Mortality
and morbidity in white water rafting in New Zealand
Abstract
Objectives This study provides the first descriptive overview
of fatal and non-fatal injury associated with white water and
other recreational river rafting in New Zealand. The current study
sought to identify the nature and causes of hospitalisable injuries
and to identify the causes of fatal injuries to white water rafters.
Design The data were obtained from the New Zealand Health Information
Service (NZHIS) mortality and morbidity files. Mortality data
for the period from 1983 to 1995 and morbidity data from 1983–1996
were used.
Participants Members of the public who took part in white water
and other recreational river rafting activities through-out the
above periods.
Results Of the 33 fatalities, over 80% were male. Almost all the
fatalities involved drowning, more than a third resulting from
the raft capsizing. Nearly half of the 215 hospitalisations resulted
from fractures, victims spending an average of 3.3 days in hospital.
The effects of submersion, and intracranial injuries were the
next most common categories.
Conclusions and implications In relation to fatalities, the potentially
modifiable risk factors involve improved resistance to raft capsizing,
and equipment and skills required to stay afloat. In relation
to injuries, the potentially modifiable risk factors relate mainly
to preventing slipping and falling through the design of footwear,
protective equipment, and procedures for entry and egress.
Keywords: Injury prevention; drowning; water recreation; injury
mortality; rafting.
Introduction
White water rafting first became popular in New Zealand after
the end of the Second World War. The commercial
development of white water rafting as an ‘adventure’ activity
has taken place in more recent times, with 21 companies currently
forming the New Zealand Rafting Association (http://www.nz-rafting.co.nz/).
The most frequently rafted river in New Zealand is the Shotover
River near Queenstown in the South Island with approximately 40,000
clients per annum1
In November 1994, following concerns about safety standards, the
New Zealand Maritime Safety Authority began a review of the commercial
white water rafting industry.2 The subsequent report noted widespread
concerns from industry representatives about inadequate safety
standards. Concerns were expressed about guide training and supervision,
the influence of commercial pressures, and unsafe operational
practices. A recent accident investigation report1 into a fatality
on the Shotover River in 1995 noted that there was a lack of statistical
information about participation in rafting in New Zealand. The
present study was undertaken to provide a descriptive overview
of fatal and non-fatal injury associated with white water and
other recreational river rafting in New Zealand.
Recreational boating fatalities
There has been only one published study of injury associated with
white water rafting.3 This was based on reports of non-fatal injuries
submitted by commercial rafting companies in West Virginia between
1995 and 1997. The most frequently injured body-part was the knee
although one-third of injuries involved the face. The majority
of facial injuries occurred in the raft itself. A number of other
studies4,5 have investigated fatalities within various forms of
recreational boating through the use of accident report data.
Although these studies have included incidents involving all types
of recreational boats and have therefore focused on a much broader
field than the present study, there are some notable findings
involving the identification of risk factors and the consideration
of fatality patterns applied to boat operators. These studies
therefore provide the most appropriate context for the analysis
of fatal injuries in the present investigation.
One such study4 investigated the risk factors associated with
fatal boating incidents. Boating Accident Report (BAR) files in
Ohio from 1983–1986 were analysed for this purpose. During the
four years of the study, 107 reported incidents resulted in 124
deaths. Of those, 99 deaths (80%) were classified as drownings,
and alcohol was mentioned in relation to 21% of the incidents.
The highest proportion of fatalities occurred within the 20–29
year age group. The number of fatalities decreased as age increased.
Experience and training also appeared to affect fatality levels.
Those boat opera-tors who were inexperienced (less than 20hrs)
or who had no training showed a higher incidence of fatalities
than those who were more experienced (greater than 100hrs) or
had undergone some training. As the number of boating incidents
resulting in fatal injuries seemed to be related to the age, training
and experience of the operator, it was concluded that youth, inexperience
and lack of training were associated with increased risk of recreational
boating fatalities.
In a similar study investigating boating fatalities in Canada
over a two-year period,5 alcohol and the non-use of personal flotation
devices (PFDs) were cited as the two primary factors relating
to boating fatalities. Of the 429 boating fatalities occurring
in the two-year period, only 22 were non-drowning deaths (these
included hypothermia, effects of sudden immersion, and collision-
or propeller-related trauma). According to the report, drowning
was the third highest cause of death related to recreational activities
in young adult men in Canada, and 40% of these drownings resulted
from boating incidents.
Research has indicated that more males than females in the USA
drown as a result of boating incidents6 (ratio of 14 to 1). In
fact, the rate (per 100,000 population) of drowning in general
is greater for males than females as well (ratio of 5 to 1). The
male drowning rate increases sharply from age 10 to peak at age
18, whereas for females, the drowning rate stays constant from
age 10 through to age 30. These differences have been attributed
to variations between sexes in ‘exposure, supervision, cultural
expectations, biological make-up, and other factors such as alcohol
use’6 (p. 177).
A temporal variation in boating fatalities has also been documented.6
Fifty percent of all drownings involving boats occur during the
summer season (May to August in the north-ern hemisphere). Not
surprisingly, drowning is reported as being one of the most seasonal
of injuries, presumably due to increased exposure to aquatic activities
during the summer months.
The present analysis of recreational river rafting in New Zealand
investigated whether fatality patterns relating to time of year,
sex differences and age, are similar to those discovered in overseas
research relating to recreational boating ingeneral. A major difference
between the data reported in these studies investigating boating
fatalities in general and the present study investigating river
rafting is important to note. The present study used data obtained
from the New Zealand Health Information Service (NZHIS) and was
there-fore based on injury and fatality data on an individual
basis. The data used in the previous studies arose from reports
submitted by commercial operators or from BAR files and were therefore
based on fatality data on an event basis (i.e., it was the incident
rather than the fatality that was reported). The present study,
based on individual records, therefore provides a much more complete
description of the injury outcomes of rafting incidents.
Method
Data were obtained from the New Zealand Health Information Service
(NZHIS). The NZHIS maintains a record of all deaths (mortality
file) and all discharges from public and private hospitals (morbidity
file). For this study however, only material from the public hospitals
were used. Also excluded were readmissions and transfers between
hospitals which, if used, would have given inflated estimates
of incidence. Private hospitals tend to be involved in follow-up
treatment and details are often incomplete, making them of limited
use in this study. This study covered a 13-year period (1983 to
1995) for mortality and a 14-year period (1983 to 1996) for morbidity.
For convenience, deaths recorded in the NZHIS mortality file will
be referred to as ‘fatalities’ and hospital discharges reported
in the NZHIS morbidity file will be referred to as ‘hospitalisations’.
In order to identify the fatalities and hospitalisations sustained
in river rafting incidents the NZHIS files were searched for incidents
involving recreational rafting. The first step of the search involved
identifying all cases with the word ‘raft’ contained in the event
description (this is a ‘one-line’ description of injury circumstances
provided in narrative fields) contained on each record on the
file. From this list, those incidents that took place on a site
other than a river (such as a lake or ocean) were deleted. From
1994 onwards, some hospitals may have recorded rafting incidents
in the event description using a default system based on the definitions
associated with the International Classification of Diseases External
Causes of Injury and Poisoning codes (E-codes).7 In such cases,
the word ‘raft’ would not be found in the event description and
therefore would not have been picked up in the search. Similarly,
prior to 1985, hospitalisations that did not span midnight (i.e.,
were not overnight stays) were not recorded in the files used
in this study. There-fore, the data from 1983 and 1984 were missing
hospitalisations lasting less than one day. Again, some rafting
incidents may not have been recorded.
The NZHIS files contain basic demographic information on the victim,
including age and gender. The NZHIS files also contain ‘external
cause’ codes corresponding to the
Injury mortality and morbidity in rafting 195
International Classification of Diseases (ICD-9-CM)7 Supplementary
Classification of External Causes of Injury and Poisoning (referred
to as E-codes). The morbidity file, only, contains ‘nature of
injury’ codes corrresponding to the ICD-9-CM injury and poisoning
diagnostic codes (referred to as N-codes). The E-codes were used
to group fatalities and hospitalisations by cause and the N-codes
were used to group hospitalisations by injury type.
Results
There were 33 fatalities and 215 hospitalisations associated with
river rafting during the period of the study, giving an average
of 2.5 fatalities (s.d. = 2.3) and 15.4 hospitalisations (s.d.
= 6.0) per annum.
Fatalities
More than three-quarters of victims (81.8%) were males. The victims’
ages ranged from 17 to 65 years (mean = 34.3yrs, s.d = 13.3yrs).
The age distributions for both fatalities and hospitalisations
are shown in Figure 1.
Twenty-two victims (66.7%) were residents of the North Island
of New Zealand; seven (21.2%) were residents of the South Island
of New Zealand; and four (12.1%) were from overseas. Table 1 describes
the external causes of fatalities based on E-code categories.
The majority of cases (66.7%) were classified as ‘accident to
watercraft causing submersion’(E-code 830). In total, 94% of fatalities
involved drowning or submersion.
Using the event descriptions on the mortality file, a number of
different events were identified that resulted in river rafting
fatalities during the period of the study. Table 2 shows the main
classifications of events that led to fatal incidents. The majority
of fatalities resulted either from the raft capsizing (36.4%)
or as a result of being thrown out of the raft and drowned (30.3%).
A further 12.1% were drowned after being trapped in some way.
Figure 1. Percentages of hospitalisations and fatalities by victims’
age group.
Hospitalisations
One hundred (46.5%) of the people hospitalised as the result of
a rafting injury were female and 115 (53.5%) were male. Victims’
ages ranged from 8 to 73 years (mean = 30.4yrs, s.d. = 11.7yrs).
Of those who were hospitalised, 145 (67.4%) were residents of
the North Island of New Zealand, 41 (19.1%) were residents of
the South Island of New Zealand, and 29 (13.5%) were overseas
residents. The annual number of hospitalisations ranged from 7
to 25 (mean = 15.4, s.d. = 6.0). The time spent in hospital ranged
from 0 to 43 days (mean = 3.3 days, s.d. = 4.1 days).
The ICD-9-CM N-codes were converted to Abbreviated Injury Scale
(AIS)8 severity scores by means of the ICDMAP9 program. The distribution
of the highest AIS scores showed that no cases received a score
above 3 (serious) which was received by only 10%, with the majority
of injuries (51%) receiving a score of 2 (moderate).
Table 3 shows the descriptions of external cause recorded for
hospitalisations based on E-code categories. These descriptions
provide some indication of the types of causes of recreational
river rafting injuries, such as falling and slip-ping (14.9% in
total) submersion due to fall from raft (14%), and being struck
by objects or persons (4.2%). The largest number of hospitalisations
were classified as ‘unspecified water transport accident’ (39.5%).
