How to use local crash data to choose community safety priorities
How to turn crash, injury and traffic data into a ranked list of local safety priorities
Picture a neighborhood safety group with three requests. Parents want help on a dark road by the middle school. Shop owners want an end to weekly fender benders at a busy signal. Others worry about speeding on a curve at the edge of town. The group can push hard on only one or two.
Data won’t make the choice for you, but it keeps the choice honest. In 2024, 39,254 people died on U.S. roads and about 2.42 million were injured, the National Highway Traffic Safety Administration (NHTSA) reports [1]. A good priority list aims your time at the places and behaviors behind the worst harm.
This guide shows how a crash data analyst builds that list, ending with a scoring worksheet and a worked example. New to organizing? Start with our guide to building a community traffic safety program.
Start with the right goal: fewer deaths and serious injuries
Police rate each person’s injury on a 5 level scale called KABCO [2]. K means the person died within 30 days. A means a suspected serious injury, such as a broken leg, a crush injury or being unconscious when taken from the scene [2]. B is a suspected minor injury, C a possible injury and O no apparent injury.
Build your list around K and A crashes, which analysts call KA crashes. That fits the Safe System idea that deaths and serious injuries are the failures that matter most. A Safe Streets and Roads for All (SS4A) action plan also starts from a baseline of fatal and serious injury crashes [3].
Total crash counts can mislead. A busy corner with 60 fender benders can outrank a dark road where 3 people on foot were badly hurt.
Two cautions apply. An officer judges an A injury at the scene. The Federal Highway Administration (FHWA) calls the KABCO scale notably inaccurate compared with medical injury scales [2]. And KA numbers are small, so they bounce from year to year. Use the 5 most recent complete years. Leave out the current year, which is still filling in.
Build a data inventory: what to gather and who to ask
NHTSA groups state traffic records into 6 systems: crash, driver, vehicle, roadway, citation and adjudication, and injury surveillance [4]. You need pieces of several, plus local sources.
Comparison compiled by Safety Behind the Wheel Foundation
| Data source | What it shows | Blind spots | Who to ask |
|---|---|---|---|
| Police crash reports (state or city dashboard) | Where crashes happened, how severe, who was involved, crash type | Unreported crashes; severity judged at the scene; some crashes mapped in the wrong spot | State highway safety office or DOT crash data unit; city or county traffic engineer |
| Fatal crash records | Every fatal crash, coded the same way nationwide | Small local numbers | NHTSA’s free online tools, such as STSI |
| Injury data (EMS, emergency department, hospital, trauma registry, death records) | Injuries judged by clinicians; crashes police never saw | Location often vague; privacy limits; usually summary tables only | State health department injury program; trauma center; EMS agency |
| Citation data | Where police wrote tickets, and for what | Shows enforcement effort, not risk | Police department; court records office |
| School route information | Where children walk, bike and wait for buses | Few numbers; depends on who joins walk audits | School district transportation office; parent groups; Safe Routes to School coordinator |
| Maintenance records | Streetlight outages, damaged signs, faded markings, sidewalk gaps | Shows conditions, not outcomes | Public works; the utility that owns the streetlights |
| Community reports | 311 requests, surveys, walk audits, near miss reports | Skewed toward people with the time and trust to report | 311 office; council offices; your own survey |
| Exposure data | Traffic counts, population, school enrollment, transit stops, walking counts | Counts may be old or estimated | State DOT traffic count map; regional planning agency; school district; transit agency |
Drive With Care’s data inventory. Record each file’s source, pull date and years covered.
What each source can and cannot tell you
Crash dashboards: your backbone
Start with your state or city crash dashboard. Our guide to local crash data and safety maps lists many. NHTSA’s State Traffic Safety Information (STSI) tool adds 10 years of state and county data on belt use, impaired driving and speeding [5]. Filter to K and A crashes first, then by road user, time and crash type.
Injury data: filling the gaps
NHTSA estimated that in 2019, 32% of injury crashes and about 60% of damage only crashes were never reported to police [6]. Ambulance, emergency department, hospital and trauma records catch some of those people. NHTSA’s traffic records program lists the National EMS Information System (NEMSIS) as a related national data system [4].
Some states link these records. Maryland’s Crash Outcome Data Evaluation System (CODES) joins crash, hospital, driver license, citation and vehicle data to guide policy [7]. For a quick start, try WISQARS, a free tool from the Centers for Disease Control and Prevention (CDC). It covers fatal and nonfatal injuries, including motor vehicle crashes, and lets you compare states [8].
Medical data show who gets hurt and how badly, such as older people on foot, but are weak at pinpointing locations. Health care providers generally must protect patient information under the federal HIPAA Privacy Rule [9], so expect summary tables.
Citations: a measure of effort, not danger
A state citation system tracks tickets, arrests and court outcomes [4]. A map of speeding tickets shows where officers patrolled. A street with few tickets may simply get little patrol. Treat citations as a record of enforcement, not risk. For speed problems, ask for a speed study.
School routes: risk that crash data can hide
Crash maps can look clean near schools because few children walk there. FHWA’s Safe Routes to School page, last updated in 2016, says about half of students walked or biked to school in 1969. Now fewer than 15% of school trips are made that way. Parents often cite traffic danger [10]. An empty map may mean the risk keeps kids off the street.
A walk audit fills that gap. NHTSA’s Walkability Checklist helps families and groups rate room to walk, crossings, driver behavior and more [11]. FHWA points to the National Center for Safe Routes to School as a hub for program ideas [10]. Ask parents to mark where they feel unsafe, then compare those marks with your KA map.
Maintenance records: conditions you can fix fast
Streetlight outage logs, sign damage reports and work orders show conditions that change quickly. Lay them over your nighttime KA crashes. FHWA credits lighting with a 42% drop in nighttime pedestrian injury crashes at intersections [12]. So repeated outages near those crashes make an easy first request.
Community reports: useful, but biased
311 requests, surveys and near miss reports capture problems before anyone gets hurt. But they mirror who has the time, internet access and trust to report. Use them for context, check them against crash and exposure data, and never rank streets by complaint volume.
Exposure: the bottom number in every rate
State transportation departments publish annual average daily traffic (AADT), an estimate of vehicles passing a point on a typical day. FHWA’s Traffic Monitoring Guide describes methods for counting vehicles and people biking and walking [13]. Add population, school enrollment and transit stops to see where people walk and wait.
Counts, rates and severity: read all three
A count shows how much harm happened, while a rate shows how risky travel is there. Our guide to understanding traffic safety data explains rates in depth. For ranking, analysts use:
- Road segments: crashes per million vehicle miles. Multiply daily traffic by road length, by 365 and by the number of years.
- Intersections: crashes per million vehicles entering.
- Severity: count KA crashes separately and figure a KA rate too.
Severity weighting. Many agencies rank sites by equivalent property damage only (EPDO), which weights each crash by its cost compared with a damage only crash [2]. In Highway Safety Manual weights updated to 2016 dollars, a fatal crash counts 568 and a serious injury crash 30. Minor and possible injuries count 11 and 6 [2].
That lets a single death dominate the score, though whether a crash turns fatal often involves chance. FHWA’s crash cost guide notes that states usually blend severity costs to soften the effect of very high fatal costs [2]. A simple fix is to give K and A crashes the same weight.
The low volume trap. A quiet road with 300 vehicles a day and 1 crash can post the city’s highest rate. Require a minimum, such as 2 KA crashes, or compare against similar roads.
People on foot. Vehicle traffic does not measure how much people walk. Use walking counts where you have them.
Finding high risk corridors and behaviors
Three ways to find risky places
- Hot spots. A place with repeated severe crashes, such as an intersection with 3 left turn KA crashes, deserves a close look.
- High injury network. An SS4A action plan must map higher risk locations as a high injury network or an equivalent, ideally covering every road no matter who owns it [3].
- Systemic analysis. FHWA notes that severe crashes tend not to cluster, yet the risk factors behind them are strikingly consistent [14]. It recommends low to moderate cost fixes where risk is highest, using crash, roadway, traffic, land use, socioeconomic and demographic data [14].
Suppose 6 of your 8 pedestrian KA crashes over 5 years happened on unlit four lane roads near bus stops. Every stretch like that becomes a candidate, even one with no crash yet. Severe crashes also shift location from year to year, FHWA notes, which limits what a hot spot list can catch [14].
Behaviors
Compare your area’s KA crashes with state figures for speeding, impairment, unbelted occupants, young drivers and motorcyclists. Nationally in 2024, speeding was a factor in 29% of deaths and alcohol impaired driving in 30% [1]. Of passenger vehicle occupants killed with known belt use, 48% were unbelted [1]. Treat distraction and drowsiness figures as minimums. Behavior patterns point to partners such as police, schools and employers.
Avoid misleading comparisons when you rank
- Different lengths. A 4 mile road collects more crashes than a 2 block street. Compare per mile or per vehicle mile.
- Different road types. Compare main roads with main roads and side streets with side streets.
- Different time windows. Use the same years everywhere, and drop partial years.
- Different sources. Don’t chart police KA counts and hospital counts together, because they count different things.
- Tickets and complaints. More citations or 311 calls can reflect patrols and reporting habits, not danger.
- Last year’s spike. A place with an unusually bad year often improves on its own, which statisticians call regression to the mean.
Add an equity check
Map KA crashes alongside where people walk, bike or ride transit, and where fewer households own cars. FHWA lists socioeconomic and demographic data among systemic analysis inputs [14]. SS4A’s current requirements, updated March 2026, name demographics as a possible factor and call for engagement that allows community representation [3]. Neighborhoods that file few 311 requests may still carry the most severe crashes.
Drive With Care's priority scoring worksheet
This is Drive With Care’s suggestion, not a federal method. Score each candidate 0 to 3 per factor, multiply by the weight and add. Severity carries the most weight on purpose.
Comparison compiled by Safety Behind the Wheel Foundation
| Factor | What to measure | Score 0 | Score 3 | Weight |
|---|---|---|---|---|
| Severe harm | KA crashes in the last 5 complete years | None | 4 or more | ×3 |
| Severe crash rate | KA rate compared with similar roads | Well below average | More than twice average | ×2 |
| Vulnerable road users | Share of KA crashes involving people walking, biking or riding motorcycles, or children and older adults | None | Most | ×2 |
| Risk features | Features such as no lighting, no sidewalk, multiple lanes at higher speeds, sharp curves, bus stops without crossings | None | 3 or more | ×1 |
| Community concern | Walk audits, school concerns, surveys, near miss reports | None | Strong and repeated | ×1 |
| Readiness | Road owner engaged, low cost fix available, project already planned | None yet | All three | ×1 |
Maximum score: 30. Scores of 1 and 2 fall between the endpoints. Use the total to start a discussion, not to end one.
Worked example: two corridors
This hypothetical example uses the same 5 years, 2021 through 2025, for both roads.
Comparison compiled by Safety Behind the Wheel Foundation
| Measure | Main Street | Mill Road |
|---|---|---|
| Road | 1.2 mile, four lane shopping street | 1.5 mile, two lane road past a middle school |
| Average daily traffic | 24,000 vehicles | 7,000 vehicles |
| Vehicle miles over 5 years | 52.6 million | 19.2 million |
| All crashes | 150 | 30 |
| All crashes per million vehicle miles | 2.9 | 1.6 |
| KA crashes | 4 (1 K, 3 A) | 4 (all A) |
| KA crashes per 100 million vehicle miles | 7.6 | 20.9 |
| EPDO score (2016 weights) | 1,124 | 231 |
| Who was badly hurt | 3 drivers and 1 motorcyclist, in left turn or red light crashes at 2 signals | 3 people walking after dark and 1 driver who ran off the road |
| Conditions noted | Left turns across traffic without a turn arrow | No sidewalk on one side; 4 streetlight outage reports; bus stops without crossings |
Drive With Care’s calculations. EPDO crash mix: Main Street 1 K, 3 A, 18 B, 28 C, 100 O. Mill Road 4 A, 6 B, 5 C, 15 O.
Total crashes, the overall rate and EPDO all point to Main Street. Yet its single fatal crash supplies 568 of its 1,124 EPDO points. The severe crash rate points the other way. Mill Road’s KA rate is nearly 3 times Main Street’s. Of its 4 badly hurt people, 3 were on foot.
Now apply the worksheet, assuming similar roads here average about 10 KA crashes per 100 million vehicle miles. Mill Road scores 29 (9, 6, 6, 3, 3, 2). Main Street scores 20 (9, 2, 2, 2, 2, 3), held back by a near average rate and fewer vulnerable victims.
An analyst would keep both on the list, with different tools. Mill Road leads with a systemic package: restore the lights now, then seek crossings, a sidewalk and a safety review. Main Street needs a spot fix at 2 signals, such as protected left turn arrows.
Protect privacy when you share findings
- Publish counts by corridor, intersection or block and by year. Leave out names, plate numbers, home addresses and crash report copies.
- When a number is tiny, such as 1 death on one block, combine years or areas so no one can be identified.
- Use health data only in the summary form an agency provides. Follow any data use agreement.
- Ask families before sharing a victim’s story, and let them decide how it is told.
Turn the list into action
Share your draft ranking with the road owner’s traffic engineer before going public. They may know of planned projects or data errors. For your top corridor, ask for a road safety audit (RSA), a formal review by an independent team that considers all road users [15]. FHWA lists RSAs among its proven safety countermeasures, with 10% to 60% fewer total crashes [12]. Our Safe System guide covers other proven fixes. Rerun the analysis yearly the same way.
Frequently asked questions
What is the difference between crash frequency and crash rate?
Frequency is the number of crashes at a place over a set period. A rate divides that number by exposure, such as vehicle miles traveled. Frequency shows where harm piles up. A rate shows where each trip carries more risk.
Why focus on serious injuries instead of all crashes?
Minor crashes far outnumber severe ones, so totals steer attention to busy places rather than dangerous ones. Ranking by fatal and serious injury crashes aims effort at life changing harm.
Can a community group get hospital or EMS data?
Usually not record by record, because federal privacy rules protect patient information [9]. Ask your state health department’s injury program for summary tables by county, age or road user. Some states link medical and crash records [7].
Why won’t our city share its list of dangerous intersections?
Agencies may worry about lawsuits. Federal law keeps some safety data compiled for federal programs out of court cases. But FHWA notes it does not shield them from public records requests [15]. Ask what the agency can share, such as KA counts by corridor.
What is a high injury network?
It is a map of the streets where a large share of a community’s deaths and serious injuries happen. SS4A action plans must include one or an equivalent [3]. If your city has one, start there, then add systemic and school route concerns.
The bottom line
Recommendation from Safety Behind the Wheel Foundation
Rank places by deaths and serious injuries over 5 complete years. Check every count against a rate and similar roads. Look for the risky features behind severe crashes, not only past hot spots. Then share findings carefully with the people who own the road.
Related articles
- How to build a community traffic safety program that lasts
- Local crash data and safety maps: how to find and read your area’s data
- Understanding traffic safety data: how to read crash statistics wisely
- Safe System approach: how road design protects you from human mistakes
- Measuring traffic safety campaign results: how to know it worked
Sources and further reading
- 1National Highway Traffic Safety Administration (NHTSA). Overview of Motor Vehicle Traffic Crashes in 2024. Traffic Safety Facts Research Note, DOT HS 813 791. April 2026. crashstats.nhtsa.dot.gov/…/813791
- 2Federal Highway Administration (FHWA). Crash Costs for Highway Safety Analysis. FHWA-SA-17-071. January 2018. highways.dot.gov/…/fhwasa17071.pdf
- 3U.S. Department of Transportation (USDOT). Safe Streets and Roads for All (SS4A): Comprehensive Safety Action Plan Requirements. Updated March 27, 2026. transportation.gov/grants/ss4a/action-plan-requirements
- 4National Highway Traffic Safety Administration (NHTSA). Traffic Records program page (accessed October 2026), and Traffic Records Program Assessment Advisory, 2018 Edition, DOT HS 812 601, August 2018. nhtsa.gov/data/traffic-records and crashstats.nhtsa.dot.gov/…/812601
- 5National Highway Traffic Safety Administration (NHTSA). Crash Data Analysis (CDAN) tools, including State Traffic Safety Information (STSI). Accessed October 2026. cdan.dot.gov and cdan.dot.gov/STSI
- 6Blincoe L, Miller T, Wang J-S, et al. The Economic and Societal Impact of Motor Vehicle Crashes, 2019 (Revised). NHTSA, DOT HS 813 403. February 2023. crashstats.nhtsa.dot.gov/…/813403
- 7National Highway Traffic Safety Administration (NHTSA). Data Integration: Linking It All Together. DOT HS 812 784. July 2019. crashstats.nhtsa.dot.gov/…/812784
- 8Centers for Disease Control and Prevention (CDC). WISQARS (Web-based Injury Statistics Query and Reporting System). Accessed October 2026. wisqars.cdc.gov
- 9Centers for Disease Control and Prevention (CDC), Public Health Law Program. Health Insurance Portability and Accountability Act of 1996 (HIPAA). Last reviewed September 10, 2024. cdc.gov/phlp/…/hipaa
- 10Federal Highway Administration (FHWA). Safe Routes to School. Last updated November 7, 2016. fhwa.dot.gov/environment/safe_routes_to_school
- 11National Highway Traffic Safety Administration (NHTSA). Walkability Checklist Guide. Accessed October 2026. nhtsa.gov/pedestrian-safety/walkability-checklist
- 12Federal Highway Administration (FHWA). Proven Safety Countermeasures: Lighting (FHWA-SA-21-050) and Road Safety Audit (FHWA-SA-21-048). Accessed October 2026. highways.dot.gov/…/lighting and highways.dot.gov/…/road-safety-audit
- 13Federal Highway Administration (FHWA), Office of Highway Policy Information. Traffic Monitoring Guide. 2022. fhwa.dot.gov/policyinformation/tmguide/tmg_2022
- 14Federal Highway Administration (FHWA). Systemic Approach to Safety and Why a Systemic Approach (updated February 18, 2026); Data-Driven Safety Analysis (DDSA) (updated September 26, 2025); Systemic Safety Project Selection Tool, FHWA-SA-13-019 (July 2013). highways.dot.gov/…/systemic, highways.dot.gov/…/why-systemic-approach and highways.dot.gov/…/data-driven-safety-analysis-ddsa
- 15Federal Highway Administration (FHWA). Road Safety Audits (RSA) (updated July 1, 2025) and RSA Legal Issues (updated February 15, 2024). highways.dot.gov/…/road-safety-audits-rsa and highways.dot.gov/…/rsa-legal-issues
This article reflects the priorities of traffic safety epidemiology, crash data analysis and GIS mapping. It is general education, not legal, engineering or other professional advice for any community or location. Data systems, program rules and web addresses change, so verify details with official sources. Reviewed October 2026.
Keep learning
Safe Driving Resource Center
Guides, self-checks and articles for drivers, parents of new drivers, and families recovering after a crash.
