War-wounded civilians are the focus of this article. Iraq's health care system, already weakened before the devastating 2003 invasion, now has to cope with a relentless stream of daily bomb blasts and shootings.

Official indifference (by the nations which invaded Iraq) to the welfare of living victims persists.

## 3. Three times as many injuries as deaths have been reported

An informative statistic for analysing and evaluating injuries is the RATIO of injuries to deaths for a given incident. This ratio can be calculated by dividing the maximum estimate of injuries by the maximum estimate of deaths. If there are equal numbers of injuries to deaths, then this ratio is 1.0. If there are twice as many injuries as deaths, this ratio is 2.0. If there are twice as many deaths as injuries, this ratio is 0.5.

18 of the 107 incidents had a injury-to-death ratio of less than 1.0, and 7 incidents had a ratio of exactly 1.0. The remaining 82 incidents had an injury-to-death ratio of greater than 1, with a maximum ratio of 69. This maximum ratio was provided by 207 reported injuries and 3 reported deaths during massive aerial bombardment of Baghdad on the night of 21-22 March.1 Although the reports of injuries were provided by Iraqi government sources, independent estimates from the Red Cross confirmed at least 100 injuries, which still represents a massive injury-death ratio of 33. This lends some support to the claims that parts of the air-war (particularly in and around Baghdad) were conducted using precision-guided munitions, where there were few deaths but many injuries from falling and flying masonry, shrapnel etc. Most of the larger ratios were indeed the result of aerial bombardment, relatively early in the campaign. The smaller ratios typically come from the later ground war and “post-war” conflict.

If one wished to answer the question “what is a typical, or average” ratio
of injuries to deaths, there are two statistical averaging procedures which
might be used. One is the **mean** ratio (the mean is the sum of all ratios
divided by the number of incidents from which ratios could be calculated).
The mean injury-death ratio is 5.0 (in other words, 5 injuries per death).

A second averaging procedure is the **median** ratio. This is
found by setting out all 107 ratios in ascending order, and picking the ratio
which occurs at the 54th position (i.e. in the middle of the series). The median
injury-death ratio is 2.85 (in other words, around three injuries per death).

Often the mean and the median of a set of scores are quite close to one another. The mean tends to differ from the median when the distribution is statistically skewed. The distribution of injury-death ratios in the IBC data base is indeed skewed, with a small number of incidents having very high injury-death ratios, which are not typical of the larger number of incidents. Only 23 of the incidents have an injury-death ratio of greater than 6, with the majority of these being below 10. However the “top” 10 incidents have injury-death ratios, in ascending order, of 10.2, 13.1, 13.9, 16.2 16.6, 17.8, 20, 24, 45, and 69. These few incidents skew the mean upwards.

In our view, the more “typical” estimate is given by the median.
This would suggest that, on average, in a typical incident in this war, there
were about 3 injuries for every death. **Multiplying the 7711 maximum
reported deaths (up to July 7th, 2003) by the median of 2.85 provides a figure
of 21,976, which might be considered a more accurate estimate of injuries that
takes into account the 43 database entries for which injuries were not reported
and other data absent from the IBC database, as noted earlier.**

Many of the reports of injuries are simply anonymous numbers. But Western journalists were sometimes able to get close to the field of battle and report their encounters with the wounded.