American Whitewater Accident Study

Covering Non-Motorized Human Powered Craft 1995-1998 

By Jennifer L. Plyler, Ph.D^.

Prologue

It's a bright sunny day. Two friends load an old aluminum canoe and head for the pastoral retreat of a local pond. The fishing poles are along for the ride. The faded orange horse-collar lifejackets will make great backrests as they bask in the sun.

In another part of the country, a group of kayakers set a car shuttle. They’ve paddled this Class II-III section of the river successfully many times. They’ll end their day by having dinner at the local eatery and discussing paddling plans for the following weekend.

Big water, big river, that’s how these friends spend their time together. They load up the rafts and head out for a weekend of fun. They’ll camp, tell stories by a warm, glowing fire, and share lots of laughs about the day’s events.

These are three very different groups, but in the past few years a few of them have ended a day on the water with a tragic, unexpected drowning.

Introduction

The goals of this study were to gain a better understanding of why drownings happen, what preventive measures could be taken to decrease them, and develop a greater understanding of where our boating safety efforts should be focused. Specifically, this study examines comparative risk factors for non-motorized human powered boating accidents for kayakers, canoeists, and rafters. To answer these questions, American Whitewater studied hundreds of accident reports from American Whitewater Journal (Journal) articles and the United States Coast Guard's (USCG) Boater Accident Reports (BAR’s) dating from 1995-1998.

American Whitewater requested the accident reports from Bruce Schmidt, Statistician with the USCG Office of Recreational Boating Safety, in the spring of 1999. When we received the electronic files, we understood why it took more than six months to collect and send this data. To ensure the confidentiality for the victims and their families, Mr. Schmidt meticulously reviewed all records and removed all references to names or other identifying information. American Whitewater commends Mr. Schmidt and the USCG for working to provide us with this information.

Accident Reports used in this Study

The data American Whitewater received was based on BAR's submitted to the Coast Guard. Under the Boater Safety Act of 1971, state' Boating Law Administrators are required to complete a BAR on all boating accidents and submit the form to the USCG. This information is then entered into a Coast Guard database. Each accident is assigned a record number and contains the date of the accident, cause of the accident, location, water conditions, personal flotation device (PFD) use, alcohol use, vessel type (kayak/canoe or inflatables), and a narrative of the accident.

American Whitewater (AW), by contrast, receives reports of whitewater accidents through a volunteer network of local paddlers. Newspaper articles, Internet postings, and personal narratives are regularly submitted. While an unknown number of incidents are missed, our database provides us with a good sampling of accidents occurring in our sport. Although they represent a fraction of the total number of the boating accident reports received by the USCG, AW reports are consistently more detailed and accurate.

State Boating Law Administrators are required by law to complete a BAR on all water-related accidents in their jurisdiction. For various reasons these reports are not always filed. From 1995 to 1998, an average of 14% (18^)[4] of all whitewater water-related fatalities were reported to American Whitewater (AW) but no BAR was submitted to the USCG (fig 1). Several of the missed accidents involved internationally known whitewater kayakers and were widely reported in the mainstream press.

Although the Coast Guard has the most comprehensive database of water related accident reports, the accuracy of their BAR’s varies widely. Information is often incomplete, contradictory, and inaccurate when cross-referenced with the detailed whitewater reports contained in the American Whitewater database. In states like Pennsylvania and Ohio, where watercraft officers receive good training in accident investigation, the quality of the BAR’s were generally high. The same cannot be said for investigations performed by people lacking specialized training in water safety. 

Here are some of the problems we encountered:

Determining the boat types from the BAR’s was surprisingly difficult. For example, under vessel type, canoes and kayaks were lumped together. "Inflatable" was a catchall category involving powered and non-powered craft. There were even instances where an accident appeared to involve a motorboat instead of a non-motorized human powered boat as listed in the USCG database. The author had to check the narratives closely to determine the exact vessel type, and was not always able to do so. In these cases, the vessel types were coded as “unknown.”

Many narratives said that the “boat capsized” but never explained why this happened.

 Many accidents occurred in cold water (below 50 degrees Fahrenheit) but this was often not listed as a “cause” per se. Consequently, it was often difficult to conclude the role that cold water played in accidents for all boat types.

Although a failure to wear PFD’s was not often listed as a cause of the accidents described in the narratives, this behavior is clearly linked with trouble. Between 1995 and 1998 over 50% (86) of all accident victims were not wearing PFDs. This problem, although not common among kayakers (whitewater, recreational or ocean touring), was a major factor in many canoe accidents that occurred on flatwater.

Descriptive terms for water conditions also caused problems. The USCG offers these descriptive options:

• Calm (waves less than 6”)
• Choppy (waves 6”-2’)
• Rough (waves 2’-6’)
• Very rough (waves > than 6’)

This listing is more applicable to ocean or lake conditions than to those found on streams and rivers. There is no separate category for moving water, which is often a factor for accidents occurring on rivers and streams. It was often difficult to determine whether current was present, creating hazards associated with swift water.

Furthermore, the narratives did not always match the water conditions checked on the form. For example, if the body of water was a lake, under “water condition” it might be described as choppy. Then in the narrative the water conditions are described as a “calm lake.” In some cases the author extrapolated the water conditions from the narratives based on her knowledge of certain rivers and lakes.

It would, of course, be ideal if the USCG and State Boating Law Agencies adopted the International Scale of Whitewater Difficulty that was pioneered by American Whitewater (See Appendix). Descriptions of the classes used to “rate” rapids are included in the “Whitewater Safety Code” (See Appendix). This scale ranges from Class I, described as occasional riffles with low, regular wave patterns, to Class VI, a rapid at the extreme limit of navigability for whitewater experts. According to Rich Bowers, AW Executive Director, “American Whitewater is committed to working with the USCG and Boating Safety Administrators to ensure that more descriptive and accurate set of categories included in the BAR's.” The Class I-VI rating system may be impractical for investigators untrained in the subtleties of river classification. However, an investigator can easily choose between “no current”, “fast current”, and “whitewater rapids.” This would clearly separate whitewater accidents without going through the BAR’s individually.

Despite the pitfalls within the USCG data, it remains the single most comprehensive database of boating accidents. The USCG database for 1995-1998, supplemented by water-related accidents reported in the Journal, and a similar study by Joan Maybe described in River Safety Report 1989-1991,^[5] forms the basis for this report and the analysis that follows.

Non-motorized Human Powered Paddlesport Trends^[6]

American Whitewater’s intent is to provide general information on boating fatalities, boater profiles, and recommendations for safety programming. Non-motorized human powered boating trends provide a context for analyzing the safety record of paddlesport participants. We can use this information to help us understand why some boating accidents occur, who is likely to be involved, and what can be done to prevent them. If we know how many people participate in canoeing, kayaking, and rafting, then we can have a better understanding of what the number of deaths tells us.

Table 1. Number of Kayak, Canoe, and Rafting Participants and Enthusiasts for 1994-1995, 1998, and 1999

	1994-1995		1998		1999
Vessel	Participants	Enthusiasts	Participants	Enthusiasts	Participants	Enthusiasts
Kayak	2.6 million	400,000	4.2 million	400,000	5.1 million	600,000
Canoe	14.1 million	3.6 million	18.1 million	2.9 million	17.5 million	2.1 million
Raft	15.2 million	3.8 million	9.7 million	1.1 million	8.8 million	1.7 million
Total	31.8 million	7.8 million	32 million	4.4 million	31.4 million	4.2 million

Sources: Cordell, Ken, McDonald. Barbara, and Briggs, J. Alden. 1995. Emerging Markets for Outdoor Recreation in the Unites States based on the National Survey on Recreation and Environment. Research conducted by the Sporting Good Manufacturers Association and the USDA Forest Service. www.outdoorlink.com/infosource/nsre.
Outdoor Recreation Council of (ORCA). 1999. Outdoor Recreation Participation Study for the , Second Edition. Highlighting the Enthusiast. Prepared by Leisure Trends Group/Gallup. Boulder, CO. 303.786.7900.

Through anecdotal information and observations we know that participation in paddle sports has grown dramatically in the last decade. Boaters who have been around a while will confirm that rivers and lakes are more crowded than ever.

We found that although the total number of participants for kayaking, canoeing, and rafting has remained relatively stable since 1995, the distribution of enthusiasts^[7] among the various segments has changed  significantly(Table 1). There are more kayak enthusiasts now than ever. The number of rafting enthusiasts increased in 1999 after a significant decline from 1994 to 1998. And although canoeists make up the largest percent of all enthusiasts, their numbers are decreasing because newcomers prefer to buy and use kayaks.

Boating demographics are also changing. More women are participating in canoeing and kayaking and more men are getting into rafting. Even though canoeists out-number kayakers six to one, current kayak sales (recreational, whitewater, and ocean touring) exceed canoe sales 

Table 3. Best-selling boat categories in 1999 according to manufacturers and retailers

Boat Category	Retail	Manufacturer
Recreational kayaks	39%	11.5%
Recreational canoes	29.2%	18.9%
Sea/touring kayaks	10.4%	12.9%
Whitewater kayaks	10.4%	4.9%
Inflatables	3.2%	8.2%

Source: Paddlesports Business. 1999. 1999 State of the Paddlesports Industry Report. Sponsored by Paddlesports Business and Canoe and Kayak Magazine.

 Canoeists had 17.5 million participants and 2.1 million enthusiasts in 1999 (ORCA, 1999). Overall manufacturer and retail sales reflect the large number of participants in this segment given that recreational canoes fall within the top two best-selling boat categories (Table 2). However, boat makers, retailers and the ORCA (1999) report confirm that recreational canoes are not rated as a hot market segment in the future (Table 3).

 In 1999, rafters had 8.8 million and 1.7 million participants and enthusiasts respectively. However, the number of enthusiasts represents an over 50% decline from 1995-1998 (3.8 million to 1.7 million-Table 1) (ORCA, 1999).

 Of the total number of paddlers in 1999, roughly 4.2 million were considered enthusiasts (ORCA, 1999). Of the 4.2 million, the number of kayaking enthusiasts was estimated at 600,000; canoe, 2.1 million; and raft, 1.7 million. According to Jason Robertson, American Whitewater Access Director, of the roughly 2.7 million canoe and kayak enthusiasts (600,000 + 2.1 million), approximately 720,000 are whitewater paddlers.

Overall Non-Motorized Human Powered Boating Related Fatality Trends

Figure 2 indicates that prior to 1995, the frequency of drownings tended to spike every two to three years but remained relatively stable over the years. However, since 1996, non-motorized human powered boating deaths have been increasing annually. This rise could reflect an increase in popularity and corresponding increase in the number of enthusiasts in the sport in the mid-1990s. Or, it might simply represent an increase in reporting as evidenced by the number of reports received on popular Internet newsgroups such as rec.boats.paddle.

The problem is that the data supplied does not breakout fatalities for participants and enthusiasts. But we can make some statements about fatality trends using the total number of fatalities for both of these groups.

Is the recent increase in fatalities significant? Yes and no. Any drowning is one too many. However, the total number of deaths related to non-motorized human powered vessels was 168 out of 32 million participants in 1998, a less than one-tenth of one percent death rate.

Is the recent increase in fatalities significant among enthusiasts? Again, yes and no. The total number of enthusiasts has decreased from 7.8 million in 1995 to 4.2 million in 1999 and the total number of fatalities has increased slightly in recent years.

The number of kayak enthusiasts has increased almost 50% from 2.6 million in 1995 to 5.1 million participants in 1999 while their death rates have remained stable. The number of canoe enthusiasts has decreased from 3.6 million in 1995 to 2.1 in 1999 but the number of fatalities has increased. Finally, the number of rafting enthusiasts has decreased from 3.8 million to 1.7 million in 1999 and their fatality numbers are decreasing (ORCA, 1999 and SGMA and the USFS, 1995). The trends demonstrate that the safety record for kayakers and rafters are good but canoeists are facing some challenges. However, the numbers (total number of enthusiasts versus the number of deaths for each vessel type) show that death rates for enthusiasts for all boat types is negligible. Even when the total number of whitewater enthusiasts (720,000) is compared to the total number of deaths (for enthusiasts and participants), the death rate is still a miniscule.

Nevertheless, this recent growth in fatalities should be a wake-up call for all paddlers given that many of these deaths were preventable by wearing PFDs, assessing water conditions more closely, or by using proper clothing.

Whitewater paddling safety trends versus other adventure sports

In September 2000, American Whitewater completed a comparative study on whitewater safety and usage.  This study was designed to determine a national whitewater fatality rate, based on measured levels of use on 30 managed rivers.  American Whitewater found that there were an average of 0.86 fatalities per 100,000 whitewater boaters (both private and commercial) over a five-year period between 1994 and 1998.  In 1998, the whitewater fatality rate for these same rivers was slightly higher at 1.15 per 100,000. Both numbers represent maximum fatality rates as the total private use numbers are almost certainly underreported.

Whitewater kayaking fatalities were higher than the overall whitewater fatality rate with 2.9 fatalities in 1998 per 100,000 participants.  In comparison with other common activities kayaking appears to be relatively safe: scuba diving (3.5), mountaineering (3.2), or driving a passenger vehicle (15.2).

Officials for 35 managed rivers were asked for use figures for private and commercial boaters, and whitewater boating fatality data for the five years between 1994 and 1998.  Of the 35 river managers contacted, 30 supplied data fitting our requirements. Since commercial rafting companies submit detailed seasonal use numbers, their totals are significantly more accurate than private use counts, which were measured under a variety of different methods with different levels of accuracy and completeness.

There were 5,732,683 commercial boaters and 1,687,880 private boaters in the five-year study period, for a total of 7,420,563 whitewater boaters. During this same period there were a total of 64 whitewater boating deaths, including 26 commercial boater deaths and 38 private boater deaths; 17 of these deaths occurred on just one river, Colorado’s Arkansas River. Eleven shore-based deaths (from activities like falling off cliffs or jumping off bridges) were also reported and subsequently omitted from the total. American Whitewater calculated 0.86 whitewater fatalities per 100,000 boaters, or 2.25 private boater deaths per 100,000 and 0.45 commercial boater deaths per 100,000.

Where and When Fatalities Occur

Most water related accidents occur in the spring or summer for all vessel types (Table 4). This fact has remained unchanged since the 1970's.

Table 5. Deaths for time of year and vessel type

Vessel Type	1995	1996	1997	1998
Kayak	Summer	Spring	Spring	Summer
Canoe	Summer	Spring	Summer	Summer
Noncommercial	Summer	Summer	Summer	Summer
Commercial	Spring	Summer	Summer	Summer
Unknown	Summer	Summer	Spring	Spring

Table 6. Most frequently reported accident locations (state) by vessel type

	1995	1996	1997	1998
Kayak	California Montana Wyoming	New Jersey Tennessee Washington	Colorado California
Canoe	Wisconsin California	Massachusetts Washington	New York Oregon Missouri	Alaska
Noncommercial	Multiple states1	North Carolina	Multiple states	California Oregon New Mexico
Commercial	California	Multiple states	Multiple states	California
Unknown	Minnesota Washington	Colorado Maryland	Maine Virginia	Florida Vermont

¹ Multiple states indicate that five or more states had the same number of accidents.

 

Table 7. Most frequently reported primary causes of accidents by vessel type

Vessel Type	1995	1996	1997	1998
Kayak	Strainer/sieve	Low head dam/undercut	Unknown/ inexperience/ pinned in water	Strainer/ unknown
Canoe	Unknown/ hyperthermia/ Large hole	Unknown/ strainer/ flood	Alcohol/no PFD	Unknown/ no PFD
Non- Commercial	Flood/pinned in water	Flood/foot entrapment	Unknown/ strainer/ capasizing	Flood
Commercial	Flood	Undercut/flood/health problems	Unknown/ strainer/ undercut	Strainer/ flood
Unknown	Unknown/ capsizing	Unknown/ Health probs/ capsizing	Unknown/ capsizing/ bad weather	Unknown/ low head dam

California, Colorado, and Washington State had the greatest number of accidents from 1995 to 1998 (Table 5). These states have more whitewater rivers, and because much of the river flow comes from snowmelt, long periods when high water seasons and warm weather coincide. Although a connection to cold water was not substantiated by the data, it is worth noting that rivers fed by snowmelt are always very cold.

Causes and Contributing Factors of Non-Motorized Human Powered Boating Fatalities-Introduction

It is important to make the distinction between contributing factors and causes throughout this report. Cold water, improper PFD use, and whitewater are contributing factors but they are not a primary cause of accidents based on the data. They do, however, make a situation worse. In whitewater accidents, failure or improper use of a PFD and exposure to cold water are most often listed as contributing factors, not primary causes. For example, a paddler may drown in cold water but the primary cause of the accident could capsizing after hitting a rock or encountering a violent storm. Although the cold water may have limited the paddler’s ability to save himself or hampered rescue attempts, it did not cause the accident.

Causes of Non-Motorized Human Powered Boating Accidents

According to the narratives, the primary causes of all non-motorized human powered boating fatalities from 1995-1998 were (Table 5):

• Strainers or an obstacle on a river that allows water to pass through but is too small to allow people or boats to pass through. Examples include fallen trees, debris, or jumbles of boulders.

• Flooded rivers or creeks that result from heavy rains or increased snowmelt for example.

• Being caught and held in a large hydraulic. A hydraulic is created when water flows over an object and creating a recirrculating upstream flow beneath the water’s surface.

In addition to these causes, many boats were described as “capsized” but it was unclear as to what caused the boat to turn over. Thus, many accident causes are really “unknown.” Based on the narratives, drownings from capsizing could have been caused by health problems, alcohol, and lack of PFD use.

Some people believe that water-related fatalities in non-motorized human-powered boats are most often associated with whitewater[8]. However, the data tells us a very different story. From 1995 to 1998, the percent of drownings on whitewater versus flatwater were nearly equal (Figure 4). Accidents are just as likely to occur on flatwater as whitewater.

 

In contrast, boating related deaths in the ocean were significantly lower than other water conditions despite the increased popularity in sea kayaking. When accidents did occur on saltwater, the majority of fatalities were caused by severe and often unexpected changes in weather conditions such as high winds, waves, or lightening.

Contributing Factors to Non-Motorized Human Powered Boating Accidents

In general, PFD misuse or no use appears to be a major contributing factor in water related fatalities. On the average, 50% (86) of all victims from 1995 to 1998 failed to wear a PFD, or used one improperly (Figure 3). In fatalities involving PFD misuse, the victims fell into one of three categories of groups:

 

·        PFDs were not found on the victim or in the vessel,

·        PFDs were in the vessel but not worn, or

·        PFDs were in such poor condition that they were useless.

Deaths resulting from PFD misuse or non-use are preventable. In fact, deaths related to PFD non-use continue to rise! According to Rich Bowers, American Whitewater Executive Director, this increase “could be attributed to better reporting of accidents¾but whatever the case, no one should ever boat in any water conditions without a PFD.”

With the advancements in cold water paddling equipment and clothing, one would expect a decline in the frequency of boating accidents in cold water (< 50 degrees Fahrenheit). However, the number of fatalities in cold water has remained unchanged since 1970’s. Figure 5 indicates that cold water conditions could have been a contributing factor in roughly one-third of all water-related accidents from 1995 to 1998. It is disturbing that this figure is essentially unchanged compared to Maybe’s (1992) earlier findings for specific boat types (i.e., kayak, canoe, and noncommercial rafts) given the advancements in cold water boating equipment. 

Impact of Reporting on Determining Causes and Contributing Factors

Sketchy reporting of the causes and contributing factors was evident in my general analysis of boating accidents. Further compounding the analysis was that many of the causes and factors could be interconnected. For example, PFD use was associated with vessel type as demonstrated with the canoe accident victim profile that follows. In other words, canoe-drowning victims in general failed to wear lifejackets. Of the victims paddling unknown vessel types from 1995 to 1998, an average of 80% (31) did not wear a PFD! Accurate reporting of vessel type could have further substantiated or refuted my analysis of the following boater profiles.

The general data presents a broad picture of boating fatalities, but as we explore further, a clearer image of accident victims emerges.

 

Non-Motorized Human Powered Boater Profiles

 Perhaps the most compelling information that the Journal, USCG data and Maybe’s (1992) report gives us is a clear description of who is involved in boating related fatalities, which in turn points to several areas to focus future safety program efforts.

 The four scenarios described in the prologue represent the unique demographics of those perishing on our waterways.

Canoeists

Clearly, canoe fatalities occur more frequently than accidents involving kayaks (Figure 6). Although deaths among canoe victims have not returned to the high point observed in 1977, they have been on the rise since 1995.  Given that there were 17.5 million canoe participants in 1999 and that recreational canoes were the top-selling boats in 1999, we can probably expect to see an increase in canoe accidents in the next several years (ORCA, 1999 and Paddlesports Business, 1999).

The inexperience of canoe victims is clear when you realize that the primary cause of canoe related accidents was failure to wear a lifejacket (Figure 7). The typical USCG narrative reads, "Vessel capsized on small pond. Victim was not wearing a PFD." An average of 50% (38) of the victims paddling canoes during 1995 to 1998, and nearly 90% (57) in 1997 and 1998 failed to wear a PFD. The narratives supported that PFD non-use was the actual cause of most drownings involving canoes, rather than just a contributing factor.

Canoe accidents usually occur on flatwater (Figures 8 and 9). The typical accident scenario involved a fishing trip; a family outing; or a day of swimming on a pond, lake, or placid river. Whitewater was essentially a non-issue among victims of canoe accidents. In fact, only a small number (11) of canoe accidents occurred in whitewater.

Exposure to cold water is often a contributing factor in canoeing fatalities. From 1995 to 1998, cold water conditions were associated with roughly 30% (13) of canoe accidents (Figure 10). Inexperienced boaters rarely invest in cold water paddling gear or may be unaware that it is available.
 

 Kayakers - The Missing Link

It is important to note that that there is a major gap in the accident reports for kayaks. We do not have information on drownings involving recreational kayaks (i.e., Keowees, Swifty, Crossover, and other flatwater kayaks) since the BAR report does not break down kayaks by type. Yet, this is an important market segment comprising the first and third best selling boat categories according to retailers and manufacturers respectively in 1999.

Charlie Walbridge, American Whitewater Safety Editor concludes that there are one or two recreational flatwater kayak deaths per year but points out that this is a fairly safe group given the estimated total number of flatwater kayakers. Thus, they are not addressed in detail but mentioned as a “missing link.”

Kayakers can be divided into two categories: sea kayakers and recreational whitewater paddlers.
 

Sea Kayakers

Sea kayaking accidents accounted for less than 5% (7) of the boating accidents from 1995-1998 and the accident rate has remained relatively stable. When sea kayakers were involved in a boating accident, the cause was usually attributed to bad weather or sudden, unexpected changes in weather conditions. Although adverse weather conditions can cause wave swells, ocean con