Cartographic Modeling Lab

The Cartographic Modeling Lab (CML) is a research unit within the Biostatistics Analysis Center (BAC) of the Perelman School of Medicine's Center for Clinical Epidemiology and Biostatistics (CCEB). The CML specializes in applying geographic information systems (GIS) software and hardware to digitally link data and geography to generate spatial databases, maps, spatial statistical analyses, and mapping applications, providing a useful way to reveal spatial and temporal relationships among data. By using GIS to visualize geographic relationships that affect health outcomes, public health risks, disease transmission, access to health care, and other public health concerns, the CML conducts spatial research, policy analysis, and develops mapping applications of value to investigators at Penn and beyond.

 

Search results

Now showing 1 - 5 of 5
  • Publication
    Vacant Properties and Violence in Neighborhoods
    (2012-01-01) Branas, Charles; Rubin, David; Guo, Wensheng
    Objectives. Violence remains a significant public health issue in the United States. To determine if urban vacant properties were associated with an increased risk of assaultive violence and if this association was modified by important neighborhood institutions (e.g., schools, parks/playgrounds, police stations, and alcohol outlets). Methods. Longitudinal ecologic study of all 1816 block groups in Philadelphia. Aggravated assault and vacant property data were compiled yearly from 2002 to 2006 and linked to block groups. A mixed effects negative binomial regression model examined the association of vacant properties and assaults between and within block groups. Results. Among all block groups, 84% experienced at least one vacant property, 89% at least one aggravated assault, and 64% at least one gun assault. Between block groups, the risk of aggravated assault increased 18% for every category shift of vacant properties (IRR 1.18, 95% CI: 1.12, 1.25, P < 0.001). Parks/playgrounds and alcohol outlets potentially modified the association between vacant properties and aggravated assaults but only at low levels of vacancy. Conclusions. Increasing levels of vacancy were associated with increased risk of assaultive violence in urban block groups.
  • Publication
    Homicide and Geographic Access to Gun Dealers in the United States
    (2009-06-23) Wiebe, Douglas J.; Nance, Michael L.; Krafty, Robert T.; Koper, Christopher S.; Branas, Charles; Elliott, Michael R.
    Background: Firearms are the most commonly used weapon to commit homicide in the U.S. Virtually all firearms enter the public marketplace through a federal firearms licensee (FFL): a store or individual licensed by the federal government to sell firearms. Whether FFLs contribute to gun-related homicide in areas where they are located, in which case FFLs may be a homicide risk factor that can be modified, is not known. Methods: Annual county-level data (1993–1999) on gun homicide rates and rates of FFLs per capita were analyzed using negative binomial regression controlling for socio-demographic characteristics. Models were run to evaluate whether the relation between rates of FFLs and rates of gun homicide varied over the study period and across counties according to their level of urbanism (defined by four groupings, as below). Also, rates of FFLs were compared against FS/S – which is the proportion of suicides committed by firearm and is thought to be a good proxy for firearm availability in a region – to help evaluate how well the FFL variable is serving as a way to proxy firearm availability in each of the county types of interest. Results: In major cities, gun homicide rates were higher where FFLs were more prevalent (rate ratio [RR] = 1.70, 95% CI 1.03–2.81). This association increased (p < 0.01) from 1993 (RR = 1.69) to 1999 (RR = 12.72), due likely to federal reforms that eliminated low-volume dealers, making FFL prevalence a more accurate exposure measure over time. No association was found in small towns. In other cities and in suburbs, gun homicide rates were significantly lower where FFLs were more prevalent, with associations that did not change over the years of the study period. FFL prevalence was correlated strongly (positively) with FS/S in major cities only, suggesting that the findings for how FFL prevalence relates to gun homicide may be valid for the findings pertaining to major cities but not to counties of other types. Conclusion: Modification of FFLs through federal, state, and local regulation may be a feasible intervention to reduce gun homicide in major cities.
  • Publication
    Exploring the Role of the Food Environment on Food Shopping Patterns in Philadelphia, PA, USA: A Semiquantitative Comparison of Two Matched Neighborhood Groups
    (2013-01-14) Hirsch, Jana A.; Hillier, Amy
    Increasing research has focused on the built food environment and nutrition-related outcomes, yet what constitutes a food environment and how this environment influences individual behavior still remain unclear. This study assesses whether travel mode and distance to food shopping venues differ among individuals in varying food environments and whether individual- and household-level factors are associated with food shopping patterns. Fifty neighbors who share a traditionally defined food environment (25 in an unfavorable environment and 25 in a favorable environment) were surveyed using a mix of close- and open-ended survey questions. Food shopping patterns were mapped using Geographic Information Systems (GIS). Stores visited were beyond the 0.5-mile (805 meters) radius traditionally used to represent the extent of an individual’s food environment in an urban area. We found no significant difference in shopping frequency or motivating factor behind store choice between the groups. No differences existed between the two groups for big food shopping trips. For small trips, individuals in the favorable food environment traveled shorter distances and were more likely to walk than drive. Socioeconomic status, including car ownership, education, and income influenced distance traveled. These findings highlight the complexities involved in the study and measurement of food environments.
  • Publication
    PhillydotMap: The Shape of Philadelphia
    (2009-10-01) Branas, Charles; Davitt, Joan K; Hillier, Amy; Cheetham, Robert; Cheetham-Richard, Rachel; Cohen, Rachel B; Decker, Joan; Heckert, Megan; Steinberg, Harris; Sullivan-Marx, Eileen; Hutchins, Jason; Tomlin, C. Dana; Kerman, Lucy; Volpe, Stella; Lechner, Zachary L; Wiebe, Douglas; McLarnon, Michael; Newlin, Heather; Scholssberg, Dina; Wormley, Diane-Louise
    This book is the outgrowth of a working group entitled, “Modeling Urban Environmental Impacts on Health, Development, and Behavior" sponsored by the University of Pennsylvania Institute for Urban Research. The purpose of the working gropu was to engage faculty from across campus and to encourage their collaborative use of GIS technology in the modeling of urban form and function. These ten chapters represent a wide range of GIS applications, from community-based social services to public history to social science research.
  • Publication
    Hospital-Reported Pneumococcal Susceptibility to Penicillin
    (2004-01-01) Branas, Charles; Metlay, Joshua P.; Fishman, Neil O.
    Geographic variation in drug susceptibility among isolates of Streptococcus pneumoniae has influenced national treatment guidelines for community-acquired pneumonia. Whether individual hospital susceptibility data provide reliable and valid information for providers is unclear. We examined the geographic and temporal variability in hospital-reported rates of pneumococcal susceptibility. We surveyed all 52 hospitals that provided acute adult care in the five counties surrounding Philadelphia and collected data on levels of penicillin susceptibility among all pneumococcal blood isolates from 1998 to 2000. In 1998, pneumococcal nonsusceptibility to penicillin varied from 0% to 67% of all blood isolates across the 33 hospitals with ≥10 isolates in that year. Hospital location did not correlate with the level of reported pneumococcal susceptibility (p = 0.8). In addition, correlations were not significant in reported pneumococcal susceptibility to penicillin within individual hospitals during the 3 years.