Introduction

Top Abstract Introduction Patients and methods Results Discussion References

The use of weapons against people or targets containing people inevitably has a direct impact on the health of those people. ^{1 2} This impact is related to factors dependent on the design of weapons and on their use. The nature of injury is closely related to the design of the weapon; wounds from bullets, fragments, and buried antipersonnel mines are distinguishable.^3-6 Factors dependent on the user, such as discipline and desire to avoid or injure civilians, determine the number and kind of people injured^5-10 and may, in the case of bullets, determine which part of the body is injured. This century has seen an increased proportion of civilians injured during war.¹⁰ This is usually ascribed to military weapons passing into the hands of those with no respect for the civilian population or the Fourth Geneva Convention, which protects civilians. In parallel, there has been an extraordinary development of the military efficiency of weapons.¹¹ This generates a provocative question: to what extent is the weapon development this century linked to the increased proportion of civilians injured? This poses a further question: does increased ease with which a weapon can be used to achieve military objectives (military efficiency) increase the potential for civilian casualties?¹²

The hallmarks of countries where most modern wars are fought are poverty, destroyed social and economic infrastructure, and availability of a variety of weapons.¹ Disciplined armies train their soldiers in the laws of war, which include respect for the civilian population; by contrast, modern wars tend to be fought by forces that are poorly trained and may even target civilians. Another feature of these modern wars is that competent medical facilities are few or non-existent. Care of those wounded during these conflicts has fallen to international aid agencies. One of the few sources of data about casualties in these wars is the hospitals run by the International Committee of the Red Cross. We examined all the data held by the Red Cross on wound injuries treated in its hospitals from January 1991 to July 1998 to explore these two questions. We also examined data from the Kabul hospital during a period when the city of Kabul was under siege.

	Patients and methods

Top Abstract Introduction Patients and methods Results Discussion References

Database
The wound database of the International Committee of the Red Cross was installed in January 1991 and originates from a system of data collection originally designed to give the organisation an indication of the activities of its independent hospitals. All patients wounded in war who have been admitted to the Red Cross hospitals of Peshawar and Quetta (Afghan border of Pakistan), Kabul and Khandahar (Afghanistan), Khao I Dang (Cambodian border of Thailand), Butare (Rwanda), Novi Atagi (Chechenia), and Lokichokio (Sudanese border of Kenya) have routinely had a data form filled out on their death or discharge from surgical wards. Age and sex, the cause of injury, and the time lapsed between injury and admission are recorded for each patient. Patients are not asked whether they are combatants.

Kabul
The Red Cross hospital in Kabul, Afghanistan, functioned independently until the fall of the communist government in mid-1992. It was the first of its kind to be in a city under siege rather than removed from the conflict over a border. Where the hospital was working was thus the same place as where patients were wounded. Patients were wounded in the city itself and at the front lines surrounding the city. Those wounded among the rebel forces besieging the city had access to the first aid posts run by the Red Cross outside the city and then were transported to the hospital by the organisation's ambulances; few reached the hospital within six hours. By contrast, those wounded in the city reached hospital usually within an hour and certainly within six hours.¹³ Patients in the city were representative of victims of urbanised, modern conflict; many were clearly civilians.

Analysis
The patients' data were analysed by age and sex and the cause of injury. As in previous studies, ^{5 6 9} women and girls, boys (under 16 years of age), and men of 50 or more were considered to be civilians. In this study bullet indicates any gunshot wound, fragment indicates injury from shell, bomb, or mortar, and mine indicates injury from an antitank or antipersonnel mine. Differences in the proportion of people injured by bullets in comparison with mortars or mines were evaluated using the ² test.

	Results

Top Abstract Introduction Patients and methods Results Discussion References

A total of 27 825 patients were registered between January 1991 and July 1998. Of these, 18 877 were injured by bullets, bombs, shells, mortars, or mines; the rest were admitted because of burns or blunt trauma or for reconstructive surgery. Of the 18 877 who were injured by weapons, 2012 were admitted to the Kabul hospital in less than six hours after injury.

Table 1 shows the numbers of people who were injured by bullets, mortars, or mines and the proportions who were civilians. Table 2 shows the same information for patients admitted to Kabul hospital within six hours of injury. In all the hospitals and in Kabul under siege a significantly greater proportion of civilians had been injured by mortars or mines than by bullets (P<0.001).

	Discussion

Top Abstract Introduction Patients and methods Results Discussion References

Limitations
These data are probably the best available means of examining the direct human impact of the use of weapons in modern conflicts. Their validity and reliability have not been ascertained by formal independent means because of the constraints imposed on collecting them under field conditions, and there is obvious scope for misclassification. Some patients lie about how they were injured to gain admission to hospital or they may not know exactly what injured them, and our means of classifying patients as combatants or civilians is a potential source of error. Nevertheless, any misclassification in this setting is likely to have underestimated the numbers of civilians. The number of men aged 16-49 who were civilians was probably greater than the combined number of women, boys, and men over 49 who were combatants. Thus the proportion of civilians is almost certainly higher than the proportions given here.

Weapon type and civilian injuries
To our knowledge, the implications for civilian injuries brought by different weapons has not been fully examined before. These data show that factors relating to both the design of weapons and the discipline or intent of the user have implications for civilian injuries.

Weapons, law, and preventive medicine
The process of making or promoting policy and law entails analysing data which clarify the nature of the problem that the policy or law is trying to avoid. International humanitarian law is no exception. These data show that the number of civilian injuries is related not only to whether weapons are in the hands of untrained and undisciplined users but also to the type of weapon in those hands. This argues for a greater need to control the transfer of weapons of increasing military efficiency and warrants urgent and serious examination of states' obligations under international humanitarian law in relation to arms transfer. Such an examination should naturally follow the precedent set by the drawing up of a treaty banning the production, stockpiling, transfer, and use of antipersonnel mines. The medical profession has a responsibility to examine the global weapon problem as a health issue¹²; this is a form of preventive medicine.

	Acknowledgments

   Contributors: HOS first found variation in age and sex of patients according to cause of injury on examining the ICRC surgical database. RMC had the core idea that the weapon type may lend itself to indiscriminate use, formally re-examined the database, and wrote the paper. RMC is guarantor for the study.

	Footnotes

Funding: No additional funding.

Competing interests: None declared.

	References

Top Abstract Introduction Patients and methods Results Discussion References

1.	Sidel VW. The international arms trade and its impact on health. BMJ 1995; 311: 1677-1680[Free Full Text].
2.	Coupland RM. The effects of weapons on health. Lancet 1996; 347: 450-451[Medline].
3.	Bowyer GW. Management of small fragment wounds in modern warfare: a return to Hunterian principles? Ann R Coll Surg 1997; 79: 175-182.
4.	Sellier KG, Kneubuehl BP. Wound ballistics. Amsterdam: Elsevier , 1994.
5.	Coupland RM, Korver A. Injuries from antipersonnel mines: the experience of the International Committee of the Red Cross. BMJ 1991; 303: 1509-1512.
6.	Coupland RM. Hand grenade injuries among civilians. JAMA 1993; 270: 624-626[Abstract/Free Full Text].
7.	Andersson N, Palha da Sousa C, Paredes S. Social cost of land mines in four countries: Afghanistan, Bosnia, Cambodia and Mozambique. BMJ 1995; 311: 718-721[Abstract/Free Full Text].
8.	Ascherio A, Biellik R, Epstein A, Snetro G, Gloyd S, Ayotte B, et al. Deaths and injuries caused by landmines in Mozambique. Lancet 1995; 346: 721-724[Medline].
9.	Jeffrey S. Anti-personnel mines: who are the victims? J Accid Emerg Med 1996; 13: 343-346[Abstract/Free Full Text].
10.	Garfield RM, Neugut AI. Epidemiological analysis of warfare: a historical analysis. JAMA 1991; 226: 688-692.
11.	Prokosch E. The technology of killing. London: Zed Books , 1995.
12.	International Committee of the Red Cross. The medical profession and the effects of weapons: the symposium. Geneva: ICRC , 1996.
13.	Coupland RM. Epidemiological approach to the surgical management of the casualties of war. BMJ 1994; 308: 1693-1697[Free Full Text].
14.	Grossman D. On killing: the psychological cost of learning to kill in war and society. Boston, MA: Little, Brown , 1995.

(Accepted 29 September 1998)