- Email: [email protected]
Musculoskeletal injuries in the Afghan war
Injury
(1992)
23,
(8), 545-548
545
Printed in Great Britain
Musculoskeletal
injuries in the Afghan war
M. K. Bhatnagar, M. J. Curtis and G. S. Smith The Johns Hopkins Hospital, Baltimore, Maryland,
USA
Among the 1274 pafienb admitted to a Pakistan border hospital from 7985 to 1987, the distribution and oukome of musculoskeletal war injuries differed from those seen in other conflicts. Serious complications horn injuries were found in approximately 50 per cent of patients, of which most were wound infections, chronic osteomyelitis, and restriction of joint motion. Guerrillas in the Afghan war had no access to acute medical treatment in the field. Many patients died before reaching the hospital, as reflected in the low proportion of paraxial injuries; very high complication rates were noted for all injuries. Although some complications, such as soft tissue infection and foreign body r&mentionare nof site spec$c, other complications such as contracture, non-union, loss of range of motion, and chronic osteomyelifis are highly related to the region injured. Early surgical management and evacuation of those with musculoskeletal war injuries can greatly improve the outcome from war trauma and reduce the subsequent disability. However, the iweasing use of hand-held antiaircraft missiles may prevent the rapid evacuation of fhewoundedin future conflicts, and may make fhe situation seen in Afghanistan more common.
Introduction The importance of early surgical and antibiotic treatment is well-recognized in the management of military casualties (Aldea et al., 1987; Cozen, 1985). Central to reducing mortality, long-term complications, and disability, has been the use of helicopters and other means of rapid evacuation of the wounded for definitive surgical care (Bellamy et al., 1986; Carey, 1987). However, with the advent and widespread use of hand-held anti-aircraft missiles, the ability of opposition forces to shoot down evacuation aircraft has raised concern that in some conflicts rapid evacuation of the wounded may not be possible. Wounded guerilla soldiers in the recent civil war in Afghanistan did not have access to modem facilities and had to rely on ground transportation, often involving walks or animal rides of up to 6 days’ duration before receiving definitive surgical care at hospitals over the Pakistan border. This lack of early surgical attention greatly increased morbidity and mortality (Bhatnagar and Smith, 1989). This paper examines the natural history of high-velocity war wounds among those Afghan guerrilla fighters who survived long enough to reach medical care, the complications following these injuries, and the impact of delayed treatment on subsequent disability.
Patients and methods All patients admitted to the Afghan Surgical Hospital on the Afghanistan-Pakistan border from 22 February 1985 to 22 0 1992 Butterworth-Heinemann 002C-1383/92/080545-04
Ltd
February 1987 were reviewed retrospectively. Medical records were abstracted by one of the authors (M.B.) for age, sex, site of injury, type of injury, complications of the injury, hospital management and residual deformity. More detailed description of the area and hospital facilities are described elsewhere (Bhatnagar and Smith, 1989). The injuries seen at the hospital do not represent all injuries sustained by the guerrilla fighters, since many of the most severely injured may have died before arrival at the hospital. Sites of injury were grouped for analysis of the data. Distal extremity includes injuries to the wrist, hand, fingers, ankle, and foot. Long bone - metaphyseal/joint, represents injuries to the shoulder, elbow, hip and knee. Long bones diaphyseal, involves injuries to the arm (humerus), forearm (radius and ulna), thigh (femur), and leg (tibia and fibula). Non-union was defined as a fracture which showed no clinical or radiographic evidence of progressive healing at 6 months after injury. Malunion was defined as a healed fracture with deformity sufficient to require surgical intervention in order to improve function. Contracture was defined as a soft tissue injury resulting in loss of range of joint motion sufficient to interfere with function. Statistical analysis was performed using multiple logistic regression, utilizing the Generalized Linear Interactive Modelling system (GLIM, release 3.77), Royal Statistical Society, London.
Results A total of 1254 patients with 1292 musculoskeletal injuries were admitted to this surgical hospital during the 2-year period; all were male. The casualties ranged in age from 2 to 74 years; 86 per cent were in the 15-50 year age group. Of these, 3 per cent were treated for injuries to both upper and lower extremities, the majority of whom had a primary upper extremity injury with less severe injuries to the lower extremities. Lower extremity injuries were seen 1.6 times more frequently than those of the upper extremity. The most commonly injured region was the leg (21.4 per cent), followed by the thigh (17.6 per cent), forearm (9.3 per cent), and knee (8.4 per cent). Only 19 patients (1.5 per cent) presented with para-axial injuries to the scapula, spine, ilium, pelvis, clavicle, or axilla (Table I). None of these patients had any other important injuries. Fractures comprised almost 30 per cent of all injuries seen; all were grade III open fractures, from high-velocity projectiles and were grossly contaminated (Gustilo and Anderson, 1976). The diaphyses were injured three times more
Injury: the British Journal of Accident Surgery (1992)Vol. 23/No. 8
546 TableI.
Para-axial injuries by site of injury, Afghan Surgical Hospital, Pakistan, 1985-1987 No.
Site of injury
Scapula
2 1 5 1 3 7
Spine Iliac Pelvis Clavicle Axilla
19
Total injuries
Table II. Neurological injuries in the Afghan war Lower extremity
Upper extremity Site of injury Brachial nerve Median nerve Radial nerve Ulnar nerve Ulnar+ median nerve Ulnar+ radial nerve Unspecified Total UE injury
No. 1 7 5 19 5 6 7 50
NO.
Site of injury
4 4 4 1 1
Sciatic nerve Peroneal nerve Tibia1 nerve Sural nerve Tibia1 +sural nerve Unspecified Total LE injury
3 17
UE, upper extremity; LE, lower extremity.
frequently than the metaphyses and joints. The most frequently fractures bones requiring surgery were the tibia and fibula (31.3 per cent), followed by the femur (27.3 per cent), humerus (13.5 per cent), and forearm (12.2 per cent). The fractures were comminuted in 84 patients, 76 of which were diaphyseal injuries; the risk of comminution was significantly less in the femur compared with other long bones (P < 0.05). There were 86 patients (6.7 per cent) with major neurological deficits as a result of their wounds (TiibkZI). Upper extremity nerve damage was significantly more common than lower extremity nerve damage (PC 0.001) and the ulnar nerve accounted for almost 45 per cent of all nerve injuries. Foreign bodies, e.g. shrapnel, etc., were seen’in approximately 6.3 per cent of the injuries; soft tissue infections were present in 13.6 per cent of all cases, the majority in the thigh and leg. There was no statistically significant difference in the presence of foreign bodies or in the rate of soft tissue infection between body regions (TiableIII). Chronic osteomyelitis was present in 11.3 per cent of the patients. The femur and tibia accounted for 61 per cent of all osteomyelitis, with the rate of osteomyelitis in diaphyseal fractures being significantly higher than for metaphyseal fractures
(P <
0.001).
Malunion was a complication in 3.6 per cent of cases, that is more than 10 per cent of the fractures; 10 per cent of elbow injuries had malunion requiring surgical correction; the radius and ulna, femur and tibia required surgical correction of malunion in approximately 6 per cent of patients. There was no statistically significant different in the rate of malunion for various fractures. There were 43 fracture non-unions, 41 of which were long bone diaphyses. The risk of developing non-union in the diaphyses was eight times higher than in either the joints or the distal extremity, P < 0.005 (TableIV). The diaphyseal non-unions were distributed between the arm, forearm, thigh, and leg; fractures of the humerus had the highest frequency of non-union (8.7 per cent), but this difference was not statistically significant. Amputations were required in 82 instances, 7.2 per cent of all patients (T&dell/), and were performed most frequently for injuries of the distal extremities, especially the foot (22 per cent of which required amputation). Injuries of the leg, ankle, and foot accounted for 55 out of 87 amputations (63.2 per cent) and injuries of the leg resulted in 35 amputations (40.2 per cent). The risk of a distal extremity injury requiring amputation was six times that for large joint injury (PC O.OOl), and 1.75 times that for a long bone injury (P < 0.02). Poor range of motion was found in 5 per cent of the patients (TableIII), and was most likely to occur with injuries of joints, most frequently the shoulders and knees. The ankle was also frequently involved, and mobility was limited in 37 per cent of ankle injuries. The risk of losing motion was more than doubled with a joint injury when compared with injuries of the long bones. Fixed flexion contractures were present in 5 per cent of the patients overall; 22.7 per cent of hand injuries and 24.4 per cent of phalangeal injuries resulted in contractures. Few contractures or difficulties with range of motion occurred
Table III. Relative risk of injury complications by body regions, Afghan Surgical Hospital, Pakistan, 1985-1987 Wound sepsis Distal extremity (N= 257) Long bone metaphyseal/joints (N= 257) Long bone diaphyseal (N=692)
Poor ROM’
Chronic osteo l
Foreign body
1
1
1
1
1.23 NS
2.38 P
1.33 NS
1.06 NS
0.81 NS
0.17 P< 0.001
3.39 P
0.79 NS
‘ROM, range of movement. NS, not statistically significant. Chronic osteo, Chronic osteomyelitis
Table IV. Relative risk of complications by body regions, Afghan Surgical Hospital, Pakistan, 1985-1987 Amputation Fracture distal extremity (N= 257) Long bone metaphyseal/joint (N=257) Long bone diaphyseal (N=692) NS, Not statistically significant
1
Non-union
Malunion
Simple Fx
1
1
1
Comminuted Fx
1
0.07 P-E O.OCll
0.20 NS
1.45 NS
0.89 NS
0.33 NS
0.35 P< 0.02
8.03 P-z 0.005
1.39 NS
5.14 P-z 0.001
6.16 P< 0.001
Bhatnagar et al.: Muscdoskeletal injuries in the Afghan war
547
with injuries of the arm, forearm, thigh, or leg. The relative chance of developing contracture was 12 5 times higher in the distal extremities than in the long bones, and more than three times higher in the distal extremities than in the joints.
arterial and 41 per cent of injuries were vascular among police officers injured by firearms (Jones et al, 1968; Berman and Salter, 1985). The lethal nature of extremity trauma with vascular complication is demonstrated by reports that these injuries were the cause of death in one of every ten United States servicemen who died in Korea and Vietnam (Russotti and Sim, 1985). Wound infections were present in 13.6 per cent of the patients treated. There was no statistically significant difference in the rate of infection according to site of injury. The wartime rate of infection is known to be higher than civilian injuries; SO per cent of wounds from the Vietnam war were culture positive and 21.6 per cent developed soft tissue infections (Seindstein and Newman, 1968). The rate of infection seen in the Arab-Israeli War was 14 per cent (Michaeli, 1979). Jackson reported an overall 18 per cent infection rate in the Falklands Campaign (Jackson et al., 1983) which increased to 25 per cent for those injuries treated more than 6 h after injury. The rate of infection seen in civilian injuries ranges from 1.5 per cent to 3 per cent (Gustilo and Anderson, 1976; Elstrom et al., 1978; Rittmann et al., 1979; Woloszyn et al., 1988). In civilian population study groups, the rate of infection approached wartime rates only in patients who did not receive antibiotics (14-20 per cent) and in whom primary treatment was delayed by more than 8-12 h (Tscheme and Gotzen, 1984). In the Afghan hospital population, the rate of chronic osteomyelitis was significantly higher in the long bones, especially the femur and tibia. The metaphyseal and joint injuries were of equal severity and contamination, but the greater vascularity of these bony regions may offer some protection against infection (Ketenjian and Shelton, 1972; Franklin et al., 1984). The greater incidence of infection of open tibia1 fractures compared with femoral fractures has been attributed to the decreased soft tissue coverage in this region (Chapman, 1986; Holden, 1972; Ryan et al., 1981). The overall rate of osteomyelitis is very similar to that observed by others (Rittmann et al., 1979). Similarly, the long bones had an eightfold greater risk of developing non-union than distal extremity injuries. The amputation rate for distal extremity injuries was twofold higher than that for long bones, and almost sixfold higher than for injuries to joints. This reflects the severe localized injuries sustained by blast injuries such as exploding land mines. Elbow and tibia1 fractures had high rates of malunion, but there was no statistically significant difference in malunion rates between the joints and long bones. The definition of malunion to include only those cases severe enough to require secondary surgery to improve function will certainly underestimate the overall incidence as estimated from radiographic measurements. Functional loss of mobility was much more common after injury to the large joints than to the distal extremities and to long bones. Of the patients, 5 per cent suffered from severe functional restriction, significantly higher than the few cases reported by Rittmann et al. (1979). This is probably due to the traditional Afghan practice of immobilizing soft tissue trauma and delay in seeking care. The Mujahidin tightly bandage and immobilize wounded extremities, consequently leading to adhesions within joints and contractures. This practice was the primary cause of residual impairment for hand and finger injuries in the absence of fracture which contrasts with a low rate of impairment reported in other countries by Duncan and Kettlekamp (1974) and Granberry (1973).
Discussion The 1254 patients reviewed include all civilian and military war casualties treated at the Afghan Surgical Hospital for management of musculoskeletal trauma. This hospital was established specifically to treat male casualties from one of the seven guerilla factions during the Afghan Civil War (Bhatnagar and Smith, 1989). Females were treated elsewhere. The main causes of injury seen in this study were fragmentation weapons (50 per cent) (Bhatnagar and Smith, 1989) similar to those in the Korean, Vietnam, and Falklands wars, but the distribution of injuries by body region differs greatly (Jones et al., 1968; Johnson et al., 1981; Jackson et al., 1983). Though lower extremity injuries were more frequent than upper extremity injuries, paralleling other conflicts, only 3 per cent of the patients had multiple injuries, suggesting that the most severely injured died before receiving treatment. Furthermore, only 19 patients (1.5 per cent) had injuries to the para-axial region, which is far less than would be expected based on other conflicts (Carey, 1987). Many para-axial injuries involve the thorax and abdomen, with accompanying trauma to vital structures, which is rapidly fatal if not treated immediately (Bellamy et al., 1986). Few such patients survive the arduous trip across the mountains to our hospital. Management of fractures and excision of wounds represented about two-thirds of the surgical cases, a similar proportion to that reported by Morris at a similar hospital on the Pakistan border (Morris et al., 1985). Serious complications from injuries were found in approximately 50 per cent of patients, of which most were soft tissue infections, chronic osteomyelitis, and restriction of joint motion. This figure is much higher than that described in other conflicts. Cornminuted fractures were significantly less common in the femur than other fractured long bones; this is likely to be due to the bulk of overlying musculature which absorbs and dissipates significant amounts of the energy from the projectiles and the fact that the direct impact of blasts from land mines is sustained by lower regions of the leg. There were 86 patients (6.7 per cent) with severe neurological deficits requiring surgery. This is fewer than that seen by Berman and Salter (1985) among police officers injured by low-velocity gunshot wounds (I4 per cent), but is similar to the prevalence of neurological injury observed by De Wind (1987) in the war in Uganda (6 per cent), and that reported by Jones et al. (1968) in Vietnam (3.6 per cent). The most commonly injured nerve was the ulnar nerve: ali neurological deficits were secondary to axonotmesis. The paucity of soft tissue covering nerves in the upper limb, in particular the ulnar nerve, accounts for the increase in nerve injury from these high-velocity injuries, relative to the lower limb which is more frequently injured overall. Similarly, there were no patients seen in our study with severe vascular injuries; arterial injuries usually involve greater wounding power, and the considerable delay in the availability of medical treatment probably resulted in these patients dying before they could reach the hospital. During the Vietnam conflict, almost 5 per cent of injuries were
548
Injury: the British Journal of Accident Surgery (1992) Vol. 231N0.8
References Aldea P. A., Aldea G. S. and Shaw W. W. (1987) A historical perspective on the changing methods of management of major trauma of the lower extremity. Surg, Gyneco]. Obstet. 165, 54. Bellamy R. F., Maningar Pa. and Vayer J. S. (1986) Epidemiology of trauma: military experience. Ann. Emwg. Med 15, 1384. Berman A. T. and Salter F. (1985) Low-velocity gunshot wounds in Police officers. Clin. O&p. 192, 13. Bhatnagar M. K. and Smith G. S. (1989) Trauma in the Afghan guerilla war: effects of lack of access to care. Surgery 105,699. Carey M. E. (1987) Learning from traditional combat mortality and morbidity data used in the evaluation of combat medical care. M&f. Med. 152, 6. Chapman M, W. (1986) The role of intramedullary fixation in open fractures. C/in. Or/hop. 212, 26. Cozen L. N. (1985) Military orthopedic surgery. C/in. Orthop. 200, 50. DeWind C. M. (1987) War injuries treated under primitive circumstances: experiences in a Ugandan mission hospital. Br.]. Surg. 74, 193. Duncan I. and Kettlekamp D. (1974) Low velocity gunshot wounds of the hand. Arch. Surg. 109,395. Elstrom J. A,, Pankovich A. M. and Egwele R. (1978) Extraarticular low-velocity gunshot fractures of the radius and ulna. 1. Bone ]oinr Surg. 60A, 335. FranWin J. L., Johnson K. D. and Hansen S. T. (1984) Immediate internal fixation of open ankle fractures. ]. Bone ]oint Sr4rg. 66A, 1349. Cranberry W. (1973) Gunshot wounds of the hand. Hand 5,220. Gustilo R. B. and Anderson J. T. (1976) Prevention of infection in the treatment of one thousand twenty-five open fractures of long bones. ]. Bone ]oint Surg. WA, 453. Holden C. E. A. (1972) The role of blood supply to soft tissue in the healing of diaphyseal fractures. ], Bone ]oint Surg. HA, 993. Jackson D. S., Batty C. G., Ryan J. M. et al. (1983) The Falklands war: army field surgical experience. Ann. R. Co/l. Surg. Engl. 6.5, 281.
Johnson D. E., Panijayanond P., Lumjiak S. et al. (1981) Epidemiology of combat injuries in Thailand. 1. Trauma 21, 484. Jones E. E., Peters A. F. and Gasior R. M. (1968) Early management of battle casualties in Vietnam. Arch. Surg. 97, 1. Ketenjian A. Y. and Shelton M. L. (1972) Primary internal fixation of open fractures: A retrospective study of the use of metallic internal fixation in fresh open fractures. ]. Truumu 12, 756. Michaeli D. (1979) Medicine on the battlefield: a review. ]. R. Sot. Med. 72,370. Morris D., Sugrue W. and McKenzie E. (1985) On the border of Afghanistan with the International Committee of the Red Cross. N.Z. Med. J 98, 750. Rittmann W. W., Schibli M., Liatter P. et al. (1979) Open fractures: long-term results in 200 consecutive cases. CIin. Orthop. 138, 132. Russoti G. M. and Sim F. H. (1985) Missile wounds of the extremities: a current concept review. Orthopedics 8, 106. Ryan J. R., Hensel R. T., Salciccioli G. G. et al. (1981) Fractures of the femur secondary to low-velocity gunshot wounds. ] Trauma 21, 160. Seindstein M. and Newman A. (1968) Some clinical factors involved in the healing of war wounds. Arch. Surg. 96, 176. Tscheme H. and Gotzen L. (1984) Fractures with .!%ffTissue Injuries. New York: Springer-Verlag. Wolosryn J. T., Uitvlugt G. M. and Castle M. E. (1988) Management of civilian gunshot fractures of the extremities. Ch. Orfhop. 226, 24.
Paper accepted 5 May 1992.
Requestssfor reprints should be addressed to: Dr Bhatnagar, Department of Orthopaedic Surgery, The Johns Hopkins Hospital, 600 N. Wolfe Street, Baltimore, Maryland 21205, USA.
(1992)
23,
(8), 545-548
545
Printed in Great Britain
Musculoskeletal
injuries in the Afghan war
M. K. Bhatnagar, M. J. Curtis and G. S. Smith The Johns Hopkins Hospital, Baltimore, Maryland,
USA
Among the 1274 pafienb admitted to a Pakistan border hospital from 7985 to 1987, the distribution and oukome of musculoskeletal war injuries differed from those seen in other conflicts. Serious complications horn injuries were found in approximately 50 per cent of patients, of which most were wound infections, chronic osteomyelitis, and restriction of joint motion. Guerrillas in the Afghan war had no access to acute medical treatment in the field. Many patients died before reaching the hospital, as reflected in the low proportion of paraxial injuries; very high complication rates were noted for all injuries. Although some complications, such as soft tissue infection and foreign body r&mentionare nof site spec$c, other complications such as contracture, non-union, loss of range of motion, and chronic osteomyelifis are highly related to the region injured. Early surgical management and evacuation of those with musculoskeletal war injuries can greatly improve the outcome from war trauma and reduce the subsequent disability. However, the iweasing use of hand-held antiaircraft missiles may prevent the rapid evacuation of fhewoundedin future conflicts, and may make fhe situation seen in Afghanistan more common.
Introduction The importance of early surgical and antibiotic treatment is well-recognized in the management of military casualties (Aldea et al., 1987; Cozen, 1985). Central to reducing mortality, long-term complications, and disability, has been the use of helicopters and other means of rapid evacuation of the wounded for definitive surgical care (Bellamy et al., 1986; Carey, 1987). However, with the advent and widespread use of hand-held anti-aircraft missiles, the ability of opposition forces to shoot down evacuation aircraft has raised concern that in some conflicts rapid evacuation of the wounded may not be possible. Wounded guerilla soldiers in the recent civil war in Afghanistan did not have access to modem facilities and had to rely on ground transportation, often involving walks or animal rides of up to 6 days’ duration before receiving definitive surgical care at hospitals over the Pakistan border. This lack of early surgical attention greatly increased morbidity and mortality (Bhatnagar and Smith, 1989). This paper examines the natural history of high-velocity war wounds among those Afghan guerrilla fighters who survived long enough to reach medical care, the complications following these injuries, and the impact of delayed treatment on subsequent disability.
Patients and methods All patients admitted to the Afghan Surgical Hospital on the Afghanistan-Pakistan border from 22 February 1985 to 22 0 1992 Butterworth-Heinemann 002C-1383/92/080545-04
Ltd
February 1987 were reviewed retrospectively. Medical records were abstracted by one of the authors (M.B.) for age, sex, site of injury, type of injury, complications of the injury, hospital management and residual deformity. More detailed description of the area and hospital facilities are described elsewhere (Bhatnagar and Smith, 1989). The injuries seen at the hospital do not represent all injuries sustained by the guerrilla fighters, since many of the most severely injured may have died before arrival at the hospital. Sites of injury were grouped for analysis of the data. Distal extremity includes injuries to the wrist, hand, fingers, ankle, and foot. Long bone - metaphyseal/joint, represents injuries to the shoulder, elbow, hip and knee. Long bones diaphyseal, involves injuries to the arm (humerus), forearm (radius and ulna), thigh (femur), and leg (tibia and fibula). Non-union was defined as a fracture which showed no clinical or radiographic evidence of progressive healing at 6 months after injury. Malunion was defined as a healed fracture with deformity sufficient to require surgical intervention in order to improve function. Contracture was defined as a soft tissue injury resulting in loss of range of joint motion sufficient to interfere with function. Statistical analysis was performed using multiple logistic regression, utilizing the Generalized Linear Interactive Modelling system (GLIM, release 3.77), Royal Statistical Society, London.
Results A total of 1254 patients with 1292 musculoskeletal injuries were admitted to this surgical hospital during the 2-year period; all were male. The casualties ranged in age from 2 to 74 years; 86 per cent were in the 15-50 year age group. Of these, 3 per cent were treated for injuries to both upper and lower extremities, the majority of whom had a primary upper extremity injury with less severe injuries to the lower extremities. Lower extremity injuries were seen 1.6 times more frequently than those of the upper extremity. The most commonly injured region was the leg (21.4 per cent), followed by the thigh (17.6 per cent), forearm (9.3 per cent), and knee (8.4 per cent). Only 19 patients (1.5 per cent) presented with para-axial injuries to the scapula, spine, ilium, pelvis, clavicle, or axilla (Table I). None of these patients had any other important injuries. Fractures comprised almost 30 per cent of all injuries seen; all were grade III open fractures, from high-velocity projectiles and were grossly contaminated (Gustilo and Anderson, 1976). The diaphyses were injured three times more
Injury: the British Journal of Accident Surgery (1992)Vol. 23/No. 8
546 TableI.
Para-axial injuries by site of injury, Afghan Surgical Hospital, Pakistan, 1985-1987 No.
Site of injury
Scapula
2 1 5 1 3 7
Spine Iliac Pelvis Clavicle Axilla
19
Total injuries
Table II. Neurological injuries in the Afghan war Lower extremity
Upper extremity Site of injury Brachial nerve Median nerve Radial nerve Ulnar nerve Ulnar+ median nerve Ulnar+ radial nerve Unspecified Total UE injury
No. 1 7 5 19 5 6 7 50
NO.
Site of injury
4 4 4 1 1
Sciatic nerve Peroneal nerve Tibia1 nerve Sural nerve Tibia1 +sural nerve Unspecified Total LE injury
3 17
UE, upper extremity; LE, lower extremity.
frequently than the metaphyses and joints. The most frequently fractures bones requiring surgery were the tibia and fibula (31.3 per cent), followed by the femur (27.3 per cent), humerus (13.5 per cent), and forearm (12.2 per cent). The fractures were comminuted in 84 patients, 76 of which were diaphyseal injuries; the risk of comminution was significantly less in the femur compared with other long bones (P < 0.05). There were 86 patients (6.7 per cent) with major neurological deficits as a result of their wounds (TiibkZI). Upper extremity nerve damage was significantly more common than lower extremity nerve damage (PC 0.001) and the ulnar nerve accounted for almost 45 per cent of all nerve injuries. Foreign bodies, e.g. shrapnel, etc., were seen’in approximately 6.3 per cent of the injuries; soft tissue infections were present in 13.6 per cent of all cases, the majority in the thigh and leg. There was no statistically significant difference in the presence of foreign bodies or in the rate of soft tissue infection between body regions (TiableIII). Chronic osteomyelitis was present in 11.3 per cent of the patients. The femur and tibia accounted for 61 per cent of all osteomyelitis, with the rate of osteomyelitis in diaphyseal fractures being significantly higher than for metaphyseal fractures
(P <
0.001).
Malunion was a complication in 3.6 per cent of cases, that is more than 10 per cent of the fractures; 10 per cent of elbow injuries had malunion requiring surgical correction; the radius and ulna, femur and tibia required surgical correction of malunion in approximately 6 per cent of patients. There was no statistically significant different in the rate of malunion for various fractures. There were 43 fracture non-unions, 41 of which were long bone diaphyses. The risk of developing non-union in the diaphyses was eight times higher than in either the joints or the distal extremity, P < 0.005 (TableIV). The diaphyseal non-unions were distributed between the arm, forearm, thigh, and leg; fractures of the humerus had the highest frequency of non-union (8.7 per cent), but this difference was not statistically significant. Amputations were required in 82 instances, 7.2 per cent of all patients (T&dell/), and were performed most frequently for injuries of the distal extremities, especially the foot (22 per cent of which required amputation). Injuries of the leg, ankle, and foot accounted for 55 out of 87 amputations (63.2 per cent) and injuries of the leg resulted in 35 amputations (40.2 per cent). The risk of a distal extremity injury requiring amputation was six times that for large joint injury (PC O.OOl), and 1.75 times that for a long bone injury (P < 0.02). Poor range of motion was found in 5 per cent of the patients (TableIII), and was most likely to occur with injuries of joints, most frequently the shoulders and knees. The ankle was also frequently involved, and mobility was limited in 37 per cent of ankle injuries. The risk of losing motion was more than doubled with a joint injury when compared with injuries of the long bones. Fixed flexion contractures were present in 5 per cent of the patients overall; 22.7 per cent of hand injuries and 24.4 per cent of phalangeal injuries resulted in contractures. Few contractures or difficulties with range of motion occurred
Table III. Relative risk of injury complications by body regions, Afghan Surgical Hospital, Pakistan, 1985-1987 Wound sepsis Distal extremity (N= 257) Long bone metaphyseal/joints (N= 257) Long bone diaphyseal (N=692)
Poor ROM’
Chronic osteo l
Foreign body
1
1
1
1
1.23 NS
2.38 P
1.33 NS
1.06 NS
0.81 NS
0.17 P< 0.001
3.39 P
0.79 NS
‘ROM, range of movement. NS, not statistically significant. Chronic osteo, Chronic osteomyelitis
Table IV. Relative risk of complications by body regions, Afghan Surgical Hospital, Pakistan, 1985-1987 Amputation Fracture distal extremity (N= 257) Long bone metaphyseal/joint (N=257) Long bone diaphyseal (N=692) NS, Not statistically significant
1
Non-union
Malunion
Simple Fx
1
1
1
Comminuted Fx
1
0.07 P-E O.OCll
0.20 NS
1.45 NS
0.89 NS
0.33 NS
0.35 P< 0.02
8.03 P-z 0.005
1.39 NS
5.14 P-z 0.001
6.16 P< 0.001
Bhatnagar et al.: Muscdoskeletal injuries in the Afghan war
547
with injuries of the arm, forearm, thigh, or leg. The relative chance of developing contracture was 12 5 times higher in the distal extremities than in the long bones, and more than three times higher in the distal extremities than in the joints.
arterial and 41 per cent of injuries were vascular among police officers injured by firearms (Jones et al, 1968; Berman and Salter, 1985). The lethal nature of extremity trauma with vascular complication is demonstrated by reports that these injuries were the cause of death in one of every ten United States servicemen who died in Korea and Vietnam (Russotti and Sim, 1985). Wound infections were present in 13.6 per cent of the patients treated. There was no statistically significant difference in the rate of infection according to site of injury. The wartime rate of infection is known to be higher than civilian injuries; SO per cent of wounds from the Vietnam war were culture positive and 21.6 per cent developed soft tissue infections (Seindstein and Newman, 1968). The rate of infection seen in the Arab-Israeli War was 14 per cent (Michaeli, 1979). Jackson reported an overall 18 per cent infection rate in the Falklands Campaign (Jackson et al., 1983) which increased to 25 per cent for those injuries treated more than 6 h after injury. The rate of infection seen in civilian injuries ranges from 1.5 per cent to 3 per cent (Gustilo and Anderson, 1976; Elstrom et al., 1978; Rittmann et al., 1979; Woloszyn et al., 1988). In civilian population study groups, the rate of infection approached wartime rates only in patients who did not receive antibiotics (14-20 per cent) and in whom primary treatment was delayed by more than 8-12 h (Tscheme and Gotzen, 1984). In the Afghan hospital population, the rate of chronic osteomyelitis was significantly higher in the long bones, especially the femur and tibia. The metaphyseal and joint injuries were of equal severity and contamination, but the greater vascularity of these bony regions may offer some protection against infection (Ketenjian and Shelton, 1972; Franklin et al., 1984). The greater incidence of infection of open tibia1 fractures compared with femoral fractures has been attributed to the decreased soft tissue coverage in this region (Chapman, 1986; Holden, 1972; Ryan et al., 1981). The overall rate of osteomyelitis is very similar to that observed by others (Rittmann et al., 1979). Similarly, the long bones had an eightfold greater risk of developing non-union than distal extremity injuries. The amputation rate for distal extremity injuries was twofold higher than that for long bones, and almost sixfold higher than for injuries to joints. This reflects the severe localized injuries sustained by blast injuries such as exploding land mines. Elbow and tibia1 fractures had high rates of malunion, but there was no statistically significant difference in malunion rates between the joints and long bones. The definition of malunion to include only those cases severe enough to require secondary surgery to improve function will certainly underestimate the overall incidence as estimated from radiographic measurements. Functional loss of mobility was much more common after injury to the large joints than to the distal extremities and to long bones. Of the patients, 5 per cent suffered from severe functional restriction, significantly higher than the few cases reported by Rittmann et al. (1979). This is probably due to the traditional Afghan practice of immobilizing soft tissue trauma and delay in seeking care. The Mujahidin tightly bandage and immobilize wounded extremities, consequently leading to adhesions within joints and contractures. This practice was the primary cause of residual impairment for hand and finger injuries in the absence of fracture which contrasts with a low rate of impairment reported in other countries by Duncan and Kettlekamp (1974) and Granberry (1973).
Discussion The 1254 patients reviewed include all civilian and military war casualties treated at the Afghan Surgical Hospital for management of musculoskeletal trauma. This hospital was established specifically to treat male casualties from one of the seven guerilla factions during the Afghan Civil War (Bhatnagar and Smith, 1989). Females were treated elsewhere. The main causes of injury seen in this study were fragmentation weapons (50 per cent) (Bhatnagar and Smith, 1989) similar to those in the Korean, Vietnam, and Falklands wars, but the distribution of injuries by body region differs greatly (Jones et al., 1968; Johnson et al., 1981; Jackson et al., 1983). Though lower extremity injuries were more frequent than upper extremity injuries, paralleling other conflicts, only 3 per cent of the patients had multiple injuries, suggesting that the most severely injured died before receiving treatment. Furthermore, only 19 patients (1.5 per cent) had injuries to the para-axial region, which is far less than would be expected based on other conflicts (Carey, 1987). Many para-axial injuries involve the thorax and abdomen, with accompanying trauma to vital structures, which is rapidly fatal if not treated immediately (Bellamy et al., 1986). Few such patients survive the arduous trip across the mountains to our hospital. Management of fractures and excision of wounds represented about two-thirds of the surgical cases, a similar proportion to that reported by Morris at a similar hospital on the Pakistan border (Morris et al., 1985). Serious complications from injuries were found in approximately 50 per cent of patients, of which most were soft tissue infections, chronic osteomyelitis, and restriction of joint motion. This figure is much higher than that described in other conflicts. Cornminuted fractures were significantly less common in the femur than other fractured long bones; this is likely to be due to the bulk of overlying musculature which absorbs and dissipates significant amounts of the energy from the projectiles and the fact that the direct impact of blasts from land mines is sustained by lower regions of the leg. There were 86 patients (6.7 per cent) with severe neurological deficits requiring surgery. This is fewer than that seen by Berman and Salter (1985) among police officers injured by low-velocity gunshot wounds (I4 per cent), but is similar to the prevalence of neurological injury observed by De Wind (1987) in the war in Uganda (6 per cent), and that reported by Jones et al. (1968) in Vietnam (3.6 per cent). The most commonly injured nerve was the ulnar nerve: ali neurological deficits were secondary to axonotmesis. The paucity of soft tissue covering nerves in the upper limb, in particular the ulnar nerve, accounts for the increase in nerve injury from these high-velocity injuries, relative to the lower limb which is more frequently injured overall. Similarly, there were no patients seen in our study with severe vascular injuries; arterial injuries usually involve greater wounding power, and the considerable delay in the availability of medical treatment probably resulted in these patients dying before they could reach the hospital. During the Vietnam conflict, almost 5 per cent of injuries were
548
Injury: the British Journal of Accident Surgery (1992) Vol. 231N0.8
References Aldea P. A., Aldea G. S. and Shaw W. W. (1987) A historical perspective on the changing methods of management of major trauma of the lower extremity. Surg, Gyneco]. Obstet. 165, 54. Bellamy R. F., Maningar Pa. and Vayer J. S. (1986) Epidemiology of trauma: military experience. Ann. Emwg. Med 15, 1384. Berman A. T. and Salter F. (1985) Low-velocity gunshot wounds in Police officers. Clin. O&p. 192, 13. Bhatnagar M. K. and Smith G. S. (1989) Trauma in the Afghan guerilla war: effects of lack of access to care. Surgery 105,699. Carey M. E. (1987) Learning from traditional combat mortality and morbidity data used in the evaluation of combat medical care. M&f. Med. 152, 6. Chapman M, W. (1986) The role of intramedullary fixation in open fractures. C/in. Or/hop. 212, 26. Cozen L. N. (1985) Military orthopedic surgery. C/in. Orthop. 200, 50. DeWind C. M. (1987) War injuries treated under primitive circumstances: experiences in a Ugandan mission hospital. Br.]. Surg. 74, 193. Duncan I. and Kettlekamp D. (1974) Low velocity gunshot wounds of the hand. Arch. Surg. 109,395. Elstrom J. A,, Pankovich A. M. and Egwele R. (1978) Extraarticular low-velocity gunshot fractures of the radius and ulna. 1. Bone ]oinr Surg. 60A, 335. FranWin J. L., Johnson K. D. and Hansen S. T. (1984) Immediate internal fixation of open ankle fractures. ]. Bone ]oint Sr4rg. 66A, 1349. Cranberry W. (1973) Gunshot wounds of the hand. Hand 5,220. Gustilo R. B. and Anderson J. T. (1976) Prevention of infection in the treatment of one thousand twenty-five open fractures of long bones. ]. Bone ]oint Surg. WA, 453. Holden C. E. A. (1972) The role of blood supply to soft tissue in the healing of diaphyseal fractures. ], Bone ]oint Surg. HA, 993. Jackson D. S., Batty C. G., Ryan J. M. et al. (1983) The Falklands war: army field surgical experience. Ann. R. Co/l. Surg. Engl. 6.5, 281.
Johnson D. E., Panijayanond P., Lumjiak S. et al. (1981) Epidemiology of combat injuries in Thailand. 1. Trauma 21, 484. Jones E. E., Peters A. F. and Gasior R. M. (1968) Early management of battle casualties in Vietnam. Arch. Surg. 97, 1. Ketenjian A. Y. and Shelton M. L. (1972) Primary internal fixation of open fractures: A retrospective study of the use of metallic internal fixation in fresh open fractures. ]. Truumu 12, 756. Michaeli D. (1979) Medicine on the battlefield: a review. ]. R. Sot. Med. 72,370. Morris D., Sugrue W. and McKenzie E. (1985) On the border of Afghanistan with the International Committee of the Red Cross. N.Z. Med. J 98, 750. Rittmann W. W., Schibli M., Liatter P. et al. (1979) Open fractures: long-term results in 200 consecutive cases. CIin. Orthop. 138, 132. Russoti G. M. and Sim F. H. (1985) Missile wounds of the extremities: a current concept review. Orthopedics 8, 106. Ryan J. R., Hensel R. T., Salciccioli G. G. et al. (1981) Fractures of the femur secondary to low-velocity gunshot wounds. ] Trauma 21, 160. Seindstein M. and Newman A. (1968) Some clinical factors involved in the healing of war wounds. Arch. Surg. 96, 176. Tscheme H. and Gotzen L. (1984) Fractures with .!%ffTissue Injuries. New York: Springer-Verlag. Wolosryn J. T., Uitvlugt G. M. and Castle M. E. (1988) Management of civilian gunshot fractures of the extremities. Ch. Orfhop. 226, 24.
Paper accepted 5 May 1992.
Requestssfor reprints should be addressed to: Dr Bhatnagar, Department of Orthopaedic Surgery, The Johns Hopkins Hospital, 600 N. Wolfe Street, Baltimore, Maryland 21205, USA.