Social inequalities in male mortality amenable to medical intervention in British Columbia

Social inequalities in male mortality amenable to medicalintervention in British Columbia

Evan Wooda, Anthony M. Sallarb, Martin T. Schechtera, b,Robert S. Hogga, b,*

aBritish Columbia Centre for Excellence in HIV/AIDS, St. Paul's Hospital, Faculty of Medicine, University of British Columbia,

608±1081 Burrard Street, Vancouver, BC, Canada V6Z 1Y6bDepartments of Medicine, Health Care and Epidemiology, Faculty of Medicine, University of British Columbia, Vancouver, BC,

Canada

Abstract

The objective of this study is to examine the rates of mortality among di�erent social classes and socioeconomicgroups of British Columbian males from causes of death amenable to medical intervention. We examined the ratesof avoidable mortality from the causes of death published by Charlton, excluding causes of death restricted to

women as well as perinatal deaths. For the purposes of our study, we determined a population at risk using 20%samples of occupational data for men from the 1981, 1986 and 1991 censuses conducted by Statistics Canada. Forthe analysis of mortality by social class, individuals were divided into ®ve social class levels based on occupation

using an adaptation of the UK Registrar General's Social Class Scale. In addition, three levels of socioeconomicanalysis were performed using the Blishen Index classi®cation system. Once individuals were assigned to a socialclass in each classi®cation system, the death rates from each amenable cause was calculated and standardized to thetotal population. For almost every cause of death examined, the rate of mortality was higher in individuals of lower

social and socioeconomic classes than in individuals of the upper social and socioeconomic classes. These resultswere consistent regardless of the social class component, education, occupation, or income was being measured. Themortality gradient was most notable in deaths due to hypertensive heart disease, tuberculosis, asthma and

pneumonia and bronchitis. Due to the fact that these causes of death were observed to be consistently higher in thelower social classes, we feel that speci®c measures aimed at improving survival from these conditions in lower socialclasses could help to amend the social class disparity. # 1999 Elsevier Science Ltd. All rights reserved.

Keywords: Avoidable mortality; Socioeconomic status; Social class; British Columbia

Introduction

The notion of avoidable mortality in the medical lit-erature derives from the work of Rutstein, who pro-

posed a list of causes of unnecessary untimely death to

be used to assess the quality of medical care (Rutsteinet al., 1976, 1983). This list identi®ed diseases forwhich death was avoidable, if appropriate medical

treatment was provided. Rutstein proposed that bydetermining areas with high rates of avoidable mor-tality one might be able to identify areas of inadequatehealth care delivery.

The concept of avoidable mortality has since been

Social Science & Medicine 48 (1999) 1751±1758

0277-9536/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.

PII: S0277-9536(99 )00081-7

* Corresponding author. Tel.: +1-604-6315516; fax: +1-

604-6315464.

E-mail address: [email protected] (R.S. Hogg)

adopted in a number of studies looking at di�erences

in avoidable mortality between geographical areas and

at the variation of avoidable mortality over time, ®rst

by Charlton and later by a number of other investi-

gators (Charlton et al., 1983; Charlton and Velez,

1986; Carr-Hill et al., 1987; Poikolainen and Eskola,

1988; Mackenbach et al., 1988b; Boys et al., 1991).

Geographical analysis has helped to determine spatialvariation in the rate of avoidable causes of death

within and between countries. These studies have

attempted to explain di�erences in avoidable mortality

rates between geographic areas by the presence or

absence of certain health care services. In studies of

avoidable mortality over time, analysis has generally

shown that the death rate from avoidable causes has

decreased over the last several decades and that thisdecrease has been larger than the decrease in the death

rate from non-avoidable causes. In most time-trend

studies, improvements in health care and its delivery

are used to explain the decrease in deaths from avoid-

able causes (Charlton and Velez, 1986; Mackenbach et

al., 1988b).

More recently, investigators have noted strong as-

sociations between mortality from conditions amenable

to medical intervention and socioeconomic variables

(Mackenbach et al., 1989). Researchers in New

Zealand showed a strong social class gradient in avoid-

able mortality, with lower socioeconomic groups

experiencing higher mortality from avoidable causes ofdeath (Marshall et al., 1993). Studies of avoidable

mortality by social class have de®ned the di�erent

social classes using a number of scales based on occu-

pation, education, and/or income (Mackenbach et al.,

1990; Marshall et al., 1993).

British Columbia, like the rest of Canada, has ex-

perienced rapid metropolitan growth and high immi-

gration rates over the last several decades. This has

required unique planning of the province's universal

health care system, which strives to improve the healthof all citizens regardless of social class or socioeco-nomic status. In this context, the study of di�erential

rates of avoidable mortality among di�erent socialclasses has particular signi®cance since it can serve asan indication of how the current organization andsupply of health care is being utilized by the di�erent

classes of society. The objective of this study is toexamine the rates of mortality among di�erent socialclasses and socioeconomic groups of British

Columbian males from causes of death amenable tomedical intervention.

Data and methods

For the purposes of our study, we determined a

population at risk using 20% samples of occupationaldata for men from the 1981, 1986 and 1991 censusesconducted by Statistics Canada. Since previous investi-

gators have determined that occupation data onwomen's death certi®cates is unreliable for social classcategorization, we restricted our analysis to men aged

15±64 years old between the years 1981±1991(Marshall et al., 1993).We examined the rates of avoidable mortality from

the causes of death published by Charlton, excluding

causes of death restricted to women as well as perina-tal deaths (see Table 1). Mortality data was providedby the Division of Vital Statistics of the Ministry of

Health. We were able to identify causes of death basedon the 3 digit disease code (ICD-9) according to the9th revision of the International Classi®cation of

Diseases (World Health Organization, 1977).In order to conduct a comprehensive analysis that

would allow us to look at the individual and combined

Table 1

Selection of amenable causes of death

Selected cause of deatha ICD code (9th revision)

Hypertensive disease 401±405

Tuberculosis 010±018, 137

Asthma 493

Chronic rheumatic heart disease 393±398

Appendicitis 540±543

Acute respiratory infections 460±466, 487

Bacterial infections 004, 034, 320, 381±383, 390±392, 680±686, 711, 730

Hodgkin's disease 201

Abdominal hernias 550±553

Acute and chronic cholecystitis 574±575

De®ciency anaemias 280±281

Pneumonia and bronchitis 480±486, 490

a After Charlton.

E. Wood et al. / Social Science & Medicine 48 (1999) 1751±17581752

in¯uences of occupation, income and education, we

divided the study population into social levels using

four classi®cation methods. For the analysis of mor-

tality by social class, individuals were divided into ®ve

social class levels based on occupation using an adap-

tation of the UK Registrar General's Social Class

Scale (O�ce of Population Censuses and Surveys,

1970). In this classi®cation system persons with pro-

fessional and technical occupations are assigned to

social class one, while unskilled manual employees are

assigned to social class ®ve. Individual's occupations

were determined from the three-digit code of the 1980

revision of the manual of occupational codes by

Statistics Canada (1971). In addition, three levels of

socioeconomic analysis were performed using the

Blishen Index classi®cation system (Blishen et al.,

1987). Individuals were placed into one of four socioe-

conomic levels based on a Blishen education score, a

Blishen median income score and a total index score.

Income level was based on pooled median employment

income for all paid labor force participants in each oc-

cupation (Blishen et al., 1987). Education level was

based on the net `proportion' of well-educated incum-

bents, namely the proportion with a university degree

or post-secondary diploma minus the proportion with-

out a high school certi®cate or diploma (Blishen et al.,

1987). For all three analyses, individuals with the low-

est education, income, or combined index scores were

assigned to socioeconomic level four, while individuals

with the highest education, income, or combined index

scores were assigned to socioeconomic level one.

Once individuals were assigned to a social class in

each classi®cation system, the death rate from eachamenable cause was calculated and directly standar-dized to the total population. Once rates were standar-

dized, rate ratios were constructed by comparingobserved mortality rates in the highest social class withthose rates exhibited by the lowest social class.

Con®dence intervals around rate ratios were calculatedusing multipliers based on Byar's approximations ofthe exact Poisson limits (Breslow and Day, 1987) . Age

standardized death rates were expressed as deaths per100,000 population. In addition, the strength of themortality trend or slope across social groupings wasdetermined by performing an aged-standardized

weighted regression of the class-speci®c mortality ratesfor each cause of death. This technique enabled us toestimate the overall social class mortality trends.

Results

Based on data from the 1996 Canadian Census themid-period population of British Columbian malesbetween the ages of 15±64 years was 974,020. Over the

study period a total of 10,978,910 person-years wereobserved and 928 deaths were attributable to one ofthe selected amenable causes of death. For a total of

9,411,945 (86%) persons-years and 766 (83%) deathsthere was satisfactory occupational data for socialclass categorization. Of the 928 deaths, 421 (45.4%)

were attributable to pneumonia and bronchitis, 122(13.1%) to hypertensive disease, 109 (11.7%) toasthma, 73 (7.9%) to rheumatic heart disease, 56 to

Table 2

Age-standardized mortality rates (per 100,000 population) for causes of death amenable to medical intervention in British

Columbia males aged 15±64 years by social class, 1981±1991a

Social class scale

Cause of death n 1 2 3 4 5 Not

coded

Pooled

rate

All

men

Slope RR 95% CI

Hypertensive disease 122 0.7 0.9 1.6 1.1 1.7 0.7 1.3 1.1 0.2 2.3 (0.9±5.8)

Tuberculosis 41 0.1 0.3 0.1 0.2 0.6 0.9 0.4 0.4 0.1 3.5 (0.5±26.5)

Asthma 109 1.0 0.7 1.2 1.0 1.5 0.7 1.1 1.0 0.2 1.5 (0.7±3.1)

Rheumatic heart disease 73 0.6 0.6 0.9 0.8 0.7 0.7 0.7 0.7 0.0 1.3 (0.5±3.7)

Acute respiratory disease 26 0.0 0.1 0.5 0.3 0.3 0.2 0.2 0.2 0.1 ± ±

Bacterial infections 31 0.1 0.8 0.0 0.1 0.3 0.5 0.3 0.3 0.0 2.1 (0.3±16.5)

Hodgkin's disease 56 0.8 0.6 0.6 0.6 0.5 0.4 0.6 0.5 ÿ 0.1 0.7 (0.2±2.2)

Hernias, appendicitis and cholecystitis 48 0.4 0.4 0.4 0.4 0.5 0.6 0.4 0.4 0.0 1.2 (0.4±4.1)

Pneumonia and bronchitis 421 2.1 3.8 4.1 2.9 5.0 5.8 3.8 3.8 0.5 2.3 (1.4±4.0)

All amenable causes 928 5.9 8.2 9.5 7.4 11.0 10.3 8.8 8.4 1.0 1.9 (1.3±2.6)

a A total of 10,978,910 persons years were attributed to men in this study of which 743,930 were for class 1, 1,298,730 for class 2,

586,635 for class 3, 2,726,103 for class 4 and 4,056,097 class 5. An additional 1,567,415 person years could not be coded into any

social class over this period.

E. Wood et al. / Social Science & Medicine 48 (1999) 1751±1758 1753

Hodgkin's disease, 48 (5.2%) to hernias, appendicitisand cholecystitis, 41 (4.4%) to tuberculosis, 31 (3.3%)to bacterial infections, 26 (2.8%) to acute respiratory

disease and 1 (0.1%) to de®ciency anaemias.Table 2 illustrates the strong social class gradient

from the causes of death amenable to medical interven-

tion in British Columbian males aged 15±64. In thesocial classes de®ned by the Social Class Scale, the

lowest mortality rates for most avoidable causes ofdeath occurred in the highest social classes, while thehighest rates of mortality occurred in the lowest social

classes. For example, the death rate from pneumoniaand bronchitis in the highest social class was 2.1 per100,000 while, in the lowest social class, this rate was

5.0 per 100,000. A similar pattern of elevated mortalityrates in the lower social classes was observed for every

Table 3


Columbia males aged 15±64 years by Blishen education index, 1981±1991a

Blishen education index

Cause of death n 1 2 3 4 Not

coded

Pooled

rate

All

men

Slope RR 95% CI

Hypertensive disease 122 0.9 0.7 1.5 2.2 0.7 1.3 1.1 0.4 2.4 (1.4±3.8)

Tuberculosis 41 0.2 0.5 0.4 0.7 0.9 0.4 0.4 0.2 4.4 (1.6±12.1)

Asthma 109 0.8 1.1 1.4 1.8 0.7 1.1 1.0 0.3 2.2 (1.3±3.8)

Rheumatic heart disease 73 0.7 1.0 0.7 1.0 0.7 0.7 0.7 0.1 1.5 (0.8±2.9)

Acute respiratory disease 26 0.2 0.3 0.2 0.2 0.2 0.2 0.2 0.0 0.8 (0.2±2.9)

Bacterial infections 31 0.4 0.0 0.2 0.4 0.5 0.3 0.3 0.0 1.3 (0.5±3.3)

Hodgkin's disease 56 0.6 0.3 0.5 0.5 0.4 0.6 0.5 ÿ0.1 0.7 (0.3±1.7)

Hernias, appendicitis and cholecystitis 48 0.4 0.7 0.4 0.4 0.6 0.4 0.4 0.0 1.1 (0.4±2.9)

Pneumonia and bronchitis 421 3.2 1.8 3.7 6.7 5.8 3.8 3.8 0.9 2.1 (1.6±2.8)

All amenable causes 928 7.4 6.6 9.1 14.2 10.3 8.8 8.4 1.7 1.9 (1.6±2.3)

a A total of 10,978,910 persons years were attributed to men in this study of which 4,005,913 were for category 1, 620,930 for cat-

egory 2, 3,258,730 for category 3 and 1,525,922 for category 4. An additional, 1,567,415 person years could not be coded into any

social class over this period.

Table 4


Columbia males aged 15±64 years by Blishen income index, 1981±1991a

Blishen Income Index


coded

Pooled

rate

All

men

Slope RR 95% CI


Tuberculosis 41 0.3 0.3 0.4 0.6 0.9 0.4 0.4 0.1 2.2 (0.7±6.7)

Asthma 109 0.9 0.8 1.5 1.8 0.7 1.1 1.0 0.4 2.0 (1.1±3.7)




Hodgkin's disease 56 0.6 0.5 0.5 0.9 0.4 0.6 0.5 0.1 1.4 (0.6±3.0)




a A total of 10,978,910 persons years were attributed to men in this study of which 1,947,123 were for category 1, 3,210,967 for

category 2, 2,594,450 for category 3 and 1,658,955 for category 4. An additional, 1,567,415 person years could not be coded into

any social class over this period.


cause of death with the exception of Hodgkin's disease

which appeared with a slightly higher incidence rate in

the highest social class than in the lowest social class.

The con®dence intervals for the risk ratios between the

lowest and the highest social class, however, revealed

that the risk ratios were only statistically signi®cant for

avoidable deaths attributable to pneumonia and bron-

chitis which had a risk ratio of 2.3 (95% CI: 1.4±4.0)

and mortality from all amenable causes which had a

risk ratio of 1.9 (95% CI: 1.3±2.6).

Table 3 summarizes the analysis of deaths by

Blishen Education Scale among males aged 15±64. For

individuals classi®ed by education, there were statisti-

cally signi®cant risk ratios between the highest and the

lowest education class for deaths attributable to

Hypertensive disease (RR=2.4; 95% CI: 1.4±3.8),

Tuberculosis (RR=4.4; 95% CI: 1.6±12.1), Asthma

(RR=2.2; 95% CI: 1.3±3.8), pneumonia and bronchi-

tis (RR=2.1; 95% CI: 1.6±2.8) and mortality from all

avoidable causes (RR=1.9; 95% CI: 1.6±2.3).

As summarized in Table 4, a similar mortality gradi-

ent was observed among males classi®ed by the

Blishen Income Scale. Signi®cant mortality gradients

were observed between individuals in the lowest and

the highest income class for deaths attributable to

hypertensive disease (RR=2.5; 95% CI: 1.4±4.7),

Asthma (RR=2.0; 95% CI: 1.1±3.7), pneumonia and

bronchitis (RR=1.9; 95% CI: 1.4±2.5) and from all

amenable causes (RR=1.8; 95% CI: 1.4±2.2).

Table 5 shows rates of avoidable mortality in British

Columbian men using the combined Blishen

Education/Income Scale. For individuals classi®ed

using education and income, there were signi®cant gra-dients in the rates of avoidable mortality between thehighest and the lowest classes for deaths attributable

to hypertensive disease (RR=2.9; 95% CI: 1.7±5.0),tuberculosis (RR=2.7; 95% CI: 1.0±7.3), asthma(RR=2.2; 95% CI: 1.3±3.8), pneumonia and bronchi-

tis (RR=2.3; 95% CI: 1.7±3.0) and all cause mortality(RR=2.1; 95% CI: 1.7±2.5).

Discussion

This study demonstrates the strong social class and

socioeconomic gradients in mortality from avoidablecauses of death in British Columbian males. Foralmost every cause of death examined, the rate of mor-

tality was higher in individuals of lower social andsocioeconomic classes than in individuals of the uppersocial and socioeconomic classes. This trend was most

notable in deaths due to hypertensive heart disease,tuberculosis, asthma and pneumonia and bronchitis.The original goal of Rutstein et al. was to develop a

list of `sentinel health events' to be used as indicators

of the quality of health care. Studies which followedindicated that reductions in mortality from avoidablecauses have contributed appreciably to recent gains in

life expectancy and that the death rate from avoidablecauses has declined faster than the rate of deaths fromnon-amenable causes (Poikolainen and Eskola, 1986;

Mackenbach et al., 1988b). Although these results giveevidence that the decrease in avoidable mortality overtime is due at least in part to improvements in health

Table 5


Columbia males aged 15±64 years by Blishen socioeconomic index, 1981±1991a

Blishen Index


coded

Pooled

rate

All

men

Slope RR 95% CI


Tuberculosis 41 0.2 0.2 0.4 0.7 0.9 0.4 0.4 0.1 2.7 (1.0±7.3)

Asthma 109 0.8 0.6 1.6 1.8 0.7 1.1 1.0 0.4 2.2 (1.3±3.8)




Hodgkin's disease 56 0.6 0.7 0.4 0.9 0.4 0.6 0.5 0.0 1.5 (0.7±3.2)




a A total of 10,978,910 persons years were attributed to men in this study of which 2,446,938 were for category 1, 2,364,615 for

category 2, 2,722,987 for category 3 and 1,856,955 for category 4. An additional, 1,567,415 person years could not be coded into

any social class over this period.


care, studies have only been able to show weak associ-ations between health care variables and reductions in

avoidable mortality (Jougla et al., 1987; Kunst et al.,1988; Mackenbach et al., 1988a; Marshall et al., 1993).At the same time, the failure to ®nd signi®cant corre-

lation between health care variables and avoidablemortality does not invalidate the use of avoidable mor-tality as a measure of the e�ectiveness of health care

(Mackenbach et al., 1990). It may be that the healthcare variables used in previous studies, such as the pre-sence of a teaching hospital, are too crude and that

future studies should attempt to make use of morespeci®c measures of health care delivery (Mackenbachet al., 1990).For example, the present study provides evidence

that the way in which the supply of health care is orga-nized and its accessibility to the di�erent levels of so-ciety, may partially explain the social gradient in

avoidable mortality. Evidence to support this sugges-tion comes from recent evaluations of the supply andutilization of health care in Canada (Barer and

Stoddart, 1992; Kephart et al., 1998). These studieshave indicated that certain communities, particularly inrural areas, su�er from a critical under-supply health

care services. In addition, studies have suggested thatboth distance as well as lower education and incomelevels are related to reduced utilization of medical ser-vices (Freeborn et al., 1977; Haynes and Bentham,

1982; Kephart et al., 1998). Therefore, it follows thataccess to medical care in British Columbia, may bemost restricted among persons of lower socioeconomic

status, for whom the travel to a larger urban centerand time away from work is less a�ordable.Another potential explanation for our results is that

there is a higher prevalence of health risk behaviorsamong the lower social classes in British Columbia.Over the last several decades, tobacco and alcohol con-sumption and obesity, have been cited as the major

determinants of premature and preventable mortality(Wiley and Camacho, 1980; Hirdes and Forbes, 1992; ,Wilson, 1994). In addition, it is generally accepted that

persons of lower socioeconomic status are signi®cantlymore likely to lead a sedentary lifestyle, to be over-weight and to smoke cigarettes (Lui and Cedres, 1982;

Winkleby et al., 1990). Therefore, it is possible thatvariation in health risk behaviors across social classesmay also have played a role in causing the observed

mortality gradient.Unlike studies which have attempted to relate health

care variables to avoidable mortality, all studies whichrelate avoidable mortality and social class variables Ð

including the present analysis Ð have found ratherstrong and consistently negative correlation (Charltonet al., 1983; Carr-Hill et al., 1987; Jougla et al., 1987;

Poikolainen and Eskola, 1988; Marshall et al., 1993).The death rates from avoidable causes in lower social

classes and lower socioeconomic groups were found tobe consistently higher than the death rates determined

for the higher social classes and higher socioeconomicgroups. These ®ndings have not been dependent on themethodology used, raising con®dence that they re¯ect

reality (Mackenbach et al., 1990).In the present analysis, the avoidable mortality rate

was related to social class, regardless of which social

class component, occupation, education, or incomewas being measured. For individuals classi®ed in eachscale there was a signi®cant gradient in mortality for

deaths due to pneumonia and bronchitis and from allamenable causes. For the most part, the avoidablecauses of death that were found to vary signi®cantlyby social class were the same for each of the three

Blishen scales with the exception of the Blishen incomescale, which did not show a signi®cant gradientbetween the highest and the lowest income groups for

deaths due to tuberculosis. The fact that previousresearch has indicated that persons of lower educationare less likely to seek medical services, may partially

explain why there was a signi®cant gradient in deathsdue to tuberculosis between the individuals classi®edby education (Freeborn et al., 1977; Kephart et al.,

1998). The consistency of our ®ndings across all classi-®cation methods make it di�cult to discern the rolesof each social class component. Our results indicatethat education may be a slightly better indicator of

avoidable mortality than income, however, the scalethat combined these characteristics provided the mostsigni®cant results. Instead, our results highlight the

interrelationship between income, education, occu-pation and health and validate the hypothesis that cer-tain components or determinants of social class, such

as income and education, may be used as indicators ofhealth status or health care needs (Kunst andMackenbach, 1994; Mackenbach et al., 1997; Hart etal., 1998; Kunst et al., 1998). Furthermore, due to the

fact that speci®c avoidable causes of death, such aspneumonia and bronchitis, were observed to be con-sistently higher in the lower social classes, regardless of

the social class categorization method, we feel thatspeci®c measures aimed at improving survival fromthese conditions in lower social classes could help to

amend the social class disparity.There are a number of limitations which accompany

this type of analysis that must be recognized when

interpreting the results. For instance, the statisticalanalysis was performed on a sample of only 928deaths, creating the potential shortcomings thataccompany statistical analysis on a small sample size.

In addition, the death rate among persons without oc-cupational data was high. The exclusion of these indi-viduals could potentially bias our study population to

underestimate individuals, such as those who areunemployed, retired or on social assistance, in one or


more of the social classes. As a result, our analysesmay actually underestimate the social gradient in

avoidable mortality. In addition, our data may bebiased due to numerator/denominator bias. This po-tential problem stems from the fact that occupational

data for the deceased was obtained from death certi®-cates, whereas the occupation of the corresponding liv-ing population was obtained from the Census (Kunst

et al., 1998). Similarly, our analysis could have beenpartially skewed by incorrect social classi®cation. It isinevitable that there will be some misclassi®cation of

individuals who for reasons such as family or spousalsupport live in a social class that is not representativeof their educational, income, or occupational indi-cators. Finally, the nature of the study does not enable

us to control confounding variables such as smoking,alcohol consumption and obesity, all of which are riskfactors for a host of illnesses including heart disease

and hypertensive disorders.In summary, the present analysis revealed that social

class inequalities existed for almost every amenable

cause of death studied. The strong and consistent cor-relation between social class and rates of avoidablemortality underscore a number of issues. In British

Columbia, inequitable access to health care acrosssocial classes may explain, in part, the social gradientin avoidable mortality. In addition, social conditionsgiving rise to disease also deserve greater attention.

Death from each of the diseases reviewed in this studyis by de®nition preventable, regardless of social classand the results of our analysis present a challenge to

British Columbian health care providers.

Acknowledgements

This work was supported by the National HealthResearch Development Program of Health Canada

through a National Health Research Scholar Award toDr. Hogg and through a National AIDS ResearchScientist Award to Dr. Schechter. We thank Dr.Richard Gallagher of the British Columbia Cancer

Agency and the Division of Vital Statistics of theBritish Columbia Ministry of Health for assistancewith this study.

References

Barer, M.L., Stoddart, G.L., 1992. Toward integrated medical

resource policies for Canada: 8. Geographic distribution of

physicians. Canadian Medical Association Journal 147,

617±623.

Blishen, B.R., Carroll, W.K., Moore, C., 1987. The 1981

socio economic index for occupations in Canada. Review

of Canadian Sociology and Anthropology 24 (4), 465±487.

Boys, R.J., Forster, D.P., Jozan, P., 1991. Mortality from

causes amenable and non-amenable to medical care: the

experience of eastern Europe. British Medical Journal 303,

879±883.

Breslow, N.E., Day, N. E., 1987. In: Statistical methods in

cancer research, Vol. II. The design and analysis of cohort

studies. International agency for research on cancer, Lyon.

Statistics Canada, 1971. Occupation Classi®cation Manual.

Ottawa, Canada.

Carr-Hill, R.A., Hardman, G.F., Russell, I.T., 1987.

Variations in avoidable mortality and variations in health

care resources. Lancet 1, 789±792.

Charlton, J.R., Hartley, R.M., Silver, R., Holland, W.W.,

1983. Geographical variation in mortality from conditions

amenable to medical intervention in England and Wales.

Lancet 1, 691±696.

Charlton, J.R., Velez, R., 1986. Some international compari-

sons of mortality amenable to medical intervention. British

Medical Journal (Clin. Res. Ed.) 292, 295±301.

Freeborn, D.K., Pope, C.R., Davis, M.A., Mullooly, J.P.,

1977. Health status, socioeconomic status and utilization

of outpatient services for members of a prepaid group

practice. Medical Care 15, 115±128.

Hart, C.L., Smith, G.D., Blane, D., 1998. Inequalities in mor-

tality by social class measured at 3 stages of the lifecourse.

American Journal of Public Health 88, 471±474.

Haynes, R.M., Bentham, C.G., 1982. The e�ects of accessibil-

ity on general practitioner consultations, out-patient atten-

dance's and in-patient admissions in Norfolk, England.

Social Science & Medicine 16, 561±569.

Hirdes, J.P., Forbes, W.F., 1992. The importance of social re-

lationships, socioeconomic status, and health practices

with respect to mortality among Ontario males. Journal of

Clinical Epidemiology 554, 175±182.

Jougla, E., Ducimetiere, P., Bouvier-Colle, M.H., Hatton, F.,

1987. Relation entre le niveau de developpement du sys-

teme de soins et le niveau de la mortalite `evitable' selon

les departements francais. Rev. Epidemiol. Sante publ. 35,

365±377.

Kephart, G., Thomas, V.S., MacLean, D.R., 1998.

Socioeconomic di�erences in the use of physician services

in Nova Scotia. American Journal of Public Health 88,

800±803.

Kunst, A.E., Groenhof, F., Mackenbach, J.P., 1998.

Mortality by occupational class among men 30±64 years in

11 European countries. EU Working Group on

Socioeconomic Inequalities in Health. Social Science &

Medicine 46, 1459±1476.

Kunst, A.E., Looman, C.W.N., P, M.J., 1988. Medical care

and regional mortality di�erences within the countries of

the European Community. European Journal of

Population 4, 223±245.

Kunst, A.E., Mackenbach, J.P., 1994. The size of mortality

di�erences associated with educational level in nine indus-

trialized countries. American Journal of Public Health 84,

932±937.

Lui, K., Cedres, L.B., 1982. Relationship of education to

major risk factors and death from coronary heart disease,

cardiovascular diseases, and all causes. Circulation 66,

1308±1314.

Mackenbach, J.P., Bouvier-Colle, M.H., Jougla, E., 1990.


`Avoidable' mortality and health services: a review of

aggregate data studies. Journal of Epidemiology and

Community Health 44, 106±111.

Mackenbach, J.P., Kunst, A.E., Cavelaars, A.E., Groenhof,

F., Geurts, J.J., 1997. Socioeconomic inequalities in mor-

bidity and mortality in western Europe. The EU Working

Group on Socioeconomic Inequalities in Health [see com-

ments]. Lancet 349, 1655±1659.

Mackenbach, J.P., Kunst, A.E., Looman, C.W., Habbema,

J.D., van der Maas, P.J., 1988a. Regional di�erences in

mortality from conditions amenable to medical interven-

tion in The Netherlands: a comparison of four time

periods. Journal of Epidemiology and Community Health

42, 325±332.

Mackenbach, J.P., Looman, C.W., Kunst, A.E., Habbema,

J.D., van der Maas, P.J., 1988b. Post-1950 mortality

trends and medical care: gains in life expectancy due to

declines in mortality from conditions amenable to medical

intervention in The Netherlands. Social Science &

Medicine 27, 889±894.

Mackenbach, J.P., Stronks, K., Kunst, A.E., 1989. The con-

tribution of medical care to inequalities in health: di�er-

ences between socio±economic groups in decline of

mortality from conditions amenable to medical interven-

tion. Social Science & Medicine 29, 369±376.

Marshall, S.W., Kawachi, I., Pearce, N., Borman, B., 1993.

Social class di�erences in mortality from diseases amenable

to medical intervention in New Zealand. International

Journal of Epidemiology 22, 255±261.

O�ce of Population Censuses and Surveys, 1970.

Occupational Mortality. The Registrar General's decennial

supplement for England and Wales, 1970±72. HM

Stationery O�ce, London, 1978.

Poikolainen, K., Eskola, J., 1986. The e�ect of health services

on mortality: decline in death rates from amenable and

non-amenable causes in Finland, 1969±81. Lancet 1, 199±

202.

Poikolainen, K., Eskola, J., 1988. Health services resources

and their relation to mortality from causes amenable to

health care intervention: a cross-national study.

International Journal of Epidemiology 17, 86±89.

Rutstein, D.D., Berenberg, W., Chalmers, T.C., Child, C.G.,

Fishman, A.P., Perrin, E.B., 1976. Measuring the quality

of medical care. A clinical method. New England Journal

of Medicine 294, 582±588.

Rutstein, D.D., Berenberg, W., Chalmers, T.C., Child, C.G.,

Fishman, A.P., Perrin, E.B., 1983. Measuring the quality

of medical care, second revision of tables of indexes. New

England Journal of Medicine 302, 1146±1150.

Wiley, J.A., Camacho, T.C., 1980. Life-style and future

health: evidence from the Alameda County Study.

Preventative Medicine 9, 1±21.

Wilson, P.S., 1994. Established risk factors and coronary

artery disease: the Framingham study. American Journal

of Hypertension 7, 75±125.

Winkleby, M.A., Fortmann, S.P., Barrett, D.C., 1990. Social

class disparities in risk factors for disease: eight year

prevalence patterns by level of education. Preventative

Medicine 19, 1±12.

World Health Organization, 1977. Manual of International

Statistical Classi®cation of Diseases, Injuries and Causes

of Death, 9th Revision. WHO, Geneva.


Documents

Social inequalities in male mortality amenable to medical intervention in British Columbia