Canadian soldiers patrol an area in the Dand district of southern Afghanistan on June 7, 2009. (Colin Perkel/The Canadian Press)Canadian soldiers have been part of the International Security Assistance Force (ISAF) in Afghanistan since 2002. It's been a dangerous posting for all of the coalition soldiers involved, with 1,480 coalition military deaths recorded as of Oct. 26, 2009, including 131 Canadians.

The number of Canadian soldiers on the ground there has always been much lower than the numbers of either American or British soldiers posted there. So one would expect that the number of Canadian military fatalities in the Afghan field would be lower than for their closest military allies — and it is.

Statistical analysis, however, also reveals that the Afghan mission is taking a much bigger toll on Canadian forces, proportionately speaking, than the other major coalition nations.

To arrive at that conclusion, the U.K. Medical Research Council's Biostatistics Unit examined coalition military death rates in Afghanistan in nine consecutive 140-day periods, beginning May 1, 2006 and ending Oct. 4, 2009.

During that 41-month period, 1,070 coalition soldiers died in Afghanistan (all fatality figures are from iCasualties.org). Of these, 590 were American, 213 were British, 116 were Canadian, and the rest were from the 40 other countries contributing soldiers to ISAF.

So the U.S. and Britain lost considerably more soldiers than Canada during that period. But since Canada had far fewer soldiers in Afghanistan throughout the entire period studied (see table below), statistical analysis by the Medical Research Council showed than Canadian soldiers regularly faced a higher fatality rate than either the Americans or the British. At one point, the observed Canadian fatality rate was more than quadruple the U.S. rate. [Story continues below graph.]

Coalition military deaths in Afghanistan
Country	May 1/06 - Sept. 7/06	Sept. 18/06 - Feb. 4/07	Feb. 5/07 - June 24/07	June 25/07 - Nov. 11/07	Nov. 12/07 - Mar. 30/08	Mar. 31/08 - Aug. 17/08	Aug. 18/08 - Jan. 4/09	Jan. 5/09 - May 17/09	May 18/09 - Oct. 4/09
U.S. deaths (troops)	54 (23,300)	18 (22,000)	50 (24,800)	58 (25,000)	25 (31,000)	88 (35,000)	53 (35,000)	54 (39,000)	190 (57,000)
U.K. deaths (troops)	33 (4,500)	6 (5,250)	15 (5,250 - 6,900)	22 (6,900)	8 (7,000)	24 (8,000)	24 (8,000)	21 (8,000)	60 (9,000)
Cdn. deaths (troops)	17 (2,250)	12 (2,250)	16 (2,500)	11 (2,500)	10 (2,500)	9 (2,500)	16 (2,500)	12 (2,500)	13 (2,500)
Source: U.K. MRC Biostatistics Unit

To be sure, the nature of the Afghan mission varies dramatically for each nation's soldiers. Canada's military efforts, for instance, are concentrated in the southern province of Kandahar. The coalition's soldiers face very different risks, depending on where they're based, the kind of work they're doing, the support they receive, the nature and degree of Taliban resistance, etc. But the study's authors note that an "explanation is needed for the consistently higher UK/Canadian than U.S. military fatality rates in Afghanistan … chance is not the explanation."

To arrive at fatality-rate figures, the study looked at troop levels for each nationality, the actual number of fatalities, and then calculated the fatality rate per 1,000 personnel years. A personnel year is defined as one soldier posted to the Afghan mission for one year. So 2,500 Canadian soldiers posted for the study period of 20 weeks or 140 days is equal to 962 personnel years.

Here's a graphic representation of what they found:

To simplify matters, we've left off any reference to confidence intervals in this graph. But statisticians point out that if one seeks to compare the true fatality rates, confidence intervals do matter. Because the Canadian contingent is smaller, the confidence interval is wider than for either the Americans or the British. But even after taking that into account, the Canadian fatality rate is still demonstrably higher.

For example, the study's authors added the first four periods (a total of 80 weeks) and found the Canadian fatality rate was 15.7 per 1,000 personnel years. With a 95 per cent confidence interval (in other words, the result one would expect 19 times out of 20 times), the Canadian death rate in that 80-week period would range from 12 to 20 per 1,000 personnel years. It's high enough that there's no overlap at all with the U.S. rate over that same period (4.2 to 5.6 deaths per 1,000 personnel years) or the British rate (7 to 11). The same condition holds for a summary of periods five through eight.

The study's authors note that the observed fatality rates for Canadian and British troops during period nine (the 20-week period ending Oct. 4, 2009) was "akin to the fatality rate for Soviet forces in Afghanistan in the 1980s."

It should also be noted that, as in Iraq, the number of civilian deaths in Afghanistan far exceeds the number of military deaths. A report by the UN Assistance Mission in Afghanistan (UNAMA) said 1,013 civilians were killed in the first six months of 2009 and acknowledged that the toll could be higher.