Coordinated by Roberta PACE, Alain PARANT

 

Spatial differences of mortality: social and medical factors

 

Farida Laoudj Chekraoui

Laboratory of the mathematics and the applications of the mathematics LMAM,

University of Jijel, Algeria

 

Abstract: The spatial differences of mortality are partially explained by an irregular dividing up of social groups between territories and by an unequal granting of medical resources between various territories. In metropolitan France, since the beginning of 1980s, the almost regular increase of the life expectancy at birth in all areas is accompanied by maintaining the regional disparities of mortality. The irregular dividing up of the social groups between regions, combined with regional specificities as far as behaviours, environment and structures of economic activities are concerned, explains a part of these regional differences. The underprivileged people are the most affected by the regional differences of mortality. In Nord-Pas-de-Calais, labours die earlier on average compared with those who live in Ile-de-France. We can see easily that what applies to France applies all the more so to Algeria, a country less advanced at socioeconomic level. In this country, the unequal granting of health structures between territories and their variable efficiency makes worse the spatial differences of both health and mortality. Reduction, not to say disappearance of these differences is obtained by an improvement of working conditions and better granting of medical capacities.

Keywords: Spatial differences of mortality, irregular dividing up of the social groups between territories, unequal granting of the medical resources

 

1.          Introduction

 

The spatial disparities of mortality are the result of a conjunction of several factors Social, environmental and political of which they matter to consider the measure in order to assess the real differences of mortality and possibly trying to correct them by establishing appropriate and target policies. The first part of this article is about metropolitan France. It is devoted to the analysis of the regional disparities of male mortality and to measurement of the incidence of Socio-professional category.

The second part concerns Algeria, a country where, in spite of considerable efforts undertaken since around ten years, the number of hospitals and medical facilities remains very insufficient to meet the demand. Furthermore, the unequal distribution of these facilities all over the territory and their variable efficiency let suppose that their effect as regards health spatial differences – assessed to one of the differences of mortality - is significant.

Health structures in Algeria suffer from underequipped surgical rooms, from often old-fashioned medical equipment or in poor condition, from incomplete medicines and regularly in shortage of stock…. The gravity of the situation varies considerably between big cities and other territories. The specialists and the other health professionals are concentrated in the Northern big cities. In this second part, after highlighting regional differences of mortality characterizing Algeria since the year 2000 (the black decade i.e. 1990s’ is not included), the accent is about the unequal distribution of both medical care equipment and taking care of the patients, the spatial disparities of the accessibility to the primary medical care and of the availability of health technologies being supposed to be explained by the least mortality in well-off territories.

2.                   The regional disparities of male mortality in metropolitan France and the incidence of the socio-professional group

The link between mortality and socio-professional group (SPC) is studied extensively in France. If various studies show the intensity of this link, however, and despite the persistence of disparities in mortality between regions, the impact of social differences on such disparities is very little studied.

The objective of the first part of this article is to measure the influence of the profession assessed by socio-professional group (SPC) on the regional variation of male mortality in Metropolitan France along the period 2005-2007[i].

The studied population is made up only of working population aged between 30 and 59[ii].

In order to measure the impact of SPC on mortality, we try to isolate the effect of the PCS structure on mortality, from the effect of all other factors, using the method developed by Hemery and Quang-Chi in 1978. This method consists, first of all, in bringing the ratio of mortality rates of active people aged between 30-59 living in the area to the mortality rate of active natives of Metropolitan France, noted d (R / FM). This ratio measures mortality regional disparities of active people aged between 30 and 59, compared with the mortality of the active population aged always from 30 to 59 of the whole Metropolis:

  d(R /FM) = 

                    

 

 (Where tR is the mortality rate of active people aged 30-59 in the region R directly standardized by age and TFM mortality rate of the working population of 30-59 years old throughout the metropolis directly standardized by age).

It is then divided into two indexes.

The first one is an index of abnormally high death rate related to the PCS, noted sR:

SR =      

Where  is the mortality rate of active people of 30-59 years old standardized directly by age and indirectly by SPC; it represents what would be the mortality rate of the region R if the mortality by SPC was identical to that observed for every PCS in the whole of metropolitan France and  the mortality rate of active people aged between 30-59 for metropolitan France standardized directly by age;

The second is an index of abnormally high death rate independent from SPC, noted aR: aR =

Where is the mortality rate of a working population having 30-59 years old in the region R, standardized directly by age and  the mortality rate of active people aged 30-59 standardized directly by age and indirectly by SPC.

These two indexes can be interpreted as follows:

- SR < 1 indicates a structure by SPC of the region R playing in favour of a sub-mortality, and SR > 1 showing an opposite situation.

 

- aR < 1 indicates that the factors other than the structure by PCS play in favour of a sub-mortality of the region R, aR > 1 showing an opposite situation.

2.1. Mortality social differences similar in all regions of the metropolis

Between 30 and 60 years old, executives and intellectual occupations show the weakest mortality rates, the middle occupations, craftsmen, storekeepers and business managers show rates considerably superior. Whereas, the farmers show higher rates; but it is especially the labours and the employees who show the most significant mortality rate (table 1). In 2005-2007, the rate (comparative rate of mortality standardized on age[iii]) of labours is approximately 4 times higher than that of the executives and the intellectual professions.

 

Table 1: Comparative rate of mortality by SPC of active people aged 30-59 (per 100.000):

Metropolitan France in 2005-2007.

Profession and socio-professional category

TCM per 100 000

Executives and superior intellectual professions

98

Middle professions

178

Craftsmen, Storekeepers, business managers

177

Farmers

231

Employees

376

Labours

372

(Ratio rate Labours/rate Executives)

3,8

Comparative rate of Mortality of working population aged 30-59

247,6

 

On the whole, the social differences of mortality of the working population aged between 30 and 59 noticed in metropolitan France are found in almost all the regions. The mortality of both executives and intellectual professions is the weakest in all the regions but those of labours and employees is the highest everywhere (figure1).

In 2005-2007, the regional disparities of mortality of the working population vary considerably according to the PCS. If the spectrum variation is relatively reduced for the favoured social classes, it is more enough open for the labours, the employees, and the farmers. Mortality weak levels for labours and employees in Alsace, Midi-Pyrénées and Ile-de-France go with very high levels in Brittany and in Nord-Pas-de-Calais.

As for the executives and intellectual professions, even if the disparity of their mortality rates between regions is much less significant, Midi-Pyrénées and Franche-Comté stand out from the rest by the weakest mortality levels.

Figure 1: Regional comparative rate of mortality of active people aged 30-59 by SPC in 2005-2007

(per 100.000).

 

 

2.2. The influence of structures by SPC on regional disparities of male mortality

In 2005-2007, the working population of 30-59 years old in Nord-Pas-de-Calais, Brittany and Upper Normandy are characterized by an abnormally high death rate compared with the metropolitan average (index superior to 100) and those of Midi-Pyrénées, Ile-de-France and Rhône-Alpes, by a clear abnormally low death rate (index inferior to 100) (figure 2). This assessment results partially from differences of social structures between regions. What would have been if in each region the distribution of active people according to the SPC were identical to that of metropolitan France?

 

Figure 2: Ratio of mortality regional comparative rates of active people aged 30-59 years to the national rate. Metropolitan France 2005-2007, standardization only on age.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The standardization on the PCS does not really modify the geography of the mortality of the working population in particular for regions in abnormally high death rate compared with the Metropolitan average (ref. figure 3). Indeed, they are the same regions that are found at extreme. The mortality rates in Nord-Pas-de-Calais, Brittany and Upper Normandy remain comparatively the highest and the mortality rates of the working population having 30-59 years old in Midi-Pyrénées remain the lowest. Compositions by SPC of Nord-Pas-de-Calais, Brittany and Upper Normandy exert an adverse effect on the mortality of active people aged 30-59 years old in these regions, but others play in the same direction, because even after correction of the effect of these compositions, the position of the regions in the national hierarchy has not improved.

At the other end of the hierarchy, the Midi-Pyrénées region occupies the best position, even after neutralizing the favourable effect of the structure by SPC. In other words, the abnormally low death rate mortality of active people aged 30-59 in Midi-Pyrénées depends essentially on other factors as in the previous three regions.

The standardization on structure by SPC brings also some specificity. Thus, the abnormally low death rate of the active members aged 30-59 years in Ile-de-France does appear virtually, the comparative rate standardized on SPC being too close to the average. The same observation applies in Rhône-Alpes where the abnormally low death rate of male population is much clearer after the standardization. In contrast, the abnormally high death rate of the active people aged 30-59 in Picardie and Low-Normandy is much less significant after standardization and relatively close to the average. Champagne-Ardenne is the only region whose position changes after standardization by SPC according to the metropolitan average. It goes from a situation of abnormally high death rate of the active members aged 30-59 years to a situation of slight abnormally low death rate. 

When, the abnormally low death rate of the active people aged 30-59 years observed in Ile de France and in Rhône-Alpes results from a SPC structure favourable to the survival. The abnormally high death rate observed in Picardie and Lower-Normandy, but mainly in Champagne-Ardenne depends on the contrary to a SPC structure unfavourable to the survival (figure 3).

 

Figure 3: Ratio of mortality regional comparative rates of active people aged 30-59 years to the national rate. Metropolitan France 2005-2007, standardization on age and on SPC.

 

 

 

 

 2.3 Regional disparities of mortality and the unequal spatial distribution of the SPC

By confrontation of mortality regional rates standardized on age and on SPC and rates standardized only on age, we have previously succeeded, with inaccuracies of the method, to identify the regions which are characterized by a favourable effect of the structure by SPC on the mortality of the active people aged 30-59 years and the ones which are characterized by an unfavourable effect of this structure.

For all that we haven't decomposed the abnormally high/low death rate of the region according to the role played respectively by the structure by SPC and by the whole of other factors (with the exception of the structure by age in which the effect is here neutralized). It is what we carried out in this part.

Among the regions with very high abnormally low death rate of active people aged 30-59 years (12% inferior to Metropolitan average, Île-de-France is distinguished by a very favourable effect of its structure by SPC which explains 12 % of the sub-abnormally high death rate, all other factors do not contributing at all (Table 2).

Midi-Pyrénées and Rhône-Alpes are characterized also by a very high abnormally low death rate of active people aged between 30 and 59. Two regions showing a favourable structure of SPC, but here the influence on the mortality is much lower than in the Île-de-France. In fact, the structure by SPC explains only 2 % of the abnormally low death rate compared with the average. In Midi-Pyrénées, the abnormally low death rate of the active population represents 19 % inferior to the average, as in Rhône-Alpes where the abnormally low death rate reaches 10%; the first cause of this established fact depends then on the whole other factors.

In the opposite, the active people aged 30-59 years show a high abnormally high death rate, always compared to the metropolitan average in Nord-Pas-de-Calais, Brittany, Upper Normandy, Lorraine, Lower-Normandy and Picardie. In these regions, the structure by SPC has an unfavourable effect on the survival of the active population, but the weight of this effect differs from one region to another. For an abnormally high death rate of 31% in Nord-Pas-de-Calais and 27% in Brittany, hardly 5% and 1% of this abnormally high death rate are respectively due to the structure by SPC. While in Upper Normandy, Lorraine, Lower-Normandy and Picardie, more than a half of the difference to the metropolitan average is explained by the structure by SPC.

In Champagne-Ardenne, Centre, Bourgogne, Limousin and Auvergne, the entire abnormally high death rate compared with the average is an unfavourable effect of the structure by SPC, the whole other factors playing in contrast towards a slight abnormally low death rate.

 

Table 2: Regional mortality difference of active men aged 30-59 years compared with the national average by SPC structure and by other factors: Metropolitan France 2005-2007.

 

Generally, in most of regions of the province with the exception of Provence-Alpes-Côte-d’Azur, Rhône-Alpes, Midi-Pyrénées, the structure by SPC is at the origin of an abnormally high death rate of active people aged 30-59 years compared with the metropolitan average. The other factors playing rather in the direction of an abnormally low death rate, the exceptions noted in this model are Brittany, Nord-Pas-de-Calais and the Upper Normandy.

 

 

3.                   Social and medical factors and territorial differences of mortality in Algeria

3.1 Mortality is decreasing in Algeria but at different rhythms depending on the wilayas

In Algeria, life expectancy at birth is increasing for both men and women, but men still die, on average, earlier than women. Life expectancy at birth for men goes from 66,74 years in 1993 to 74,70 years in 2007(i.e. an increase of 12%) and that of women goes from 68,12 years to 76,80 years (an increase of 12.8%)[iv]. However, other than this general improvement, large spatial disparities remain; the opposition was strongly marked between the north and the south of the country[v],[vi]. Between wilayas Tamanrasset, in the region of Hoggar Tassili, and Boumerdes, in the north-central region, the difference comes to 16 and 10 years respectively for men and women, (Appendix).

Infant mortality-mortality during the first year of life- decreased very low down or even worse in some wilayas, it goes down in most wilayas but again at very different rates, and the same observation for an underprivileged compared to a favoured North, the extreme situation is noticed in the wilaya of Illizi. In this region where people are very vulnerable and living conditions particularly difficult, boys’ infant mortality increased from 36.6 deaths per 1,000 living births in 2005 to 40.8 ‰ in 2010 that of girls increased along the same period from 30.2 ‰ to 31.8 ‰ (Appendix).

3.2 The unequal distribution of health professionals, source as an explanatory factor of territorial inequalities in mortality

These inequalities in mortality are partly explained by the unequal distribution of health facility and health professionals on the Algerian territory. Data provided by the Ministry of Health indicate that in 2010, in Hoggar Tassili region, there were on average 2,728 people for one doctor or specialist; the ratio was at the same date 630/1 in the north-central region.

For 2005, the level of mortality is clearly related to the ratio of population per general practitioner (general practitioner or specialist or dentist or pharmacist). Figure 4 shows clearly that wilayas for which this ratio is low are also those where infant mortality tends to be the lowest, Algiers is distinguished particularly. Conversely, the wilayas where health practitioners are responsible for a larger population are characterized by relatively high levels of infant mortality.

The allocation of health professionals is spatially unequal in Algeria, with the result that the least well off wilayas, especially those in the south of the country, pay the highest price in terms of survival of their population.

 

Figure 4: Distribution of 48 wilayas depending on the male and female infant mortality rate (per 1,000) and the number of inhabitants per health practitioner (GPs, specialists, dentists and pharmacists in thousands) in 2005 semi logarithmic scale.

 

 

 

 

4.         Conclusion

The analysis of spatial disparities in mortality highlights the importance of factors related to the structure of economic activity and the composition of the population being old enough to work according to SPC, as well as socio-economic context largo sensu.

In France, the weak regional heterogeneity in mortality of active men aged 30 to 59 years, executives or members of the superior intellectual professions, compared to the high variability of the mortality of employees and labours, can be interpreted as the result of a greater tendency to geographical mobility of executives, the high variability of the mortality of employees and labours can, in turn, maintain their lower propensity to migrate (Rican et al., 2003). If the differences are sometimes very pronounced from a region to another for some socio-professional groups, in particular labours and employees, due to working conditions extremely variable, the influence of differences in distributions of SPC on mortality regional disparities of active people aged 30-59, without being negligible, is nevertheless relatively limited. It is in the Ile-de-France that the distribution of active people by SPC manifest the strongest effects towards an abnormally low death rate, whereas in most other regions induces an abnormally high death rate. This is particularly true in areas devoid of influential metropolis (Champagne-Ardenne, Poitou-Charentes, Franche-Comté, Lower-Normandy, Bourgogne, Centre, Auvergne, Limousin), but much less, if at all, in those equipped with (Rhône-Alpes, Midi-Pyrénées, Languedoc-Roussillon, Provence-Alpes-Côte-d’Azur, ...). The effect of residual factors on regional abnormally high/low death rate is relatively significant. On the whole, in most provincial areas; that these other factors play in favour of abnormally low death rate or a contrario in favour of abnormally high death rate, such as in Brittany, Nord-Pas-de-Calais and Upper Normandy, etc.

Notwithstanding the unequal spatial distribution of SPC policies that aim, in France, to reduce health inequalities should pay particular attention to improving working conditions.

In Algeria, despite efforts made to fight against health inequalities, real differences remain; the policies implemented primarily being the benefit to individuals most favoured by their education or the fact of living in the most developed areas.

The health of an individual is a result of several factors (living conditions, environment, and job) on which it is possible to act for reducing the inequalities.

Actions must mobilize all Ministries (Transport, Health, Education, Housing) and be undertaken at national and local levels. Effective interventions are upstream of the healthcare system and involve the support at early childhood as well as renovation of districts and housing for poor families, improving the transportation network in order to make health facilities, usually located in city centres, accessible to all, the development of prevention and care systems in under-served areas, ...

The granting of resources should be based more on needs defined in terms of health status and precarious of populations living in targeted areas to reflect the principle "equal need, equal resources."

Nomadic population particularly underprivileged should be able to benefit from "mobile" prevention campaign in order to inform them on existing health risks and encouraging them to seek treatment. For more efficiency, the establishment of support networks mobilizing neighbourhoods and promoting mutual aid and the exchange is required, these groups of primary care are highlighted as major parties of the program in underprivileged areas.

 

 

Bibliography

Hemery, S., Quang-Chi, D., « Facteurs sociodémographiques des disparités de la mortalité infantile », In  Les disparités démographiques régionales, Colloques nationaux du CNRS, Paris 1978.

Kouaouci, A., Saadi R., « La reconstruction des dynamiques démographiques locales en Algérie (1987-2008) par des techniques d’estimation indirecte », Cahiers québécois de démographie, vol. 42, n° 1, 2013, p.101-132.

Meslé, F., Vallin, J., Évolution et variations géographiques de la surmortalité masculine : du paradoxe français à la logique russe, Population, 53e année, n° 6, 1998, p.1079-1101.

Mesrine, A., « Les différences de mortalité par milieu social restent fortes », In La société française : données sociales, 1999, p.228-235.

Rican, S., Jougla, É., Salem, G., « Inégalités socio-spatiales de mortalité en  France », BEH n° 30-31, 2003.

PRESSAT, R., Manuel d’analyse de la mortalité, chapitre III, 1985, INED.

PRESSAT, R., Dictionnaire de démographie, 1978, Presses Universitaire de France.

Salem, G, Rican, S, Jougla, E [Coord.], Atlas de la santé en France, Vol 1 : Les causes de décès, Paris, John Libbey Eurotext, 2000.

 

 

Appendix


Life expectancy at birth (2007) and infant mortality rate (2005, 2010) by sex, region and willayas in Algeria

Sources: Kouaouci and Saadi for life expectancy at birth; National Statistical Office (NSO) for infant mortality rates.

 

 



[i] Changes took place in the composition of SPC groups applying to the civil status of the years framing the census 1982, 1990 and 1999 make delicate not only such calculation but also such a study of evolution. More particularly, it is not certain that, for these years, the deaths gathered are classified by SPC in the same way as people taken in census.

[ii] The study is limited only to the working population of 30-59 years old because, because of small number of non-working population between 30 and 60 years old, and of their concentration in the group of 55-59 years old, a standardization involving non-working population would be almost redundant standardization by age. This limitation in the scope of study is due to the fact that the non-working population is very heterogeneous in terms of mortality, and it is impossible to control this heterogeneous: the former PCS of retired is not available in the classification of death, even the PCS of non working spouse, the reason that led to inactivity even less.

 [iii] Here, the comparative rates are the result of a direct standardization by age, the reference structure being the composition by five-yearly age of the working population of 30-59 years old, assembled sexes, in metropolitan France in 2006.

[iv] A. Kouaouci, R. Saadi, « La reconstruction des dynamiques démographiques locales en Algérie (1987-2008) par des techniques d’estimation indirecte », Cahiers québécois de démographie, vol. 42, n° 1, 2013, pp.101-132.

[v] To avoid any misinterpretations related to the decade of violence, territorial comparison is made only for the year 2007.

[vi] The number of deaths and living births are data provided by the National Statistical Office (NSO).