio-economic conditions, but rather a measure of the conditions seen at the neighbourhood level. The index could be used in an etiological analysis, but it cannot replace an individual measure, which is the only way of portraying individual or family education, for example. Therefore, in an etiological analysis, these two types of measures should be considered simultaneously, through multilevel modelling.⁷⁹ This is now possible thanks to a new file combining a sample from the 1991 Census of Canada with mortality data from 1991 to 2001.⁸⁰

Combating social inequalities in health has become a major challenge for health systems, both in Canada⁸¹ and around the world.⁸² The availability of tools to measure inequalities is a prerequisite to any planning to reduce them. In Quebec, the deprivation index is now used at every stage of the health planning process, including the measurement and monitoring of inequalities,^{36b-38, 42, 43} the development of strategic goals, ⁸³ the evaluation of both provincial and local services ^{40b, 44} and resource allocation to the regions.⁸⁴

A recent study by the Canadian Institute for Health Information (CIHI)⁸⁵ demonstrated the existence of clear gradients in hospital admissions and in self-reported health in 15 CMAs, based on this deprivation index. The relevance and usefulness of a measure often become apparent only when the measure is put to use. The Canadian index of material and social deprivation is therefore available for trial by researchers and managers in the health sector. It is also associated with a variety of products now available on the Institut national de santé publique du Québec (INSPQ) website.⁸⁶

Acknowledgements

We would like to thank Russell Wilkins of Statistics Canada for facilitating access to the Canadian Mortality Data. However, analysis of these data and the opinions expressed in this text are not those of Statistics Canada. We would also like to thank Robert Choinière of the Institut national de santé publique du Québec, as well as two anonymous reviewers, for their input on the initial version of this text.

Footnotes

 ^{*^} Imputation of income was by the nearest-neighbour method, based on the Euclidian distance between the other indicators (other than income) included in the deprivation index, using the SAS FASTCLUS procedure.

 ^{†^} Families include couples with or without children and single parents with at least one child. ⁵⁰

References

1. ^ a,b Canadian Institute for Health Information. The Health Indicators Project: The next 5 years. Report from the Second Consensus Conference on Population Health Indicators. Ottawa (ON): Canadian Institute for Health Information; 2005.
2. ^ Statistics Canada (Health Division). A review of national health surveys in Canada, 1978-1987. In: Health Reports-Prototype. Ottawa; 1988. p. 28-50. Cat. No. 82-003-XPE.
3. ^ Swain L, Catlin G, Beaudet MP. The National Population Health Survey—its longitudinal nature. Health Rep. 1999;10(4):69-82.
4. ^ Statistics Canada. Canadian Community Health survey profiles 2003 [Internet]. Ottawa (ON): Statistics Canada; [modified 2005 Jun 30]. Available from: http://www.statcan.gc.ca/pub/ 82-576-x/82-576-x2005001-eng.htm.
5. ^ a,b,c Gordon D. Area-based deprivation measures: A U.K. perspective. In: Kawachi I, Berkman L, editors. Neighborhoods and health. New York: Oxford University Press; 2003. p. 179-210.
6. ^ Kunst AE, Mackenbach JP. Measuring socioeconomic inequalities in health. Copenhagen: World Health Organization; 1994.
7. ^ a,b Krieger N, Chen JT, Waterman PD, Soobader MJ, Subramanian SV, Carson R. Geocoding and monitoring of US socioeconomic inequalities in mortality and cancer incidence: does the choice of area-based measure and geographic level matter?: the Public Health Disparities Geocoding Project. Am J Epidemiol. 2002 Sep 1;156(5):471-82.
8. ^ Krieger N, Williams DR, Moss NE. Measuring social class in US public health research: concepts, methodologies, and guidelines. Annu Rev Public Health. 1997;18:341-78.
9. ^{^ a,b,c} Choinière R. Évolution des disparités de la mortalité selon le revenu à Montréal. Journées annuelles de santé publique, editor. 2002. Unpublished Work.
10. ^ Mustard CA, Derksen S, Berthelot JM, Wolfson M. Assessing geographic proxies for household income: a comparison of household and neighbourhood level income measures in the study of population health status. Health Place. 1999 Jun;5(2):157-71.
11. ^ Roos NP, Mustard CA. Variation in health and health use by socioeconomic status in Winnipeg, Canada: does the system work well? Yes and no. Milbank Q. 1997;75(1):89-111.
12. ^ Wilkins R, Berthelot JM, Ng E. Trends in mortality by neighbourhood income in urban Canada from 1971 to 1996. Health Rep, Supplement;2002;13:1-28.
13. ^ Wilkins R, Ng E, Berthelot JM, Mayer F. Provincial differences in disability-free life expectancy by neighbourhood income and education in Canada, 1996. Ottawa (ON): Statistics Canada; 2002. Unpublished Work.
14. ^ a,b Marmot M, Wilkinson RG, editors. Social Determinants of Health. 2nd ed. Copenhagen: World Health Organization; 2003.
15. ^ a,b Leclerc A, Fassin D, Grandjean H, Kaminski M, Lang T. Les inégalités sociales de santé. Paris: Éditions La Découverte et Syros; 2000.
16. ^ Adams J, Ryan V, White M. How accurate are Townsend Deprivation Scores as predictors of self-reported health? A comparison with individual level data. J Public Health (Oxf). 2005 Mar;27(1):101-6.
17. ^ Carstairs V, Morris R. Deprivation: explaining differences in mortality between Scotland and England and Wales. BMJ. 1989 Oct 7;299(6704):886-9.
18. ^ Carstairs V. Deprivation indices: their interpretation and use in relation to health. J Epidemiol Community Health. 1995 Dec;49 Suppl 2:S3-S8.
19. ^ Jarman B. Identification of underprivileged areas. Br Med J (Clin Res Ed). 1983 May 28;286(6379):1705-9.
20. ^ a,b Jarman B, Townsend P, Carstairs V. Deprivation indices. BMJ. 1991 Aug 31;303(6801):523.
21. ^ Morris R, Carstairs V. Which deprivation? A comparison of selected deprivation indexes. J Public Health Med. 1991 Nov;13(4):318-26.
22. ^ Noble M, McLennan D, Wilkinson K, Whitworth A, Barnes H. The English Indices of Deprivation 2007. London (UK): Department for Communities and Local Government Publications; 2008.
23. ^ Townsend P, Phillimore P, Beattie A. Health and Deprivation: Inequalities and the North. London: Croom Helm; 1988.
24. ^ Bajekal M, Jan S, Jarman B. The Swedish UPA score: an administrative tool for identification of underprivileged areas. Scand J Soc Med. 1996 Sep;24(3):177-84.
25. ^ a,b,c Tello JE, Jones J, Bonizzato P, Mazzi M, Amaddeo F, Tansella M. A census-based socio-economic status (SES) index as a tool to examine the relationship between mental health services use and deprivation. Soc Sci Med. 2005 Nov;61(10):2096-105.
26. ^ Benach J, Yasui Y, Borrell C, Pasarin MI, Martinez JM, Daponte A. The public health burden of material deprivation: excess mortality in leading causes of death in Spain. Prev Med. 2003 Mar;36(3):300-8.
27. ^ a,b Challier B, Viel JF. Relevance and validity of a new French composite index to measure poverty on a geographical level. Rev Epidemiol Santé Publique. 2001 Feb;49(1):41-50.
28. ^ a,b Krieger N, Chen JT, Waterman PD, Soobader MJ, Subramanian SV, Carson R. Choosing area based socioeconomic measures to monitor social inequalities in low birth weight and childhood lead poisoning: The Public Health Disparities Geocoding Project (US). J Epidemiol Community Health. 2003 Mar;57(3):186-99.
29. ^ Hogan JW, Tchernis R. Bayesian factor analysis for spatially correlated data, with application to summarizing area-level material deprivation from census data. J Am Stat Assoc. 2004 Jun 1;99(466):314-24.
30. ^ Kunst AE. Commentary: Using geographical data to monitor socioeconomic inequalities in mortality: experiences from Japanese studies. Int J Epidemiol. 2005 Feb;34(1):110-2.
31. ^ a,b,c Salmond C, Crampton P, Sutton F. NZDep91: A New Zealand index of deprivation. Aust N Z J Public Health. 1998 Dec;22(7):835-7.
32. ^ Frohlich N, Mustard C. A regional comparison of socioeconomic and health indices in a Canadian province. Soc Sci Med. 1996 May;42(9):1273-81.
33. ^ Matheson FI, Moineddin R, Dunn JR, Creatore MI, Gozdyra P, Glazier RH. Urban neighborhoods, chronic stress, gender and depression. Soc Sci Med. 2006 Nov;63(10):2604-16.
34. ^ Pampalon R, Raymond G. A deprivation index for health and welfare planning in Quebec. Chronic Dis Can. 2000;21(3):104-13.
35. ^ a,b,c Schuurman N, Bell N, Dunn JR, Oliver L. Deprivation indices, population health and geography: an evaluation of the spatial effectiveness of indices at multiple scales. J Urban Health. 2007 Jul;84(4):591-603.
36. ^{^ a,b} Dupont MA, Pampalon R, Hamel D. Deprivation and cancer mortality Among Quebec Women and Men, 1994-1998. Québec (QC). Institut national de santé publique du Québec. 2004.
37. ^ Hamel D, Pampalon R. Trauma and deprivation in Québec. Québec (QC). Institut national de santé publique du Québec; 2002.
38. ^ Martinez J, Pampalon R, Hamel D. Deprivation and stroke mortality in Quebec. Chronic Dis Can. 2003 Spring-Summer;24(2-3):57-64.
39. ^{^} Pampalon R, Rochon M. Health Expectancy and Deprivation in Québec, 1996-1998. Paper presented at: 13th Annual Meeting of the International Network on Health Expectancies (REVES), Vancouver (BC).
40. ^ a,b Pampalon R, Raymond G. Indice de défavorisation matérielle et sociale: son application au secteur de la santé et du bien-être. Santé, Société et Solidarité. 2003;(1):191-208.
41. ^ Pampalon R, Philibert M, Hamel D. Inégalités sociales et services de proximité au Québec: Développement d’un système d’évaluation issu d’une collaboration entre chercheurs et intervenants. Santé, Société et Solidarité. 2004;(2):73-88.
42. ^ Pampalon R, Hamel D, Gamache P. Évolution récente de la mortalité prématurée au Québec selon la défavorisation matérielle et sociale. In: Frohlich K, De Koninck M, Bernard P, Demers A, editors. Les inégalités sociales de santé au Québec. Montréal (QC): Les Presses de l’Université de Montréal; 2008.
43. ^ Pampalon R, Hamel D, Gamache P. Recent changes in the geography of social disparities in premature mortality in Québec. Soc Sci Med. 2008 Oct;67(8):1269-1281.
44. ^ Philibert M, Pampalon R, Hamel D, Thouez JP, Loiselle C. Material and social deprivation and health and social services utilisation in Québec: a local-scale evaluation system. Soc Sci Med. 2007;64:1651-64.
45. ^{^} Ministère de la Santé et des Services sociaux du Québec [Ministry of Health and Social Services, Quebec]. Variations nationales de l’indice de défavorisation en 2001.[Internet]. [Québec (QC)]: Ministère de la Santé et des Services sociaux du Québec: 2009. Available from: http://www.msss.gouv.qc.ca/statistiques/ atlas/atlas/index.php?id_carte=11
46. ^ a,b Townsend P. Deprivation. J Soc Pol. 1987 Apr;16:125-46.
47. ^ a,b Wagle U. Rethinking poverty: definition and measurement. Int Soc Sci J. 2002;54(1/171):155-65.
48. ^ Congdon P. The epidemiology of suicide in London. J Roy Stat Soc Ser A Sta. 1996;159:515-33.
49. ^ a,b,c Curtis S, Copeland A, Fagg J, Congdon P, Almog M, Fitzpatrick J. The ecological relationship between deprivation, social isolation and rates of hospital admission for acute psychiatric care: a comparison of London and New York City. Health Place. 2006 Mar;12(1):19-37.
50. ^ a,b Statistics Canada. 2001 Census Dictionary. Ottawa (ON): Statistics Canada; 2003.
51. ^ a,b Krieger N, Zierler S, Hogan JW, Waterman P, Chen J, Lemieux K, Gjelsvik A. Geocoding and Measurement of Neighborhood socioeconomic Position: A U.S. Perspective. In: Kawachi I, Berkman L, editors. Neighborhoods and Health. New York: Oxford University Press; 2003. p. 147-78.
52. ^ a,b Niggebrugge A, Haynes R, Jones A, Lovett A, Harvey I. The index of multiple deprivation 2000 access domain: a useful indicator for public health? Soc Sci Med. 2005 Jun;60(12):2743-53.
53. ^{^} Muecke C, Hamel D, Bouchard C, Martinez J, Pampalon R, Choinière R. Doit-on utiliser la standardisation directe ou indirecte dans l’analyse de la mortalité à l’échelle des petites unités géographiques? Québec (QC): Institut national de santé publique du Québec: 2005.
54. ^ Gilthorpe MS. The importance of normalisation in the construction of deprivation indices. J Epidemiol Community Health 1995 Dec;49 Suppl 2:S45-S50.
55. ^ Index 99 Team. Index of Deprivation 1999 Review. Final consultation. Report for Formal Consultation. Stage 2: Methodology for an Index of Multiple Deprivation. Oxford: University of Oxford; 1999.
56. ^ Barnett S, Roderick P, Martin D, Diamond I, Wrigley H. Interrelations between three proxies of health need at the small area level: an urban/rural comparison. J Epidemiol Community Health. 2002 Oct;56(10):754-61.
57. ^ Barnett S, Roderick P, Martin D, Diamond I. A multilevel analysis of the effects of rurality and social deprivation on premature limiting long term illness. J Epidemiol Community Health. 2001 Jan;55(1):44-51.
58. ^ Martin D, Brigham P, Roderick P, Barnett S, Diamond I. The (mis)representation of rural deprivation. Environ Plann A. 2000;32:735-51.
59. ^ Gilthorpe MS, Wilson RC. Rural/urban differences in the association between deprivation and healthcare utilisation. Soc Sci Med. 2003 Dec;57(11):2055-63.
60. ^ Haynes R, Gale S. Deprivation and poor health in rural areas: inequalities hidden by averages. Health Place. 2000 Dec;6(4):275-85.
61. ^ Davey-Smith G, Dorling D, Mitchell R, Shaw M. Health inequalities in Britain: continuing increases up to the end of the 20th century. J Epidemiol Community Health. 2002 Jun;56:434-5.
62. ^ James PD, Wilkins R, Detsky AS, Tugwell P, Manuel DG. Avoidable mortality by neighbourhood income in Canada: 25 years after the establishment of universal health insurance. J Epidemiol Community Health. 2007 Apr;61(4):287-96.
63. ^ Leclerc A, Chastang JF, Menvielle G, Luce D. Socioeconomic inequalities in premature mortality in France: have they widened in recent decades? Soc Sci Med. 2006 Apr;62(8):2035-45.
64. ^ Mackenbach JP, Bos V, Andersen O, Cardano M, Costa G, Harding S, Reid A, Hemström O, Valkonen T, Kunst AE. Widening socioeconomic inequalities in mortality in six Western European countries. Int J Epidemiology. 2003;32:830-7.
65. ^ Singh GK, Siahpush M. Increasing inequalities in all-cause and cardiovascular mortality among US adults aged 25-64 years by area socioeconomic status, 1969-1998. Int J Epidemiol. 2002 Jun;31:600-13.
66. ^ Turrell G, Mathers C. Socioeconomic inequalities in all-cause and specific-cause mortality in Australia: 1985-1987 and 1995-1997. Int J Epidemiol. 2001;30:231-9.
67. ^ Allison PD. Logistic regression using SAS system: theory and application. North Carolina: SAS Institute Inc.; 1999.
68. ^ Tiwari HK, Elston RC. The approximate variance of a function of random variables. Biometrical J. 1999;41(3):351-7.
69. ^ Boström G, Rosén M. Measuring social inequalities in health: politics or science? Scand J Public Healt. 2003;31:211-5.
70. ^ Regidor E. Measures of health inequalities: part 1. J Epidemiol Community Health. 2004;58:858-61.
71. ^ Regidor E. Measures of health inequalities: part 2. J Epidemiol Community Health. 2006;58:900-3.
72. ^ Wilkinson RG, Pickett KE. Income inequality and population health: a review and explanation of the evidence. Soc Sci Med. 2006 Apr;62(7):1768-84.
73. ^ Health Canada. Healthy Canadians: A federal report on comparable health indicators 2004. Ottawa (ON): Health Canada; 2004.
74. ^ Chen J, Ng E, Wilkins R. The health of Canada’s immigrants in 1994-95. Health Rep. 1996;7(4):33-50.
75. ^ Wilkins R. Neighbourhood income quintiles derived from Canadian postal codes are apt to be misclassified in rural but not urban areas. Ottawa (ON): Statistics Canada; 2004. Unpublished Work.
76. ^ Ellaway A. Are single indicators of deprivation as useful as composite indicators in predicting morbidity and mortality: results from the Central Clydeside Conurbation. Health Bull (Edinb). 1997 Sep;55(5):283-4.
77. ^ Gordon D. Census based deprivation indices: their weighting and validation. J Epidemiol Community Health. 1995 Dec;49 Suppl 2:S39-S44.
78. ^ Carr-Hill R, Chalmers-Dixon P. The public health observatory handbook of health inequalities measurement. Oxford: South East Public Health Observatory; 2005.
79. ^ Subramanian SV, Jones K, Duncan C. Multilevel Methods for Public Health Research. In: Kawachi I, Berkman L, editors. Neighborhoods and Health. New York: Oxford University Press; 2003. p. 65-111.
80. ^ Wilkins R. Linking two of Canada’s national treasures: mortality follow-up from a census-based sample of the Canadian population. Paper presented at: Annual Meeting of the Canadian Population Society. Congress of the Humanities and Social Sciences; 2005 May 31–Jun 1; London (ON).
81. ^ Butler-Jones D. The Chief Public Health Officer’s report on the state of public health in Canada in 2008: addressing health inequalities. Ottawa (ON): Public Health Agency of Canada; 2008. Cat. No.: HP2-10/2008E.
82. ^ Commission on social determinants of health. Closing the gap in a generation. Health equity through action on the social determinants of health. Geneva: World Health Organization; 2008.
83. ^ Ministère de la Santé et des Services sociaux du Québec. Troisième rapport national sur l’état de santé de la population du Québec. Riches de tous nos enfants: La pauvrete et ses repercussions sur la santé des jeunes de moins de 18 ans. Québec (QC): La directions des communications du ministère de la Santé et des Services sociaux du Québec; 2007.
84. ^ Ministère de la Santé et des Services sociaux du Québec. Nouveau mode d’allocation des ressources 2007-2008. Québec (QC): Direction de l’allocation des ressources, MSSS; 2007. Unpublished Work.
85. ^ Canadian Institute for Health Information. Reducing gaps in health: a focus on socio-economic status in urban Canada. Ottawa (ON): Canadian Institute for Health Information; 2008.
86. ^ Institut national de santé publique du Québec. Indice canadien de défavorisation. [Internet]. Québec (QC): Institut national de santé publique du Québec; 2008. Available from: http://www.inspq.qc.ca/santescope/ liens.asp?comp=9&nav=M

Only feature articles are peer reviewed. Authors retain responsibility for the content of their articles; the opinions expressed are not necessarily those of the CDIC editorial committee nor of the Public Health Agency of Canada.