From: Paul Trester
To: IT Service Desk; City Clerk
Subject: Public Comment: Oral Communications – Pedestrian Crosswalk Safety Date: Tuesday, August 25, 2026 2: 01: 45 AM
Attachments: Crosswalk Safety-- recommendation for making an education video for presentation at Encinitas schools-. docx Crosswalk Safety -- Recommendations to improve horse- pedestrian crosswalk safety at 2705 El Camino del Norte. docx
CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
City Council Clerk, Staff and Council members 8-25-26
I have two attachments below. Please prepare to project them at the City Council meeting to be held on 8-26-26 at 4 pm. I will attend, and I will fill out a slip to request permission to speak for 3 minutes on both topics regarding crosswalk safety in Encinitas.
Thank you,
Paul Trester
Olivenhain resident
Word document for presentaon. Sent to it-service- desk@encinitasca. gov and to cityclerk@encinitasca. gov for Public Comment: Oral Communicaons – Pedestrian Crosswalk Safety.
Copy below of my email to editor of Encinitas Advocate and author of the 8-7-26 arcle on Safe Streets Encinitas advocacy group.
Paul Trester <paultrester43@gmail. com> Aug 8,
2026,
7:26 PM
to KARENB, editor, bcc: me
Aenon: Karen B.
8-8-26
Aer reading your arcle about the rally and efforts of Safe Streets Encinitas advocacy group in the 8-7-26 edion of the Encinitas Advocate , I keep thinking of ways to advance accident prevenon for pedestrians who need to cross a street. At this me, I would suggest an effort by possibly a City Council formed group to prepare a 10-15 minute safety video for our school districts to show to all students about being seriously alert, focused, and defensive when preparing both to cross and during the crossing of a street. I have observed too oen a casualness or worse, not paying aenon defensively, in persons making a crossing.
A video might include techniques like:
no smartphone interacon;
no talking except to themselves;
pressing the buon at a crosswalk;
watching for and making eye contact with the car drivers who have stopped; making mulple head swings le, right, le, right to detect an oncoming vehicle, which might not be slowing;
waving an up stretched arm when beginning a brisk walk out onto the street; knowing to halt if a driver begins to move forward;
looking for e-bikes who may be zooming carelessly through in a bike lane; looking for a fast moving bicyclist.
In total, being on the defensive alert must be learned and pracced by all pedestrians.
A city council -prepared (and with appreciable reviews) educaonal video now would be mely, and most likely acceptable by our school districts, I believe.
Please consider and forward this email to persons who may have an interest.
Thank you,
Paul Trester
Olivenhain resident
To: Dan Nuer -- Director -Encinitas City Engineer
From: Paul Trester -resident of Olivenhain
8-14-26
Subject: Recommendaons to improve horse/ pedestrian crosswalk safety on 2705 El Camino del Norte
Recommendaons to improve horse/ pedestrian crosswalk safety on El Camino del Norte at site of recent pedestrian fatality:
Add two diamond warning signs (design type with ashing yellow lights on sign perimeter when crosswalk-warning buon is pressed by pedestrian or horseback rider) at the following locaons:
one sign on Westbound roadside near West end of bridge’s guard rail, which is approximately 300 from the crosswalk
one sign on Eastbound side, before the modest rise in road; a driver seated low within a car medium or smaller size cars) cannot see momentarily the crosswalk which is being approached.
Reasoning supporng Recommendaons:
When an eastbound car at moderate speed ascends the road rise there is only 8 seconds typically unl reaching the crosswalk.
When a westbound car is crossing the bridge, visibility of the exisng diamond warning sign is obscured by vegetaon outgrowth.
During summer months, the morning sunrise low-over- the- horizon beams westward onto/ down the road; thus, drivers eastbound are staring into the sun’s glare.
During summer months, the late aernoon the low-toward-the horizon sun beams eastward onto/down the road; thus, drivers westbound are staring into the sun’s glare. An addional ashing warning sign near the bridge would provide an alert for crosswalk acvity ahead. Cars driving westbound are oen moving at 50 mph or faster.
During summer months, a pedestrian or a horseback rider planning to use the crosswalk safely would also be subject to temporary sun glare when looking for oncoming traffic both in the early morning and in the late aernoon.
NOTE: My above request to the Encinitas Website’s Department of City Road Engineering was entered as REQ - 10640
From: Ron Medak
To: Peter Weichers
Subject: Special City Council Meeting 8/ 26, Item 1
Date: Tuesday, August 25, 2026 11: 28: 27 AM
Attachments: IMG_ 3793. mov
Physical Evidence Fin Crash. pdf
S Coast Hwy separated bikeway case report v0. 91- DRAFT. pdf Manuscript- Cicchina study on separated bike lanes. pdf CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
Hi Pete,
Please publish the following:
I want to begin by thanking the present City Council for prioritizing public safety, including traffic safety, and City Staff for carrying out our City Council' s priorities. All of the five E's that Mr. Nutter refers to are important and our City Council' s efforts to accomplish all of them need to be recognized. Organizations who claim that our City Council is being remiss or uncaring about this matter are doing a disservice to the people of Encinitas because their actions seem mainly to focus on undermining credibility in our present city government.
Be that as it may, I do think we can and must improve traffic safety in Encinitas. For many reasons, among them the forced buildout throughout SD County, there is definitely an increase in traffic through our town and with it as well an increase in irresponsible driving. It can be quantified with traffic and speed counts, but all of us who drive, cycle, or walk in Encinitas are aware of this.
There are a number of roads that lend themselves to this dangerous behavior, among them is El Camino del Norte. I know it well and have cycled it almost every week for the last 20 years, yet I confess I didn't think of it as a speed corridor. A few days after the crash that took Fin Fadick' s life my son and I went to the trail crossing and tape measured what appears to be the physical evidence as marked on the street by Sheriff deputies. Please see Attachment 1 to see that data. Admittedly, we did not have access to the vehicle in order to be able to verify that the skid marks came from that Nizzan Z, nor did we have the Electronic Data Recorder ( EDR) from the vehicle. The Sheriff fatal crash investigation team does, however, and they will hopefully provide the results of that data to the District Attorney if they decide a prosecution is warranted. As a concerned layperson, however, I also returned to the crash site within a few days of the crash at the exact same time and took a video of a pedestrian crossing in the crosswalk at the exact same time as the crash. That video is Attachment 2. At the August 4 Sheriff E-Bike Safety presentation one of the sheriffs made the important point that legal speeds are not necessarily the maximum speeds allowed by signage ( on ECDN that is 35 mph), but rather legal speed is what is safe for the present conditions.
Anybody who watches the attached video will readily see that the safest allowable speed at the time of the crash was well below 35 mph, and if those very fresh, black skid marks were from the Nissan Z that vehicle was traveling far in excess of a safe speed.
Although the City Council and Staff are doing a yeoman's job of addressing traffic safety in Encinitas, we can and must use more proven effective methods to improve safety here. I call this Vision Better. Unfortunately, because of that subset of drivers I referred to above we can expect to see continued irresponsible and dangerous driver behavior that must be addressed as best we can. Everybody agrees that speed is an important factor as a cause of crashes and severity of injuries. To that end I recommend that Council direct Staff to determine which roadways are most at risk of speeding in Encinitas and then design and construct traffic slowing devices to physically slow motor vehicles to safe speeds for the safety of our citizens. That will involve considering devices such as speed cushions, raised crosswalks, raised intersections, speed tables, rumble strips, and roundabouts as well as devices to encourage safe speed driving. All of these traffic calming devices can only be used when they are designed in such a way that our emergency medical service vehicles can do the job they are tasked with. Therefore, I propose that a committee be formed and led by a member of City Council for this purpose. The committee should include at least one member of City Council as Chair, Chief Gordon, Captain Shane Watts, Dan Nutter, Traffic Engineer staff representative, City Manager Campbell.
I would like to call the City Council's attention to another matter of traffic infrastructure in Encinitas. There is dangerous infrastructure that needs to be removed as soon as possible. I am referring, of course, to the dangerous bollards and wheel blocks that have been used throughout the City in order to create Class IV cycle tracks. With all due respect to Mr. Nutter's presentation, the City has been sadly lacking in evaluation of the installation of these dangerous Class IV Cycle Tracks. When the Southbeach 101 cycle tracks were put in place there was no before and after traffic count of cycle usage and volume. Granted that was during the Blakespear- led City Council. However, as recent as last year Commissioner Duncan requested that Chief Traffic Engineer Bandegan provide his data on the number of cycle crashes on the Southbeach 101 bollard built cycle tracks. Chief Traffic Engineer Bandegan reported one month later that his data showed five (sic!) crashes during the time the bollard lanes had been in place.
Please refer to Attachment 3 to see that the actual number of off Class IV cycle tracks was 27 using official data and over 40 if one adds the well-documented crashes from Serge Issakov's " Please Bring Safe Cycling Back to Cardiff" webpage. These cycle tracks have been the direct cause of two cyclist fatalities in Encinitas in recent years: Capt.
Curry who was killed when striking bollards and wheel blocks on Southbeach Hwy 101, and Dr. Jennings Worley who died in a right hook crash enabled and caused by the Class IV cycle track installed on Leucadia Blvd at Moonstone Court. I am attaching Steve Linke's (along with Serge Issakov and myself as co-authors) Case report of the Southbeach 101 cycle track and an excellent 2020 study by Cicchina et al documenting a 100% increased risk of cycle crashes in bollard created uni-directional Class IV cycle tracks and a 20X (sic!) increase in cyclist crashes in bi-directional Class IV cycle tracks.
To that end, a subcommittee has been formed at the MTSC composed of John Eldon and myself for the purpose of evaluation of Class IV cycle tracks in Encinitas and a plan for the removal of all Class IV cycle tracks deemed potentially dangerous and/or unnecessary. Any informed person in the city should support this effort.
Thank you,
Ron Medak
Olivenhain
Dry pavement
Approximately 3% uphill grade
71 feet of fresh, intermittent, dark ABStype braking marks ending at the first point of pedestrian contact
26-foot pedestrian throw
Vehicle came to rest 66 feet from the point
of impact (straight-line measurement)
No visible tire marks after impact
7/ 17/ 2025 DRAFT v0.91
1 Case report showing elevated bicycle crashes and injuries in a separated bikeway in Encinitas, California
Steven P. Linke, PhD (splinke.traffic@gmail. com), Ron Medak, MD, and Serge Issakov
ABSTRACT
There is substan al inconclusive and contradictory informa on published on the alleged safety of separated bikeways. This report is an inves ga on of crashes on a South Coast Highway 101 corridor in the southern part of Encinitas, California. A “ before-a er ” analysis of police report and emergency medical services data showed that crashes increased by 59% a er conversion of a por on of the roadway to a separated bikeway configura on, seemingly due mainly to the ver cal components and reduced space to maneuver. In contrast, no meaningful change in crash numbers occurred in an adjacent control corridor during the same me periods. Although no bicycle counts were collected by the city to assess crash rates, data from adjacent areas indicated that the increased crashes were largely not atributable to increased volumes. Addi onal crash data collected using crowdsourcing during the a er” me period further highlighted separated bikeway hazards. The combined police report, emergency medical services, and crowdsourced crash data included a fatality and numerous severe injuries atributable to the separated bikeway infrastructure. In addi on, the crash data showed that reliance on police report data alone is likely to miss a majority of crashes, par cularly in separated bikeways. Although this report addresses only one separated bikeway and control corridor, when considered in combina on with other published observa ons, it raises serious concerns about the actual safety of separated bikeways.
INTRODUCTION
Over the past couple of decades, there has been an accelera ng movement in transporta on planning to separate bicyclists from motorists by crea ng bikeways with ver cal components, such as flexible post delineators, wheel stops, curbs/ medians, parked cars, etc. In California, these are legally called “ Class IV separated bikeways,” but they also are known as “ separated bicycle lanes,” “ cycle tracks,” or “protected bicycle lanes.” The ra onale is that they may increase ridership due to an increased percep on of safety, and that they also may increase actual safety.1
Based largely on a 2023 self-published report,2,3the Federal Highway Administra on (FHWA) is now claiming that such separated bikeways are established as “ proven safety countermeasures” that “reduce crashes by up to 53%.”4,5 In addi on, the Na onal Associa on of City Transporta on Officials ( NACTO) is now claiming that separated bikeways are “ appropriate for all streets in all contexts.” 6 However, there is substan al countervailing evidence that crashes are likely to increase with vehicles at intersec ons on 7/ 17/ 2025 DRAFT v0.91
2 separated bikeway corridors and with fixed objects, debris, other bicyclists, pedestrians, etc. throughout such corridors, which could lead to net increases in crashes and injury severity.7 The 2023 FHWA report and other studies repor ng separated bikeway safety tend to exhibit remarkable limita ons, including weak study designs, small sample sizes/ lack of sta s cal significance, unreliable bicycle counts, insufficient accoun ng for confounding variables, exclusion ofa majority of crashes ( e.g., those that occur at and around intersec ons and those not involving moving motor vehicles), etc. The findings have been largelyinconclusive or contradictory between studies. 7 Over the past few years in the north coastal region of San Diego County, California, there have been at least three widely reported fatal crashes in separated bikeways, and evidence indicates that the separated bikeway components themselves ( plas c post delineators and wheel stops) were primary causes.
1. On November 28, 2020, a bicyclist was killed in the westbound separated bikeway on Leucadia Boulevard at Moonstone Court in Encinitas, when he struck the side of a right-turning truck right hook crash). 8 The City of Encinitas eventually removed some of the plas c posts in the area and setled a lawsuit that alleged9:
These] dangerous condi ons created a trap that a reasonably careful person would not no ce or an cipate as being dangerous, since the lack of dashed lines, and presence of bollards/ stanchions created a false sense of safety while opera ng a bicycle in the bikeway while at the same me, depriving automobiles adequate opportunity to merge into the bikeway to turn right onto Moonstone Court.
2. On March 17, 2024, a bicyclist was killed in the northbound separated bikeway near the 2800 block of South Coast Highway 101 in Encinitas. 10,11 Officialpolice records indicate that there was no vehicle involvement, but that the bicyclist had collided with a fixed object ( California Crash Repor ng System12 Collision ID 4622805). Images of the crash site show that it was adjacent to plas c post delineators and wheel stops.
3. On October 12, 2024, a bicyclist was killed in the northbound separated bikeway near the 1900 block of Jimmy Durante Parkway in Del Mar.13,14 The City of Del Mar eventually removed some of the plas c posts in the area, presumably because they were suspected of causing the crash.
Just since 2024, there have been several other reports of bicyclist fatali es and severe injuries in separated bikeways across North America, for example, two fatali es in Cambridge, Massachusets, 15 a life-altering trauma c brain injury in Seatle, Washington, 16 and a hospitaliza on in Montreal, Canada. 17 In each of these cases, the ver cal components ( plas c posts, curbs, and/ or parked vehicles) appear to have contributed to the vehicle drivers (typically in larger SUVs or trucks) being unaware of the bicyclists traveling in the same direc on (blind spot), and to the bicyclists being unaware that the right turns were happening un l it was too late—virtually iden cal to the “right hook crash” on Leucadia Boulevard in 7/ 17/ 2025 DRAFT v0.91
3 Encinitas. The Seatle crash also led to a lawsuit against the city for the dangerous separated bikeway design, 16 and the legal status is unknown in the other ci es.
In the South Coast Highway 101 bikeway, along with the fatality described above, bicyclists have been repor ng many other crashes, which were either directly or indirectly caused by the ver cal components, and several have resulted in serious injuries. The intent of this case report is to assess the crash history of this bikeway.
RESULTS
Street corridors
The “ separated bikeway corridor,” which is the subject of this report is the approximately 1.3-mile por on of South Coast Highway 101 stretching from the southern city limit of Encinitas ( at its border with Solana Beach) up to and including the Chesterfield Drive intersec on ( Figure 1, orange line). The ver cal components include plas c delineator posts and wheel stops. In addi on, an approximately 1.4- mile “control corridor” includes the por on of South Coast Highway immediately north of the separated bikeway corridor, stretching from north of Chesterfield Drive up to and including the K Street intersec on Figure 1, yellow line).
Figure 1. Study corridors: Separated Bikeway Corridor (orange) and Control Corridor (yellow).
Official crash data
Official crash data were firstcollected from the online California Highway Patrol’s “ California Crash Repor ng System” ( CCRS) database, which compiles local police reports. 12 However, it is well-known that this under-reports bicycle crashes, par cularly those that do not involve motor vehicles and/ or that occur off-roadway ( separated bikeways are legally off-roadway). 7 Therefore, addi onal official crash data 7/ 17/ 2025 DRAFT v0.91
4 were obtained from the Encinitas Fire & Marine Safety Department’ s emergency medical service (EMS) dispatch logs.
All crash data were filtered for loca on on the corridors and to include only bicycle- involved incidents.
Some crashes were only present in the CCRS database, some were only present in the EMS database, and some appeared in both. Table 1 demonstrates this in a subset of the analyzed crashes that occurred in 2021 through 2022.
Table 1. Official crash data sample ( 2021 through 2022).
Longitudinal ( before-a er) approach
The wheel stops crea ng the separated bikeway were installed by April 2020, and CCRS data was available through December 2024. Therefore, equal 57-month analysis periods before (July 2015 through March 2020) and a er (April 2020 through December 2024) conversion to the separated bikeway were compared. This “ longitudinal” approach ( crashes before and a er on the same roadway) minimizes contamina on with poten ally confounding variables rela ve to a “cross-sec onal” approach ( crashes on different roadways compared during the same me period). Confounding variables in this context are differences in roadways other than the bikeway separa on that could contribute to crashes.
Large increase in crashes a er separated bikeway conversion Table 2 shows the “before” and “ a er” bicycle-involved crash numbers for both the separated bikeway and control corridors. When only the police report-based crashes ( CCRS database) are considered for the separated bikeway corridor, there were only 7 crashes before conversion and 10 crashes a er (litle meaningful difference). However, the EMS dispatch logs add an addi onal 27 crashes that did not generate police reports. In total, there were 17 crashes in the separated bikeway corridor reported in these official recordsbefore conversion, and 27 a er conversion, represen ng a 59% increase. 7/ 17/ 2025 DRAFT v0.91
5 In contrast, the control segment had 14 total crashes before and 13 a er (no meaningful difference).
Unfortunately, the City of Encinitas did not collect any bicycle counts in these corridors either before or a er conversion of the bikeway. Therefore, it is impossible to compare crash “ rates” that could have accounted for possible changes in volume. However, the inclusion of the control corridor data helps control for bicycle volumes and other factors that may have changed between the “before” and “a er” me periods.
Table 2. Crash numbers before and a er separated bikeway conversion. The numbers are split into crashes that appeared only in the CCRS database, only in the EMS dispatch logs, or both.
EMS data demonstrates significant under- repor ng of bicycle crashes in police reports It is remarkable that 63% (45 of 72) of the total crashes in both corridors were recorded only in the EMS dispatch logs—no police reports generated for them to appear in the CCRS database, despite the fact that the injuries were severe enough to call for emergency assistance.
Bicycle volume data
Although the City of Encinitas has collected no bicycle counts in the separated bikeway corridor itself, the San Diego County Associa on of Governments ( SANDAG) has a con nuous automated bicycle counter located in Solana Beach at the intersec on of South Coast Highway 101 and Lomas Santa Fe Drive (about
0.7 miles south of the separated bikeway).
Apart from a modest temporary increase in 2020 during the pandemic, bicycle counts remained rela vely stable on South Coast Highway 101 over the course of the analysis period (Figure 2).
Cumula ve bicycle traffic was about 7% higher in the a er period rela ve to the before period at the SANDAG counter, which would not account for the 59% increase in crashes in the separated bikeway corridor.
7/ 17/ 2025 DRAFT v0.91
6 Figure 2. Average daily bicycle counts: Counts obtained from the SANDAG “ Coast Highway & Coastal Rail Trail” con nuous bicycle counter.
Exploratory analysis of crowdsourced crash data (101 Crash List) Since conversion of the South Coast Highway 101 corridor to a separated bikeway, Serge Issakov has maintained a “crowdsourced” crash list (101 Crash List) compiled from reports from vic ms, family members or friends of the vic ms, or eyewitnesses. 18 From April 2020 through May 2025, there were 37 crashes added to the list, 10 of which also appeared in the CCRS database and/ or EMS dispatch logs.
Table 3 includes a sampling of the crashes from 2021 through 2022 side-by-side with the CCRS and EMS crashes, giving credence to the crowdsourcing approach. 7/ 17/ 2025 DRAFT v0.91
7 Table 3. Crowdsourced “ 101 Crash List” sample (2021 through 2022). 101 Crash List crashes directly atributable to the bikeway Through review of the 101 Crash List reports, it was determined that a minimum of 19 of the crashes on the list could be directly atributed to the existence of the ver cal components of the bikeway (e.g., solo crashes from directly hi ng a plas c post or wheel stop, or as a result of the inability to conduct an avoidance maneuver). Among these 19 crashes, 16 did not appear in either the CCRS or EMS crash lists, and 10 of those resulted in visible or worse injuries, further demonstra ng that even the combined CCRS and EMS crash data does not fully capture all bicyclist injuries, let alone all bicyclist crashes.
Recent individual case reports ( supplemental data)
It is also important to note that the crashes and injuries are happening to both experienced/ vehicular bicyclists, as well as more casual/ recrea onal bicyclists. Below are 8 example individual case reports.
On May 3rd, 2020, two different children under the age of 16 hit wheel stops, and one of them was thrown into the #2 vehicle lane.
On May 26th, 2020, experienced randonneur “ Len” hit a wheel stop trying to avoid surfers and suffered fractures.
On March 13th, 2025, bicycle commuter “ Annie” was forced to ride in the #2 vehicle lane, because there were pedestrians in the cycle track. She was crowded by a close-passing car and atempted to re-enter the cycle track, struck a wheel stop, and damaged her bike.
On May 11th, 2025, casual e-bicyclist “ Priscilla” lost control in sand, swerved le, and fell, hi ng her head on a wheel stop. She suffered a concussion, mul ple scrapes, and EMS was dispatched and took her to the hospital.
7/ 17/ 2025 DRAFT v0.91
8 On August 14th, 2023, experienced bicyclist “ Fred” collided with a wheel stop, fell, and hit his head. 7/ 17/ 2025 DRAFT v0.91
9 On October 5th, 2024, experienced bicyclist “ Ken,” ac ng as sweeper for a local recrea onal cycle club ride, hit a plas c post, crashed, and broke his clavicle and five ribs. He also had a concussion and brain bleed and was taken to the hospital unconscious.
7/ 17/ 2025 DRAFT v0.91
10 On April 29th, 2025, experienced bicyclist “ Herb” hit a plas c post and wheel stop, crashed, and suffered a concussion. EMS was dispatched and took him to the hospital. 7/ 17/ 2025 DRAFT v0.91
11 On March 17, 2024, USMC Captain “ Ryan” was killed at the loca on shown below near the northbound entrance at the southern city limit.10 The police report determined the cause to be a collision with a fixed object, presumably a wheel stop, a plas c post, or the curb.12
DISCUSSION
The primary finding in this reportwas the 59% increase in crashes in the South Coast Highway 101 separated bikeway corridor in the 57-month period a er its conversion compared to the 57-month period before. This longitudinal ( before-a er) approach is superior to a cross-sec onal approach to minimize poten al confounders. Another strength of this report is the inclusion of before-a er data from the con guous, very similar, control corridor.
The main limita on is the fact that this is a case report of a single separated bikeway corridor, so it is not possible to make any broad sta s cal conclusions, although that was not a goal. Another limita on is the fact that the City of Encinitas did not collect any bicycle count (exposure) data to account for possible changes in ridership. However, the stable bicycle counts observed directly south of the studied bikeway, as well as the stable crash numbers observed in the control corridor directly north, are strongly 7/ 17/ 2025 DRAFT v0.91
12 indica ve that bicycle counts did not change considerably in the separated bikeway corridor in the before vs. a er me periods—at least not enough to account for the 59% increase in crashes. In addi on, there is substan al evidence that many of the crashes were directly or indirectly caused by the ver cal components themselves, which could not have occurred but for the conversion to a separated bikeway.
Although the crowdsourced data was not available in the before me period, and because it could not be collected in a uniform manner for all crashes, it is considered exploratory. However, the data was thoroughly documented and corroborated, and it provides valuable, complementary informa on, as described below.
A poten al minor limita on of this report is the fact that the construc on of the separated bikeway occurred over a few months, so a few crashes with ver cal components occurred before all ver cal components had been installed. Also, a two-way cycle track was added to part of the control corridor in 2023, but the main roadway with its standard bicycle lanes, remained in opera on, and there is not enough data yet to assess the safety of the cycle track.
This report highlights the fact that police report crash data, while cri cally important in any crash report/ study, is far from comprehensive. When considered in the context of EMS dispatch logs, the police report data missed a majority of the crashes ( 63% of officially reported crashes appeared only in the EMS data). Even the combina on of these two official data sources was far from complete, though.
Crashes compiled through crowdsourcing overlapped to an extent with the official sources but also included many unique crashes, typically solo incidents.
It is not surprising that solo bicycle crashes would not appear in police reports, because police reports are heavily focused on motor vehicle-involved crashes. However, several of the solo events led to injuries and did not appear in the EMS logs either. While minor (visible) injuries would not necessarily lead to an EMS call, some were more severe, but the vic ms used alterna ve means to get medical aten on, so these injury crashes did not appear in either official source. Researchers in other studies have analyzed records for emergency department visits and hospitaliza ons, which could help capture these addi onal crashes without relying on crowdsourcing. That was beyond the scope of this report, but studies likely need to include all of these sources to be comprehensive.
In conclusion, despite substan al marke ng on the alleged safety of separated bikeways, there seem to be many contexts in which their installa on makes condi ons more hazardous for bicyclists, as demonstrated in this report and discussed in detail elsewhere. 7 Further, this report confirms the need to assess crashes more comprehensively to reach safety conclusions. Police report data alone and data that fails to account for solo bicyclist, off-road, or intersec on crashes is not sufficient to make defini ve conclusions on safety. Future separated bikeway studies should focus on before-a er results with accurate and complete bicycle exposure and crash data, and local decision- makers should proceed with extreme cau on when considering installa ons.
7/ 17/ 2025 DRAFT v0.91
13 DETAILED METHODS
Street corridor characteris cs
Separated bikeway corridor. Generally, there are two vehicle drive-lanes going in each direc on (outside lane 11 feet wide, inside lane 10 feet wide) with single one-way bikeways on each side (7 feet of rideable width with 3-foot buffers from vehicles), along with a 5-foot sidewalk on the far west side. In areas with parking, there is also an 8-foot parking lane on the west side with a 2-foot buffer from the drive lane.
Black wheel stops with white strips and green plas c delineator posts were installed in the 3-foot bicycle buffer between about February and May 2020. The wheel stops are generally about 10 feet long with 10- foot gaps between, and the plas c delineators are generally installed in every other wheel stop gap.
Prior to conversion, generally, there were 5-foot unbuffered bicycle lanes.
Control corridor. Generally, there are one or two vehicle drive lanes going in each direc on with 5-foot unbuffered bicycle lanes. In 2023, a two-way cycle track was installed on the west side, but the main configura on remained. Insufficient me has elapsed to assessits safety impacts.
Intersections. All intersec ons within the corridors were included, and areas within 250 feet of the bounding Chesterfield Drive and K Street intersec ons were included, consistent with industry standards.
Bicycle counts
Bicycle counts were obtained from the SANDAG Open Data Portal, 19 based on an automated inpavement Eco Counter Zelt Evo at the “Coast Highway & Coastal Rail Trail” loca on (near the intersec on of Coast Highway and Lomas Santa Fe Drive), about 0.7 miles south of the Cardiff cycle track.
Crash data
California Crash Reporting System ( CCRS) database. The most recent ten years of crash records in the State of California are publicly available from the California Highway Patrol’s California Crash Repor ng System ( CCRS) through the California Open Data Portal. 12 This database was formerly known as the Statewide Integrated Traffic Records System ( SWITRS). The years 2015 through 2024 were available for download at the me this study was conducted ( March 2025), and there are three comma- separated value (CSV) files for each year that can be linked with a “ CollisionId” field common to each file: Crashes crash details), Par es (par es involved in the crashes), and InjuredWitnessPassengers ( numbers and severi es of injuries). Note that data from 2023 and 2024 are considered “ provisional.” The Crash files were first filtered for “City Name” = “ Encinitas” and combined into one master database for all years. Second, the database was filtered for all collisions that occurred on Coast Highway 101 as the “PrimaryRoad” ( this was a manual process, because there were several spelling varia ons. Third, the specific loca ons of the collisions along Coast Highway 101 were determined by using the block numbers, or by using distance (“ SecondaryDistance”) and direc on (“SecondaryDistance”) from the listed cross-streets (“ SecondaryRoad”).
7/ 17/ 2025 DRAFT v0.91
14 Emergency Medical Services ( EMS) dispatch logs. EMS dispatch logs, filtered for loca on and bicycle involvement, were obtained from the Encinitas Fire & Marine Safety Department for both the main study area and the control area for 2015 through 2024. Injuries for dispatches that resulted in transport to a hospital were assumed to be severe, and all others were assumed to be non-severe. Note that one crash was excluded, because a police report defini vely established that it was a solo motor vehicle incident. 101 Crash List
The 101 Crash List is crowdsourced in chronological order based on informa on about crashes provided to Mr. Issakov by the vic ms themselves, family members or friends of the vic ms, or eyewitnesses. The informa on was transmited in person; via email, text or phone; or on social media (e.g,, Facebook).
Social media posts that could not be independently corroborated were excluded. Injury severity was assessed by Dr. Medak based on the accounts as complaint of pain/ probable, visible/ minor, or severe/ serious.
ABOUT THE AUTHORS
Dr. Steven Linke, PhD (corresponding author is a re red scien st and former traffic commissioner for the City of Carlsbad, California.
Dr. Ron Medak, MD, is a physician and traffic commissioner for the City of Encinitas, California.
Serge Issakov is a lifelong cyclist who has been ac vely advoca ng for more and safer cycling for more than 20 years. He is a board member of the San Diego Bicycle Club, the San Diego County Bicycle Coali on and the California Associa on of Bicycling Organiza ons. He is also a League ( of American Bicyclists) Cer fied Instructor, a ride leader, and a former member of the San Diego Mobility Board.
REFERENCES
1) Winters, M.; Teschke, K. Route Preferences among Adults in the near Market for Bicycling: Findings of the Cycling in Ci es Study. Am. J. Health Promot. 2010, 25 (1), 40–47. htps:// doi.org/ 10.4278/ ajhp.081006-QUAN- 236.
2) Dixon, K.; Avelar-Moran, R.; Seyedeh, M. M. Developing Crash Modification Factors for Separated Bicycle Lanes; FHWA- HRT-23-078; Federal Highway Administra on, 2023. htps:// highways. dot.gov/ media/ 33856 ( accessed 2025- 06-23). 3) Federal Highway Administra on. Office of Research, Development, and Technology. Developing Crash Modification Factors for Separated Bicycle Lanes TechBrief; 2023. htps:// highways. dot.gov/ sites/ wa.dot.gov/ files/ FHWA- HRT-23-025.pdf (accessed 2025- 06-23). 4) Federal Highway Administra on. Proven Safety Countermeasures: Bicycle Lanes; FHWA- SA-21-051; Federal Highway Administra on, Office of Safety Research and Development, McLean, Virginia, 2023. htps:// highways. dot.gov/ media/ 19726.
7/ 17/ 2025 DRAFT v0.91
15 5) Federal Highway Administra on. Facebook: Separated Bike Lanes. htps:// www.facebook. com/ FederalHighwayAdmin/ posts/ conver ng-a-tradi onal-bike -lane-to-aseparated- lane-with-low-cost-flexible-de/ 842694937893039/ ( accessed 2025-07-09). 6) Na onal Associa on of City Transporta on Officials. Urban Bikeway Design Guide: Protected Bike Lanes. htps:// nacto.org/ latest/ urban-bikeway- design-guide-protected- bike -lanes/ ( accessed 2025- 06-23).
7) Linke, S. P. Unproven Separated Bicycle Lane Safety: Overreaching Conclusions and Marketing; Carlsbad, CA, 2025.
8) The Coast News Group (staff). Cyclist’ s death inspires bike safety advocates in Encinitas. htps:// thecoastnews. com/ cyclists- death-inspires- bike-safety- advocates- in-encinitas/ ( accessed 2025-07-09).
9) The Coast News Group (Kaila Mellos). Settlement talks begin in lawsuit over cyclist’ s death in Encinitas. htps:// thecoastnews. com/ setlement-talks-begin-in-lawsuit-over-cyclists- death-inencinitas/ ( accessed 2025-07-04). 10) The Coast News Group (Jordan Ingram). Sheriff IDs cyclist found dead in apparent solo crash in Encinitas. htps:// thecoastnews. com/ sheriff-ids-cyclist-found-dead-in-apparent- solo-crash-inencinitas/ ( accessed 2025-07-09). 11) NBC 7 San Diego (Omari Fleming). Cycling advocates want changes to protective bike path in North County. htps:// www.nbcsandiego. com/ news/ local/ protec ve-bike -path-north-county/ 3467315/ accessed 2025-07-09).
12) California Highway Patrol. California Crash Reporting System ( CCRS) - Dataset - California Open Data. htps:// data. ca.gov/ dataset/ ccrs (accessed 2025- 07-09). 13) San Diego Union Tribune ( Rob Nikolewski). Man dies in solo bicycle crash in Del Mar. htps:// www.sandiegouniontribune. com/ 2024/ 10/ 12/ man- dies-in-solo-bicycle- crash- in-del-mar/ accessed 2025-07-09).
14) Times of San Diego (Jennifer Vigil). Bicyclist Dies in Hospital After Del Mar Crash. htps:// mesofsandiego. com/ crime/ 2024/ 10/ 12/ bicyclist-dies-in-hospital- a er-del-mar-crash/ accessed 2025-07-09).
15) Streetsblog Massachusets. Truck Drivers Have Caused Two Deaths In Two Weeks In Cambridge - Streetsblog Massachusetts. htps:// mass.streetsblog. org/ 2024/ 06/ 24/ truck-drivers-have- causedtwo-deaths-in-two-weeks-in-cambridge ( accessed 2025-07-04). 16) KOMO News, J. H. Cyclist sues Seattle over bike lane design after brain injury in Green Lake crash. htps:// komonews. com/ news/ local/ cyclist-green-lake-neighborhood- king-county-superior- courtaviv-litov-26-years-old-bike-lanes-life-altering-injuries-stritmater-firm-configura on- tesla (accessed 2025- 07-05).
17) CTV News. Cyclist recovering after being run over by semi truck in Old Montreal. htps:// www.ctvnews. ca/ montreal/ ar cle/ cyclist-recovering- a er-being-run-over-by-semi-truck-inold-montreal/ ( accessed 2025-07-12). 18) Issakov, S. Cardiff 101 Crash List - Google Docs. htps:// nyurl.com/ cardiff101crashlist ( accessed 2025-07-16).
19) SANDAG. SOC - Bike Counts (SANDAG) 2024 | SANDAG Open Data Portal. htps:// opendata. sandag.org/ Transporta on/ SOC-Bike -Counts-SANDAG- 2024/ fcefavc6/ data_preview (accessed 2025-07-10).
Contents lists available at ScienceDirect
Accident Analysis and Prevention
journal homepage: www.elsevier.com/locate/aap
Not all protected bike lanes are the same: Infrastructure and risk of cyclist collisions and falls leading to emergency department visits in three U.S. cities
Jessica B. Cicchino a,*, Melissa L. McCarthy b, Craig D. Newgard c, Stephen P. Wall d, Charles J. DiMaggio e, Paige E. Kulie f, Brittany N. Arnold c, David S. Zuby a a
Insurance Institute for Highway Safety, Arlington, VA, United States b George Washington University Milken Institute School of Public Health, Washington, DC, United States c Center for Policy and Research in Emergency Medicine, Department of Emergency Medicine, Oregon Health & Science University, Portland, OR, United States d Ronald O. Perelman Department of Emergency Medicine, Department of Population Health, New York University School of Medicine, New York, NY, United States e Department of Surgery, Division of Trauma and Critical Care, New York University School of Medicine, New York, NY, United States f Department of Emergency Medicine, George Washington University Medical Center, Washington, DC, United States
ARTICLE INFO
Keywords:
Bicycle
Separated bike lane
Cycle track
Bike path
Streetcar tracks
Bicycle facilities
Injury
ABSTRACT
Objective: Protected bike lanes separated from the roadway by physical barriers are relatively new in the United States. This study examined the risk of collisions or falls leading to emergency department visits associated with bicycle facilities ( e.g., protected bike lanes, conventional bike lanes demarcated by painted lines, sharrows) and other roadway characteristics in three U.S. cities.
Methods: We prospectively recruited 604 patients from emergency departments in Washington, DC; New York City; and Portland, Oregon during 2015– 2017 who fell or crashed while cycling. We used a case- crossover design and conditional logistic regression to compare each fall or crash site with a randomly selected control location along the route leading to the incident. We validated the presence of site characteristics described by participants using Google Street View and city GIS inventories of bicycle facilities and other roadway features.
Results: Compared with cycling on lanes of major roads without bicycle facilities, the risk of crashing or falling was lower on conventional bike lanes ( adjusted OR= 0.53; 95 % CI= 0.33, 0.86) and local roads with ( adjusted OR = 0.31; 95 % CI = 0.13, 0.75) or without bicycle facilities or traffic calming ( adjusted OR =0. 39; 95 % CI = 0.23, 0.65). Protected bike lanes with heavy separation ( tall, continuous barriers or grade and horizontal separation) were associated with lower risk ( adjusted OR = 0.10; 95 % CI= 0.01, 0.95), but those with lighter separation ( e.g., parked cars, posts, low curb) had similar risk to major roads when one way ( adjusted OR = 1.19; 95 % CI = 0.46, 3.10) and higher risk when they were two way ( adjusted OR = 11.38; 95 % CI = 1.40, 92.57); this risk increase was primarily driven by one lane in Washington. Risk increased in the presence of streetcar or train tracks relative to their absence ( adjusted OR= 26.65; 95 % CI= 3.23, 220. 17), on downhill relative to flat grades ( adjusted OR =1. 92; 95 % CI= 1. 38, 2. 66), and when temporary features like construction or parked cars blocked the cyclist’ s path relative to when they did not ( adjusted OR = 2.23; 95 % CI = 1.46, 3.39).
Conclusions: Certain bicycle facilities are safer for cyclists than riding on major roads. Protected bike lanes vary in how well they shield riders from crashes and falls. Heavier separation, less frequent intersections with roads and driveways, and less complexity appear to contribute to reduced risk in protected bike lanes. Future research should systematically examine the characteristics that reduce risk in protected lanes to guide design. Planners should minimize conflict points when choosing where toplace protected bike lanes and should implement countermeasures to increase visibility at these locations when they are unavoidable. https:// doi.org/ 10.1016/ j.aap.2020. 105490
Received 30 July 2019; Received in revised form 25 February 2020; Accepted 27 February 2020 Corresponding author.
E-mail address: jcicchino@iihs. org ( J.B. Cicchino).
Accident Analysis and Prevention 141 (2020) 105490
0001-4575/ © 2020 Elsevier Ltd. All rights reserved.
T
1. Introduction
Bicycling popularity in urban areas has increased in the United States during the 21st century. U.S. workers who reported commuting to work by bicycle increased more than 60 % from 2000 to 2008–12, and the proportion of adults who cycle to work nearly doubled during this period in the largest 50 U.S. cities ( McKenzie, 2014). With this increase in cycling exposure has come an increase in fatalities and injuries among adult bicyclists. The number of bicyclists age 20 and older fatally injured in U.S. crashes with motor vehicles increased by nearly 50 % during 2000– 2017 ( Insurance Institute for Highway Safety, 2018).
Age- adjusted emergency department visit rates for bicycling- related injuries in the United States have similarly risen in recent years Sanford et al., 2015).
Growing cycling popularity and rising cycling- related injuries and deaths have encouraged U.S. cities to install infrastructure for bicyclists along more of their roads. Conventional bike lanes demarcated by painted lines have long existed in the United States, and over the past decade U.S. cities have begun to incorporate protected bike lanes.
Protected bike lanes, also called cycle tracks or separated bike lanes, are bicycle facilities separated from motor vehicle traffic byaphysical barrier such as parked cars, curb, grade, landscaping, posts, or a combination of these or other features. Protected bike lane mileage in the United States increased from about 40 miles in 2008 to about 400 miles in 2018 ( People for Bikes, 2018).
Recent North American evaluations of the effects ofconventional bike lanes have had inconsistent results, with some finding them tobe associated with fewer bicyclist crashes or injuries overall ( Bhatia et al., 2016; Hamann and Peek- Asa, 2013; Park et al., 2015; Pulugurtha and Thakur, 2015; Teschke et al., 2012; Wall et al., 2016) or in specific circumstances ( Kondo et al., 2018), and others reporting no change in crashes or increases associated with them ( Chen et al., 2012; Raihan et al., 2019; Wei and Lovegrove, 2013). Different findings inpartreflect disparities in how evaluations were conducted. For example, studies of bike lane efficacy vary inhow and ifcycling exposure was accounted for, which is important given that constructing facilities for cyclists can increase ridership ( Buehler and Pucher, 2012; Dill and Carr, 2003).
Research on the effects ofprotected bike lanes onbicyclist crashes and injuries in North America is sparser than that for conventional bike lanes. Teschke et al. ( 2012) used a case- crossover design to compare infrastructure at locations where cyclists treated in Toronto and Vancouver, Canada emergency departments were injured with infrastructure at randomly selected locations along the routes cyclists took prior to their injuries, and found that injury risk in protected bike lanes was one tenth of that on major roads with parked cars. Bicyclist injury rates per kilometer traveled were 28 % lower on Montreal protected bike lanes compared with similar nearby streets without cycling infrastructure ( Lusk et al., 2011). A later Montreal study found that injury rates were lower in protected bike lane segments than on comparison streets but were not always lower at intersections, with effects varying among the lanes examined ( Nosal and Miranda- Moreno, 2012).
In the United States, an evaluation in New York City reported that bicyclist injury rates in crashes with motor vehicles were 23 % lower on roads with protected bike lanes compared with roads without cycling infrastructure using pedestrian activity as a proxy for bicyclist exposure, although the finding wasnotstatistically significant (Wall et al., 2016). However, severity was higher for injuries sustained in protected bike lanes than those sustained on roads without cycling facilities.
Simple before- after examinations of police- reported bicyclist- motor vehicle crash rates in protected bike lanes in New York City and Washington, DC, that accounted for exposure but did not use controls have produced mixed findings, withdecreases atNewYorkintersections after the installation of protected bike lanes and increases along the initial lanes constructed in Washington ( Goodno et al., 2013; Sundstrom et al., 2019). A cross- sectional study examining data from 12 U.S. cities found that the density of protected bike lanes at the city and block level, but not of conventional bike lanes, was associated with fewer police-reported fatalities and serious injuries to all road users Marshall and Ferenchak, 2019).
With their growing prevalence, more needs to be known about the safety of protected bike lanes in the United States. Existing U.S. evaluations have focused on crashes involving motor vehicles, but other incidents such as falls or collisions with pedestrians or other cyclists cause many cyclist injuries treated in emergency departments ( e.g., Beck et al., 2016; de Rome et al., 2014; Schepers et al., 2015; Stutts and Hunter, 1999; Teschke et al., 2012).
The current study examined the risks associated with infrastructure characteristics, including protected bike lanes, of bicyclist crashes or falls leading to emergency department visits in the U.S. cities of Washington, DC; New York City; and Portland, Oregon. These cities were chosen because of their combination of high cycling rates relative to other U.S. cities ( Portland, Washington), large amount of protected cycling infrastructure ( New York), varied bike facility design, and mix of other roadway characteristics ( e.g., streetcar tracks in Portland and Washington, extensive traffic calming onPortland’ slocalroads). There were approximately 5 miles of protected bike lanes in use in Portland, 10miles in Washington, and 100 miles in New York by the end of 2018 People for Bikes, 2018), and during 2017, 6.3 % of adult workers in Portland, 5.0 % in Washington, and 1.3 % in New York biked to work United States Census Bureau, 2018).
We used a case- crossover design similar to Teschke et al. ( 2012).
Infrastructure characteristics at the location where adult cyclists crashed or fell were compared with those at a randomly selected location along the route leading to their incidents. Because cyclists served as their own controls and comparisons made between case and control sites were within trip, the design accounts for exposure to roadway features while matching rider and general trip ( e.g., weather) characteristics between sites.
2. Methods
2.1. Patients
We enrolled 604 adults who sought treatment after falling or crashing while riding a bicycle at the emergency departments of George Washington University Hospital in Washington, Oregon Health and Sciences University in Portland, and Bellevue Hospital and the Ronald O. Perelman Center for Emergency Services of NYU Langone Medical Center in New York City. Bellevue and the Washington and Oregon hospitals are Level 1 trauma centers, and NYU Langone is a universitybased quaternary hospital juxtaposed to Bellevue Hospital. Trained research staff interviewed patients intheemergency department. The research teams enrolled patients during set hours ( 9 a.m. – 10 p.m. in Washington, 8 a.m. – 11 p.m. in Portland, 8 a.m. – midnight in New York); patients who visited the emergency department outside of coverage hours were not enrolled. Data collection began in different months in each city ( April 2015 in Washington, November 2015 in Portland, April 2016 in New York) and lasted through September 2017.
The final sample included 354patients from Washington, 131from Portland, and 119 from New York. The protocol was approved by each hospital’ s institutional review board.
Adult cycling patients were eligible if they crashed or fell while riding a bike; could remember the route leading to their incidents, understand consent, and communicate with emergency department staff (inEnglish inWashington andPortland andinEnglish orSpanish in New York); and if their incident occurred within a week of the interview; their trip was 0.10 mile or longer and was within the hospital’ s catchment area; and they were not trick riding, racing, or riding with more than one person on a bicycle during their trip. There were 982 adult cyclists who presented to the emergency departments in Washington and Portland during coverage hours in the study period. Of these, 676 ( 69 %) were eligible, 254 ( 26 %) were ineligible, and 52 ( 5 J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 2 left the emergency department before research assistants could screen them. The research assistants enrolled 485 patients in Washington and Portland, which was 72 % of the eligible screened patients in those cities. The most common reasons for being ineligible in Washington and Portland were being unable to remember their route 44 cyclists) and being injured outside of the hospital’ s catchment area 40 cyclists). Data on cycling patients not enrolled were not collected consistently in New York.
2.2. Interview and injury coding
Research staff used astructured questionnaire tointerview participants. The primary purpose of the interview was to record the route the participant took during the trip leading to their crash or fall and collect information that could not be obtained from site inspections. The research assistant mapped each participant’ s route electronically using the website www. mapmyride. com and selected a control site along the route by multiplying a random proportion between 0.01–0.99 by the length of the entire route and placing the control site at the resulting distance from the start of the trip. For instance, if the trip was 7.5 miles and the random proportion was 0.61, the control location was marked
4.58 miles ( 7.5× 0.61) from the trip’s starting point. The probability of selecting a control location with a certain type of infrastructure was proportional to the cyclist’ s distance- based exposure to the infrastructure during their trip.
We adapted additional interview questions from Teschke et al. 2012) to assess circumstances leading to the incident, trip purpose, personal characteristics, what type of route the cyclist was riding at the case and control sites, which lane of the roadway the cyclist was in if they were riding on the road, and temporary site characteristics blocking the cyclist’ s path such as construction or parked cars. The research assistants showed the participants a Google Street View image of the case and control sites as they answered questions about them.
Following emergency department or hospital discharge, one research assistant at each site reviewed the medical record of each subject and coded each injury sustained using the Abbreviated Injury Scale Gennarelli and Wodzin, 2008). The AIS score ranges from 0 to 6, with 0 representing no injury and 6 representing nonsurvivable injuries. A score of 2 indicates a “ moderate” injury and 3 or greater indicates a serious” injury.
2.3. Site feature identification
We characterized route types at the case and control locations into one of 10 categories. If a site was an intersection, characteristics were recorded for the route type the cyclist was riding on prior to reaching the intersection.
1 Major road: Arterial or collector roads as classified bythefunctional class system, where cyclists were not in a conventional or protected bike lane or lane with shared lane markings. Following Teschke et al. ( 2012), who found that injury risks were higher on major roads with parked cars than on other route types, major roads were the reference in analyses.
2 Bike lane on major road: Conventional bike lanes with painted separation from moving motor vehicles on arterial or collector roads.
This classification includes bikelaneswithbuffers (i.e., painted space between bike lane and road) if there was not also vertical physical separation. Few ( 7 %) bike lane sites had painted buffers.
Nearly two-thirds ( 62 %) were located next to a parking lane, 35 % were next to the edge of the road, and 3 % were between two lanes of moving traffic (e.g., between athrough lane andturnlane) Fig. 1).
3 Sharrows on major road: Shared lane markings on arterial or collector roads ( Fig. 1).
4 Local road: Local roads as classified bythefunctional class system, driveways, and parking lots, without traffic calming and where cyclists were not in lanes with bicycle facilities. Few ( 3 %) sites identified aslocalroads wereonprivate property.
5 Local road with bike lane, sharrows, or traffic calming: Local roads that or had traffic calming orwhere cyclists were riding inbike lanes lanes with sharrows. Traffic calming included nearby speed bumps and Portland’ s neighborhood greenways, which are local roads that give priority to bicyclists and pedestrians through speed bumps, traffic diverters, andsharrows.
6 Sidewalk: Paths next to roadways designed for pedestrian use. 7 Off-road/ trail: Off-road areas other than sidewalks with mixed use, such as multiuse trails or roadways shut down to motor vehicle traffic. Nearly all (99 %) off- road sites were paved.
Protected bike lanes were defined asbikelanes physically separated from motor vehicle travel lanes with vertical barriers. All were located on major roads. We considered three categories of protected bike lanes based on the type of separation used and direction of travel. “ Heavy separation” ( Fig. 2) included tall, continuous barriers ( e.g., bridge rails, tall concrete barriers or walls), or lanes at sidewalk- level that were also separated horizontally from the road. Protected lanes located on bridges all had heavy separation. “ Light separation” ( Fig. 3) was transient ( parked cars), noncontinuous ( posts, parking stops), short ( continuous low curb), and/ or did not provide horizontal separation ( raised lane immediately adjacent to the road). The categories of protected bike lanes included:
8 One- way protected bike lane, light separation
9 Two-way protected bike lane, light separation
10 Protected bike lane, heavy separation: All of these lanes in this study were two way.
We classified route types based on acombination ofpatient reports and site reviews. The research assistants asked the participants if the Fig. 1. Examples of a bike lane ( left) and sharrows ( right).
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 3 routes they were riding on at the case and control sites were roads, bike lanes, sidewalks, or off-road locations. Wecross- referenced participants’ reports with Google Street View and GIS inventories of cycling facilities maintained by the study cities to validate that the named route types were present at the sites. Off-roadsitesthatwerenotviewable on Google Street View were visually assessed using Google Earth satellite view. New York’ s GIS inventory of cycling facilities included facility installation date, and Washington and Portland’ s included installation year. When it was ambiguous if a facility was installed before or after a trip from the main data sources, we consulted installation dates of new facilities obtained from the city (Washington) or from publicly available information ( New York, Portland). We further broke down route types from the initial four categories using these tools and roadway functional class maps maintained by states and the District of Columbia. If bike lanes or sharrows were present, we determined from questionnaire responses if participants were riding in a lane with these markings, another lane, or other route type ( e.g., sidewalk) without special markings for bicyclists.
The named route type was not present at 10 % of locations. The participant misnamed the route type ( e.g., called a multiuse trail or road with sharrows a bike lane) in more than half of these. At most remaining locations, the participant named the route type they were approaching rather than the route type they came from at an intersection ( 13 sites), or said they were in a bike lane when none was present ( 26 sites). If the patient said they were in a bike lane when none was present, we assumed they were riding in the road. There were no locations where a participant said they were riding on a sidewalk at a site without one. However, there was one location where sharrows were present but the cyclist said they were traveling in an unmarked lane.
Other features identified from sitereview included grade andthe presence of streetcar ( tram) or train tracks. Grade was determined through measuring elevation in Google Earth at the case or control site and 0.05 miles before the site and calculating the rise over run.
Elevation could not be measured on bridges and overpasses, and grade for sites with these elements was unknown. Grades greater than 1 % were considered uphill, less than 1 % downhill, and between 1 % and 1 % flat. Intersections were defined aslocations where two ormore roads meet; junctions with alleys, driveways, or entrance/ exit ramps were not considered intersections.
2.4. Analyses
In the primary analysis, we used conditional logistic regression to examine the association between environmental characteristics and site type, with a binary indicator for site type ( 1= case, 0= control) as the dependent variable. Conditional logistic regression takes into account the paired study design ( each pair is one subject, at case and control sites) by stratifying by pair and maximizing a conditional likelihood function that avoids estimating stratum parameters. For further information see SAS Institute ( 2011). Independent variables included route type, grade, and the presence of streetcar or train tracks and temporary features that blocked the cyclist’ s path. Results are presented unadjusted by individual variable and adjusted with all covariates included. Unadjusted results were produced using conditional logistic regression models with a single predictor; the result is also known as a matched- pair odds ratio. Because crashes and falls leading to emergency department visits are rare events, odds ratios are good approximations of relative risks and so results from logistic regression models are interpreted as changes in risk.
or Because infrastructure may differentially affectrisk ofcrashes falls depending on the cause or injury severity, we conducted four sensitivity analyses of the primary adjusted model restricting the analysis to incidents of specific circumstances orinjury severities: 1) crashes and falls involving moving vehicles, 2) crashes and falls not involving moving vehicles, 3) incidents where cyclists sustained minor or no injuries ( AIS 0 or 1), and 4) incidents where cyclists sustained moderate or more severe injuries ( AIS 2+). Another sensitivity analysis excluded patients who reported riding on a route type at the case or control site that was not observed to be present upon site inspection.
The primary analysis was also conducted separately for each city.
An additional conditional logistic regression model was constructed that included intersection presence, the independent variables from the primary model ( route type, grade, streetcar or train tracks, temporary features), and interaction terms between intersection and the other independent variables. This allowed the crash or fall risk associated with various characteristics to be computed separately at intersections Fig. 3. Examples of protected bike lanes with light separation, from left to right: 15th Street NW, Washington DC, separated with posts and parked cars; Pennsylvania Ave NW, Washington DC, separated with parking stops; 7th Avenue, New York, separated with continuous low curb and grade, immediately adjacent to the road; 1st Street NE, Washington DC, separated with continuous low curb.
Fig. 2. Examples of protected bike lanes with heavy separation: Hudson River Greenway in Battery Park, New York ( left); Pulaski Bridge, New York ( center); Southwest Moody Ave, Portland ( right).
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 4 and away from intersections.
We classified circumstances leading tocrashes andfalls based on responses to an open- ended question asking patients to describe the circumstances of their accident and forced- choice questions asking what, if anything, they collided with or fell to avoid colliding with. We categorized the proportion of incident circumstances occurring on each route type as collisions with or falls to avoid moving motor vehicles ( cars, SUVs, pickups, motorcycles, trucks, buses), stopped or parked motor vehicles ( including doors), other cyclists, pedestrians, infrastructure e.g., curb, pole, fence), or surface features ( e.g., potholes, uneven pavement, streetcar tracks); falls due to other causes ( e.g., slippery surface, avoiding adverse surface conditions, clothing caught in chain), or other/ unknown causes. We computed relative proportions and associated 95 % confidence intervals toassess therateofeach circumstance on each route type relative to the rate for the reference category of major roads. Only case sites ( and not control sites) were included in these analyses. For relative proportions including a route type where a type of circumstance never occurred, exact 95 % confidence intervals were computed that could handle zero values using the Farrington- Manning relative risk score statistic ( Chan and Zhang, 1999).
3. Results
3.1. Cyclist, trip, and injury characteristics
Table 1 summarizes the cycling and trip characteristics of the study sample. Participants were mostly male, and about half were under age
40. More than 80 % were regular cyclists who reported biking on most days during the months of the year when they ride. Two thirds of trips were shorter than 3 miles, about half were commuting trips, and most occurred on weekdays, with clear conditions, and during daylight. Most cyclists ( 97.7 %) presenting at an emergency department were injured Table 2), but fewer than half sustained at least one moderate or severe AIS 2+) injury. Among the 254 participants with AIS 2+ injuries, almost 70 % sustained injuries to the extremities.
3.2. Risk of crashing or falling
Table 3 displays results of the unadjusted and adjusted conditional logistic regression models comparing characteristics at case locations to those at control locations. Relative to major roads, risks of crashing or falling were significantly lower onlocalroads withandwithout bike infrastructure or traffic calming, bike lanes, andprotected bike lanes with heavy separation in both models, and on off-road locations intheunadjusted model only. Risks were significantly higherinbothmodels on two-way protected bike lanes with light separation relative to major roads, on downhill grades relative to flatgrades, whentemporary features were blocking the path relative to when they were not present, and when streetcar or train tracks were present relative to when they were absent.
3.2.1. Sensitivity analysis
Risk was examined separately for crashes with or falls to avoid moving vehicles ( Table A1, Appendix A) and for other crashes and falls Table A2). Reductions in risk for bike lanes and off-roadlocations relative to major roads were strongest in the vehicle model, while temporary features that blocked the cyclist’ s path only increased risk of Table 1
Characteristics of cyclists and their trips ( unknown values excluded, total sample N= 604).
Characteristic Number of cyclists with
nonmissing values
Percent
Male 604 72.3
Age 603
18–29 33.3
30–39 29.0
40–49 16.3
50–59 14.3
60–69 5.3
70+ 1.8
Regular cyclist 601 82.0
Completed college degree or higher 599 68.9
Income >= $ 50,000 495 66.7
Race/ ethnicity
White, non- Hispanic 597 66.2
Hispanic 13.9
Black, non- Hispanic 11.4
Asian/ Pacific Islander 4.2
Other 4.4
Trip purpose 603
To/ from work/ school 54.4
Exercise or recreation 19.4
Personal business ( e.g., errands) 10.1
Social reasons ( e.g., movies, visit friends) 10.0
During work 5.3
Other 0.8
Weekday 604 82.1
Daylight 604 84.4
Clear weather 597 88.1
Trip distance 604
1 mile 33.8
1 to < 3 miles 32.8
3 to < 5 miles 11.1
5 to < 10 miles 13.3
10+ miles 9.1
Helmet used 603 62.5
Shared or rental bike 602 7.3
Table 2
Injury severity among all cyclists and injured body regions of cyclists with moderate or severe ( AIS 2+) injuries.
Injury characteristic Percent
Maximum injury severity ( AIS) N = 604
AIS 0 2.3
AIS 1 54. 8
AIS 2 35. 1
AIS 3+ 7.0
Maximum severity unknown 0.8
Injured body regions with AIS 2+ injuries, among cyclists with at least one AIS 2+ injury N = 254 Head 14. 6
Face 8.3
Neck 0.4
Thorax 11. 4
Abdomen 3.2
Spine 7.9
Extremities 72. 4
Note: Some of the 254 cyclists sustained multiple AIS 2+ injuries.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 5 crashes or falls not involving moving vehicles. Risk was significantly higher on two-way protected bike lanes with light separation compared with major roads when examining crashes or falls not involving vehicles, but effects onthose involving vehicles could notbereliably estimated for these lanes.
Results when analyses were limited to cyclists that sustained no or AIS 1 (Table A3) and AIS 2 or more severe injuries ( Table A4) yielded similar results for most infrastructure types; protected bike lanes with light separation were associated with nonsignificant increases inrisk relative to major roads when the analysis was limited to moderate or more serious injuries. When the 46 cyclists who reported being in a route type that was not present at the case or control site where the lapse could not be explained by misnaming ( e.g., calling sharrows or a multiuse trail a bike lane) were excluded, the results were very similar to the primary analysis ( Table A5).
3.2.2. Intersections
Most incidents occurred away from intersections, but risks were higher at intersections ( unadjusted OR = 5.17; 95 % CI= 3.60, 7.43). Table 4 describes the risks of crashing or falling associated with various route types and other characteristics at intersections and away from intersections. Risks by roadway segment type were similar to those observed in the primary analysis. At intersections, however, risk was higher on bike lanes relative to major roads ( p = 0.0535). Interactions between intersection presence and route type indicated that cyclists were significantly morelikelytocrashor fall at intersections on bike lanes ( p = 0.0018) and on local roads with bike lanes, sharrows, or traffic calming (p = 0.0098) than at nonintersections on these facilities relative to major roads. Similarly, an interaction between intersection presence and grade indicated that cyclists were more likely to crash or fall at nonintersections than intersections when grade was unknown relative to when it was flat (p = 0.0402); this effect likely reflects the types of sites where grade could not be measured ( bridges, overpasses).
No other interactions were significant. Risk washigher fortwo- way protected bike lanes with light separation relative to major roads at both intersections ( p = 0.0731) and nonintersections ( p= 0.0921).
3.3. Incident circumstances
Table 5 summarizes the circumstances of crash or fall incidents by route type. Overall, less than half ( 40.2 %) of cyclists collided with or fell to avoid moving motor vehicles.
Circumstances varied by the type of route where the incident occurred.
Table 6 presents the relative proportions of incident circumstances by route type compared with the proportion that occurred on major roads. A smaller proportion of cyclists crashed with or fell to avoid moving motor vehicles at off-road locations than onmajor roads, buttheproportion involved in motor vehicle crashes on other route types didn’ t differ significantly from major roads. Collisions with or falls to avoid stopped or parked vehicles were less likely on off-road locations, sidewalks, orlocalroads without bike infrastructure or traffic calming. Theproportion ofcyclists whocollided with or fell to avoid other cyclists was higher at off-road locations andon both types of two-way protected bike lanes than on major roads, and the proportion who collided with or fell to avoid pedestrians was higher at offroad locations and on protected bike lanes with light separation. Relative to major roads, collisions with or falls to avoid infrastructure and other falls were more likely at off-road locations andonsidewalks; collisions withor falls to avoid infrastructure were additionally more likely on protected bike lanes with heavy separation.
The majority of cyclists who collided with or fell to avoid moving vehicles did so at intersections ( 58.9 %). The definition ofintersection in this study did not include junctions with driveways, alleys, or exit/ entrance ramps, and crashes or falls in protected bike lanes were reviewed to determine if they occurred at these additional junction types. In protected bike lanes, 60.0 % of incidents involving moving vehicles occurred at intersections, 26.7 % at junctions with driveways or alleys, 6.7 % at junctions with exit ramps, and 6.7 % at midblock ( not at junctions). Most incidents involving pedestrians in protected bike lanes occurred midblock ( 66.7 %) and those involving other cyclists were evenly distributed between intersections and nonintersections; none of these incidents occurred at junctions with driveways, alleys, or exit ramps.
Table 3
Comparison of route types and other characteristics at case and control sites and associated crash/ fall risk estimates ( N= 604).
Characteristic # of case sites/ # of control sites Unadjusted OR ( 95 % CI) Adjusted OR ( 95 % CI) Route type
Major road ( ref) 244/ 187 1.00 1.00
Bike lane on major road 92/ 109 0.52*( 0.33, 0.82) 0.53*( 0.33, 0.86) Sharrows on major road 16/ 17 0.68 ( 0.29, 1.61) 0.57 ( 0.23, 1.43) Local road, no bike infrastructure/ traffic calming 50/ 79 0.37*( 0.23, 0.61) 0.39*( 0.23, 0.65) Local road with bike lane, sharrows, or traffic calming 17/27 0.28*( 0.12, 0.67) 0.31*( 0.13, 0.75) Sidewalk 60/ 61 0.61+( 0.36, 1.05) 0.70 ( 0.40, 1.22)
Off-road/ trail 83/ 93 0.49*( 0.29, 0.83) 0.60+( 0.35, 1.04) One- way protected bike lane, light separation 18/ 13 1.07 ( 0.42, 2.72) 1.19 ( 0.46, 3.10) Two- way protected bike, light separation 21/ 9 8.40*( 1.08, 65.53) 11.38*( 1.40, 92.57) Protected bike lane, heavy separation 3/ 9 0.08*( 0.01, 0.73) 0.10*( 0.01, 0.95) Grade
Flat ( ref) 277/ 309 1.00 1.00 1.00
Downhill 225/ 167 1.66*( 1.24, 2.23) 1.92*( 1.38, 2.66) Uphill 75/ 103 0.82 ( 0.57, 1.18) 0.81 ( 0.55, 1.19)
Unknown 27/ 25 1.27 ( 0.68, 2.36) 1.50 ( 0.77, 2.89)
Temporary features
No ( ref) 483/ 520 1.00 1.00
Yes 114/ 75 1.94*( 1.33, 2.84) 2.23*( 1.46, 3.39)
Unknown 7/ 9 0.88 ( 0.28, 2.83) 0.73 ( 0.20, 2.63)
Streetcar or train tracks
No ( ref) 582/ 600 1.00 1.00
Yes 22/ 4 19.00*( 2.54, 141.93) 26.65*( 3.23, 220.17)
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks). OR= odds ratio; CI = confidence interval.
p < 0.05.
p < 0.10.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 6 3.4. Regional differences
The main analyses examining crash or fall risk by route type were conducted separately by city and are presented in Tables A6–A8 in the Appendix A. Patterns of results for conventional bike lanes, local roads, downhill grades, and temporary features were consistent across cities, although sample sizes were small in New York and Portland, which limited the power to achieve statistical significance. However, weobserved some differences among cities. Forexample, riding onsidewalks was associated with a significantly lowerriskinPortland relative to major roads, while it was associated with an elevated risk in New York. in The direction of effects forone- wayprotected bikelanesdiffered Washington and New York, and were associated with increased risk in Washington and decreased risk in New York relative to major roads, but or neither effect wasstatistically significant. Mostsiteswithstreetcar train tracks were located in Portland.
Characteristics of the protected bike lanes that served as case and control sites and their locations are described in Table 7. Nearly all incidents on two-way protected bike lanes with light separation occurred in Washington, while all but one protected bike lane site with heavy separation were in New York. An approximately 0.67-mile section of protected bike lane along two-way vehicle traffic on15thStreet NW in Washington between Massachusetts Avenue and Pennsylvania Avenue stood out as particularly risky, accounting for 11 of the 21 crashes or falls on two-way protected bike lanes with light separation in the study and only one control site. On average, protected bike lanes with light separation each were crossed by driveways, alleys, exit ramps, or intersecting roads about 19 times per mile, although the nature of these crossings varied by city; lanes in Washington were crossed more often by driveways and alleys than those in New York, and those in New York were crossed by intersecting roads more often than lanes in Washington. Protected bike lanes with heavy separation that were grade- separated were crossed by driveways, alleys, exit ramps, or intersecting roads an average of 6 times per mile, and those on bridges an average of twice per mile.
The characteristics of incidents on protected bike lanes with light separation appeared to differ between Washington andNew York, although the number of incidents was small in each city. Table 8 summarizes the circumstances and relation to junction of incidents on the type of protected bike lane with light separation, broken down by city, direction of travel of the bike lane ( one or two way), and proximity to the curb ( curbside or in the center of the road). More than half of incidents in curbside lanes in Washington occurred at junctions with intersecting roads or driveways/ alleys and nearly half involved moving vehicles. In New York, about a quarter occurred at junctions and less than a quarter involved moving vehicles. About a quarter of incidents in curbside lanes in Washington occurred at junctions with driveways or alleys, while less than 10 % did in New York. The types of separation used also varied between these cities ( not in table). Protected lanes in New York with light separation were almost exclusively separated from the road with parked cars, while protected lanes in Washington used a greater variety of separation types.
4. Discussion
Protected bike lanes are the facility most preferred by cyclists Winters and Teschke, 2010), with some reporting that they feel safer riding in them than on other types of infrastructure ( Monsere et al., 2014; Winters et al., 2012). Cycling levels increased in cities that have built them ( Buehler and Dill, 2016). Their rising popularity in North America has led to increased interest in knowing if they live up to expectations and protect cyclists more than other infrastructure types. This study demonstrates that risks of crashes or falls leading to emergency department visits can vary among protected bike lanes, with the lowest risks seen on those with heavy separation from the road and few junctions.
Table 4
Comparison of route types and other characteristics at case and control sites at intersections and nonintersections and associated crash/ fall risk estimates ( N= 604).
Nonintersection Intersection
Characteristic # of case sites/ # of control
sites
Adjusted OR ( 95 % CI) # of case sites/ # of control
sites
Adjusted OR ( 95 % CI)
Route type
Major road ( ref) 168/ 157 1.00 76/ 30 1.00
Bike lane on major road 49/ 102 0.39*( 0.21, 0.72) 43/ 7 3.87+( 0.98, 15. 32) Sharrows on major road 8/ 16 0.45 ( 0.14, 1.46) 8/ 1 6.37 ( 0.52, 78. 41) Local road 28/ 68 0.30*( 0.15, 0.58) 22/ 11 0.59 ( 0.21, 1.65) Local road with bike lane, sharrows, or traffic
calming
4/ 24 0.07*( 0.01, 0.34) 13/ 3 9.09 ( 0.32, 260. 81)
Sidewalk 36/ 53 0.48*( 0.24, 0.97) 24/ 8 1.20 ( 0.41, 3.53) Off-road/ trail 74/ 90 0.66 ( 0.35, 1.23) 9/ 3 1.24 ( 0.21, 7.16) One- way protected bike lane, light separation 13/ 12 1.19 ( 0.36, 3.91) 5/ 1 Two- way protected bike, light separation 11/ 8 7.80+( 0.71, 85.17) 10/ 1 13. 38+( 0.78, 228. 26) Protected bike lane, heavy separation 2/ 9 0.04*( 0.00, 0.55) 1/ 0 Grade
Flat ( ref) 188/ 280 1.00 89/ 29 1.00
Downhill 129/ 147 1.74*( 1.16, 2.61) 96/ 20 2.17 ( 0.93, 5.07) Uphill 51/ 91 1.00 ( 0.61, 1.63) 24/ 12 0.39+( 0.14, 1.10) Unknown 25/ 21 1.92+( 0.92, 400) 2/ 4 0.21 ( 0.03, 1.60) Temporary features
No ( ref) 312/ 466 1.00 171/ 54 1.00
Yes 77/ 65 3.40*( 2.00, 5.78) 37/ 10 1.38 ( 0.50, 3.78) Unknown 4/ 8 0.89 ( 0.17, 4.65) 3/ 1 0.45 ( 0.03, 6.11) Streetcar or train tracks
No ( ref) 381/ 536 1.00 201/ 64 1.00
Yes 12/ 3 10. 00*( 1.16, 86.28) 10/ 1
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks), the intersection indicator, and the interactions between these and the intersection indicator. OR= odds ratio; CI = confidence interval. p < 0.05.
p < 0.10.
model could not produce reliable estimates.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 7 Table5
Di stributionoffalorcrashcircumstancesbyroutetypepercentinparenthesesN604 Cr ashwithorfaltoavoid
Facilitytype
Mo vingvehicleStoppedparkedvehicleOthercyclistPedestrianInfrastructureSurfacefeatureOtherfalOtherspecificsunknownTotal Ma jorroad
111455441806252087294518425103416
244100
Bi kelaneonmajorroad
465002021733322211110109887222
92100
Sharrowsonmajorroad74383188000425021250
16100
Localroad
22440240480120480112205100120
50100
Localroadwithbikelanesharrowsortrafficcalming9529000159529421180 17100
Si dewalk
244001171172338133711716267117
60100
Of froadtra
il91080
161939108111331214525301112
83100
On ewayprotectedbikelanelightseparation7389211104222156211121110 18100
Twowayprotectedbikelanelightseparation7333148314352382951482950 21100
Pr otectedbikelaneheavyseparation133301333013330003100Al l 24340273121
3456254136609716187144915
604100
Table6Falorcrashcircumstancesrelativetomajorroadand95confidenceintervalsbyroutetypeN604 Ro utetype
Co lisionwithmoving
vehicle
Co lisionwithstoppedparked
vehicle
Co lisionwithcyclistfalto
avoidcyclist
Co lisionwithpedestrianfaltoavoid
pedestrian
Co lisionwithinfrastructureOtherfal
Ma jorroadref
1. 00
1. 00
1. 00
1. 00
1. 00
1. 00
Bi kelaneonmajorroad
1. 10086141
1. 21075193
1. 33034519
2. 650381855
0. 38005304
0. 85040181
Sharrowsonmajorroad
0. 96054170
1. 040362980000980000039530000835
1. 22032470
Localroad
0. 97069136
0. 22006089
3. 250951111
2. 440232639
2. 79085917
0. 87039243
Localroadwithbikelanesharrowsortraffci
calming
1. 16073186
0+000126000091700003732
2. 050271572
1. 15030445
Si dewalk
0. 88063123
0. 09001066
0. 68008552
4. 070582829
4. 651751231
2. 60149456
Of froadtra
i l
0. 230130450000025
7. 843171937
13232925999
4. 621851153
2. 94179482
On ewayprotectedbikelanelightseparation085047155
0. 620162340000888
271153213816
1. 940251489
1. 08028422
Twowayprotectedbikelightseparation073039136
0. 26004182
5. 811562158
290559914076
3. 200741498
0. 93024366
Pr otectedbikelaneheavyseparation0730153650000384
13562288077000017575
116220067470000692
No tep005p010Circumstanceswherenoroutetypesdifferedfrommajorroadssurfacefeatureotherspecificsunknowndonotappearintable J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 8
4.1. Protected bike lanes
Protected bike lanes are designed to prevent bicycle- vehicle collisions. While most bicyclist fatalities are due to crashes with vehicles Schepers et al., 2015), injuries leading to emergency department treatment are often the result of other types of falls and collisions. The risk of crashing or falling was elevated in some protected lanes with light separation in this study, and this increase in risk was especially apparent in crashes and falls not involving moving vehicles. Research from Copenhagen has reported that protected bike lanes change the distribution of crash types, with the frequency of some types increasing ( e.g., crashes involving pedestrians, two bicyclists, turning vehicles) and others decreasing ( e.g., rear-ends by motor vehicles, crashes with parked cars) when protected bike lanes are built ( Jensen, 2008a). It appears that some U.S. protected lanes may also introduce new non-vehicle hazards.
Pedestrians were involved in nearly a quarter of incidents in protected bike lanes with light separation in the current study but were not involved in many incidents on roads or conventional bike lanes.
Surveys, observational studies, and naturalistic cycling studies have noted that pedestrians can be frequent obstacles in protected bike lanes Basch et al., 2018; Conway et al., 2013; Goodno et al., 2013; Schleinitz et al., 2015; van der Horst et al., 2014). For example, Basch et al. (2018) observed about two pedestrians obstructing Manhattan, New York City, protected bike lanes per mile, including one pedestrian about every 2 miles pushing an object or walking a dog in the protected bike lane, and more than half of cyclists who use Washington’ s Pennsylvania Avenue protected bike lane surveyed by Goodno et al. ( 2013) reported nearcrashes with pedestrians in that facility. Other cyclists were also involved in incidents in two-way protected bike lanes, and two-cyclist conflicts have similarly been observed inother two- wayprotected bike lanes involving head-on, same-direction, and crossing configurations Schleinitz et al., 2015; van der Horst et al., 2014).
Most incidents in protected bike lanes involving pedestrians in this study occurred midblock, which can result from pedestrians using the lane for travel, crossing midblock, exiting a vehicle parked adjacent to it, or waiting for a taxi or other vehicle. It is unclear why protected bike lanes may be more susceptible to pedestrian obstructions than conventional bike lanes, but possibilities include that they can stand between pedestrians exiting parked cars and the sidewalk ( Vandenbulcke et al., 2014) or that some pedestrians treat protected bike lanes as sidewalks because they are buffered fromtraffic. Therewerenocrashes or falls due to pedestrians on the protected bike lanes with heavy separation, and the more substantial barriers and fewer intersections on these facilities likely gave pedestrians fewer openings to enter and cross them. Countermeasures to deter pedestrians from using protected bike lanes need to be developed. These results also highlight the risk of comingling cyclist and pedestrian routes when protected bike lanes are altered for construction or other purposes, and of blocking access to sidewalks that run adjacent to protected bike lanes. Data collection for this study concluded before the arrival of shared e-scooters, and future work should monitor if safety problems arise from these road users sharing protected bike lanes with cyclists.
Table 7
Locations and descriptions of protected bike lanes at case and control sites.
Road City Separation Intersections or ramps
crossing lane per mile
Driveways or alleys
crossing lane per mile
Painted
buffer
Side of
street
Direction of
adjacent vehicle
traffic
case sites/#
control sites
One way, light
separation
L Street NW DC Posts, parking stops 11 17 Y L One way 2/ 1 M Street NW DC Posts, parked cars 11 13 Y R One way 3/ 1 R Street NE* DC Parked cars 17 0 Y L One way 0/ 1
1st Ave NYC Parked cars 18 1 Y L One way 5/ 3
2nd Ave NYC Parked cars 18 3 Y L One way 2/ 1
6th Ave NYC Parked cars 21 0 Y L One way 3/ 2
7th Ave NYC Continuous low curb,
grade
21 0 N L One way 1/ 0
Broadway NYC Planters, Parked cars 20 2 Y L One way 1/ 1 8th Ave NYC Parked cars 20 < 1 Y L One way 0/ 1
9th Ave NYC Parked cars 20 1 Y L One way 0/ 1
Columbus Ave NYC Parked cars 17 0 Y L One way 0/ 1
Hawthorne Blvd Portland Posts 13 0 Y R One way 1/ 0
Two way, light
separation
15th Street NW DC Posts, parked cars 15 14 Y L One way 2/ 4 15th Street NW DC Posts, parked cars 7 6 Y S Two way 11/ 1 Pennsylvania Ave NW DC Parking stops 12 0 Y C Two way 5/ 2 1st Street NE DC Continuous low curb,
posts, parking stops
7 9 N S Two way 1/ 1
Kent Ave NYC Parked cars 10 10 Y L One way 2/ 0
Kent Ave NYC Posts 6 14 Y S Two way 0/ 1
Heavy separation
Williamsburg Bridge NYC Bridge rail 0 0 N S Two way 2/ 3 Manhattan Bridge NYC Bridge rail 1 0 N S Two way 1/ 1
Hudson River
Greenway
NYC Grade, trees,
landscaping
4 2 N S Two way 0/ 1
Pulaski Bridge NYC Concrete barrier
topped with rail
4 0 N S Two way 0/ 1
Queens Boulevard NYC Concrete barrier
topped with rail
8 0 N S Two way 0/ 1
Queensboro Bridge
Greenway
NYC Grade, concrete wall,
trees, landscaping
14 0 N C Two way 0/ 1
Southwest Moody Ave Portland Grade, railing 8 3 N S Two way 0/ 1 Note: Not all separation types were used concurrently for entire lane. DC = Washington DC; NYC= New York City; L= left, R= right, C = center, S= side of twoway street. denotes contraflow lane; other lanes onone- way streets followed direction oftraffic.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 9 The risk of crashes and falls involving moving vehicles in two of the three protected bike lane types could not be assessed in this analysis.
However, these incidents were not eliminated, particularly those occurring at intersections or junctions with driveways or alleys. This was especially the case in Washington, DC, where curbside- protected bike lanes were more frequently intersected by driveways and alleys.
Crashes with vehicles in protected bike lanes occurred less often in New York, where there were no alleys and most lanes were seldom intersected by driveways. Increased density of junctions increase the risk of bicyclist- motor vehicle crashes because they introduce additional opportunities for conflict (Li et al., 2017; Pulugurtha and Thakur, 2015; Siddiqui et al., 2012; Vandenbulcke et al., 2014; Wei and Lovegrove, 2013). Protected bike lanes with heavy separation in this study had fewer junctions than those with light separation, which likely contributed to their lower risk.
Intersections and other junctions can be particularly challenging for vehicles turning across contraflow ortwo- way protected bike lanes, because drivers look most frequently in the direction of traffic andthus may be less likely to detect cyclists approaching from the opposing direction ( Räsänen and Summala, 1998; Schepers et al., 2011; Summala et al., 1996). Two- way protected bike lanes alongside two-way vehicle traffic addadditional complexity asturning drivers need tomonitor bike both oncoming vehicle traffic andtwo- waybicycle traffic inthe lane. The riskiest protected bike lane segment in this study was a twoway lane with light separation along a two-way street.
Cities should consider the density of driveways and other junctions when choosing where to place protected bike lanes ( Federal Highway Administration, 2015; National Association of City Transportation Officials, 2014). Raised cycle crossings that lower vehicle speeds have been effective treatments atEuropean intersections withprotected bike lanes ( Gårder et al., 1998; Schepers et al., 2011) and are recommended for consideration in the Massachusetts Department of Transportation’ s Separated Bike Lane Planning and Design Guide ( 2015) at driveways and local street crossings. In 2011, the U.S. Federal Highway Administration issued interim approval for the use of green pavement in bike lanes, their extension through intersections, and other conflict areas.
Evidence on the effectiveness ofcolored bike lanes through intersections has been mixed ( Hunter et al., 2000; Jensen, 2008b; Schepers et al., 2011), with a simulator study suggesting that extending bike lanes with white dotted lines through intersections better captures drivers’ attention than green coloring ( Warner et al., 2017). Design guides for protected bike lanes recommend using high- visibility markings at junctions with driveways, as well as restricting parking 20–30 feet prior to the driveways and using signage to alert drivers exiting driveways of potential conflicts (Federal Highway Administration, 2015; Massachusetts Department of Transportation, 2015; National Association of City Transportation Officials, 2014).
Additional countermeasures have been recommended at intersections with protected bike lanes. Dedicated cyclist signals with a leading or partially protected phase and bike boxes can reduce conflicts atintersections ( Dill et al., 2012; Ledezma- Navarro et al., 2018). Two- stage turn queue boxes that allow for left turns from the rightmost lane without merging with traffic andlateral shifting oflanes atintersections to allow turning traffic tocross thebike lane arefeatured indesign guides ( Federal Highway Administration, 2015; Massachusetts Department of Transportation, 2015; National Association of City Transportation Officials, 2014) but have not been formally evaluated.
Yield to cyclist signage, smaller curb radii, and protected intersection designs with islands also showed promise at improving driver behavior around cyclists at intersections in a simulator study ( Warner et al., 2017). Some of these countermeasures are used by the study cities.
It is possible that some protected bike lanes might increase the risk of minor injuries while simultaneously reducing the risk of the most serious injuries. If this were the case, these lanes would still have a beneficial effect onsafety. Most injuries inthisstudy were minor, and theTathere were few crashes involving vehicles at nonjunctions, which is b
l
e
8 Di
s
t
r
i
b
u
t
i
o
n
o
f
j
u
n
c
t
i
o
n
t
y
p
e
s
a
n
d
i
n
c
i
d
e
n
t
c
i
r
c
u
m
s
t
a
n
c
e
s
b
y
l
o
c
a
t
i
o
n
a
n
d
t
y
p
e
o
f
p
r
o
t
e
c
t
e
d
b
i
k
e
l
a
n
e
w
i
t
h
l
i
g
h
t
s
e
p
a
r
a
t
i
o
n
p
e
r
c
e
n
t
o
f
e
a
c
h
p
r
o
t
e
c
t
e
d
b
i
k
e
l
a
n
e
t
y
p
e
i
n
p
a
r
e
n
t
h
e
s
e
s
Ci
t
y
Ty
p
e
Ju
n
c
t
i
o
n
t
y
p
e
In
c
i
d
e
n
t
c
i
r
c
u
m
s
t
a
n
c
e
s
In
t
e
r
s
e
c
t
i
o
n
A
l
l
e
y
o
r
d
r
i
v
e
w
a
y
J
u
n
c
t
i
o
n
w
i
t
h
e
x
i
t
r
a
m
p
N
o
n
j
u
n
c
t
i
o
n
M
o
v
i
n
g
v
e
h
i
c
l
e
P
e
d
e
s
t
r
i
a
n
O
t
h
e
r
c
y
c
l
i
s
t
O
t
h
e
r
c
i
r
c
u
m
s
t
a
n
c
e
Wa
s
h
i
n
g
t
o
n
O
n
e
w
a
y
c
u
r
b
s
i
d
e
N
5 2
4 0
0 2
4 0
0 0
1 2
0 0
3 6
0 0
1 2
0 0
0 1
2 0
0 Tw
o
w
a
y
c
u
r
b
s
i
d
e
N
1 4
6 4
2 9
3 2
1 4
0 5
3 5
7 6
4 2
9 3
2 1
4 3
2 1
4 2
1 4
3 Tw
o
w
a
y
c
e
n
t
e
r
o
f
r
o
a
d
N
5 4
8 0
0 0
0 1
2 0
0 1
2 0
0 2
4 0
0 0
2 4
0 0
Ne
w
Y
o
r
k
O
n
e
w
a
y
c
u
r
b
s
i
d
e
N
1 2
3 2
5 0
0 0
9 7
5 0
3 2
5 0
3 2
5 0
0 6
5 0
0 Tw
o
w
a
y
c
u
r
b
s
i
d
e
N
2 0
1 5
0 0
0 1
5 0
0 0
0 0
2 1
0 0
0 Po
r
t
l
a
n
d
O
n
e
w
a
y
c
u
r
b
s
i
d
e
N
1 0
0 1
1 0
0 0
0 1
1 0
0 0
0 0
0 J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 10 scenario leading to the majority of bicyclist fatalities ( Insurance Institute for Highway Safety, 2018). However, this study was not able to conclusively disentangle if protected bike lanes with less separation differentially affect lessand more injurious crashes andfalls. There were not enough cyclists who sustained serious or more severe ( AIS 3+) injuries to conduct a separate analysis at that severity level; 42 cyclists in the study sustained AIS 3+ injuries, and only one of these was injured in a protected bike lane.
Despite the increase in risk associated with some protected bike lanes, others are doing a good job at reducing injuries. Systematic research is needed to determine the characteristics of protected bike lanes that decrease and increase risk of crashes and falls of all types so that more robust design guidance can be developed. Our study points to the type of separation, direction of travel in bike lanes and adjacent vehicle travel lanes, and number of junctions as possible contributors. These characteristics, however, often co-occurred with one another and varied among the cities examined. We choose to categorize protected lanes with light separation as one-way and two-way, but could have classified them inother ways ( e. g., many junctions vs. fewer junctions, separated by parking vs. other light separation) that would have pointed to similar increases in injury risk in some lanes. These or other unreported characteristics of the protected lanes in Washington, DC may have driven the increase in risk observed on protected lanes there.
4.2. Other facilities
Conventional bike lanes were associated with lower risks than major roads overall and at nonintersections, but intersections were problematic for these facilities. Bike lanes to the right of travel lanes make cyclists susceptible to right- hook crashes, where a vehicle turns right in front of cyclist traveling straight ( Hurwitz et al., 2015). Many of the treatments recommended for use with protected bike lanes at intersections also apply to conventional bike lanes.
Results for some other infrastructure characteristics support findings from Teschke et al. ( 2012) and elsewhere. Local streets with and without bicycle facilities or traffic calming were associated with low crash or fall risks ( Aldred et al., 2018; Minikel, 2012), and downhill grade increased risk ( Allen- Munley et al., 2004; Klop and Khattak, 1999), likely because it increased cyclist speed. Streetcar or train tracks increased risk substantially, which is consistent with findings from Toronto, Vancouver, and Brussels ( Teschke et al., 2016; Vandenbulcke et al., 2014) and should be a consideration for cities expanding or implementing a streetcar network.
Findings for other infrastructure characteristics were consistent with Teschke et al. ( 2012) but differ withother previous research. In the current study and Teschke et al. ( 2012) risks associated with sharrows and multiuse trails or off-road locations were lower than those for major roads, although not always significantly so. Sharrows have been associated with positive changes in driver and cyclist behavior Furth et al., 2011; Hunter et al., 2010) but with increases in injury severity or crash rates in prior studies ( Ferenchak and Marshall, 2016; Wall et al., 2016). Others have reported increased risks associated with multiuse trails and other off-road locations (Aultman- Hall and Hall, 1998; de Rome et al., 2014; Moritz, 1998; Reynolds et al., 2009).
4.3. Limitations
While a case- crossover design evaluates the relative risks associated with infrastructure at a point in time, it cannot explain if the installation of a protected bike lane made a roadway safer or less safe. Protected bike lanes are typically installed on major thoroughfares where more protection for cyclists is warranted. Since they also often are constructed in city centers, they may have more exposure to intersections than the other route types that we investigated. Thus, results could in part reflect highrisk characteristics in the locations of protected bike lanes unrelated to their physical separation. It is crucial that controlled before- after studies of protected bike lanes are performed in the United States to inform policy decisions of if these lanes should be built.
Similarly, while the case- crossover design accounted for cyclist activity, the current study did not incorporate motor vehicle and pedestrian volumes because they were not consistently available. Higher motor vehicle and pedestrian volumes would make crashes with these road users more likely and having this information could better elucidate why crash types occurred at particular sites. Patients who died or who could not remember their route due to head injuries were excluded, so by design we did not include the most severely injured patients.
A cyclist’ s recollections of characteristics of their route may not have always been correct. Although Google Street View has been validated as a reliable alternative to in-person site visits for determining infrastructure features ( Mooney et al., 2016; Nesoff etal., 2018), our method was not able to capture temporary alterations to facilities, such as changes due to construction, that may have happened between the Google Street View capture and the trip dates. Large effect sizes with wide confidence intervals werereported forsomefacilities withsmall numbers of case or control sites ( e.g., two-way protected bike lanes with light separation, streetcar or train tracks), and these should not be interpreted as precise estimates of risk on these facilities.
4.4. Conclusions
Protected bike lanes increase ridership, but designs vary in the type of separation from the roadway and amount of potential conflict points.
Some designs may introduce new hazards that increase the risk of a crash or fall resulting in emergency department attendance without eliminating crashes with motor vehicles. Planners should consider the number of intersections with roads, driveways, and alleys when choosing where to place protected bike lanes and should implement countermeasures to maximize the visibility of cyclists at these conflict points when they are unavoidable. Designs with heavier separation and few conflict points appear todiminish hazards andcarry alowrisk of crashes or falls. Future work should more systematically examine the features that lead to higher and lower risk to guide design.
CRediT authorship contribution statement
Jessica B. Cicchino: Writing - original draft, Methodology, Formal analysis, Supervision. Melissa L. McCarthy: Methodology, Project administration, Writing - review & editing. Craig D. Newgard: Project administration, Writing - review & editing. Stephen P. Wall: Project administration, Writing - review & editing. Charles J. DiMaggio:
Project administration, Writing - review & editing. Paige E. Kulie:
Software, Investigation. Brittany N. Arnold: Investigation. David S.
Zuby: Conceptualization, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This work was supported by the Insurance Institute for Highway Safety. The authors would like to thank Grace Reagan for assistance with site characteristic coding; Kamil Narayan, Patricio Castillo, and Ada Rubin for managing data collection in the emergency departments; and all of the research assistants who enrolled and interviewed the participants.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 11 Appendix A
Table A2
Comparison of route types and other characteristics at case and control sites and associated crash/ fall risk estimates in crashes or falls not involving moving vehicles ( N= 361).
Characteristic # of case sites/ #
of control sites
Adjusted OR ( 95 % CI)
Route type
Major road ( ref) 133/ 113 1.00
Bike lane on major road 46/ 51 0.80 ( 0.41, 1.55)
Sharrows on major road 9/ 10 0.55 ( 0.16, 1.87)
Local road, no bike infrastructure/
traffic calming
28/ 47 0.48+( 0.22, 1.03)
Local road with bike lane, sharrows,
or traffic calming
8/ 17 0.27*( 0.08, 0.90)
Sidewalk 36/ 35 1.00 ( 0.47, 2.16)
Off-road/ trail 74/ 67 1.33 ( 0.67, 2.62)
One- way protected bike lane, light
separation
11/ 9 1.22 ( 0.34, 4.37)
Two- way protected bike, light
separation
14/ 5 11. 33*( 1.33, 96.30)
Protected bike lane, heavy
separation
2/ 7 0.12+( 0.01, 1.28)
Grade
Flat ( ref) 171/ 188 1.00
Downhill 122/ 96 1.90*( 1.23, 2.94)
Uphill 45/ 58 0.92 ( 0.55, 1.53)
Unknown 23/ 19 1.76 ( 0.83, 3.75)
Temporary features
No ( ref) 278/ 313 1.00
Yes 78/ 43 3.49*( 1.96, 6.21)
Unknown 5/ 5 1.00 ( 0.16, 6.24)
Streetcar or train tracks
No ( ref) 344/ 357 1.00
Yes 17/ 4 23. 96*( 2.74, 209. 22)
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks). OR = odds ratio; CI= confidence interval.
p < 0.05.
p < 0.10.
Table A1
Comparison of route types and other characteristics at case and control sites and associated crash/ fall risk estimates in crashes with moving vehicles or falls to avoid them ( N= 243).
Characteristic # of case sites/ # of control sites Adjusted OR ( 95 % CI) Route type
Major road ( ref) 111/ 74 1.00
Bike lane on major road 46/ 58 0.36*( 0.16, 0.77)
Sharrows on major road 7/ 7 0.70 ( 0.17, 2.98)
Local road, no bike infrastructure/ traffic calming 22/ 32 0.31*( 0.14, 0.70) Local road with bike lane, sharrows, or traffic calming 9/ 10 0.58 ( 0.14, 2.44) Sidewalk 24/ 26 0.47 ( 0.19, 1.16)
Off-road/ trail 9/ 26 0.04*( 0.00, 0.28)
One- way protected bike lane, light separation 7/ 4 1.66 ( 0.35, 7.80) Two- way protected bike, light separation 7/ 4 §
Protected bike lane, heavy separation 1/ 2 §
Grade
Flat ( ref) 106/ 121 1.00
Downhill 103/ 71 2.04*( 1.17, 3.55)
Uphill 30/ 45 0.57+( 0.29, 1.11)
Unknown 4/ 6 0.88 ( 0.18, 4.38)
Temporary features
No ( ref) 205/ 207 1.00
Yes 36/ 32 1.07 ( 0.53, 2.13)
Unknown 2/ 4 1.04 ( 0.15, 7.26)
Streetcar or train tracks
No ( ref) 238/ 243 1.00
Yes 5/ 0 §
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks). OR= odds ratio; CI = confidence interval.
p < 0.05.
p < 0.10.
model could not produce reliable estimates.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 12 Table A4
Comparison of route types and other characteristics at case and control sites and associated crash/ fall risk estimates among cyclists with moderate or more severe AIS 2+) injuries ( N= 254).
Characteristic # of case sites/ # of control sites Adjusted OR ( 95 % CI) Route type
Major road ( ref) 94/ 72 1.00
Bike lane on major road 38/ 41 0.59 ( 0.25, 1.42)
Sharrows on major road 4/ 3 0.53 ( 0.06, 4.82)
Local road, no bike infrastructure/ traffic calming 25/ 38 0.40*( 0.18, 0.87) Local road with bike lane, sharrows, or traffic calming 9/ 17 0.15*( 0.04, 0.65) Sidewalk 25/ 29 0.48 ( 0.19, 1.16)
Off-road/ trail 42/ 44 0.55 ( 0.23, 1.31)
One- way protected bike lane, light separation 7/ 4 2.18 ( 0.38, 12.38) Two-way protected bike, light separation 10/ 4 4.23 ( 0.39, 45.35) Protected bike lane, heavy separation 0/ 2 §
Grade
Flat ( ref) 101/ 121 1.00
Downhill 108/ 73 2.76*( 1.58, 4.82)
Uphill 31/ 48 0.80 ( 0.42, 1.52)
Unknown 14/ 12 1.74 ( 0.67, 4.54)
Temporary features
No ( ref) 209/ 223 1.00
Yes 43/ 29 2.19*( 1.10, 4.38)
Unknown 2/ 2 0.75 ( 0.09, 6.16)
Streetcar or train tracks
No ( ref) 242/ 252 1.00
Yes 12/ 2 §
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks). OR= odds ratio; CI = confidence interval.
p < 0.05.
model could not produce reliable estimates.
Table A3
Comparison of route types and other characteristics at case and control sites and associated crash/ fall risk estimates among cyclists with no or minor ( AIS 0 or 1) injuries ( N= 345).
Characteristic # of case sites/ # of control sites Adjusted OR ( 95 % CI) Route type
Major road ( ref) 147/ 113 1.00
Bike lane on major road 53/ 67 0.45*( 0.24, 0.83)
Sharrows on major road 12/ 14 0.54 ( 0.19, 1.48)
Local road, no bike infrastructure/ traffic calming 24/ 39 0.36*( 0.17, 0.76) Local road with bike lane, sharrows, or traffic calming 8/ 10 0.52 ( 0.16, 1.69) Sidewalk 35/ 32 0.88 ( 0.41, 1.89)
Off-road/ trail 41/ 49 0.62 ( 0.29, 1.29)
One- way protected bike lane, light separation 11/ 9 0.80 ( 0.24, 2.62) Two- way protected bike, light separation 11/ 5 §
Protected bike lane, heavy separation 3/ 7 0.16 ( 0.02, 1.50) Grade
Flat ( ref) 175/ 187 1.00
Downhill 113/ 91 1.54*( 1.02, 2.66)
Uphill 44/ 54 0.81 ( 0.49, 1.35)
Unknown 13/ 13 1.42 ( 0.55, 3.64)
Temporary features
No ( ref) 272/ 294 1.00
Yes 68/ 45 2.19*( 1.28, 3.75)
Unknown 5/ 6 1.42 ( 0.55, 3.64)
Streetcar or train tracks
No ( ref) 335/ 343 1.00
Yes 10/ 2 10.01*( 1.17, 85.68)
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks). OR= odds ratio; CI = confidence interval.
p < 0.05.
model could not produce reliable estimates.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 13 Table A6
Comparison of route types and other characteristics at case and control sites in Washington, DC, and associated crash/ fall risk estimates ( N= 354).
Characteristic # of case sites/ #
of control sites
Adjusted OR ( 95 % CI)
Route type
Major road ( ref) 137/ 101 1.00
Bike lane on major road 44/ 66 0.41*( 0.22, 0.76)
Sharrows on major road 8/ 9 0.53 ( 0.16, 1.79)
Local road 32/ 47 0.49*( 0.26, 0.91)
Local road with bike lane,
sharrows, or traffic calming
8/ 10 0.53 ( 0.16, 1.73)
Sidewalk 51/ 60 0.46*( 0.23, 0.92)
Off-road/ trail 50/ 50 0.73 ( 0.38, 1.41)
One-way protected bike lane,
light separation
5/ 3 1.62 ( 0.26, 9.96)
Two- way protected bike, light
separation
19/ 8 9.36*( 1.15, 76.07)
Protected bike lane, heavy
separation
0/ 0 §
Grade
Flat ( ref) 152/ 173 1.00
Downhill 146/ 111 1.69*( 1.14, 2.49)
Uphill 45/ 56 0.89 ( 0.54, 1.47)
Unknown 11/ 14 0.73 ( 0.29, 1.89)
Temporary features
No ( ref) 269/ 289 1.00
Yes 84/ 58 1.79*( 1.10, 2.90)
Unknown 1/ 7 §
Streetcar or train tracks 1/ 0 §
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks). OR = odds ratio; CI= confidence interval.
p < 0.05.
model could not produce reliable estimates.
Table A5
Comparison of route types and other characteristics at case and control sites and associated crash/ fall risk estimates, among patients who reported being in route types that were present at the sites ( N= 558).
Characteristic # of case sites/ #
of control sites
Adjusted OR ( 95 % CI)
Route type
Major road ( ref) 221/ 168 1.00
Bike lane on major road 87/ 101 0.53*( 0.32, 0.89)
Sharrows on major road 14/ 16 0.51 ( 0.20, 1.32)
Local road 47/ 71 0.42*( 0.24, 0.72)
Local road with bike lane,
sharrows, or traffic calming
16/ 26 0.31*( 0.13, 0.75)
Sidewalk 78/ 90 0.57 ( 0.32, 1.01)
Off-road/ trail 54/ 55 0.67+( 0.37, 1.22)
One- way protected bike lane,
light separation
17/ 13 1.08 ( 0.40, 2.88)
Two- way protected bike, light
separation
21/ 9 11.37*( 1.39, 92.68)
Protected bike lane, heavy
separation
3/ 9 0.09*( 0.01, 0.90)
Grade
Flat ( ref) 250/ 288 1.00
Downhill 214/ 155 2.04*( 1.45, 2.88)
Uphill 67/ 93 0.81 ( 0.53, 1.22)
Unknown 27/ 22 1.84+( 0.92, 3.68)
Temporary features
No ( ref) 447/ 478 1.00
Yes 104/ 73 2.02*( 1.31, 3.14)
Unknown 7/ 7 0.97 ( 0.24, 3.89)
Streetcar or train tracks
No ( ref) 537/ 554 1.00
Yes 21/ 4 25.25*( 3.03, 210. 62)
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks). OR = odds ratio; CI= confidence interval.
p < 0.05.
p < 0.10.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 14 Table A8
Comparison of route types and other characteristics at case and control sites in Portland, OR and associated crash/ fall risk estimates ( N= 131).
Characteristic # of case sites/ #
of control sites
Adjusted OR ( 95 % CI)
Route type
Major road ( ref) 42/ 30 1.00
Bike lane on major road 37/ 28 0.74 ( 0.20, 2.59)
Sharrows on major road 0/ 1 §
Local road 12/ 23 0.16*( 0.04, 0.64)
Local road with bike lane,
sharrows, or traffic calming
8/ 13 0.13*( 0.02, 0.98)
Sidewalk 26/ 25 0.14*( 0.02, 1.00)
Off-road/ trail 5/ 10 0.66 ( 0.13, 3.27)
One- way protected bike lane,
light separation
1/ 0 §
Two- way protected bike, light
separation
0/ 0 §
Protected bike lane, heavy
separation
0/ 1 §
Grade
Flat ( ref) 43/ 50 1.00
Downhill 57/ 42 3.62*( 1.32, 9.95)
Uphill 17/ 30 0.53 ( 0.18, 1.59)
Unknown 14/ 9 3.22 ( 0.79, 13. 08)
Temporary features
No ( ref) 115/ 119 1.00
Yes 14/ 12 2.47 ( 0.54, 11. 37)
Unknown 2/ 0 §
Streetcar or train tracks 21/ 4 66.44*( 5.06, 872. 75)
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks). OR = odds ratio; CI= confidence interval.
p < 0.05.
model could not produce reliable estimates.
Table A7
Comparison of route types and other characteristics at case and control sites in New York City and associated crash/ fall risk estimates ( N= 119).
Characteristic # of case sites/ #
of control sites
Adjusted OR ( 95 % CI)
Route type
Major road ( ref) 65/ 56 1.00
Bike lane on major road 11/ 15 0.32 ( 0.06, 1.64)
Sharrows on major road 8/ 7 0.45 ( 0.09, 2.32)
Local road 6/ 9 0.18+( 0.03, 1.04)
Local road with bike lane,
sharrows, or traffic calming
1/ 4 0.08 ( 0.00, 1.75)
Sidewalk 6/ 8 0.68 ( 0.12, 3.68)
Off-road/ trail 5/ 1 6.66 ( 0.56, 78.78)
One- way protected bike lane,
light separation
12/ 10 0.81 ( 0.22, 2.96)
Two- way protected bike, light
separation
2/ 1 §
Protected bike lane, heavy
separation
3/ 8 0.12 ( 0.01, 1.76)
Grade
Flat ( ref) 82/ 86 1.00
Downhill 22/ 14 2.18 ( 0.83, 5.73)
Uphill 13/ 17 0.67 ( 0.24, 1.85)
Unknown 2/ 2 0.64 ( 0.05, 8.12)
Temporary features
No ( ref) 99/ 112 1.00
Yes 16/ 5 5.77*( 1.57, 21.20)
Unknown 4/ 2 2.38 ( 0.15, 37.75)
Streetcar or train tracks a 0/ 0
Note: Adjusted model included variables listed in table as covariates ( route type, grade, temporary features, streetcar or train tracks). OR = odds ratio; CI= confidence interval.
p < 0.05.
p < 0.10.
model could not produce reliable estimates.
a variable not included in model.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 15 References
Aldred, R., Goodman, A., Gulliver, J., Woodcock, J., 2018. Cycling injury risk in London:
a case- control study exploring the impact of cycle volumes, motor vehicle volumes, and road characteristics including speed limits. Accid. Anal. Prev. 117, 75–84. https:// doi.org/ 10.1016/ j.aap. 2018. 03.003.
Allen- Munley, C., Daniel, J., Dhar, S., 2004. Logistic model for rating urban bicycle routesafety. Transp. Res. Rec. 1878, 107– 115. https:// doi.org/ 10.3141/ 1878- 13.
Aultman- Hall, L., Hall, F.L., 1998. Ottawa- Carleton commuter cyclist on-and off-road incident rates. Accid. Anal. Prev. 30, 29–43. https:// doi.org/ 10.1016/ S0001- 4575( 97)00059- 6.
Basch, YorkCity: C.H., Ethan, D., Basch, C.E., 2018. Bike lane obstructions in Manhattan, New implications for bicyclist safety. J. Community Health. https:// doi.org/ 10. 1007/ s10900- 018- 00596- 4. Advanced online publication.
Beck, B., Stevenson, M., Newstead, S., Cameron, P., Judson, R., Edwards, E.R., Bucknill, A., Johnson, M., Gabbe, B., 2016. Bicycling crash characteristics: an in-depth crash investigation study. Accid. Anal. Prev. 96, 219– 227.
Bhatia, D., Richmond, S.A., Loo, C.K.J., Rothman, L., Macarthur, C., Howard, A., 2016.
Examining ofToronto. the impact of cycle lanes on cyclist- motor vehicle collisions in the city J. Transp. Health 3, 523– 528. https:// doi. org/ 10. 1016/ j.jth.2016. 04.002.
Buehler, R., Dill, J., 2016. Bikeway networks: a review of effects oncycling. Transp. Rev. 36, 9–27. https:// doi. org/ 10. 1080/ 01441647. 2015. 1069908.
Buehler, evidenceon R., Pucher, J., 2012. Cycling to work in 90 large American cities: new the role of bike paths and lanes. Transportation 39, 409– 432. https:// doi.org/ 10. 1007/ s11116- 011-9355- 8. journal article.
of Chan, I.S.F., Zhang, Z., 1999. Test- based exact confidence intervals forthe difference two binomial proportions. Biometrics 55, 1202– 1209. https:// doi.org/ 10.1111/ j. 0006- 341X. 1999. 01202. x.
Chen, L., Chen, C., Srinivasan, R., McKnight, C.E., Ewing, R., Roe, M., 2012. Evaluating the safety effects ofbicycle lanes inNew York City. Am. J. Public Health 102, 1120– 1127. https:// doi.org/ 10. 2105/ ajph. 2011. 300319.
Conway, A., Cheng, J., Peters, D., Lownes, N., 2013. Characteristics of multimodal conflicts inurban on- street bicycle lanes. Transp. Res. Rec. 2387, 93– 101. https:// doi. org/ 10.3141/ 2387- 11.
de Rome, L., Boufous, S., Georgeson, T., Senserrick, T., Richardson, D., Ivers, R., 2014.
Bicycle crashes in different riding environments inthe Australian capital territory.
Traffic Inj. Prev. 15, 81– 88. https:// doi.org/ 10.1080/ 15389588. 2013. 781591.
Dill, buildthem, J., Carr, T., 2003. Bicycle commuting and facilities in major U.S. cities: if you commuters will use them. Transp. Res. Rec. 1828, 116– 123. https:// doi. org/
10.3141/ 1828- 14.
Dill, J., Monsere, C.M., McNeil, N., 2012. Evaluation of bike boxes at signalized intersections. Accid. Anal. Prev. 44, 126– 134. https:// doi. org/ 10.1016/ j.aap. 2010. 10. 030.
Federal Highway Administration, 2011. MUTCD — Interim Approval for Optional Use of Green Colored Pavement for Bike Lanes. ( IA-14). Retrieved from. Federal HighwayAdministration, Washington, DC. https:// mutcd. fhwa. dot. gov/ resources/ interim_ approval/ ia14/ index. htm.
Federal Highway Administration, 2015. Separated Bike Lane Planning and Design Guide.( FHWA- HEP- 15-025). Retrieved from. Federal Highway Administration, Washington, DC. https:// www. fhwa. dot. gov/ environment/ bicycle_ pedestrian/ publications/ separated_ bikelane_ pdg/ separatedbikelane_ pdg. pdf.
Ferenchak, N.N., Marshall, W.E., 2016. The relative ( in) effectiveness ofbicycle sharrowson ridership and safety outcomes. In: Paper Presented at the 95th Annual Meeting of the Tranpsortation Research Board. Washington, DC,.
Furth, P.G., Dulaski, D.M., Bergenthal, D., Brown, S., 2011. More than sharrows: laneatthe within- a-lane bicycle priority treatments in three U.S. cities. In: Paper Presented 90th Annual Meeting of the Tranpsortation Research Board. Washington, DC,.
Gårder, P., Leden, L., Pulkkinen, U., 1998. Measuring the safety effect ofraised bicycle crossings using a new research methodology. Transp. Res. Rec. 1636, 64–70. https:// doi. org/ 10.3141/ 1636- 10.
Gennarelli, T.A., Wodzin, E., 2008. Abbreviated Injury Scale 2005: Update 2008.
Association for the Advancement of Automotive Medicine, Des Plaines, IL.
Goodno, M., McNeil, N., Parks, J., Dock, S., 2013. Evaluation of innovative bicycle facilities in Washington, DC: Pennsylvania Avenue median lanes and 15th street cycle track. Transp. Res. Rec. 2387, 139– 148. https:// doi.org/ 10.3141/ 2387- 16.
Hamann, C., Peek- Asa, C., 2013. On-road bicycle facilities and bicycle crashes in Iowa, 2007– 2010. Accid. Anal. Prev. 56, 103– 109. https:// doi.org/ 10.1016/ j.aap. 2012. 12. 031.
Hunter, W., Harkey, D., Stewart, J., Birk, M., 2000. Evaluation of blue bike-lane treatment in Portland, Oregon. Transp. Res. Rec. 1705, 107– 115. https:// doi. org/ 10.3141/ 1705- 16.
Hunter, W.W., Thomas, L., Srinivasan, R., Martell, C.A., 2010. Evaluation of Shared LaneMarkings. ( FHWA- HRT- 10-041). Retrieved from. Federal Highway Administration, McLean, VA. https:// www. fhwa. dot. gov/ publications/ research/ safety/ pedbike/ 10041/ 10041. pdf.
Hurwitz, EffectiveDesign D., Jannat, M., Warner, J., Monsere, C.M., Razmpa, A., 2015. Towards Treatment for Right Turns at Intersections with Bicycle Traffic. ( FHWA- ORRD-16-06). Retrieved from. Oregon Department of Transportation and Federal Highway Administration, Salem, OR and Washington, DC. https:// pdxscholar. library. pdx. edu/ cgi/ viewcontent. cgi?article= 1317& context= cengin_ fac.
Insurance Institute for Highway Safety, 2018. December). Fatality facts: Bicyclists, 2017.
Retrieved from. www. iihs.org/ topics/ fatality- statistics/ detail/ bicyclists.
Jensen, S.U., 2008a. Bicycle tracks and lanes: a before- after study. In: Paper Presented at the 87th Annual Meeting of the Tranpsortation Research Board. Washington, DC.
Jensen, S.U., 2008b. Safety effects ofblue cycle crossings: abefore- after study. Accid.
Anal. Prev. 40, 742–750. https:// doi.org/ 10.1016/ j.aap. 2007. 09.016.
Klop, J.R., Khattak, A.J., 1999. Factors influencing bicycle crash severity ontwo-lane, undivided roadways in North Carolina. Transp. Res. Rec. 1674, 78–85. https:// doi. org/ 10. 3141/ 1674- 11.
Kondo, M.C., Morrison, C., Guerra, E., Kaufman, E.J., Wiebe, D.J., 2018. Where do bike lanes work best? A Bayesian spatial model of bicycle lanes and bicycle crashes. Saf.
Sci. 103, 225– 233. https:// doi. org/ 10.1016/ j.ssci. 2017. 12. 002.
Ledezma- oflevel Navarro, B., Stipancic, J., Andreoli, A., Miranda- Moreno, L., 2018. Evaluation of service and safety for vehicles and cyclists at signalized intersections. In:
Paper Presented at the 97th Annual Meeting of the Tranpsortation Research Board.
Washington, DC.
Li, H., oncycle Graham, D.J., Liu, P., 2017. Safety effects ofthe London cycle superhighways collisions. Accid. Anal. Prev. 99, 90–101. https:// doi.org/ 10.1016/ j.aap. 2016.
11.016.
Lusk, A.C., Furth, P.G., Morency, P., Miranda- Moreno, L.F., Willett, W.C., Dennerlein, J.T., 2011. Risk of injury for bicycling on cycle tracks versus in the street. Inj. Prev. 17, 131– 135. https:// doi.org/ 10. 1136/ ip.2010. 028696.
Marshall, W.E., Ferenchak, N.N., 2019. Why cities with high bicycling rates are safer for all road users. J. Transp. Health. https:// doi.org/ 10.1016/ j.jth.2019. 03.004.
Advanced online publication.
Massachusetts andDesign Department of Transportation, 2015. Separated Bike Plane Planning Retrievedfrom. Guide. Boston, MA: Massachusetts Department of Transportation. https:// www. mass. gov/ lists/ separated- bike- lane- planning- design- guide.
McKenzie, B., 2014. Modes Less Traveled: Bicycling and Walking to Work in the United States, 2008– 2012. ( ACS- 25). U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau, Washington, DC.
Minikel, E., 2012. Cyclist safety on bicycle boulevards and parallel arterial routes inBerkeley, California. Accid. Anal. Prev. 45, 241– 247. https:// doi.org/ 10. 1016/ j.aap. 2011. 07.009.
Monsere, C., Dill, J., McNeil, N., Clifton, K.J., Foster, N., Goddard, T., Berkow, M., Gilpin, J., Voros, K., van Hengel, D., 2014. Lessons From the Green Lanes: EvaluatingProtected Bike Lanes in the US. ( NITC- RR- 583). Retrieved from. National Institutefor Transportation and Communities, Portland, OR. https:// trec. pdx. edu/ research/ project/ 583/ Lessons_ from_the_Green_ Lanes:_ Evaluating_ Protected_ Bike_Lanes_ in_the_ U.S.
Mooney, S. J., DiMaggio, C. J., Lovasi, G.S., Neckerman, K.M., Bader, M.D.M., Teitler, J.O., Sheehan, D.M., Jack, D.W., Rundle, A.G., 2016. Use of Google Street View to assessenvironmental contributions to pedestrian injury. Am. J. Public Health 106, 462–469. https:// doi.org/ 10.2105/ ajph. 2015. 302978.
Moritz, W.E., 1998. Adult bicyclists in the United States: characteristics and riding experience in 1996. Transp. Res. Rec. 1636, 1–7. https:// doi. org/ 10.3141/ 1636- 01.
National Association of City Transportation Officials, 2014. NACTO Urban BikewayDesign Guide. Retrieved from. https:// nacto. org/ publication/ urban- bikewaydesign- guide/.
Nesoff, E. D., Milam, A. J., Pollack, K. M., Curriero, F. C., Bowie, J. V., Gielen, A. C., FurrHolden, D.M., 2018. Novel methods for environmental assessment of pedestrian injury: creation and validation of the Inventory for Pedestrian Safety Infrastructure. J.
Urban Health 95, 208–221. https:// doi.org/ 10.1007/ s11524- 017-0226- 2. journal article.
Nosal, cyclingin T., Miranda- Moreno, L.F., 2012. Cycle- tracks, bicycle lanes, and on-street Montreal, Canada: a preliminary comparison of the cyclist injury risk. In: PaperPresented at the 91st Annual Meeting of the Tranpsortation Research Board.
Washington, DC,.
Park, J., Abdel- Aty, M., Lee, J., Lee, C., 2015. Developing crash modification functions to assess safety effects ofadding bike lanes for urban arterials with different roadwayand socio- economic characteristics. Accid. Anal. Prev. 74, 179– 191. https:// doi.org/
10.1016/ j.aap. 2014. 10.024.
People for Bikes, 2018. Inventory of Protected Bike Lanes. December Retrieved from. https:// peopleforbikes. org/ green- lane- project/ inventory- protected- bike- lanes/.
Pulugurtha, S.S., Thakur, V., 2015. Evaluating the effectiveness ofon- street bicycle laneand assessing risk to bicyclists in Charlotte, North Carolina. Accid. Anal. Prev. 76, 34–41. https:// doi.org/ 10.1016/ j.aap. 2014. 12.020.
Raihan, M.A., Alluri, P., Wu, W., Gan, A., 2019. Estimation of bicycle crash modification factors ( CMFs) on urban facilities using zero inflated negative binomial models.
Accid. Anal. Prev. 123, 303– 313. https:// doi.org/ 10.1016/ j.aap. 2018. 12.009.
Räsänen, M., Summala, H., 1998. Attention and expectation problems in bicycle– carcollisions: an in-depth study. Accid. Anal. Prev. 30, 657– 666. https:// doi. org/ 10. 1016/ S0001- 4575( 98)00007- 4.
Reynolds, C.C.O., Harris, M.A., Teschke, K., Cripton, P.A., Winters, M., 2009. The impactof transportation infrastructure on bicycling injuries and crashes: a review of theliterature. Environ. Health 8, 47. https:// doi.org/ 10.1186/ 1476- 069X- 8-47.
Sanford, T., McCulloch, C.E., Callcut, R.A., Carroll, P.R., Breyer, B.N., 2015. Bicycle trauma injuries and hospital admissions in the United States, 1998– 2013. JAMA 314, 947– 949. https:// doi.org/ 10.1001/ jama. 2015. 8295.
SAS Institute, 2011. The LOGISTIC Procedure. Retrieved from. https:// support. sas. com/ documentation/ cdl/ en/ statug/ 63962/ HTML/ default/ viewer. htm# statug_ logistic_ sect051. htm.
Schepers, J.P., Kroeze, P.A., Sweers, W., Wüst, J.C., 2011. Road factors and bicycle– motor vehicle crashes at unsignalized priority intersections. Accid. Anal. Prev. 43, 853– 861. https:// doi. org/ 10.1016/ j.aap. 2010. 11.005.
Schepers, P., Agerholm, N., Amoros, E., Benington, R., Bjørnskau, T., Dhondt, S., de Geus, B., Hagemeister, C., Loo, B.P.Y., Niska, A., 2015. An international review of the frequency of single- bicycle crashes ( SBCs) and their relation to bicycle modal share.
Inj. Prev. 21, e138– e143. https:// doi.org/ 10. 1136/ injuryprev- 2013- 040964.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 16 Schleinitz, Conflictpartners K., Petzoldt, T., Franke- Bartholdt, L., Krems, J.F., Gehlert, T., 2015. and infrastructure use in safety critical events in cycling – results from a naturalistic cycling study. Transp. Res. Part F: Traffic Psychol. Behav. 31, 99– 111. https:// doi.org/ 10.1016/ j.trf.2015. 04.002.
Siddiqui, andbicycle C., Abdel- Aty, M., Choi, K., 2012. Macroscopic spatial analysis of pedestrian crashes. Accid. Anal. Prev. 45, 382– 391. https:// doi. org/ 10. 1016/ j.aap. 2011.
08.003.
Stutts, J.C., Hunter, W.W., 1999. Motor vehicle and roadway factors in pedestrian and bicyclist injuries: an examination based on emergency department data. Accid. Anal.
Prev. 31, 505– 514. https:// doi. org/ 10.1016/ S0001- 4575( 99) 00007- X.
Summala, H., Pasanen, E., Räsänen, M., Sievänen, J., 1996. Bicycle accidents and drivers’ visual search at left and right turns. Accid. Anal. Prev. 28, 147– 153. https:// doi.org/
10.1016/ 0001- 4575( 95)00041- 0.
Sundstrom, C.A., Quinn, S.M., Weld, R., 2019. A bicyclist crash comparison of mixing zone and fully split phase signal treatments at intersections with protected bike lanesin New York City. In: Paper Presented at the 98th Annual Meeting of theTranpsortation Research Board. Washington, DC.
Teschke, K., Harris, M.A., Reynolds, C.C.O., Winters, M., Babul, S., Chipman, M., Cusimano, M.D., Brubacher, J.R., Hunte, G., Friedman, S. M., 2012. Route infraPublicHealth structure and the risk of injuries to bicyclists: a case- crossover study. Am. J. 102, 2336– 2343. https:// doi. org/ 10. 2105/ AJPH. 2012. 300762.
Teschke, K., Dennis, J., Reynolds, C.C.O., Winters, M., Harris, M.A., 2016. Bicycling crashes on streetcar ( tram) or train tracks: mixed methods to identify prevention measures. BMC Public Health 16, 617. https:// doi.org/ 10.1186/ s12889- 016- 3242- 3.
United States Census Bureau, 2018. Commuting Characteristics by Sex, 2017 AmericanCommunity Survey 1-year Estimates. Retrieved from. https:// factfinder. census. gov/ faces/ tableservices/ jsf/ pages/ productview. xhtml? pid= ACS_17_1YR_S0801& prodType= table.
van der Horst, A.R.A., de Goede, M., de Hair-Buijssen, S., Methorst, R., 2014. Traffic conflicts onbicycle paths: asystematic observation ofbehaviour from video. Accid.
Anal. Prev. 62, 358– 368. https:// doi.org/ 10.1016/ j.aap. 2013. 04.005.
Vandenbulcke, G., Thomas, I., Panis, L.I., 2014. Predicting cycling accident risk inBrussels: a spatial case– control approach. Accid. Anal. Prev. 62, 341– 357. https:// doi.org/ 10.1016/ j.aap.2013. 07.001.
Wall, S.P., Lee, D.C., Frangos, S.G., Sethi, M., Heyer, J.H., Ayoung- Chee, P., DiMaggio, C. J., 2016. The effect ofsharrows, painted bicycle lanes and physically protectedpaths on the severity of bicycle injuries caused by motor vehicles. Safety 2, 26. https:// doi. org/ 10.3390/ safety2040026.
Warner, J., Hurwitz, D.S., Monsere, C.M., Fleskes, K., 2017. A simulator- based analysis of engineering treatments for right- hook bicycle crashes at signalized intersections.
Accid. Anal. Prev. 104, 46–57. https:// doi.org/ 10.1016/ j.aap. 2017. 04.021.
Wei, F., Lovegrove, G., 2013. An empirical tool to evaluate the safety of cyclists: community based, macro- level collision prediction models using negative binomial regression. Accid. Anal. Prev. 61, 129– 137. https:// doi.org/ 10. 1016/ j.aap. 2012. 05. 018.
Winters, M., Teschke, K., 2010. Route preferences among adults in the near market forbicycling: findings ofthe Cycling inCities Study. Am. J. Health Promot. 25, 40– 47. https:// doi. org/ 10.4278/ ajhp. 081006- QUAN- 236.
Winters, M., Babul, S., Becker, H., Brubacher, J.R., Chipman, M., Cripton, P., Cusimano, M.D., Friedman, S.M., Harris, M.A., Hunte, G., 2012. Safe cycling: how do risk perceptions compare with observed risk? Can. J. Public Health 103, S42– 47.
J.B. Cicchino, et al. Accident Analysis and Prevention 141 ( 2020) 105490 17 From: Elizabeth
To: City Clerk
Subject: Special City Council Meeting - Roadway Safety and Traffic Enforcement - Item 1 Date: Tuesday, August 25, 2026 1: 19: 51 PM
CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
Dear Mayor and City Councilmembers,
As a parent of two young children in EUSD, I’m deeply concerned about the fatalities and serious injuries occurring on our roads. Every person should be able to walk, bike, drive, and send their children out into our community without accepting preventable death or serious injury as an inevitable part of living here.
I’m asking the City Council to treat roadway safety with the urgency, priority, transparency, and accountability it deserves. That means moving beyond expressions of concern and ensuring there is a clear, coordinated strategy for reducing serious crashes and fatalities, with measurable goals, defined responsibility, transparent reporting, and meaningful followthrough.
When people are being killed on our roads, we need to know what is being done, who is responsible for doing it, whether those measures are working, and what will change when they are not.
Please make the prevention of traffic deaths and serious injuries a clear City priority and establish the accountability necessary to turn that commitment into measurable results.
Thank you for your service to our community and for taking this issue seriously.
Elizabeth Lawlor
From: shawn johnson
To: City Clerk
Subject: Special City Council Meeting- Roadway Safety and Traffic Enforcement- Item 1 Date: Tuesday, August 25, 2026 2: 07: 53 PM
CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
Hello Encinitas City Council,
I write as a citizen and constituent of Encinitas in support of Safe Streets which for me includes slowing drivers down and enforcing the law. Specifically, I am concerned with speeding occurring on Vulcan between the stop signs at Santa Fe and E Street. The speed limit south of Santa Fe is 25 on San Elijo yet increases to 30 north of Santa Fe on Vulcan. This is a residential street with many pedestrians crossing to use the walking path that runs parallel between the train tracks and Vulcan. Drivers all day every day are far exceeding the speed limit, driving up to 50 miles an hour along this stretch and using it as an alternative to Coast Highway. I pray each day that someone is not hit and to prevent a tragedy Iimplore you to lower the speed limit to 25 to be consistent with San Elijo, or consider adding speed bumps or additional cross walk/s, and enforcing the speed limits on our roads.
Thank you for your attention to the safety of our community, Shawn Johnson
951 Stratford Dr
949 533 1153
From: Mona Angel
To: City Clerk
Subject: Roadway Safety and Traffic Enforcement
Date: Tuesday, August 25, 2026 9: 10: 41 PM
CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
To members of the Encinitas City Council,
I am compelled to share my concerns about exiting my neighborhood.
IOceanCove want to specifically address the dangerous conditions at Manchester Avenue and Drive. The problem is not simply that some drivers are speeding. The problem isthat drivers exiting Ocean Cove Drive are dealing with limited visibility in both directions,( East and West) because of the blind curves on Manchester Avenue, while vehiclesapproaching from both directions can be traveling at high speeds.
Easton This creates a dangerous situation while exiting Ocean Cove Drive. A driver can lookManchester, look West on Manchester and have very little time to react to a vehiclecoming around both curves.
notsimply I am asking the city to treat this as an intersection design and sight- distance probleman enforcement problem.
There needs to be a formal traffic engineering study that measures actual sight distance forvehicles exiting Ocean Cove, actual vehicle speeds approaching from both directions, whether CoveDrive the existing roadway design provides adequate reaction time exiting Ocean onto Manchester.
People should not have to pull out of Ocean Cove Drive onto Manchester and gamble thatthey Pleasefix can see an approaching vehicle in time. This is a preventable safety problem. this problem before someone is seriously injured or killed. There are near misses daily.
Thank you
Mona Angel
2457 Caminito Ocean Cove
Cardiff- by-the-Sea
From: J C
To: City Clerk
Subject: Special City Council Meeting - Roadway Safety and Traffic Enforcement - Item 1 Date: Tuesday, August 25, 2026 10: 53: 58 PM
CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
Caution: The sender name ( J C) is different from their email address ( blancofelis8@gmail. com), which may indicate an impersonation attempt. Verify the email' s authenticity with the sender using your organization' s trusted contact list before replying or taking further action.
Secured by Avanan
Greetings Council and Mayor,
How many pedestrians and bicyclists have to be mowed down on our streets before effective measures
are set up to reduce the carnage?
Please stay on track to do the difficult work of finding methods to better enforce existing laws to cite and penalize those who
ignore speed limits, and who run red lights. Please make an effort to research how other cities manage to make
their cities safer for their citizens.
Respectfully,
J Cannon
From: Aaron Hebshi
To: City Clerk
Subject: Public comment for Agenda Item 1 - Special City Council Meeting - Roadway Safety and Traffic Enforcement August 26 2026
Date: Wednesday, August 26, 2026 9:32:42 AM
Attachments: Encinitas Bike Walk report on Recommended Guidance on Bicycle Safety Infrastructure. pdf CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
Dear Mayor and City Council,
Encinitas Bike Walk would like to submit the attached report to you and city staff as an analysis of the utility of bollards and protected bicycle infrastructure. While we value the expertise and training of city staff, we've witnessed staff recommendations hemmed in by nonprofessional direction from City Council, such as the removal of bollards from Leucadia Blvd and the recommendations to remove separation from Santa Fe Dr. Our organization, Encinitas Bike Walk, is composed of long-time advocates for bicycle safety, and a traffic engineer who is an expert in effective safe street design. Therefore, we feel qualified to add these recommendations with the hope that they will positively contribute to our community' s discussions on bicycle safety.
Respectfully,
Aaron Hebshi
Attachment 1 - EBW report)
From: Aaron Hebshi
To: City Clerk
Subject: Additional Public comment for Agenda Item 1 - Special City Council Meeting - Roadway Safety and Traffic Enforcement August 26 2026
Date: Wednesday, August 26, 2026 10:24:28 AM
Attachments: ECR Walk Audit_ FINAL compressed. pdf
CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
Dear Mayor and City Council,
I would like to submit the attached Walk Audit report produced by Strong Towns Encinitas in conjunction with Jesse Hanwit, President of Village Encinitas and multiple participants from our city. This Walk Audit was conducted in an area of Encinitas following the recent deaths of Emery Chelakian and Brodee Champlain- Kingman. This report provides suggested safety improvements for the audit area and an approach that prioritizes immediate action along with longer term safety improvements. We hope that the City of Encinitas maintains the focus on traffic safety in our town and pursues the objectives of Vision Zero.
Respectfully,
Aaron Hebshi
Attachment - ECR Walk Audit)
Recommended Guidance on Bicycle Safety Infrastructure
August 2026
Report compiled by Encinitas Bike Walk Steering Committee Members:
Judy Berlfein, Aaron Hebshi, Tom Bertulis, Chris Giles, Marley Jaoudi, and Travis Newhouse Corresponding Author: Aaron Hebshi - aaron.hebshi@gmail. com
BACKGROUND
Recent actions by the Encinitas City Council have been taken to remove separated bike infrastructure in favor of unseparated bike lanes. These include:
Re-designing the Santa Fe Dr streetscape to remove the eastbound parking-separated and grade-separated bike way and replace it with a buffered bike lane adjacent to the motor vehicle lane. (Encinitas City Council, Nov 5 2025) Eliminating the re-installation of flex posts (plastic bollards) used to visually separate the bike lanes on Leucadia Blvd following re-paving between I-5 and Saxony Ave. (Encinitas City Council, Dec 10 2025)
The reason stated for these decisions was that separated infrastructure is less safe than on-street buffered bike lanes. During the December 10 2025 council meeting, Councilmember Jim O’hara stated, "removing bollards is proven to be safer than having them there," and "all the traffic philosophy...all the traffic engineering will tell you that paint is the most effective.” However, what we know about facility design is that it is much more nuanced than "paint is the most effective". At the December 10th meeting, despite voting to remove the bollards, Mayor Bruce Ehlers argued to give discretion to the engineering department to place bollards where they make sense, such as where right-turning cars cheat into a bike lane well before the dashed green conflict striping begins. And Councilmember Marco San Antonio stated that he appreciated the "visual acuity" that the bollards provide. Councilmember Joy Lyndes stated that a broader conversation is needed on when it is appropriate and when it is ill-advised to use bollards.
Encinitas Bike Walk has assembled this document to:
1) Summarize published research on the safety and utilization rates of separated bike lanes 2) Summarize Federal, State, and County guidelines on the use of separated infrastructure 3) Identify and refute common talking points regarding the safety hazards of separated bicycle infrastructure
4) Discuss the Encinitas context for pragmatic use of separated infrastructure, including additional designs and technology available, with the goal of increasing safety and ridership.
1. Current published research on separated bike lane safety and ridership rates In 2023 the Federal Highway Administration published Crash Modifications Factors (CMF) for Separated Bike Lanes and found, “the use of flexible delineator posts consistently resulted in a decrease in total crashes.” The Full Report outlines the study of six separated bikeways (three with flexposts, and three with flexposts and other barriers) in five cities which found converting a Class II traditional bicycle lane to a Class IV separated bicycle lane (SBL) averaged a ~50% reduction in bicycle crashes.
Research in 2025 from the National Cooperative Highway Research Program (NCHRP) evaluated more bikeway upgrades in five additional cities and found “converting a painted bike lane to a separated bicycle lane led to a 90% reduction in crashes, while converting a buffered bike lane to a separated bike lane resulted in a 97% reduction in bicyclist crashes. The main study findings are summarized as follows”:
Installing separated bicycle lanes at sites with:
no bicycle facilities can potentially result in an over 98% reduction in bicyclist crashes. bicycle lanes can potentially lead to a 90% reduction in total crashes and a near 100% significant reduction in Fatal, Severe, Other Injury, or Complaint of Pain crashes. buffered bicycle lanes can result in an 84–97% significant reduction in total crashes and an 86% reduction in Fatal, Severe, Other Injury, or Complaint of Pain crashes.
Installing buffered bicycle lanes at sites with:
no bicycle facilities can potentially result in a 55–77% reduction in total crashes and a 52–77% significant reduction in Fatal, Severe, Other Injury, or Complaint of Pain crashes.
bicycle lanes can potentially reduce the total and Fatal, Severe, Other Injury, or Complaint of Pain bicyclist crashes by 5–25%.
Installing conventional bicycle lanes at sites with:
no facilities can result in a 38–68% reduction in total crashes and a 37–68% reduction in Fatal, Severe, Other Injury, or Complaint of Pain crashes.
Protected bike lanes not only reduce collisions. They also provide a more comfortable riding atmosphere, thereby encouraging more people to use bicycles for transportation and/or recreation. Recent national research of multiple cities that found "installed protected bicycle lanes experienced bicycle commuter increases 1.8 times larger than standard bicycle lanes." This article by People for Bikes provides a compendium of additional studies demonstrating the safety of protected bike lanes.
2. Federal, State, and San Diego guidelines for the use of separated infrastructure The following four graphics are drawn from Federal, State, and San Diego guidelines for appropriate infrastructure. The bottom line is that on roads with fast moving motor vehicle traffic, separation is beneficial.
3. Refutation of common talking points regarding the safety hazards of separated bicycle infrastructure
Below are common talking points used to negate the value of separated infrastructure, and an explanation of how these talking points have been misapplied and overgeneralized 1) Creates a collision hazard
Former Carlsbad Mobility Commissioner Steve Linke and current Encinitas Mobility and Traffic Safety Commissioner Ron Medak have presented on the crash rates of the South Coast Cycletrack. Their key finding is an increase in crash rate of 59% after construction of a cycle track between Chesterfield and the border between Encinitas and Solana Beach, compared to a control, including a fatality, Captain Ryan Curry, at the northbound entrance to the cycle track at the Solana Beach/Encinitas border. The design deficiency with the southbound South Coast Cycletrack is its narrowness and the fact that it is de-facto shared with pedestrians due to the narrow adjacent walkway. The key here is not that the cycletrack is inherently unsafe, it is that design aspects should be improved by widening the cycle track and providing adequate pedestrian facilities. Passing space for cyclists is important to have, when possible. Cyclists going 15mph or 20mph can easily pass slower cyclists in wide cycle tracks. The reason why the class IV cycle track between Chesterfield and K Street is more widely accepted is because there is sufficient width for the ped facility (aka the sidewalk) and the Class IV cycle track.
Between Chesterfield and Solana Beach, the facility becomes so narrow it becomes a shared facility for people biking and walking. And – as per the Caltrans definition – it therefore becomes a Class I facility, which are facilities that people walking and people biking share. Instead of a Class I cycle track, we need a Class IV cycle track and a wide sidewalk in Cardiff all the way to Solana Beach, and there are many ways to do that. And we can add a Class II bike lane as well.
Basically, we could copy the cross-section that we see north of Chesterfield. Ryan Currie’s death likewise does not establish that cycle tracks are dangerous. The protected bikeway begins abruptly partway down a hill near the Encinitas–Solana Beach boundary because it was not continued south to the signal in Solana Beach. Beginning a separated facility on a downhill segment created a sudden and unforgiving transition for bicyclists traveling at speed. This was a serious continuity and transition-design failure, not evidence against properly designed, continuous cycle tracks.
Under foggy conditions, cyclists need to slow down and be more vigilant, no matter the type of infrastructure around. There was an incident on the Cardiff cycle track not long ago where a cyclist hit a stop sign. Nobody called for the removal of that stop sign. That would be silly. It’s the same with flexposts. By using common sense and having plenty of cycle track width, cyclists avoid flexposts.
2) Separation locks bicyclists in and precludes escape from hazards, debris, flooding, etc.
Again, there is technology to help us here. In 2020 there was a proposal in Encinitas to purchase a “Compact Street Sweeper” as seen in this link. It never happened, but it would make sense that we move forward with that. And it shouldn’t hold us back from building cycle tracks as there are numerous ways to remove debris from cycle tracks. 3) Increases risk of “right hooks” at intersections and driveways The argument here is that vehicles, separated from the bicycle lane, up until an intersection or driveway, will be forced to make a sudden right turn across the bicycle lane. Bollards, the argument goes, keep motor vehicle traffic from merging safely with bicycle traffic in advance of the intersection. At the December 10th 2025 City Council meeting, Steve Linke and Ron Medak presented information on a 2020 cyclist mortality that occurred on westbound Leucadia Blvd from a PODS truck that turned right across the flex-post separated bike lane at the intersection of Moonstone Court. They stated that the crash report and subsequent lawsuit and settlement strongly implicated the bollards as a source of the collision. The principal safety problem was the design of the intersection: an unnecessarily wide entrance to a small cul-de-sac encouraged a faster right turn and made it more difficult for the driver to see and yield to a bicyclist. A narrow, properly designed entrance would have slowed the truck and improved visibility.
Removing the cycle track would not, by itself, have eliminated the underlying turning conflict.
Encinitas has deployed the use of green conflict striping leading up to intersections, indicating merge zones. We are unaware of any formal analysis looking at the effectiveness of the merge zones, but anecdotally we have observed how motorists inconsistently handle these merge zones, sometimes merging across the bike lane and sometimes remaining to the left of the bike lane until the intersection.
Alternatively, Caltrans formally endorsed NACTO design guides in April 2014 which states: “Do not drop the bikeway or allow cars into the bike lane before the intersection. Continue bikeway buffers and protection all the way to the intersection,” as seen at this link. 4) Prevents cyclists from positioning themselves visibly (and visually relevantly) at driveway cuts and intersections.
Again, this is thinking from the 20th century. We’re now in the 21st century and times have changed. I wish it was as simple as removing flexposts and motorists would magically keep to the right as they make a right turn. Then we’d see right hooks across the country disappear. But motorists all too often don't do that, despite the law requiring them to do that. Alas, the vast majority of right hooks at intersections occur when there are no flexposts. Instead of victim-blaming, it would be better to make motorists, by design, drive in a safe manner. And the easiest way to do that is with Continuous Sidewalks, as you can see from this article and this video.
It’s actually a mistake - and a safety hazard - to remove bollards and flexposts as you approach an intersection or driveway. It won't magically reduce fatalities. In fact, removing bollards and flexposts will only lead to more fatalities. NACTO’s Don’t Give Up At The Intersection guidance explains that more flexposts, a tighter opening, raised crossings, and several other tech improvements benefit safety outcomes.
4. Appropriate use of safety infrastructure in Encinitas It's important to recognize the importance of site-specific design for Class IV separated bikeways and research has determined that the odds ratio of cyclist collisions and falls increase with the presence of certain factors in or along the bikeway, these should inform this bikeway facility type’s implementation. They include, but are certainly not limited to, adjacent land uses, potential conflicts and connections with intersecting bikeways and roadways, presence of parking and its utilization, bikeway separation barrier types and visibility, the existing and estimated volume of users and their design user profiles.
For example, one of our most dangerous sites is the El Camino Real and Encinitas Blvd intersection. The image below shows injury crashes from 2012-2023 (SANDA Traffic Safety Dashboard). This is the convergence of two high-speed, high-volume arterials traversing a complicated urban landscape with many driveways and shopping areas, and proximal to one of our two high schools, our two middle schools, and two of our elementary schools. In such an instance, it may be most appropriate to use multiple street-design techniques to slow motor vehicle traffic, rather than simply installing a protected bikeway.
Example corridors are highlighted here in which bollards and other separation treatments may be beneficial.
Leucadia Blvd uphill eastbound from Hymettus to Orpheus. Motor vehicles waiting at the eastbound stoplight at Leucadia and Orpheus will often queue in the bike lane, preventing bicycle passage. Flexposts to delineate the bikelane would be helpful here in precluding the space’s use by motor vehicles, while providing little danger to bicyclists due to their slow uphill speed.
Encinitas Blvd uphill eastbound from Saxony to Balour. Fast motor vehicle speeds and slow-paced bicyclists make for large speed differentials and reduced reaction time when a motor vehicle driver first observes an upcoming bicyclist in the adjacent bike lane. Visual or physical barriers would enhance safety and comfort here. Moreover, motor vehicle traffic uses the bike lane as a queue for St Johns school pickup and dropoff. Physical separation would maintain the bike lane’s utility as a bike lane.
La Costa ave east of Interstate 5. This corridor in Carlsbad is now famous for a brutal bicycle crash caught on camera where a cyclist crashed into motor vehicles pulled over in the bicycle lane and was thrown into traffic. High speed traffic corridors, as mentioned in part 2 of this paper, are best served with separated, protected bikeways.
North Coast Hwy 101 Streetscape. The image below depicts a vehicle idling in the southbound bike lane on the 500 block of North Coast Hwy 101 in Leucadia. This scenario is commonplace at peak meal times when on-street parking spots available in front of the dining establishments fill up. Idling vehicles force bicycle riders unpredictably into the street or on the sidewalk, causing potential interactions with motor vehicles or pedestrians. This is a scenario in which better curb management, through demarcated loading/unloading zones, time limits, or metered parking, or even bike lane cameras to ticket violators, would probably be a preferred treatment than adding protection/separation. Protection/separation would eliminate the ability of the curb to support parking, and reconfiguring the arrangement to place the parking between the bike lane and motor vehicle traffic would be expensive due to the installed hardscape planters and bulbouts. Moreover, speeds along this stretch of North Coast Hwy 101 are fairly low due to the constricted road and approaching roundabout, decreasing the risk of interactions resulting in a severe crash.
Summary
Substantial evidence exists demonstrating the utility of protected bicycle infrastructure in enhancing safety and ridership in certain contexts. Abandoning the appropriate use of protected infrastructure because of anecdotal crashes or known design flaws is an inappropriate approach to safe street design. Rather, we recommend analyzing and adjusting specific design flaws while maintaining the effective components of protected infrastructure, particularly adjacent to high speed motor vehicle traffic and in strategic locations to exclude illegally parked or idling motor vehicles. In addition, initiating a new street design with inexpensive and removable features (such as paint and flex posts rather than concrete and steel bollards or guardrails), in order to tweak the treatment if necessary, can ultimately identify safe solutions at lower costs, and can be iterated upon when these treatments are shown to be effective in particular street contexts. Effective treatments can then be hardened when budgets allow and/or during street repaving cycles.
Strong Towns Encinitas
Date:
Saturday, March 21, 2026
Location:
El Camino Real Corridor, Encinitas, CA
El Camino Real
Walk Audit
This report reflects a community-led effort to evaluate the El Camino Real corridor through the Strong Towns lens,
which prioritizes people’s safety and mobility over vehicle throughput. Because street design affects long-term fiscal health and quality of life, this audit identifies opportunities to turn a car-centered corridor into a more productive, walkable community.
Built for automobiles, El Camino Real has expansive
parking lots, narrow sidewalks, and fast-moving traffic.
This walk audit identifies small, low-cost design changes that can quickly improve pedestrian safety, comfort, and access from nearby neighborhoods.
Introduction: Prioritizing
Safety, Mobility, and Fiscal
Health
The Strong Towns Approach
to Walkability
The principles guiding this audit focus on creating a
financially resilient and human- scaled Encinitas where:
Safety is prioritized over vehicle speed.
Walkability drives value, as multimodal, mixed-use areas have historically generated higher per-acre tax revenue than car-dependent corridors.
Low-cost, incremental changes are more effective for
testing and iterating on safety improvements than
expensive, large-scale capital outlays.
About This Walk Audit
On March 21, 2026, Strong Towns Encinitas, in
partnership with members of the Village
Encinitas, AARP Livable Communities, and
Strong Towns Oceanside led 20 residents on a
hands-on assessment of two key routes along
Encinitas Boulevard and El Camino Real to
identify barriers to pedestrian safety and
accessibility.
Routes Evaluated
Route 1
Village Square Dr
unofficially named
Emery’s Crossing”) to
Mountain Vista Dr.
Routes Evaluated
Route 2
Village Square Dr to
Santa Fe Dr.
General Issues
Lack of shade
Narrow sidewalks
Lack of Leading Pedestrian
Interval (LPI) pedestrian signals
Long pedestrian crossing
distances
Adjacent traffic traveling at high
speeds, separated from the
sidewalk by only narrow painted
buffers ( bike lanes)
Key Findings
The crosswalk
paint is faded and
asphalt are in
disrepair.
A 20-second pedestrian
signal at Encinitas
Boulevard is insufficient
for people with mobility
issues.
Village Square Dr. & Encinitas Blvd.
The sidewalk is narrow,
only wide enough for 2
people to walk side by
side.
The asphalt in the bike lane
is of poor quality, which
may encorage cyclists to
ride on the sidewalk.
Bike lanes are narrow and
poorly defined, providing
only a minor physical
buffer between
fast-moving vehicles and
pedestrians on the
sidewalk.
Encinitas Blvd. between Village Square Dr. and El Camino Real A high number of
driveways and visual
distractions contribute to
right-hook conflicts.
The Jack in the Box sign
blocks visibility between
the sidewalk and cars
exiting the driveway.
Encinitas Blvd. between Village Square Dr. and El Camino Real WB Encinitas Blvd motorists
turning right (north) on El Camino
Real queue up behind bikes going
straight.
Lane configuration frustrates
motorists and jeopardizes cyclists'
safety.
The intersection is expansive, and
crossing distances vary from 100 to
120 feet, making pedestrian crossing
challenging, especially for older adults
and people with mobility issues.
Intersection of Encinitas Blvd and El Camino Real
The bus stop on the
southbound ECR near
Kohl's entrance has no
bench, no shade, and no
protection from the road.
The bus stop on the
northbound ECR has a
trash can positioned in
the middle of the sidewalk
near the bus stop south
of Via Molina.
The telephone pole is in
the middle of the
sidewalk, but the paved
area is widened to (likely)
accommodate mobility
devices.
Bus Stops
El Camino Real between Encinitas Blvd and Via Molena
Signal timing at the staggered crossing is unclear, and the signal transitions from countdown to a blinking hand early, leaving pedestrians unsure whether they have enough time to cross both segments.
Z-Crossing near Encinitas Village Square
El Camino Real between Encinitas Blvd and Via Molena
No crosswalk or warning for drivers exiting the
parking lot that pedestrians may be crossing at the
entrance. There’s a stop line, but it's so far back they can’t see traffic, so they don’t stop there.
Northwest Entrance To Encinitas Village Square Parking Lot El Camino Real between Encinitas Blvd and Via Molena
There is no crosswalk on
the south side of the
intersection.
Leading Pedestrian
Indicators (LPI) would be
helpful in all directions
here.
El Camino Real and Mountain Vista Dr.
The sidewalk is
taken over by
weeds in places
Cars are blindly turning out
of businesses and entering
the bike lane
El Camino Real South of Encinitas Blvd
LA Fitness Shopping Center West Entrance
2-minute wait for the pedestrian signal across El
Camino Real
Cyclists crossing east need to cross to the opposite
side of Crest Dr.to access the crosswalk and pedestrian signal
No walk signal across Crest Dr. on the west side of El
Camino Real
El Camino Real South of Encinitas Blvd
Crest & El Camino Real
Downhill traffic on the SB El
Camino Real turn lane to WB
Santa Fe Dr. is traveling fast
with a continued downhill
slope across the crosswalk
A large height differential between the
sidewalk and the turning lane at Santa
Fe Dr. limits motorists' visibility of
children and other small pedestrians.
El Camino Real South of Encinitas Blvd
El Camino Real and Santa Fe Dr.
Significant portions of the asphalt
and paint are in disrepair
There is a steep transition between the
sidewalk and the curb ramps
El Camino Real South of Encinitas Blvd
Driveway at Sprouts Shopping Center
Opportunities
Next Steps
Immediate Actions
Long-Term &
Community
Opportunities
Short-Term
Recommendations
Opportunities & Next
Steps
Maintenance: Clear weeds from sidewalks and repair
hazardous, uneven pavement.
Visibility: Repaint faded crosswalks at intersections, paint crosswalks at unmarked crossings and install "Slow" or
pedestrian-warning signage at shopping center entrances.
Signal Adjustments: Implement LPIs at all major
intersections and extend pedestrian crossing times to
accommodate older adults.
Immediate Actions (0–3 months)
Opportunities & Next
Steps
No Right-Turn-On-Red: Install a no-right-on-red restriction at Encinitas Blvd and El Camino Real intersection.
Automated Encforcement: Install red-light, speed, and bike-lane cameras to ticket reckless drivers.
Warning Speed Bumps: The city can work with property managers to install speed tables, raised crosswalks, or indicator rumble strips at driveways where pedestrians cross.
Short-Term Recommendations (
3–12 months)
Opportunities & Next
Steps
Lighting: Install or reposition street lighting at high-conflict driveway zones and intersections to improve nighttime visibility.
Traffic Calming: Install planters or attractive barriers along the outside of sidewalks to provide a physical buffer from traffic.
Transit Improvements: Coordinate with NCTD to install benches and shade structures at bus stops.
Lane Width Reduction: Narrow lanes through re-striping to reduce speeds, and use the extra space to buffer the bike lane or add bollards for additional rider separation.
Short-Term Recommendations Cont.(
3–12 months)
Opportunities & Next
Steps
Right-of-Way Re-allocation: Reclaim a lane of traffic or portions of the vegetated right-of-way to widen sidewalks and bike lanes.
Public-Private Partnerships: Work with shopping center
owners to construct dedicated walkway and cycleway
connections through parking lots, separating pedestrians from the main road.
Redevelopment: Encourage the transformation of parking lots into mixed-use spaces to reduce the need for long-distance car trips.
Long-Term & Community Opportunities (
1–3 years)
Conclusion
This walk audit serves as a roadmap for
transforming the El Camino Real corridor into a
safer, more financially productive environment. It
also aligns and supports the efforts of the
Encinitas Age-Friendly Initiative. We hope our
findings and recommendations are taken into
consideration by the City of Encinitas, City Staff and
City Council.
We are providing this report to improve the safety
and livelihoods of all Encinitas residents, especially
those who are unable to voice their concerns
formally or have historically been disenfranchised,
such as children, people with disabilities, and older
adults.
Chris Giles
Co-founder, Strong Towns Encinitas LC
Aaron Hebshi
Co-founder, Strong Towns Encinitas LC
Jesse Hanwit
President, Village Encinitas
Former Encinitas Senior Commission member
Authors
We’d like to thank the following community members for
participating in the walk audit and sharing their
observations for inclusion in this report:
Anne Larocca, Jennifer Moore, Jim Clancy, Celeste Clancy, Tom Golaszewski, Denny Cope, Charley Murtagh, Jill Levan, Russell Levan, Srividya Chandrashekhar, Brad Lefkowitz, Suzanne Stone, Judy Berlfein, Amy Stuck, and Arielle Golden.
We’d also like to thank Encinitas District 2 City Council Representative Jim O’Hara and Mobility & Traffic Safety Commission (MTSC) member Leigh Ratcliffe for walking with the community and discussing observations and potential solutions in real time.
Lastly, a special thanks to Israel Hernandez, Associate Director, Advocacy & Community Engagement, AARP
California for guidance on this project.
Acknowledgments
From: joan bockman
To: City Clerk
Subject: Item 1 Roadway Safety and Traffic Enforcement in the City of Encinitas.
Date: Wednesday, August 26, 2026 10: 09: 38 AM
CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
This is the first time you received an email from this sender ( joanbockman@sbcglobal. net). Exercise caution when clicking links, opening attachments or taking further action, before validating its authenticity.
Secured by Avanan
Dear City Council,
Please make a rule that every single thing done to improve the "life" of cars must be considered to be a detriment to pedestrians and bikes. Examples are sidewalks along roads rather than being set back with street tree strips, wider lanes ( harder to cross on foot), higher speeds, more lanes ( harder to cross on foot), lack of street trees that could stop a car before it hits the sidewalk, parking spaces ( encourage car use while making it dangerous for pedestrians to get to the front of a store or across a vast paved surface). Every single thing done for cars is something against people. It is not about enforcement, cameras or lights. It is pure infrastructure. Streets are either designed FOR people OR for cars. Spaces are either FOR people OR cars. You need to pick people. Our region is too dense to continue coddling cars. It is time to put cars back where they belong and put people FIRST.
Regards,
Joan Bockman
in memory of Emery
From: LESLIE ENGEL
To: City Clerk
Cc: LESLIE ENGEL
Subject: Special Meeting 8/ 26/ 26/ Roadway Safety Item # 1 Date: Wednesday, August 26, 2026 10: 32: 55 AM
CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
To the council members,
First, I am writing to beg council to put together a task force to make our streets safe for pedestrians immediately! This task force would include citizens, Fire, Sheriff, etc. to create a plan similar to that of the traffic calming measures implemented in Hoboken NJ.
Second, the dangerous conditions on the section of Manchester Avenue WEST of I5:
This problem affects ~ 500 residents of MacKinnon Ranch and Cardiff Cove.
Because of excessive speeds by motorists even after the MPH was reduced and we had speed feedback signs installed, this issue gets worse. The blind curve to the west and the intersection to the east near the entry/exit of both communities, makes it extremely dangerous for residents to exit our communities safely.
In Cardiff Cove there is limited visibility in BOTH directions because of the blind curve to our west and the close proximity to the overpass/intersection to our east. It is nearly impossible to exit because cars travel at high speeds in both directions at the same time.
Residents leaving Ocean Cove Drive have very little time to react to vehicles careening around both blind curves.
We are asking that this be another of the focused areas by the city to treat this as a priority and to authorize a formal traffic engineering study.
Cyclists, and drivers should be able to exit safely.
This is a preventable safety issue. Please agree to fix this before someone is seriously injured or killed!
Sincerely
Leslie Engel
From: Susan Lesan
To: City Clerk; Marco San Antonio
Cc: Brad Lefkowits
Subject: Special City Council Meeting - Roadway Safety and Traffic Enforcement - Item 1 Date: Wednesday, August 26, 2026 11:18:52 AM
CAUTION: External Email. Do not click any links or open attachments unless you recognize the sender, verified their email address, and know the content is safe.
To Mayor Ehlers, Councilmember San Antonio, and the City Council:
I am very concerned about the safety of our streets in Encinitas, for school kids, pedestrians, bicyclists, e-bikers and drivers.
My biggest area of concern is El Camino from Manchester to Encinitas Blvd, where I live.
People pay no attention to speed limits and regularly run red lights. Neighborhood kids have started back to school at Ocean Knoll, Oak Crest and SDA. They’re crossing this nightmare corridor and it’s just a matter of time before another tragedy occurs.
We also can’t ignore our aging population, including my husband and me. We’re at greater risk from this behavior.
I was driving through San Marcos this week and noticed something Encinitas doesn’t have:
speed limits painted on the street. Our speed limit signs in Encinitas are not effective: there are too few, some are not visible, and changing numbers means drivers are not sure how fast they can go.
Please consider painting speed limits on the street at intersections and at speed zone areas on El Camino. Thank you.
Sincerely,
Susan Lesan, D-4