Stefania Bertazzon

Comparison of distance measures in spatial analytical modeling for health service planning

BackgroundSeveral methodological approaches have been used to estimate distance in health service research. In this study, focusing on cardiac catheterization services, Euclidean, Manhattan, and the less widely known Minkowski distance metrics are used to estimate distances from patient residence to hospital. Distance metrics typically produce less accurate estimates than actual measurements, but each metric provides a single model of travel over a given network. Therefore, distance metrics, unlike actual measurements, can be directly used in spatial analytical modeling. Euclidean distance is most often used, but unlikely the most appropriate metric. Minkowski distance is a more promising method. Distances estimated with each metric are contrasted with road distance and travel time measurements, and an optimized Minkowski distance is implemented in spatial analytical modeling.MethodsRoad distance and travel time are calculated from the postal code of residence of each patient undergoing cardiac catheterization to the pertinent hospital. The Minkowski metric is optimized, to approximate travel time and road distance, respectively. Distance estimates and distance measurements are then compared using descriptive statistics and visual mapping methods. The optimized Minkowski metric is implemented, via the spatial weight matrix, in a spatial regression model identifying socio-economic factors significantly associated with cardiac catheterization.ResultsThe Minkowski coefficient that best approximates road distance is 1.54; 1.31 best approximates travel time. The latter is also a good predictor of road distance, thus providing the best single model of travel from patients residence to hospital. The Euclidean metric and the optimal Minkowski metric are alternatively implemented in the regression model, and the results compared. The Minkowski method produces more reliable results than the traditional Euclidean metric.ConclusionRoad distance and travel time measurements are the most accurate estimates, but cannot be directly implemented in spatial analytical modeling. Euclidean distance tends to underestimate road distance and travel time; Manhattan distance tends to overestimate both. The optimized Minkowski distance partially overcomes their shortcomings; it provides a single model of travel over the network. The method is flexible, suitable for analytical modeling, and more accurate than the traditional metrics; its use ultimately increases the reliability of spatial analytical models.

Residence location and likelihood of kidney transplantation

Background: In a universal, public health care system, access to kidney transplantation should not be influenced by residence location. We determined the likelihood of kidney transplantation from deceased donors among Canadian dialysis patients living in 7 geographic regions. Within each region we also determined whether distance from the closest transplant centre was associated with the likelihood of transplantation. Methods: A random sample of 7034 subjects initiating dialysis in Canada between 1996 and 2000 was studied. We used Cox proportional hazards models to examine the relation between residence location and the likelihood of kidney transplantation from deceased donors over a median period of 2.4 years. Results: There were significant differences in the likelihood of kidney transplantation from deceased donors and predicted waiting times between the different geographic regions. For example, the adjusted relative likelihood of transplantation in Alberta was 3.74 (95% confidence interval [CI] 2.95–4.76) compared with the likelihood in Ontario (p < 0.001). These differences persisted after further adjustment for differences in the rate of deceased organ donation. Within regions, patients who resided 50.1–150 km, 150.1–300 km and more than 300 km from the closest transplant centre had a similar adjusted likelihood of receiving a kidney transplant as those who lived less than 50 km away. Interpretation: The adjusted likelihood of undergoing a kidney transplant from a deceased donor varied substantially between geographic regions in Canada. In contrast, the likelihood of transplantation within regions was not affected by distance from the closest transplant centre.

Cruising in the Mediterranean: structural aspects and evolutionary trends

In recent years the Mediterranean has grown so markedly within the global cruise market that it now ranks second in the world. Demand growth rates are constantly positive. Supply is steadily growing; major world companies are deploying more vessels in the area; many passenger terminals and ports are undergoing infrastructural modernization. Overall, the entire Mediterranean cruise sector holds a far greater appeal than in the recent past. Vertical integration processes have played a major role in increasing the dynamism of the sector and in affecting the competition among ports. In this context, Barcelona and Civitavecchia have emerged as the top ranking ports. Continued growth in the Mediterranean market can be expected in the near future; the main challenges are continued terminal and service modernization, geopolitical stability, particularly in its central-eastern part, and effective marketing strategies integrating port activity with inland resources.

Accounting for spatial effects in land use regression for urban air pollution modeling.

In order to accurately assess air pollution risks, health studies require spatially resolved pollution concentrations. Land-use regression (LUR) models estimate ambient concentrations at a fine spatial scale. However, spatial effects such as spatial non-stationarity and spatial autocorrelation can reduce the accuracy of LUR estimates by increasing regression errors and uncertainty; and statistical methods for resolving these effects--e.g., spatially autoregressive (SAR) and geographically weighted regression (GWR) models--may be difficult to apply simultaneously. We used an alternate approach to address spatial non-stationarity and spatial autocorrelation in LUR models for nitrogen dioxide. Traditional models were re-specified to include a variable capturing wind speed and direction, and re-fit as GWR models. Mean R(2) values for the resulting GWR-wind models (summer: 0.86, winter: 0.73) showed a 10-20% improvement over traditional LUR models. GWR-wind models effectively addressed both spatial effects and produced meaningful predictive models. These results suggest a useful method for improving spatially explicit models.

Ambient Ozone Concentrations and the Risk of Perforated and Nonperforated Appendicitis: A Multicity Case-Crossover Study

Background: Environmental determinants of appendicitis are poorly understood. Past work suggests that air pollution may increase the risk of appendicitis. Objectives: We investigated whether ambient ground-level ozone (O3) concentrations were associated with appendicitis and whether these associations varied between perforated and nonperforated appendicitis. Methods: We based this time-stratified case-crossover study on 35,811 patients hospitalized with appendicitis from 2004 to 2008 in 12 Canadian cities. Data from a national network of fixed-site monitors were used to calculate daily maximum O3 concentrations for each city. Conditional logistic regression was used to estimate city-specific odds ratios (ORs) relative to an interquartile range (IQR) increase in O3 adjusted for temperature and relative humidity. A random-effects meta-analysis was used to derive a pooled risk estimate. Stratified analyses were used to estimate associations separately for perforated and nonperforated appendicitis. Results: Overall, a 16-ppb increase in the 7-day cumulative average daily maximum O3 concentration was associated with all appendicitis cases across the 12 cities (pooled OR = 1.07; 95% CI: 1.02, 1.13). The association was stronger among patients presenting with perforated appendicitis for the 7-day average (pooled OR = 1.22; 95% CI: 1.09, 1.36) when compared with the corresponding estimate for nonperforated appendicitis [7-day average (pooled OR = 1.02, 95% CI: 0.95, 1.09)]. Heterogeneity was not statistically significant across cities for either perforated or nonperforated appendicitis (p > 0.20). Conclusions: Higher levels of ambient O3 exposure may increase the risk of perforated appendicitis.

Fine particulate air pollution, nitrogen dioxide, and systemic autoimmune rheumatic disease in Calgary, Alberta

OBJECTIVE To estimate the association between fine particulate (PM2.5) and nitrogen dioxide (NO2) pollution and systemic autoimmune rheumatic diseases (SARDs). METHODS Associations between ambient air pollution (PM2.5 and NO2) and SARDs were assessed using land-use regression models for Calgary, Alberta and administrative health data (1993-2007). SARD case definitions were based on ≥2 physician claims, or ≥1 rheumatology billing code; or ≥1 hospitalization code (for systemic lupus, Sjogrens Syndrome, scleroderma, polymyositis, dermatomyositis, or undifferentiated connective tissue disease). Bayesian hierarchical latent class regression models estimated the probability that each resident was a SARD case, based on these case definitions. The sum of individual level probabilities provided the estimated number of cases in each area. The latent class model included terms for age, sex, and an interaction term between age and sex. Bayesian logistic regression models were used to generate adjusted odds ratios (OR) for NO2 and PM2.5. pollutant models, adjusting for neighbourhood income, age, sex, and an interaction between age and sex. We also examined models stratified for First-Nations (FN) and non-FN subgroups. RESULTS Residents that were female and/or aged >45 had a greater probability of being a SARD case, with the highest OR estimates for older females. Independently, the odds of being a SARDs case increased with PM2.5 levels, but the results were inconclusive for NO2. The results stratified by FN and non-FN groups were not distinctly different. CONCLUSION In this urban Canadian sample, adjusting for demographics, exposure to PM2.5 was associated with an increased risk of SARDs. The results for NO2 were inconclusive.

A preliminary spatial assessment of risk: Marine birds and chronic oil pollution on Canada's Pacific coast

Chronic oil pollution poses substantial risks to marine birds and other marine wildlife worldwide. On Canadas Pacific coast, the negative ecological consequences to marine birds and marine ecosystems in general remain poorly understood. Using information relating to oil spill probability of occurrence, areas of overall importance to marine birds, and the at-sea distribution and density of 12 marine bird species and seven bird groups, including multiple Species at Risk, we undertook a spatial assessment of risk. Our results identify two main areas important to marine birds potentially at higher risk of exposure to oil. For individual bird species or species groups, those predicted to have elevated bird densities near the mainland and the northeast coast of Vancouver Island were identified as being at higher potential risk of exposure. Our results, however, should be considered preliminary. As with other anthropogenic stressors, in order to better understand and subsequently mitigate the consequences of chronic oil pollution on marine birds, improved information relating to marine birds and the occurrence of oil spills on Canadas Pacific coast is needed.

Spatial analysis in ecological risk assessment: Pollutant bioaccumulation in clams Tapes philipinarum in the Venetian lagoon (Italy)

Exposure characterization is a central step in Ecological Risk Assessment (ERA). Exposure level is a function of the spatial factors linking contaminants and receptors, yet exposure estimation models are traditionally non-spatial. Non-spatial models are prone to the adverse effects of spatial dependence: inflated variance and biased inferential procedures, which can result in unreliable and potentially misleading models. Such negative effects can be amended by spatial regression modelling: we propose an integration of geostatistics and multivariate spatial regression to compute efficient spatial regression parameters and to characterize exposure at under-sampled locations. The method is applied to estimate bioaccumulation models of organic and inorganic micropollutants in the tissues of the clam Tapes philipinarum. The models link bioaccumulation of micropollutants in clam tissue to a set of environmental variables sampled in the lagoon sediment. The Venetian lagoon case study exemplifies the problem of multiple variables sampled at different locations or spatial units: we propose and test an effective solution to this common and serious problem in environmental as well as socio-economic multivariate analysis.

Spatial Analysis of Wildlife Distribution and Disease Spread

Many of the interactions between organisms depend on the distance or the ease of movement (accessibility) between them which can be based on the concept of the neighbors or of the neighborhoods of given individuals. A number of different statistical approaches have been developed (Fortin and Dale 2005; Perry 1995) to address the definitions of neighbors and neighborhoods in order to implement measures of those characteristics that are most important to the interactions under study. In particular, the numbers of neighbors (however defined) and their distances can be combined into measures of aggregation, dispersion or crowding (Lloyd 1967), which can have clear effects on important demographic processes, such as the spread of disease, beyond the simple effect of distance to the nearest organisms of the same or different kinds.

Alternative Distance Metrics for Enhanced Reliability of Spatial Regression Analysis of Health Data

We present a spatial autoregressive model (SAR) to investigate the relationship between the incidence of heart disease and a pool of selected socio-economic factors in Calgary (Canada). Our goal is to provide decision makers with a reliable model, which can guide locational decisions to address current disease occurrence and mitigate its future occurrence and severity. To this end, the applied model rests on a quantitative definition of neighbourhood relationships in the city of Calgary. Our proposition is that such relationships, usually described by Euclidean distance, can be more effectively described by alternative distance metrics. The use of the most appropriate metric can improve the regression model by reducing the uncertainty of its estimates, ultimately providing a more reliable analytical tool for management and policy decision making.