FAIR SHARES FOR THE ZONE: ALLOCATING HEALTH-CARE RESOURCES FOR THE NATIVE POPULATIONS OF THE SIOUX LOOKOUT ZONE, NORTHERN ONTARIO

The purpose of this paper is to explore issues of healthcare resource allocation for a Canadian native population, that of Sioux Lookout Zone, within the context of the total provincial population. We argue that a fair share of provincial health resources for any population must be based on a population needs approach. The paper begins by establishing a method to measure a fair share of healthcare resources. It goes on to establish why a fair share is important in the light of native health conditions in general and with respect to this specific population. The fair share is then calculated and illustrated with reference to particular health-care programs. Finally, there is discussion of the methodological and policy issues that arise from this research.     

INTRAURBAN MORTALITY VARIATION and INCOME DISPARITY: A CASE STUDY OF HAMILTON-WENTWORTH REGION

This paper characterizes and explains mortality variation among census tracts of Hamilton-Wentworth Region, Ontario, during 1980–82. Analysis is based on the age-specific death rate of the 55 to 64 age group, computed for each sex by categories of major cause. Mosk and lohansson's (1986) four stages of mortality/income relationship and Omran's (1977) theory of epide-miological transition provide context for understanding mortality variations. Logit models and the maximum quasi-likelihood estimation method allow for statistical inference. Research findings show that spatial patterns of mortality are more fragmented and less explainable for females than for males; that mortality is dependent (1) negatively on median family income and the percentage of individuals who are married and (2) positively on unemployment rate and the percentages of individuals who are single, widowed, and divorced or separated; among socioeconomic variables derived from the 1981 census, median family income is the most powerful explanator of intraurban mortality variation and, in this context, other explanatory variables are not statistically significant; and, in terms of the theoretical frameworks of Mosk and johansson (1986) and Omran (1977), mortality patterns for females are shown to have progressed further than for males. Cet essai caractérise et explique les variations de mortalité apparues en étudiant les tracts de recensement de la Région d'Hamilton-Wentworth, Ontario, en 1980–82. L'analyse est basée sur le taux de mortalité spécifique au groupe des 55 à 64 ans, en prenant en considération les deux sexes et les différentes catégories de cause majeure de mortalité. La relation mortalité/revenu en 4 étapes de Mosk et Johansson offre un contexte permettant de mieux comprendre les variations de mortalité. Les modeles Logit et la méthode d'estimation de quasi-probabilité permettent de considérer une inférence statistique. On a trouvé tout d'abord, que les modèles spatiaux de mortalité sontplus fragmentés et moins expliquables pour les femmes que pour les hommes; ensuite, sur la base d'une série d'analyses bi-variées, la mortalité prouve dépendre (1) négativement du revenu familial moyen et du pourcentage d'individus mariés et (2) positivement d'un taux de chomage et de pourcentages d'individus célibataires, veufs, divorcés et séparés; troisièmement, parmi toutes les variables socio-économiques dérivées du rencensementde 1981, le revenu familial moyen est la variable expliquant le mieux les variations de mortalité intra-urbaine et, dans ce même contexte, les autres variables explicatives ne sont pas suffisamment significatives sur le plan des statistiques; et (4) en termes des cadres théoriques de Mosk et lohansson et Omran, cette étude démontre que les modèles de mortalité ont progressé davantage pour les femmes que pour les hommes.     

Snow-tracking and GIS: using multiple species-environment models to determine optimal wildlife crossing sites and evaluate highway mitigation plans on the Trans-Canada Highway

Snow-tracking data were collected for cougars (  Felis concolor ), lynx (  Lynx canadensis ), martens ( Martes americana ) and wolves ( Canis lupus ) and combined with remotely sensed imagery in a geographic information system (GIS) to identify wildlife crossing sites on the Trans-Canada Highway in Banff National Park, Alberta. We used logistic regression to assess the dependent (species presence/absence) relative to measures of topography and vegetation. The exponent form of each logistic regression equation was used to predict crossing sites in a GIS, which were then contrasted with mitigation sites proposed by Parks Canada. We found that: (1) cougars were influenced positively by normalized difference vegetation index (NDVI); negatively by northness and distance to ruggedness, (2) lynx were influenced positively by wetness, greenness, rugged terrain, eastness and distance to rugged terrain; negatively by slope, (3) martens were related positively to wetness, elevation, eastness, and distance to rugged terrain; negatively to northness, (4) wolves were influenced positively by distance to ruggedness; negatively by brightness, elevation, eastness and terrain ruggedness. There were few sympatric crossing sites for all species, supporting the use of species-specific mitigation or wide structures that capture multiple species needs. Inconsistencies were observed between the crossing sites predicted in this study and the Parks Canada proposal. The usefulness of GIS and track data to enhance mitigation projects is illustrated.