Pedestrian-Cyclist Accidents in New York State: 2007-2010

Principal Investigators: Peter Tuckel Department of Sociology Hunter College [email protected] William Milczarski Department of Urban Planning & Affairs Hunter College [email protected]

September, 2011

Introduction
While considerable attention has been paid to pedestrians involved in motor vehicle accidents, little attention has been accorded to pedestrians involved in accidents resulting from collisions with cyclists. This lack of scrutiny is surprising given the growing popularity of cycling in America. With so many more individuals adopting cycling as an alternative means of transportation, the potential for collisions with pedestrians has grown markedly. An abundance of anecdotal evidence already exists suggesting that crashes between pedestrians and cyclists is a common phenomenon. Newspapers, radio, and television are replete with stories about individuals being hit or being nearly hit by cyclists. Rogue cyclists those who flout traffic laws by running red lights, riding in the opposite direction of traffic, riding on sidewalks, etc have been prominently featured in the media. Yet little systematic data has been collected to either support or challenge this anecdotal evidence. What is the incidence of crashes between pedestrians and cyclists? Is it a common occurrence or is it a relatively rare event? What is the profile of pedestrians most likely to be involved in these collisions? And where do these collisions happen? The present study attempts to answer these questions based upon empirically-gathered data.

Methodology
This study rests upon data furnished by the Statewide Planning and Research Cooperative System (SPARCS) under the auspices of New York States Department of Health. A basic description of SPARCS is provided on its website: SPARCS currently collects patient level detail on patient characteristics, diagnoses and treatments, services, and charges for every hospital discharge, ambulatory surgery patient, and emergency department admission in New York State. 1 A key feature of the SPARCS data set includes a diagnosis for each patient using the International Classification of Disease (ICD) code. One of the values of this ICD code is for patients who were involved in a pedestriancyclist accident (E826.0). This value of the code pertains only to the pedestrians who necessitated medical treatment not the cyclists.

The data set for this study is based upon both in- and out-patients who were assigned this value on the ICD code for the four years from 2007 to 2010. In addition, all non-identifying information (i.e., non-personal information) attached to each patients record was included in the data set. This additional information included both individual-level variables such as age, sex, race, ethnicity (Hispanic origin/non-Hispanic origin) and geographic variables such as county, city, and zip code. All together, the data set consisted of 4,121 individuals of whom 3,779 were out-patients and 342 were in-patients.

Findings
1) The frequency of pedestrian-cyclist accidents over time Table 1 below displays the number of out- and in-patients involved in pedestrian-cyclist accidents for each year spanning the period under investigation. Table 1. Frequency of Accident by Year
Type of patient outinpatient patient Year 2007 2008 2009 2010 Total 1012 1013 911 843 3779 85 99 74 84 342 1097 1112 985 927 4121

Total

As the data indicate, there are approximately 1,000 patients involved in pedestrian-cyclist accidents per year, the vast proportion of whom are outpatients. Though there appears to be a downward trend in the number of

accidents over time, this inference must be treated with caution as the number of data points (e.g., years) are few.2 2) The frequency of pedestrian-cyclist accidents by month In Table 2 below are presented the number of patients involved in pedestrian-cyclist accidents in each month aggregated over the four years covered by this investigation. Not unexpectedly, the data reveal a seasonal pattern. Most accidents occur in the summer months June, July, and August. The fewest number of accidents occur in the winter months December, January, and February. Table 2. Frequency of Accident by Month: 2007-2010
Month January February March April May June July August September October November December Total Frequency 94 64 181 382 525 591 672 599 434 303 159 100 4104 Percent 2.3 1.6 4.4 9.3 12.8 14.4 16.4 14.6 10.6 7.4 3.9 2.4 100.0

2) Demographic profile of accident victims The data in Table 3 reveal there is a sex imbalance in the number of patients. Overall, three-fifths of the patients are male and two-fifths are female. However, this sex imbalance vanishes among in-patients. Here the frequency of males and females is almost the same. This indicates that

among more serious cases necessitating hospital admission, there is no gender gap. Table 3. Sex of patients: 2007-2010

Sex Female Male Total

Type_of patient out-patient in-patient 39.4% 48.8%

60.6% 100.0%

51.2% 100.0%

The age of patients differs significantly depending upon their status as outor in-patients (see Table 4). Out-patients tend to be much younger than inpatients. Among out-patients, the vast majority (72.3%) are under the age of 40. By contrast, only about a third of in-patients (34.5%) are under the age of 40. Compared to the entire New York State population, individuals involved in pedestrian-cyclist accidents who were out-patients are much less younger. Almost 24 percent of out-patients were under the age of 10 and an additional 26.5 percent were between the ages of 10 to 20. By comparison, only 12 percent of New York States population is under the age of 10 and only 13.4 percent are between the ages of 10 to 20.

Table 4. Age of patients vs. Age of New York State Residents

Type_of patient out-patient in-patient 23.9% 11.1% 26.5% 7.9% 21.9% 15.5% 14.7% 18.7% 13.0% 46.8% 100.0% 100.0% Total 22.8% 25.0% 21.4% 15.0% 15.8% 100.0% N.Y. State 12.0% 13.4% 27.4% 21.8% 25.4% 100.0%

Age

less than 10 yrs. 10 to 20 yrs. 21 to 39 yrs. 40 to 55 yrs. 55 plus yrs.

Total

Importantly, there is also a sex-age interaction effect on the incidence of accidents among both out- and in-patients (see Table 5). For both types of patients, females are far more likely to be found among those in the highest age category 55 years of age or older. For example, among out-patients in the highest age category, fully 62 percent are female compared to just 38 percent who are male. Among in-patients in the highest age category, the same story unfolds: 59.4 percent are female and 40.6 percent are male.

Table 5. Age and Sex of Accident Victims by Patient Status

Type of_patient less than 10 39.1% 60.9% 100.0% 31.6% 68.4% 100.0% 10 to 20 26.7% 73.3% 100.0% 22.2% 77.8% 100.0%

Age 21 to 39 38.4% 61.6% 100.0% 45.3% 54.7% 100.0% 40 to 55 44.6% 55.4% 100.0% 46.9% 53.1% 100.0% 55 thru highest 62.0% 38.0% 100.0% 59.4% 40.6% 100.0%

out-patient

sex Total sex Total

F M F M

in-patient

In terms of race, almost three-fifths of the patients (58%) are white, 16.7 percent are black, 4.1 percent are Asian, and the remainder are of other races (see Table 6). Compared to New York State as a whole, whites and Asians are underrepresented. Looked at in terms of ethnicity, three-quarters of patients are non-Hispanic and one-quarter are of Hispanic origin. According to the 2010 U.S. census, individuals who are Hispanic or of Latin origin comprise 17.6 percent of the population of New York State. Thus, Hispanics are overrepresented among individuals involved in pedestrian-cyclist accidents in New York State. Table 6. Race of Patients Vs. Race of New York State Residents
New York State Residents (%)* 65.7 15.9 0.6 7.3 0.0 7.4 (missing) 4.8 100.0 (19,378,102)

Race White Black or African American Native American Asian Native Hawaiian Other race Unknown Two or more races Total (%) Total (N)

Patients (%) 58.0 16.7 0.6 4.1 0.0 19.2 1.4 (missing) 100.0 (4121)

*Statistics for Unknown in New York State and Two or more races for patients were not available.

3) Geographic residence of accident victims Respecting place of residence, 55 percent of all patients come from New York City. A higher proportion of New York City residents tend to be inpatients than residents from elsewhere in the state. Table 7 below exhibits the New York City boroughs in which accident victims live. Altogether, Brooklyn has the largest number of victims, followed by Manhattan in second place and Queens in third place. The Bronx occupies a distant fourth place with Staten Island trailing far behind in last place. Interestingly, this ranking of place of residence of individuals involved in pedestrian-cyclist accidents parallels exactly the ranking of the location of all pedestrian deaths in New York City. That is to say, most pedestrian deaths occur in Brooklyn followed by Manhattan, Queens, the Bronx, and Staten Island. Within this context, it should be noted that this ranking is not simply attributable to population size. For example, Queens has many more residents than Manhattan (2,230,722 vs. 1,585,873) but lags behind Manhattan in the number of pedestrian deaths. The ranking of the boroughs with respect to place of residence of victims, though, changes according to the status of the patient. Among in-patients, Manhattan has the greatest number of victims. In part, this is due to the fact that Manhattan has a disproportionately high number of victims who are 40 years of age or older who, we have seen, are more likely to be in-patients.

Table 7. Borough of accident victims: 2007-2010

Borough Bronx Brooklyn Manhattan

Type of patient out-patient in-patient 13.1% 34.7% 26.7% 9.9% 32.1% 39.5%

Total

12.8% 34.4% 28.1%

Queens Staten Island Total

18.6% 6.8% 100.0%

16.0% 2.5% 100.0%

18.4% 6.4% 100.0%

In Table 8 below are presented the top five zip codes in each of New York Citys boroughs ranked according to the number of patients involved in pedestrian-cyclist accidents spanning the time period from 2007 to 2010. Table 8. Top Five Zip Codes in Each of New York Citys Boroughs In Terms of Number of Patients Involved in PedestrianCyclist Accidents: 2007-2010 Bronx Zip No.
10456 10467 10453 10457 10458 29 26 23 21 19

Brooklyn Zip No.

11206 11211 11220 11221 11212 44 43 39 33 32

Manhattan Zip No.

10029 10002 10009 10128 10014 52 44 43 27 26

Queens Zip No.

11102 11373 11106 11368 11377 27 25 23 21 21

Staten Is. Zip No.

10301 10314 10303 10310 10304 24 20 18 16 15

Overall, the zip code with the highest number of patients is located in Manhattan number 10029. This zip code is situated in the neighborhood of East Harlem. Its boundaries extend North-South from E. 116th Street to E. 96th Street and East-West from FDR Drive to 5th Avenue. Two other zip codes in Manhattan with a large number of patients are situated in the Lower East Side numbers 10002 and 10009. Map 1 below thematically shades the frequency distribution of patients by Manhattan zip codes.

Map 1. Manattan Zip Codes of Patients Involved in Pedestrian-Cyclist Accidents: 2007-2010

In Brooklyn, the zip code with the highest frequency of patients is number 11206. This zip code is situated in the Bedford-Stuyvesant neighborhood. The second and third place zip codes in Brooklyn are numbers 11211 and 11220 which are situated in the Greenpoint and Sunset Park neighborhoods, respectively. The distribution of patients by zip codes in Brooklyn is displayed in Map 2.

Map 2. Brooklyn Zip Codes of Patients Involved in Pedestrian-Cyclist Accidents: 2007-2010

Summary and Conclusions

Several important findings have emerged from this study. First and foremost, the incidence of pedestrians involved in accidents with cyclists who needed to receive medical treatment at a hospital far exceeds previous estimates. Earlier research, based on a sample of hospitals nationwide, estimated that there were approximately 1,000 pedestrians hit by a cyclist each year in the United States who needed to obtain medical treatment at a hospital. This present study, based on every hospital in New York State, has found that in New York State alone, there were approximately 1000 pedestrians struck by cyclists each year necessitating medical treatment at a

hospital. Even this figure grossly underestimates the frequency with which pedestrians are injured in collisions with cyclists. This figure of 1,000, for example, excludes pedestrians involved in cycling accidents who received medical attention at a walk-in clinic or from their own personal physician. Moreover, it excludes countless number of pedestrians involved in cycling accidents who either did not sustain injuries at all or whose injuries were not serious enough to warrant seeking professional medical attention. A second finding emerging from this study is that the profile of pedestrians injured in cycling accidents is a distinctive one. Compared to the population for New York State as a whole, victims tend to be much younger and to be members of minority racial/ethnic groups. In particular, Hispanics are overrepresented among pedestrians involved in cycling accidents. There are also notable demographic differences between out-patients and inpatients who have been involved in pedestrian-cyclist accidents. In-patients, who have sustained more serious injuries, are more apt to be female and older when compared to out-patients. Finally, the geographic distribution of pedestrians involved in cycling accidents tends to be concentrated in New York City as compared to the rest of the state and also to be concentrated in specific locales within the city. Brooklyn has the highest number of patients and Manhattan has the second highest number of patients. Even within these boroughs, accident victims tend to cluster within smaller geographical units. These boroughs also lead in terms of the total number of pedestrian fatalities. In sum, this study supports the wide body of anecdotal data that pedestrians are at risk not only from motor vehicle drivers but also from cyclists. In any collision between a pedestrian and a cyclist, blame might be assigned to the pedestrian, to the cyclist, or to both parties. This study has not furnished any evidence which could be used to apportion responsibility. But it has underscored the fact that pedestrian-cyclist accidents are far more common occurrences than previously thought and that government and public health officials need to pay more attention to this phenomenon.

Endnotes 1. New York State, Department of Health, SPARCS Overview, p. 1, http://www.health.state.ny.us/statistics/sparcs/operations/overview. htm 2. These results indicate that estimates concerning the incidence of pedestrian-cyclist accidents generated by an earlier research study understate the magnitude of the problem. This earlier study estimated that there were approximately 1,000 pedestrian-cyclist accidents necessitating emergency-room visits per year nationwide. The previous study was based upon data culled from a random sample of approximately one hundred hospitals with emergency rooms from across the county. (See The Cyclist- Pedestrian Wars, J. David Goodman, New York Times, September 18, 2010, http://cityroom.blogs.nytimes.com/2010/09/18/spokes-thecyclist-pedestrian-wars.) The authors of this earlier study also helped to analyze the data for this current research.