Faculty Mentor: Andy Hong (City & Metropolitan Planning, University of Utah)

Abstract

Fall injuries are a public concern, and as individuals age falls become both more common and more injurious. This study assesses outdoor fall injuries among older adults and uses quantitative methods to understand how the built environment impacts fall risk and identify specific environmental fall risk factors. We analyzed data from the National Emergency Medical Services Information System (NEMSIS) to identify fall-related injuries, linking these incidents to environmental data by ZIP codes. A zero-inflated negative binomial regression model was employed to assess the association between environmental risk factors and fall-related injuries among older adults in the United States. The analysis identified six statistically significant environmental variables: urbanization, the percentage of senior population, slope, area deprivation index, the number of harsh winter days, and pedestrian intersection density. These findings provide a foundation for further investigation into how the built environment interacts with the risk of pedestrian falls.

Introduction

Fall injuries are a significant public health concern, particularly among older adults, who are more susceptible to environmental hazards while walking outdoors. Outdoor falls most commonly result from tripping or slipping, which are functions of the built environment (Timsina et al, 2017; Li et al, 2006; Lai et al, 2009). Further, the vast majority of pedestrian injuries result from falls, rather than vehicle collisions (Rundle et al, 2024).

Our study assesses outdoor fall injuries among older adults to understand how the built environment impacts fall risk and identify specific environmental fall risk factors. Past research on pedestrian falls has primarily relied on personal accounts of falls, rather than on quantitative environmental data. Our study uses National Emergency Medical Services Information System (NEMSIS) incident data to identify fall-related injuries. After linking fall incidents to environmental data by ZIP code, we use zero inflated negative binomial regression model to evaluate environmental risk factors associated with fall-related injuries among older adults in the United States.

Method

Data

For this study, we used 2018-2022 data from NEMSIS, a national database of EMS event records. No IRB oversight was necessary for our analysis since the records are released as a de-identified, publicly available dataset. Over 12 million events were recorded by over 10,000 EMS agencies in all fifty states and the District of Columbia. The only geographic identifiers available in the public dataset are Census Region and Division. However, NEMSIS makes some geographic variables available to researchers as masked, de-identified data elements.

Our analysis involved nine environmental variables captured at the ZIP Code level, which is the most granular geographic identifier provided by NEMSIS. ZIP Code Tabulation Area (ZCTA) boundaries were used for U.S. Census Bureau data and variable creation requiring spatial manipulation. The total population and the senior (65+) population were obtained from the 2018-2022 American Community Survey 5-Year estimates and the ZIP code area and urbanized status were obtained from the U.S. Census Bureau. The Area Deprivation Index (ADI) was obtained from the Center for Health Disparities Research at the University of Wisconsin School of Medicine and Public Health and pedestrian network and intersection densities were obtained from the U.S. EPA’s Smart Location Database. The average road slope was sourced from Jed Kolko’s research on broadband networks and the number of days with inclement winter weather was calculated using NOAA Global Historical Climatology Network daily (GHCNd) data.

Measures

Pedestrian falls were identified using NEMSIS data ICD-10 (2018 – 2022). The variables used to identify pedestrian falls are(Cause of injury), eInjury_04(injury risk factor), eResponse_05(type of service requested), eSituation_09(Primary symptom), and eScene_09(incident location type).

The rule of thumb for identifying pedestrian falls is the fall happens on a street or sidewalk, and the fall results from common pedestrian behaviors such as tripping, stumbling, or slipping. For eResponse_05, injuries coded with inter-facility transport for the type of service requested are dropped from the analysis. For eSituation_09, injuries that show symptoms of diseases that can cause pedestrian falls are dropped, such as heatstrokes and seizures. For eInjury_01, data with injuries that are often related to pedestrian behaviors such as fall on ice/snow are kept for analysis. For eScene_09, the location where pedestrian falls are most likely to happen. For eInjury_04, injuries with risk factors that are not pedestrian-related are dropped, such as Motorcycle Crash > 20 MPH. Applying the five criteria for identifying pedestrian falls, a total of 701,308 pedestrian falls were identified from 12,341,436 reported injuries. Of these, 246,176 falls involved patients aged over 65 years old.

Data Analysis

To analyze the data, we aggregated the pedestrian falls of older adults (aged > 65 years) by ZIP codes. We then fitted the data using five different models: the linear model, the Poisson model, the Negative Binomial model, the Zero-Inflated Poisson model, and the Zero-Inflated Negative Binomial (ZINB) model. Ultimately, the ZINB model was selected based on data characteristics and model performance.

Data Characteristics: Approximately one-third of the ZIP codes reported zero senior pedestrian falls, suggesting that a zero-inflated model is appropriate for handling the excess zeros in the data. Additionally, the variance of senior pedestrian falls by ZIP codes was 536.49, while the mean was 8.96. The variance is much larger than the mean indicating over dispersion, which violates the assumption of the Poisson model that the mean and variance are equal.

Model Performance: The Akaike Information Criterion (AIC) was used to evaluate model performance. Among the models tested, the ZINB model had the lowest AIC value, indicating the best balance between model complexity and goodness of fit.

The ZINB model consists of two components: the count model, which predicts the counts greater than zero, and the zero model, which predicts the occurrence of zeros. Based on the investigation of correlations among environmental variables, the following independent variables were selected for the count model: senior population percentage (calculated as the senior population aged 65+ divided by the total population), pedestrian intersection density, average road slope, number of days with inclement winter weather, urbanization, and the Area Deprivation Index. For the zero model, the independent variables selected were urbanization and population density (calculated as total population divided by area).

Results

With the zero inflated negative binomial model employed on pedestrian fall of older adults, five environment variables are revealed statistically significant in the count model.

Senior percentage and urbanization have the highest impact. a one-unit increase in the senior population percentage results in about a 3.55 times higher expected count of falls, and the expected count of falls is about 6.63 times higher in urbanized areas compared to non- urbanized areas. Senior percentage, Pedestrian intersection density, and urbanization are positively correlated with predicted senior pedestrian fall count, while area deprivation index and the number of days with inclement winter weather are negatively correlated with predicted senior pedestrian fall count.

Conclusion

The findings from our study suggest that the elderly population is at high risk for pedestrian falls, necessitating increased attention and targeted interventions. Our analysis identified several key environmental factors contributing to the likelihood of falls among older adults. Notably, areas with higher pedestrian intersection density are associated with higher fall counts. This finding suggests a need for further investigation into the underlying mechanisms of this phenomenon.

In addition to understanding these mechanisms, improving urban planning is crucial. Strategies should focus on reducing conflicts between pedestrians and pedestrians/vehicles at intersections, potentially through better traffic management, enhanced pedestrian crossings, and clearer signage. Other identified risk factors include the percentage of elderly residents, road slope, the number of inclement weather days, urbanization, and area deprivation. These factors should be prioritized when designing public safety measures and urban infrastructure.

Limitation

Our study has several limitations that should be acknowledged:

Data Limitations: We collected data on nine environmental variables. While these variables provide valuable insights, the results could be more robust if additional environmental factors were considered. Including a broader range of factors could help capture the complexity of the built environment and its impact on pedestrian falls.

Model Fitting: Our analysis primarily focused on the correlation between individual environmental variables and fall risk. However, environmental factors can interact in complex ways that were not explored in this study. Further investigation is needed to understand how these variables interact and influence pedestrian fall risk collectively.

Model Interpretation: We need more literature and knowledge in specific areas to better interpret these results. A comprehensive theoretical framework for interpreting the effects of each environmental factor on pedestrian falls will be helpful. Without a solid theoretical foundation, our interpretations are limited to observed correlations, and the causal mechanisms underlying these relationships remain unclear. Further research is necessary to develop a deeper understanding of the role each environmental factor plays in pedestrian fall risk.

Bibliography

Amin, K., Skyving, M., Bonander, C., Krafft, M., & Nilson, F. (2022). Fall- and collision-related injuries among pedestrians in road traffic environment – A Swedish national register-based study. Journal of Safety Research, 81, 153-165. https://doi.org/10.1016/j.jsr.2022.02.007.

Ceccato, V., & Willems, O. (2019). Temporal and spatial dynamics of falls among older pedestrians in Sweden. Applied Geography, 103, 122-133. https://doi.org/10.1016/j.apgeog.2018.12.007.

Chippendale, T., & Boltz, M. (2015). The neighborhood environment: Perceived fall risk, resources, and strategies for fall prevention. The Gerontological Society of America, 55(4), 575-583. https://doi.org/10.1093/geront/gnu019.

Chow, K. P., Fong, D. Y. T., Wang, M. P., Wong, J. Y. H., & Chau, P. H. (2018). Meteorological factors to fall: A systematic review. International Journal of Biometeorology, 62, 2073-2088. https://doi.org/10.1007/s00484-018-1627-y.

Curl, A., Thompson, C. W., Aspinall, P., & Ormerod, M. (2016). Developing an audit checklist to assess outdoor falls risk. Proceedings of the Institution of Civil Engineers: Urban Design and Planning, 169(3), 138-153. https://doi.org/10.1680/udap.14.00056.

Elvik, R., & Bjørnskau, T. (2019). Risk of pedestrian falls in Oslo, Norway: Relation to age, gender, and walking surface condition. Journal of Transport & Health 12, 359-370. https://doi.org/10.1016/j.jth.2018.12.006.

Gallagher, E. M., & Scott, V. J. The STEPS Project: Participatory action research to reduce falls in public places among seniors and persons with disabilities. Canadian Journal of Public Health, 88(2), 129-133. https://www.jstor.org/stable/41992693.

Gyllencreutz, L., Björnstig, J., Rolfsman, E., & Saveman, B. (2015). Outdoor pedestrian fall-related injuries among Swedish senior citizens – injuries and preventive strategies. Scandinavian Journal of Caring Sciences, 29, 225-233. https://doi.org/10.1111/scs.12153.

Kelsey, J. L., Procter-Gray, E., Hannan, M. T., & Li, W. (2012a). Heterogeneity of falls among older adults: Implications for public health prevention. American Journal of Public Health, 102(11), 2149- 2156. https://doi.org/10.2105/AJPH.2012.300677.

Kelsey, J. L., Procter-Gray, E., Berry, S. D., Hannan, M. T., Kiel, D. P., Lipsitz, L. A., & Li, W. (2012b). Reevaluating the implications of recurrent falls in older adults: Location changes the inference. Journal of the American Geriatrics Society, 60(3), 517-524. https://doi.org/10.1111/j.1532- 5415.2011.03834.x.

Lai, P., Low, C. T., Wong, M., Wong, W. C., & Chan, M. H. (2009). Spatial analysis of falls in an urban community in Hong Kong. International Journal of Health Geographics, 8(14). https://doi.org/10.1186/1476-072x-8-14.

Lai, P., Wong, P., Low, C., Wong, M., & Chan, M. (2011) A small-area study of environmental risk assessment of outdoor falls. Journal of Medical Systems, 35, 1543-1552. https://doi.org/10.1007/s10916-010-9431-

Lee, S., Lee, C., Ory, M. G., Won, J., Towne Jr, S. D., Wang, S., & Forjuoh, S. N. (2018). Fear of outdoor falling among community-dwelling middle-aged and older adults: The role of neighborhood environments. Gerontologist, 58(6), 1065-1074. https://doi.org/10.1093/geront/gnx123.

Lee, S., Lee., C., & Ory, M. G. (2019). Association between recent falls and changes in outdoor environments near community-dwelling older adults’ homes over time: Findings from the NHATS study. International Journal of Environmental Research and Public Health, 16(3230). http://dx.doi.org/10.3390/ijerph16183230.

Lee, S., Lee, C., & Rodiek, S. (2020). Outdoor exposure and perceived outdoor environments correlated to fear of outdoor falling among assisted living residents. Aging & Mental Health, 24(12), 1968-1976. https://doi.org/10.1080/13607863.2019.1647139.

Lee, S., Ye, X., Nam, J. W., & Zhang, K. (2022). The association between tree canopy cover over streets and elderly pedestrian falls: A health disparity study in urban areas. Social Science & Medicine, 306. https://doi.org/10.1016/j.socscimed.2022.115169.

Li, W., Keegan, T. H. M., Sternfeld, B., Sidney, S., Quesenberry Jr., C. P., & Kelsey, J. L. (2006). Outdoor falls among middle-aged and older adults: A neglected public health problem. American Journal of Public Health, 96(7), 1192-1200. https://doi.org/10.2105/ajph.2005.083055.

Li, W., Procter-Grey, E., Lipsitz, L. A., Leveille, S. G., Hackman, H., Biondolillo, M., & Hannan, M.T. (2014). American Journal of Public Health, 104(9), e30-e37. https://doi.org/10.2105/AJPH.2014.302104.

Morency, P., Voyer, C., Burrows, S., & Goudreau, S. (2012). Outdoor falls in an urban context: Winter weather impacts and geographical variations. Canadian Journal of Public Health, 103(3), 218-223. https://doi.org/10.1007/BF03403816.

Naumann, R. B., Dellinger, A. M., Haileyesus, T., & Ryan, G. W. (2011). Older adult pedestrian injuries in the United States: Causes and contributing circumstances. International Journal of Injury Control and Safety Promotion, 18(1), 65-73. https://doi.org/10.1080/17457300.2010.517321.

Nyman, S. R., Ballinger, C., Phillips, J. E., & Newton, R. (2013). Characteristics of outdoor falls among older people: A qualitative study. BMC Geriatrics, 13(125). http://www.biomedcentral.com/1471-2318/13/125.

Rantakokko, M., Mänty, M., Iwarsson, S., Timo Törmäkangas, Leinonen, R., Heikkinen, E., & Rantanen, T. (2009). Fear of moving outdoors and development of outdoor walking difficulty in older people. Journal of the American Geriatrics Society, 57, 634-640. https://doi.org/10.1111/j.1532- 5415.2009.02180.x.

Rubenstein, L. Z. (2006). Falls in older people: Epidemiology, risk factors, and strategies for prevention. Age and Ageing, 35(S2): ii37-ii41. https://doi.org/10.1093/ageing/afl084.

Rundle, A. G., Crowe, R. P., Wang, H. E., & Lo, A. X. (2023). A methodology for the public health surveillance and epidemiologic analysis of outdoor falls that require an emergency medical services response. Injury Epidemiology, 10(4). https://doi.org/10.1186/s40621-023-00414-z.

Rundle, A. G., Crowe, R. P., Wang, H. E., Beard, J. R., & Lo, A. X. (2024). A national study on the comparative burden of pedestrian injuries from falls relative to pedestrian injuries from motor vehicle collisions. Journal of Urban Health, 101(1), 181-192. https://doi.org/10.1007/s11524-023-00815-x.

Schepers, P., den Brinker, B., Methorst, R., & Helbich, M. (2017). Pedestrian falls: A review of the literature and future research directions. Journal of Safety Research, 62, 227-234. http://dx.doi.org/10.1016/j.jsr.2017.06.020.

Strath, S., Isaacs, R., & Greenwalk, M. J. (2007). Operationalizing environmental indicators for physical activity in older adults. Journal of Aging and Physical Activity, 15, 412-424. https://doi.org/10.1123/japa.15.4.412.

Timsina, L. R., Willetts, J. L., Brennan, M. J., Marucci-Wellman, H., Lombardi, D. A., Courtney, T. K., & Verma, S. K. (2017). Circumstances of fall-related injuries by age and gender among community-dwelling adults in the United States. PLoS ONE 12(5), 1-21. https://doi.org/10.1371/journal.pone.0176561.

Twardzik, E., Clarke, P., Judd, S., & Colabianchi, N. (2021). Neighborhood participation is less likely among older adults with sidewalk problems. Journal of Aging and Health, 33(1-2), 101-113. https://doi.org/10.1177/0898264320960966.

Unguryanu, T., Grjibovski, A. M., Trovik, T. A., Ytterstad, B., & Kudryavtsev, A. V. (2020). Weather conditions and outdoor fall injuries in northwestern Russia. International Journal of Environmental Research and Public Health, 17(6096). https://doi.org/10.3390/ijerph17176096.

Watkins, A., Curl, A., Mavoa, S., Tomintz, M., Todd, V., & Dicker, B. (2021). A socio-spatial analysis of pedestrian falls in Aotearoa New Zealand. Social Science & Medicine, 288. https://doi.org/10.1016/j.socscimed.2020.113212.

Yiannakoulias, N., Rowe, B. H., Svenson, L., Schopflocher, D. P., Kelly, K., & Voaklander, D. C. (2003). Zones of prevention: The geography of fall injuries in the elderly. Social Science & Medicine, 57, 2065-2073. https://doi.org/10.1016/S0277-9536(03)00081-9.

Zhang, C., Zhang, R., Baker, J. S., Hagger, M. S., & Hamilton, K. (2022). A qualitative investigation exploring neighbourhood environment, risks and fear of falling, and fall prevention strategies among urban-dwelling older adults in a high-density city. Ageing & Society, 1-22. https://doi.org/10.1016/S0277-9536(03)00081-9.