This note was originally posted to cdn-firearms on 24/11/1995 with the title: $6.5 billion annual cost of gunshot wounds critique. Minor editing and conversion to html format on 6/9/2001.
(C) Dr. Boris Gimbarzevsky 1995-2001

No paper better illustrates the deficiencies of peer review in the
medical literature (Suter, 1994) as the paper by Miller, (1995) entitled
"Costs associated with Gunshot wounds in Canada in 1991". This paper
suffers from false assumptions, biased selection of data and its
publication in the CMAJ demonstrates the intellectual bankrupcy of
the CMAJ editorial board.

We are told in the paper that: "1991 was chosen as the study year
because this was the most recent year for which incidence data were
available". Coincidentally, it was also the year in which there were the
highest number of firearms homicides since 1974. Recent homicide numbers
are given below:

Year Mtot Gtot Mother
1990 660 196 464
1991 756 271 485
1992 732 246 486
1993 630 193 437
1994 596 196 400
Here:
Mtot - total murders
Gtot - murders comitted with firearms
Mother - murders comitted with other means

The other untenable assumption is that firearms suicides are equivalent
to firearms homicides. Suicides are very difficult to prevent, and in
males, restriction of access to one method results in substitution of
equally lethal methods (Rich et al, 1990, Sloan et al, 1990). The number
of non-firearms suicides is over twice as large as firearm suicides. For
1991 and 1992 the numbers are:

Year Stot Sgun Sother
1991 3593 1118 2475
1992 3709 1050 2659
where:
Stot - tot # of suicides
Sgun - suicides with firearm
Sother - suicides with other means

Some of the assumptions are laughable: each death is considered to have
a lifelong productivity loss of $1,016,220 and a "quality of life loss"
of $3,094,552. How the following ludicrous assumption:
"For valuing quality-of-life and productivity losses, it was
assumed that people who died of gunshot wounds would have the same
expected lifespan as other Canadians of comparable age and sex"
got past the reviewers is a mystery. Stats Canada notes that 40% of
homicide victims have consumed alcohol and drugs at the time they were
killed (often in very large quantities judging from the people I have
seen come into emergency). To not account for the high risk lifestyle
that perhaps 50% of more of homicide victims carry out is astounding.
US data has shown significantly higher rates of criminal convictions and
ongoing criminal activity among homicide victims. The CMAJ reviewers of this
paper put themselves in good company with their assumptions.

The fallacy of such assumptions can best be seen in which deaths result
from legal intervention - the 13 deaths resulting from such means in 1991
are considered to have cost Canada $60.6 million dollars. There have
been cases where innocent people have been killed by police, but even if
we assume that 50% of cases in which criminals are shot and killed by
police are found to be totally justified, the net result is a GAIN for
society, not a loss as is assumed.

Using the $4,130,346 cost of a homicide or suicide, some very interesting
"costs" are shown in the table below. If guns are bad for Canada, then
knives, bridges and high buildings are even more devastating. All costs
in table below in millions of dollars:

Homicide Suicide Guns Other
Year Guns Other Guns Other TOTAL Total Total
1991 1177 2003 4617 10222 18019 5794 12225
1992 1016 2007 4336 10982 18431 5352 12989

Using the same ludicrous assumptions that have been made for guns, we
find that homicide and suicide in Canada cost $18.0 BILLION in 1991 and
$18.4 Billion in 1992. For comparison, the 1301 AIDS deaths in Canada in
1992 cost $5.37 billion, MVA deaths would cost $14.84 billion in 1992,
and the 535 heroin overdose deaths in BC in 1993 and 1994 cost Canada
$2.2 billion.

Even in the year in which guns had the greatest "cost", 1991, using
the ludicrous assumptions for the worth of criminal life, guns were
responsible for only 32% of the societal "cost" of homicide and suicide.
In 1992 this had decreased to 29%. If Stalin Rock borrowed Bouchard's
magic wand and wished away all guns in Canada, there would still be over
$13 billion dollars in costs to society from homicide and suicide.
(Note that I don't get a total of $6.5 billion yearly above because I
only consider fatal events, not all firearms injuries. Using the
assumptions for these non-fatal events, one would have an astronomical
cost of non-fatal knifings, but I digress).

Sloan et al (1990) is quoted in this paper, but what is not mentioned
is that Sloan et al, in their comparison of Seattle and Vancouver,
found that the Canadian suicide rate was higher than the US rate, but
pistols were used much more frequently in Seattle for suicide. Rich,
(1990) showed that people jump off tall buildings or bridges if they
can't get guns. The conclusions as presented are simply wrong and can
only be considered to be motivated by ideologic considerations rather
than any reasonable standards of scientific proof.

This paper is actually far worse than I thought, near the end the
following, totally unjustified, comments are made:
"The preliminary estimates from this Canadian cost analysis study
are a first step toward understanding the economic burden of
gunshot wounds on the already limited resources of Canada's health
care system"
and:
"Several studies have suggested that controlling the ready access
to guns could reduce the number of suicides, which account for the
largest share of costs for gunshot wounds".

The publication of this paper in the CMAJ is totally irresponsible.
For several days I have been hearing about the $6.5 billion dollar cost
to Canada from gunshot wounds on the local media. This is true only if
one assumes that members of the editorial board of the CMAJ have the
same societal worth as the drug-dealing biker casualties of the ongoing
conflict between the Montreal Hells Angels and their business rivals.
I'll leave it to readers of this group to decide if such comparisons
are justified.

I'm too pissed off to write a short letter to the CMA, but if anyone
else wishes to, go ahead. Incidentally, one can get hold of the
author, Ted Miller at miller@pire.org - drop him a line and ask why he
made these incomprehensible assumption.

Boris Gimbarzevsky, MSc, MD.

References:
Miller, TR. Costs associated with gunshot wounds in Canada in 1991.
Canadian Medical Association Journal. 153(9):1261-1268

Rich, C. L. et al. Guns and Suicide: Possible Effects of Some Specific
Legislation. American Journal of Psychiatry 147(3):342-346, (1990)

Sloan, J. H. et al. Firearm Regulations and Rates of Suicide - A
Comparison of Two Metropolitan Areas.
New England Journal of Medicine 322(6):369-373, (1990)

Suter, EA. Guns in the medical literature -- a failure of peer review
Journal of the Medical Association of Georgia. 83(3):133-148(1994)

Stats Canada Juristat 1993 and other publications used for
homicide/suicide stats. There are some minor discrepancies between my
numbers and those in paper - likely these are due to suicides being
reclassified as homicides or vice-versa.