• Introduction
• Very Basic Boat Propeller Injury and Death Statistics
• U.S. Coast Guard Recreational Boating Accident Database
  • Event 1, Event 2, Event 3
• Discussion of the U.S. Coast Guard Recreational Boating Accident Database (BARD)
• Some Problems With the BARD database
  • Under Reporting of Accidents
• Direct Access to the BARD Database
• Statistics Are Just Numbers, Real People Are Being Injured and Killed
• One Last Comment About the U.S. Coast Guard BARD Accident Database
• Other Potentially Useful Accident Databases
• Questions & Feedback

Introduction

Discussing the actual number of U.S. propeller deaths and injuries can get pretty complex so we are going to present the subject gradually in increasing levels of detail. Just keep reading and stop when you reach the point that meets your needs.

Very Basic Boat Propeller Injury and Death Statistics

Recent U.S. Coast Guard annual recreational boating safety reports (2001 to 2006) indicate 28 to 47 people are killed annually by “struck by boat motor or propeller” accidents in the United States. The same reports include about 185 to 265 accidents reported to the U.S. Coast Guard annually as boat motor or propeller strikes. The Coast Guard is fairly certain the fatality numbers are accurate, but admits many boating accidents in general are not reported. They do not provide an estimate the percentage of propeller accidents going unreported.

Houseboat Propeller Accident Statistics – if you are interested in statistics for houseboat propeller accidents, please read this page first, then go to our Houseboat Propeller Accident Statistics Page .

U.S. Coast Guard Recreational Boating Accident Database

The U.S. Coast Guard maintains an accident database called the Recreational Boating Accident Database (BARD). This database is the one most commonly referred to by articles or technical papers discussing the frequency of boating accidents. Before blindly using the data, please be aware the Coast Guard acknowledges it is under reported, perhaps as much as 90 percent in terms of injuries (only about 10 percent of boating injury accidents may actually be reported).

Annual summaries from BARD for recent years are available from the Coast Guard Boating Safety web site by year and are often referred to as Boating Statistics annual reports.

Event 1, Event 2, Event 3

Boating accidents typically occur as sequence of events, such as one boat strikes another boat, then someone falls overboard, and they are eventually struck by the propeller. The Coast Guard logs these events in sequence as Event 1, Event 2, and Event 3. The Boating Statistics annual reports indicate the number of accidents of a specific type occurring as Event 1, Event 2, Event 3, and the total number of those events that occur. In the example mentioned, collision with another boat would be Event 1, falling overboard would be Event 2, and being struck by the motor/propeller would be Event 3. As can be seen below, many propeller strikes are Event 2 or Event 3, except in the case of striking swimmers already in the water. Over time, they have slightly changed the reporting methods used in their Boating Statistics annual reports, so not all statistics can be directly compared (without going to the database itself).

The actual Coast Guard description of propeller accidents is “Struck by Motor/Propeller”.

The table below combines data from several recent USCG Boating Statistics Annual summaries.

The Coast Guard data changes reporting format from time to time and we attempted to do our best in reporting the data accurately in the table below. If you spot any errors, please contact us.

One last explanation of how to read the table above – in the year 2004 the table indicates:

• There were 64 accidents in which one or more people being struck by a propeller was the first event that occurred. These are typically swimmers, water skiers, tubers, divers, etc. OR they represent accidents where the sequence of event was not recorded and they were just turned in as propeller injuries.
• 5 of those people died (those in which the first event was being struck by the propeller) and 61 of were injured. (Note this totals to 66 people indicating there were some accidents in which more than one person was injured by a prop as a first event).
• There were 97 accidents in which being injured by a propeller was a second event (these are often people who fell overboard – going overboard could have been Event 1)
• 25 accidents listed being injured by a propeller as a third event. We would anticipate some of these are the result of a boat striking another boat or floating debris (Event 1), falling overboard (Event 2), then being struck by a propeller (Event 3).
• There were 186 accidents involving being struck by the motor or propeller (note the discussion on accident under reporting elsewhere on this page)
• 31 people died as a result of being struck by the motor or propeller.

All data in the table above comes directly from the Annual Boating Statistics reports. We will try to “fill out” the remaining cells from the actual database when we get some time.

Discussion of the U.S. Coast Guard Recreational Boating Accident Database (BARD)

What is in and is Not in the Database? – From USCG Boating Statistics 2004, Federal regulations require the operator of any vessel, numbered or used for recreational purposes to file a Boating Accident Report (BAR) when as a result of an occurrence that involves the vessel or its equipment:

1. A person dies
2. A person is injured and requires medical attention beyond first aid
3. Damage to vessels or other property exceeds $2,000 or there is complete loss of any vessel, or
4. A person disappears from the vessel under circumstances that indicate death or injury

and that accident occurs in the District of Columbia, Puerto Rico, Guam, the Virgin Islands, American Samoa, the Commonwealth of Northern Mariana Islands, and all States.

Minimal reporting requirements above are set by Federal Regulations. States can have stricter reporting requirements. The accidents in this publication (the Coast Guard annual Boating Statistics Report) cover only accidents meeting the Federal minimum reporting requirements above.

The data does not include deaths from natural causes while on a vessel, accidents that occurred when a person was swimming to retrieve an object or a vessel that was adrift from a moored location (swimming from the dock or from a moored vessel), accidents or damages resulting from moored vessels during storms, unusual tides, swell conditions, or accidents occurring when a vessel gets underway in those conditions in an attempt to rescue other persons (such as accident on boats involved in Hurricane Katrina rescues). The data also does not include swimming or diving deaths occurring when a vessel was not underway (note several people have come into contact with propellers on boats that are still docked or moored and been cut by non rotating propellers – those injuries are not included). It also does not include accidents that involved commercial, military or other non-recreational activities.

Our comments – Historically BARD did not include accidents occurring beyond three miles offshore. We are a bit confused on exactly what is currently reported offshore. Accidents on private waters (like private lakes and ponds) do not have to be reported, but some are included. There also seems to still be some confusion about reporting accidents from guided fishing trips and fishing charters (some consider those commercial vessels). They seem to still be working on the definition of a “recreational vessel” or at least on getting a definition widely accepted.

Note, more than one person can be injured in a BARD reported accident. Plus those injured can be injured in different ways (such as one person crushed by an oncoming boat and one another person hit by prop in the same accident, or in some occasions, two people injured or killed by propellers in the same accident).

Some Problems With the BARD database

FIRST – This is a great database, we are not “picking” on the Coast Guard. But, just like our site, with some more manpower and money, we could do a better job. With some more resources they could too. Plus the Coast Guard’s boating safety area, and other Coast Guard programs are facing strong competition for funds by the increasing role the Coast Guard is playing in Homeland Security. The USCG boating safety operation is not only limited by budget and manpower, it is also at the mercy of:

1. Boat operators to report accidents when many of them are not aware they are required to, or don’t want to go through the process.
2. The boat operator or investigating agency supplying the data (and supplying it accurately) to fill out the
   BAR (Boat Accident Report)
3. The states or agencies involved to forward the completed Boating Accident Report to the USCG.

The failure of any link in that chain means they don’t even know about the accident. Its no wonder only a small portion of injury accidents are actually reported.

Some of the more specific problems with BARD are discussed below such as when the Coast Guard failed to included fatalities from accidents worked by their own Search and Rescue operations, and under reporting in general. Again, this is a great database and we are very grateful the Coast Guard maintains it. But, like similar data sets in other industries, it has some problems, and should not be blindly relied upon. Some improvements have already been made, and more are in progress. But, like many government operations, things tend to move a bit slowly. One major development in recent years is the development of a web based accident reporting system (boating accident reports can be filed online by state agencies). We wish them success in their efforts to build a more complete database of recreational boating accidents and in using that information to help make boating safer.

In early December 2002, we proposed an effort to marine industry companies to develop improved propeller accident data sets as part of a Boating Industry Consortium Proposed to Address Propeller Injuries. The proposal fell on deaf ears. The industry did not respond.

Audit by the Office of Inspector General – fatality count audit problems

The Office of Inspector General published an Audit Report titled, “Audit of the Performance Measure of the Recreational Boating Safety Program. United States Coast Guard. Report Number: MA-2000-084 on April 20, 2000. This report found when the U.S. Coast Guard staff responded to accidents involving fatalities they did not report them to the states in which they occurred, or to the U.S. Coast Guard Office of Boating Safety which maintains statistics.

Accidents identified in the U.S. Coast Guard Search and Rescue Management Information System (SARMIS) were not reported to the Office of Boating Safety. Boat US Foundation of Boating Safety was contracted to investigate the problem. They found the following number of fatalities per year in SARMIS that had not been reported in BARD.

The audit also pointed out there is no clear, precise definition of a boating fatality. Sometimes it is unclear if an accident is linked to recreational boating or to some other activity in or around water.

In addition, the audit pointed out the Coast Guard had no criteria for monitoring the state programs it was dependent upon receiving data from.

The Coast Guard responded that it was the state’s job to report these deaths not theirs. The SARMIS database is a workload database (indicates the workload of their search and rescue operation), not an accident database. If they tried to populate the data fields of BARDS from the data in SARMIS they would only have about 20 percent of the data they need. They state BARD accident counts are understated because recreational boats involved in fatal accidents did not report the incident to the proper State Reporting Authority as required by law.

No mention is made in this report of the corresponding potential under reporting of non-fatal accidents that were probably also listed in SARMIS, but not appearing in BARD.

Note – the Coast Guard did not adjust their historical data in future reports to reflect the omission of the SARMIS data. This can be seen in their Boating Statistics 2004 report which still lists the left column above the death total for those years (800 in 1993, 784 in 1994, etc. it does not include the additional deaths found in the SARMIS data.)

Under Reporting of Accidents

The Coast Guard recognizes many accidents are not reported because boaters are not aware they are required to file a Boating Accident Report (BAR) or they fail to comply with the regulations. In general, the Coast Guard believes that nearly all fatal accidents are included in this report, but acknowledge “only a small fraction” of the non-fatal boating accidents in the United States are reported to the Coast Guard, State, or local law enforcement agencies. The Coast Guard says, “Overall, the more serious the accident, the more frequently the reporting.”

Percent of fatal accidents reported –

• 2004 – they feel they have captured all or almost all of the fatal accidents
• 2002 – they estimate their fatal accident non reporting at 1 percent
• 2001 – they estimate their fatal accident non reporting dropped from 6 percent to 1 percent based on improvements made as a result of the Office of Inspector General Audit discussed elsewhere on this page

Percent of nonfatal accidents reported –

For the last several years, the Boating Statistics report includes a note stating the “reporting rates of subgroups of accidents, personal watercraft, propeller strikes, collisions, or white water, probably differ greatly depending on unspecified variables.” This means that even we were able to estimate the overall under reporting percentage, it could not be broadly applied to individual accident types. For example, if less than 10 percent of all boating accidents are being reported, that does not necessarily mean less than 10 percent of propeller injury accidents are being reported.

The 2004 Boating Statistics report just says they acknowledge only a small fraction of the injuries are actually reported. Some of the older reports make some estimates of that fraction.

In late 2007 we identified an accident listed in BARD as “Struck by Boat” that is obviously a propeller accident. It is identified as a propeller accident by two newspapers, by four television stations, by officials on site, as well as by the BARD Database itself in its Injury Table. However it is not classified as a propeller accident by BARD. See our Some Propeller Accidents are Misclassified page for more information on this type of under reporting.

Several studies have attempted to estimate the amount of under reporting in BARD.

• A 2007 presentation at the American Public Health Association (APHA) Annual Meeting November 3-7, 2007 in Washington D.C., Incidence of Recreational Boating Injuries in the U.S., 2002 by Bruce Lawrence and Ted Miller of Pacific Institute for Research & Evaluation compared 2002 BARD data to mortality reports, hospital discharge (HD) data and emergency department discharge (ED) data for fatalities, hospital admissions, and non-admitted medically treated cases for several states. They estimated BARD captured all but 1 to 2 percent of the fatalities, missed about 20 percent of the hospitalizations, and missed 90 percent of non-admitted medically treated injuries. They also noted the quality of reporting varies significantly from state to state. The presentation used to be online as a Power Point file. This study was sponsored by the U.S. Coast Guard.
• A July 2006 study, Recent Research on Recreational Boating Accidents and the Contribution of Boating Under the Influence by Bruce Lawrence and Ted Miller of Pacific Institute for Research & Evaluation and Daniel Maxim of USCGAux used several databases to estimate BARD reporting frequencies for fatalities, hospitalizations, and non-admitted accidents. Their study calls for more work, but appears to confirm the Coast Guard’s opinion of most fatalities being reported, then reporting frequency falls off as severity falls off. See the 2007 presentation above formore information.
• In 2001, the Coast Guard established a contract with the Emergency Room Nurses Association (ERNA) to monitor emergency rooms at participating hospitals and create a database of boating accidents from patients seeking treatment. The National Association of State Boating Law Administrators (NASBLA) Boating Accident Investigation, Reporting and Analysis Committee Interim Committee Meeting Minutes of April 21, 2005 reports 18 hospitals collected data in the first year with 171 patients being treated for injuries sustained in boating accidents. NONE of those patients filed a Boating Accident Report (BAR) even though they were told they were required to do so. Their reporting percentage was ZERO after being instructed they were required to file.
• The “2002 National Recreational Boating Survey Report” prepared for the U.S. Coast Guard by Strategic Research Group, dated 30 November 2003 is based on 25,547 returned surveys from boat operators of their experiences between September 2001 and September 2002. Their survey projects 271,470 boating accidents injuring one of more people that required medical treatment beyond first aid occurred during this period.
• A Study of Propeller Accidents at FOUR Texas Lakes in 1997, Boat-Propeller-Related Injuries — Texas, 1997. MMWR 47(17);354-356 Publication date: 8 May 1998 reports on how they monitored hospitals near FOUR Texas lakes from May 24, 1997 to Sept 1, 1997. Those hospitals reported 13 persons sustaining propeller injuries of which 3 died. Our Comments – For this same time interval BARD contains only ONE Texas propeller injury (was at Lake Travis on 22 August 1997 and it was not fatal). This indicates a max of 1 of these 13 propeller accidents was actually reported to BARD. What about propeller accidents in the dozens of other Texas Lakes? What about the three propeller deaths reported from these lakes? None of them made it to BARD (at least not as a recreational propeller accidents).
• A 1994 paper: Motorboat Propeller Injuries in Wisconsin: Enumeration and Prevention. Hargarten, Karlson, Vernick and Aprahamian. Journal of Trauma. Vol.37 No.2. Aug 1994. Pgs.187-190. cites data from a 1987 paper, Motorboat propeller injuries. C.T. Price and C.W. Moorefield. Journal of the Florida Medical Assoc.Vol.76 No.4. June 1987. Pgs.399-401. That data as presentedin the 1994 paper reports that “Over a five year period the U.S. Coast Guard reported only 30 propeller injuries in Florida, whereas a statewide survey by orthopedic surgeons found more than 195 nonfatal injuries and five fatalities during the same period.” See bottom of the first page of the actual article, which is the third page of this document.
• A 1992 study by members of the Johns Hopkins University Injury Prevention Center and the Institute for Injury Reduction found that when adjusted for under reporting “the true number of propeller injuries and fatalities may be closer to …. 2,000 to 3,000 per year.” vs. the approximately 100 strikes per year reported by the Coast Guard from 1976 to 1990. (S.P. Baker and Jon S. Vernick & Associates, “Motorboat Propeller Injuries” Baltimore, Maryland: The Johns Hopkins Injury Prevention Center
  and The Institute for Injury Reduction; September 1992.)
• CDC Study – “A Study of Boat and Boat Propeller Injuries in the United States 1991-1992” by the National Center for Injury Prevention and Control Centers for Disease Control and Prevention. This study used the National Electronic Injury Surveillance System (NEISS) which does not typically collect boat accident data. It also used the Florida Department of Health and Rehabilitative Services statewide trauma reporting registry. From September 1, 1991 through August 31, 1992, boat propellers were found to be responsible for an estimated 1,155 injuries. (43 of those were reported by the trauma reporting system in the state of Florida, and seven of those in Florida died. They estimated 670 of the nationwide injuries occurred when the engine was off, the engine was on for 146 of the injuries and the status of the engine could not be determined for 340 of the nationwide injuries. In Florida – all 43 injuries occurred in the water, the engine was off in two of the accidents, on in 38 accidents, and status of the engine could not be determined in 3 of the accidents.
• Red Cross survey – A National Boating Survey taken by telephone covering the 1988-1989 recreational boating period asked respondents to identify critical incidents or accidents while boating in the previous year. (See page 8 of the CDC study for more info) 6,060 accidents were reported to the Coast Guard during this period. The Red Cross survey found 236,599 reportable incidents (per Coast Guard criteria) and another 413,074 incidents that did not meet Coast Guard reporting criteria. The Red Cross survey separated “struck by propeller” from “struck by boat” and found 5,032 incidents of propeller injuries or propeller damage. The overall reporting percentage (6,060 reported of 236,599) was found to be about 2.5 percent.
• In a March 11, 2002 letter to the U.S. Dept of Transportation in response to a request for comments on a Notice of Proposed Rulemaking involving propeller injury avoidance measures, the National Marine Manufacturers Association (NMMA) said, “The USCG defines a reportable injury as one requiring medical treatment beyond first aid. In regards to propeller injuries, one would assume that there would not be many minor injuries that would go unreported. Being hit by a propeller, spinning at any speed, would logically result in a major event requiring some form of professional medical attention, beyond first aid. This type of event would be reported.”HOWEVER – the CDC study mentioned above said (page 3) “Of the reported propeller-related injuries, 21 % or fewer were serious enough to require hospitalization.”

The bottom line is only a small portion of the total number of boating accidents are actually reported. The portion of propeller injury accidents reported is unknown.

Under reporting of boating accidents still remains a problem from National Parks. As of 2007, the National Park Service has still not solved the problem of closing the link on accidents that are actually reported to them, NOT being reported to the Coast Guard. See Boating and Water Use Activities in the 23 March 2007 Federal Register Pages 13694-13706 Section 3.5 on Boating Accidents. Following are two quotes are from this document.

• “The NPS is in consultation with the USCG regarding their requirements related to the reporting of boat accidents and we will ensure that necessary USCG information is captured in our new reporting system.”
• “Therefore, each park unit is encouraged to develop standard operating procedures that include sharing of accident information with their state boating enforcement agencies.”

Effect of Under Reporting on Year to Year Comparisons

The Coast Guard would like to be able to set accident reduction goals (for the total of all types of boating injuries) and gage its progress toward attaining those goals. However, with only a small percentage of accidents being reported each year, slight changes in reporting rates can make considerable swings in accident counts for specific types of events, making it impossible to accurately measure progress toward their goals. Also, one would hope that over time at least a slightly larger percentage of accidents are being reported. The swings based on the percentage of accidents being reported overall and for specific types of accidents could mask any progress the Coast Guard might make.

A comment from us – Logically, it seems like a proper goal would be to focus on greatly increasing the accident reporting rate before trying to set goals based on poor data (except perhaps on the fatality data). Recently the Coast Guard proposed some safety goals based on the TOTAL of recreational fatalities PLUS commercial fatalities PLUS commercial injuries. That goal seems a bit nebulous to us. It spans commercial and recreational activities while including injuries from one side and not from the other.

Some auto injury and death statistical studies have shown deaths and injuries do not move in lockstep. For instance, a community that has a few auto injury deaths a year and several hundred auto injuries a year may be able to install safety measures (signal lights, stop signs, railroad crossing gates, etc) at a couple “death traps” and significantly impact the local annual traffic fatality rate, while the total auto accident/injury rate for the community remains relatively unchanged. They still have the same rate of minor collisions and accidents around town at slower speeds that do not typically result in death. Although the community did a great and wonderful thing, their actions may not be reflected in the injury statistics. This concept applies to the Coast Guard setting goals based on a total of injuries PLUS deaths. Leaving injuries off the recreational side and adding in commercial injuries and deaths creates a variable that does not seem to relevant to the situation at hand.

For those interested in a more technical discussion of the issues of BARD, along with other Dept. of Transportation safety databases, see the Appendix (Adobe pgs 16-18) of the DOT document Performance Data and Performance Measurement, part of the U.S. Department of Transportation 2003 Performance Plan 2001 Performance Report.

Direct Access to the BARD Database

For those wishing direct access to some of the data, the Department of Transportation maintains an old copy of BARD on RITA site. It has the database tables for 1995 up to and including 2001. You will need a spreadsheet or database program and some experience at manipulating data tables to utilize it. The link to BARD on RITA changes very frequently so you will have to find it on your own.

The database is also sold by some third party vendors and access is provided to the media by IRE.org and the Missouri School of Journalism.

Several previous years were previously available in print form as supplements to United States Coast Guard Docket 10299 as:

• Report: Recreational Fatal Boating Accident Data; Propeller Strikes 1988 – 1993
• Report: Recreational Boating Accident Data; Propeller Strikes 1988 – 1993
• Report: Recreational Boating Accident Data; Propeller Strikes 1994
• Boating Accident Report Coding Instructions

Comments on those reporting on Propeller Guard Injuries

You will occasionally see reports from some groups claiming there are only very small number of propeller strikes per year (far less than the 30 plus deaths annually typically logged by the Coast Guard database). These groups are typically only reporting the First Event numbers from the Coast Guard, talking about strikes by a particular kind of craft, or just making up their own numbers.

We have a further discussion of this problem and some of our frustrations in trying to get them to correctly report the USCG stats at Propeller Deaths and Injuries: Another “Inconvenient Truth”.

Statistics Are Just Numbers, Real People Are Being Injured and Killed

To better understand individual accidents and the disruption to human lives resulting from them, please visit our Media Coverage of Propeller Accidents by Year page.

One Last Comment About the U.S. Coast Guard BARD Accident Database

They have a great and wonderful accident database. You have to create something before people start throwing rocks at it. We are certainly not throwing rocks. We appreciate them and their efforts in maintaining this database. As mentioned earlier, its not perfect, neither are we. Thanks again to the USCG for all the work they do in maintaining this database, promoting safe boating and rescuing those in trouble on the water.

Other Potentially Useful Accident / Injury Databases

In addition to the U.S. Coast Guard BARD database, some other ongoing databases hold useful information, but tend to be much harder to access. Among them are:

• Other U.S. Coast Guard Databases and Information
  • Commercial vessel accidents in BARD – In recent times, the Coast Guard logs several commercial vessel accidents in the BARD database, but those records are not available in the Recreational BARD database it makes available.
  • Search and Rescue Management Information System (SARMIS) – a Coast Guard database that logs their work activity relating to Search and Rescue. It does not have all the fields needed for BARD. A few years ago the Coast Guard found many accidents in SARMIS were not it BARD and has since tried especially to bring the fatalities across.
  • Marine Safety Information System holds data 2002 to present (MSIS queries are normally limited to retrieving a single vessel)
  • Marine Safety Management System holds data through 2001 (MSMS is on the Coast Guard Intranet and available online as a series of data tables from the Bureau of Transportation Statistics)
  • Marine Information for Safety and Law Enforcement (MISLE) along the MISLE Analysis and Reporting System (MARS) is also on the Coast Guard Intranet. MISLE is the replacement for MSIS.
  • USCG Maritime Information Exchange Incident Investigation Reports reportable marine casualties investigated since November 2002
• Marine Index Bureau – a database of marine insurance claims. It includes both commercial and recreational vessels, as well as accidents by those working on off shore rigs, loading and unloading cargo, etc.
• National Electronic Injury Surveillance System All Injury Program (NEISS-AIP) – NEISS is a program which codes accidents of people coming to 100 hospital emergency rooms across the country IF those accidents are related to a consumer product. The program is ran by the Consumer Product Safety Commission in an effort to identify consumer product problems before they get out of hand, as well as to allow scaling up to estimate national accident frequencies and archive historical accident data for observing trends. NEISS does NOT log boating accident data. They collect data on just about any kind of consumer product related accident EXCEPT those related to boats, automobiles, motorcycles, planes, trains, firearms, food, illegal drugs and medical devices.NEISS-AIP is a subset of 66 of those 100 hospitals. At these 66 hospitals ALL injuries are reported. A series of codes are used to describe the accident and injury, plus two data fields hold free text descriptions of the accident. The system captures about half a million injury related emergency department visits a year.
• General Death Records
  • State Death Certificate databases
  • National Center for Health Statistics Mortality Data
  • Consumer Product Safety Commission DTHS database (death certificate information)
• National Park Service Incident Management Analysis and Reporting System (IMARS) (many national parks have lakes for boating and rental boats). IMARS is the new system they are still trying to get up and running. The old system is thousands of paper reports filed at each park. They do have “blast” type alert for major fatalities in their Daily Digest online news site.
• Department of Interior Safety Management Information System (SMIS) captures injuries to Dept of Interior personnel (many of them work on or near the water).
• Other Bureau of Transportation Statistics (BTS) Maritime Databases
• U.S. Military Safety Databases and Accident ReportingThe Navy has a PMV fatality report that lists private motor vehicle fatalities with a brief description of the event and the Army provides summary counts from a a similar POV (Privately Owned Vehicle) fatality report. All services have indepth safety databases, but accessing them beyond summaries, is very difficult. The Navy also provide a narrative of off duty Recreational fatalities.
  • U.S. Navy Safety Center
  • U.S. Army Combat Readiness Safety Center
  • Marines Safety Division
  • U.S. Coast Guard probably has a similar group (reports accidents to their own people) but we have not yet found it.
• Injury Data and Resources Center for Disease Control and Prevention
• National Trauma Data Bank (NTDB) appears to focus more on the injury itself and the patient’s treatment and outcome than what caused it.
• Consumer Product Safety Commission’s Injury/Potential Injury Incident File (IPII)
• Consumer Product Safety Commission Better Data Needed book on Google Books. (In general the CPSC does not cover boating accidents).
• European Injury Database public access is limited to canned querries. You have to be granted “restricted access” to view data at the record level.

Questions & Feedback

If you have any comments or questions about the statistics of other information on this site, or feedback about the site, please contact us (see Contact Us in top menu).