2. Scoping  Creation of logic model –Specify how policy and infrastructure changes will eventually impact health outcomes –Helps in focusing the impact assessment Quantitative Qualitative

3. Logic Model Build sidewalks and crosswalks Safety Physical activity Obesity Mortality Lung disease CVD Cancer Diabetes Traffic Land-use People outside Depression, anxiety, stress Hypertension Osteoporosis Air and noise pollution Permit mixed-use zoning  floor/area ratio  dwelling units  pop. density Change 60 ft. easement to 40 ft. (thin wall arcade, buildings built closer to sidewalk, Oreo deck) Social capital Parking requirements Injury I-85 traffic  connectivity Bus ridership PolicyProximal Intermediate Health Impacts Impacts Outcomes  parking Injuries and fatalities Pop. density

4. Risk Assessment  Qualitative –Traffic –Pollution –Social capital –Crime and safety –Economic development –Gentrification  Quantitative –Injury –Physical Activity

5. Determining Affected Population  The individuals who live in the study area (N. Druid Hills to Clairmont) –5 census blocks –Only counted those that lived ½ mile from highway –14,000 people  Individuals who drive through study area –ADT (23,034) x people per car (1.63) –37,545 people –No demographic data available

7. Demographics for Study Area Study Area Atlanta % Male60.049.4 Age 0-1718.926.6 18-2928.318.1 30-3923.318.4 40-4910.915.7 50+ 8.621.2

8. Demographics for Study Area Study Area Atlanta Race White 47.363.0 Black20.828.8 Asian 4.8 3.3 Ethnicity Hispanic 49.8 6.5

9. Demographics for Study Area Study Area Atlanta Foreign-born61.110.3 Non-resident 199526.6 4.1 Poverty15.8 9.2 Avg. income $45,511 $51,948

10. Household Demographics  Average family size is 3.4  Most families (70%) have 2 or more workers  12% of households have no car and 48% have 1 car  17% take transit to work and 3% walk

11. Pedestrian Data for All Crashes in DeKalb County, GA  67% of pedestrians hit were males  77% of pedestrian fatalities were males  Of the 62 fatally injured pedestrians: –47% Black –36% Hispanic –17% White DeKalb Board of Health (2003)

12. Severity of Injuries in DeKalb on Buford Highway SeverityN % Fatalities1216.2 Serious Injuries1723.0 Visible Injuries2939.2 Complaints of Injuries1216.2 No Injuries 4 5.4 * DeKalb Board of Health

13. Number of Injuries and Deaths on Buford Highway DeKalbStudy Area (8 miles)(2.37 miles) Injuries/year 18.6 6.7 Deaths/year 3.61.8 DeKalb Board of Health (2003)

16. Estimating Changes in Injury  No studies could be located to determine injury reduction based on proposed changes  Hired senior traffic engineers (Hamilton & Associates) to calculate expected changes

17. Estimating Crash Reduction CRFt = CRF1 + (CRF2 ) (1-CRF1) + … (CRFn) (1-CRF1) (1 – CRF2)…(1-CRFn-1) –Where CRFt = CRF of combined measures –CRF1 = CRF for the first countermeasure –CRF2 = CRF found the second countermeasure –CRFn = CRF for the nth countermeasure

18. Improvement MeasureAll Collisions CRF Pedestrian Collision CRF Replacement of two-way left-turn lane with raised median 25% - 45%55% Sidewalks1%65% - 75% Added/improved pedestrian crosswalks13% - 25%19% Reduced speed limit1% - 3%15% - 30% Access control: service road/frontage road5% - 12%10% - 30% Combined measuresRange39% - 65%89% - 94% Best-guess point estimate60%91% Collision Reduction Factors Hamilton & Associates (2004) *ranges represent upper and lower bound estimates from studies

19. Injuries and Fatalities: Study Area Current Expected After Reduction Pedestrian Injuries/Year 6.7.91 (.89-.94) 0.4 Pedestrian Deaths/Year 1.8.91 (.89-.94) 0.1 Automobile Injuries/Year 120.60 (.39 -.65) 46

20. Physical Activity

21.  Hard to find study that had good measures of physical activity and the built environment – we chose study with best measure of physical activity  Saelens et al. (2003) found a 72.5 minute difference in total walking per week between neighborhoods in San Diego  Saelens et al. (2004) found 124 minute difference in walking for transport In low- income neighborhoods in Seattle

22. Walkability Audit Results Neighborhood Score Grade San Diego High Walkable1.4A- San Diego Low Walkable2.0B Buford Before4.1D Buford After2.4B- 1=A to 6=F

25. Estimating Increases in Walking  Since there were only 2 data points to serve as the source for the effect parameter there is uncertainty with respect to the shape of the relationship –Linear increase (204 minute increase/week) –Dichotomous function (76.2 minute increase/week) –Curvilinear relationship (11 minute increase/week) –No effect

26. Buford Highway Post-Project – Seattle Estimate = 200 mins/week

27. Qualitative Analysis  Traffic –Probably decrease along Buford Highway with increases along other Atlanta Highways  Noise Pollution –Increase during construction then decrease afterwards due to slower and less traffic  Air Pollution –Small decreases in local area but not significant enough to affect entire Atlanta area

28.  Social capital –Increase in social capital due to increased walkability and greenspace  Crime and safety –Literature to mixed to make any predictions on direction  Economic development and Gentrification –Due to central location inside Atlanta next to exclusive neighborhoods economic development as well as gentrification are likely Qualitative Analysis

29. Assumptions for Estimating PA  Their were several limitations in the Saelens et al (2003) study which may affect the expected increases in pa predicted in this HIA  Walkability and not other factors explain the differences found in the Saelens et al (2003) study  The built environment aspects that were correlated with differences in pa in the San Diego neighborhoods will have the same effect along Buford Highway

30. Assumptions for Estimating PA  The Ft. Collins LOS instrument was able to capture the elements of the built environment related to physical activity  Increases in walking that are predicted represent increases in total pa and not a substitution  The relationship between the built environment and pa is not only correlational but causal  For the CEA analysis it was assumed that the walking bouts will be at least 10 minutes in length so that they will have health impacts

31. Assumptions for Estimating Injury  Traffic calming measures used in other parts of the county will have the same effect along Buford Highway  The effects of the crash reduction factors are additive  The best available estimates for CRFs were used, which included personal communication with local DOTs, and the predictive certainty of most of the CRFs are unknown

32. Assumptions for Estimating Injury  Traffic may be diverted onto other streets and there may be a change in injuries along those streets  The residents will use the medians and crosswalks  For the CEA It was assumed that the same number of people will be driving and walking along Buford Highway despite the projected increases in population

33. Evaluation of Impact  Sections of Buford Highway (Shallowford to I-285) will be redeveloped starting in the spring of 2005 –Changes will not be as extensive as those proposed by Georgia Tech –Decision made before HIA was completed –Discussing placing full medians with sidewalks in the southern section of Buford Highway

34. Key Challenges of HIA  Uncertainties (data, models, policy)  Timeliness  Relevance to stakeholders and decision makers –Political context –Importance relevant to other factors  Capacity to conduct HIAs

35. Next Steps for HIA  Adapting HIA to the unique policy-making environment of the U.S.  Moving from research to practice –Methods to sort through bills/initiatives to find those for which HIA is most suitable –Standardizing and streamlining impact estimation –Determine feasibility of different types of tools in various settings –Training

36. Summary  HIA is a new and evolving science in the U.S., however it is a promising new approach to quantify health impacts of a wide variety of policies and projects  HIA provides only one piece of information (health) in complex decisions and stakeholders may have different priorities  HIA provides an outlet for health to be appropriately factored into complex decisions