Introduction and Background

The Middle East region is frequently associated with wars and military conflicts of which the Arab–Israeli wars and recent Arab uprisings are good examples [Reference Fawcett1]. There is a strong association between military conflicts and infectious diseases and historically classic ‘war pestilence’ such as plague, cholera, typhoid, typhus, dysentery, and smallpox are cited as being responsible for most of the deaths during wartime. However, improvements in sanitation and public health infrastructure along with better practices relevant to troop deployment have led to a substantial decline in mortality associated with such infections [Reference Murray2]. The last century has also witnessed a significant drop in mortality resulting from infections caused by battle wounds [Reference Smallman-Raynor and Cliff3]. Efforts to prevent or minimize morbidity and mortality resulting from combat-related injuries include better personal protective equipment, training of medical personnel to provide lifesaving procedures on the battleground, and setting medical care facilities with surgical capabilities close to point of injury. The ability to enhance survival rates in soldiers and civilians with war injuries has, nevertheless, resulted in an increase in the risk of wound infections in this group [Reference Murray4]. Experience gained during the Great War (World War I) helped in reducing the incidence and subsequent complications of battle wound infections. Taking patient mortality as a measure, aggressive surgical debridement, delayed primary closure, early surgical intervention, flaps, and external fixators, although on a limited scale, all contributed to better patient outcomes in World War II compared to World War I. The main lesson gained from World War II, where up to 86% of hospitalized patients experienced an infection, was the importance of nosocomial acquisition of bacteria thus emphasizing the importance of infection control and aseptic techniques [Reference Murray4, Reference Pruitt5].

The microbial aetiology of combat-related injuries differs between various stages of management. Generally, bacterial transition occurs over time, from in the early stages an even balance of Gram-positive and Gram-negative flora, to the later predominance of antibiotic resistant Gram-negative species during treatment [Reference Murray6]. It is believed that antibiotics administered early at the time of injury exert a selection pressure leading to antimicrobial resistance [Reference Murray4]. The first large-scale utilization of antibiotics in battles was during D-Day in 1944 where penicillin was effectively used to protect injured soldiers from developing Clostridium infection and subsequent tetanus, but later in the Korean war, penicillin combined with streptomycin was the most widely used antibiotic regimen. Antibiotics were often administered prophylactically, but at a cost that only became apparent in retrospect, as resistant bacteria were increasingly reported from infected war wounds 3–5 days after injury. The significance of Pseudomonas aeruginosa and resistant Gram-negative and Gram- positive isolates was most obvious during the Vietnam war [Reference Manring7, Reference Kovaric8].

War injuries are not restricted to military personnel and since 1945, more than 100 million people, with 25 million deaths, have been recorded in military conflicts worldwide [Reference Aras9]. Characterizing and defining war-zone injuries is a complicated process due to difficulty of access to those affected, limited healthcare personnel and facilities at site of injury, and the widespread use of antimicrobial drugs. Complete patient histories are also difficult to obtain in crisis settings, limiting the healthcare worker's ability to describe all prior interventions. Furthermore, injuries sustained during war are at high risk of infection due to environmental contamination [Reference Teicher10].

The aim of this review is to describe the microbiology of war-related wound infections and factors affecting their incidence from conflict areas in Iraq, Syria, Israel, and Lebanon.

Iraq

Operation Iraqi Freedom/Operation Enduring Freedom resulted in a large number of injuries among both civilians and the military. The value of surveillance cultures was evaluated in a published report on 213 combat-related open Gustilo and Anderson type III diaphyseal tibia fractures among 192 US military personnel between March 2003 and September 2007 [Reference Burns11]. Surveillance cultures were positive in 64% of extremities and 93% of them grew Gram-negative isolates. By contrast, at 9 months following injury, 27% developed a wound infection with mainly Gram-positive bacteria. In those with initial surveillance culture-positive wounds, 38·7% developed an infection compared to 11·5% with negative surveillance culture wounds. Moreover, in subjects who subsequently developed a wound infection and had initial cultures taken, 26·2% showed a match of at least one of the initial surveillance organisms. The contradiction between initial surveillance cultures and subsequent infected wound microbiology was attributed to the use of early antibiotic treatment, which suppressed the growth of susceptible flora and selected for more resistant organisms. The study concluded that patients culture positive on surveillance cultures were more likely to develop a wound infection and osteomyelitis. Even though positive surveillance cultures were not predictive of the infecting organism in subsequent infection, they were associated with development of wound infection, osteomyelitis, and ultimate need for amputation [Reference Burns11].

A smaller study of 49 US military casualties in Baghdad, Iraq during spring 2004, with acute open traumatic injuries that underwent aerobic culture of their wounds, found a bacterial contamination rate of 49% of cultures [Reference Murray12]. Gram-positive bacteria, mostly skin-commensal flora, constituted 93% of positive cultures with only two isolates identified as methicillin-resistant Staphylococcus aureus (MRSA) while Gram-negative bacteria were quite rare (5%). These findings suggested that early wound bacteriology is composed predominantly of Gram-positive organisms of low virulence and pathogenicity, and that the use of broad-spectrum antibiotics against multidrug resistant (MDR) Gram-negative bacteria is unnecessary in early wound management [Reference Murray12].

MDR is defined as non-susceptibility to at least one agent in ⩾3 antimicrobial categories [Reference Magiorakos13]. The role of MDR bacteria in causing wound infections and even bacteraemia was apparent in US military personnel in Iraq. Highly resistant pathogens mainly Acinetobacter baumannii-calcoaceticus complex (ABC), P. aeruginosa, extended spectrum β-lactamases (ESBLs) producing Klebsiella pneumoniae, Escherichia coli, and MRSA were frequently isolated from wound and bloodstream infections in American soldiers [Reference Calhoun, Murray and Manring14–Reference Petersen17]. Data from combat support hospitals in Iraq during 2003 and 2004 revealed that coagulase-negative staphylococci accounted for 34%, S. aureus for 26%, and Streptococcus spp. for 11% of isolates. By contrast, cultures from the predominantly Iraqi population showed mainly K. pneumoniae (13%), ABC (11%), and P. aeruginosa (10%) [Reference Yun18]. The total number of Gam-positive and Gram-negative isolates is not reported, but both showed resistance to a broad array of antibiotics [Reference Yun18]. In another study of US soldiers in Iraq and Afghanistan, the isolation of Candida spp. from penetrating wound injuries was associated with an increased risk of mortality [Reference Blyth19]. However, a retrospective study of 48 British soldiers with an open fracture of the femur reported that four patients developed deep wound infections requiring surgical treatment and despite the low numbers of cases, the degree of bone loss rather than deep tissue infection was the factor most associated with outcome [Reference Bennett20].

Little has been published about infections associated with war-related trauma in Iraqi soldiers and civilians. A study by the non-governmental organization Doctors Without Borders, found sepsis to be the most common cause of death in 1169 Iraqi patients admitted to a burns unit and 92% and 63% of Gram-positive and Gram-negative isolates, respectively, were classified as MDR [Reference Ronat21]. The most common Gram-negative organisms were P. aeruginosa (34%), K. pneumoniae (12%), ABC (9%), and Enterobacter cloacae (8%), with S. aureus (26%) and S. epidermidis (11%) being the predominant Gram-positive species [Reference Ronat21].

Syria

The Syrian civil war, now in its fifth year, has resulted in >200 000 deaths, 500 000 wounded, and at least 9 million refugees [22, Reference Guha-Sapir23]. This war is an example of a military conflict that affects both civilians and soldiers and which has led to the destruction of the healthcare infrastructure in a limited-resources environment. A study by Doctors Without Borders of 61 Syrian orthopaedic patients with suspected infections found that 74% had at least one positive wound culture, 13% of which were polymicrobial. Gram-negative organisms accounted for 56% of cultures with P. aeruginosa in 23%, E. coli in 19%, and ABC in 14%; Gram-positive bacteria, including MRSA, represented 44% of isolates. Overall, 69% of patient harboured MDR organisms with MRSA representing 42% of staphylococcal isolates [Reference Teicher10].

Since February 2013, the Syrian civil war has resulted in an influx of 1300 wounded, both fighters and civilians, into Israeli borders seeking medical treatment [Reference Biswas24]. Among such patients, the incidence of MDR isolates ranged from 47% to 66%, the most common being ESBL-producing and/or carbapenem-resistant Enterobacteriaceae (CRE), MRSA and ABC; two of the CRE isolates produced New Delhi metallo-β-lactamase [Reference Peretz25, Reference Al-Assil, Mahfoud and Hamzeh26]. Data for the first 100 patients from Syria admitted to the Ziv Medical Centre between February and October 2013 show that sepsis, due to delayed definitive wound care, was the main determinant of the clinical outcome. No correlation between time from injury to arrival to the trauma room and the incidence of septic complications was evident but an increased probability of sepsis was noted in wounds that had undergone primary closure to control haemorrhage and/or wrapped in soiled blankets before definitive care. For this reason, primary closure wounds warranted reopening, debridement, and vacuum-assisted closure systems [Reference Biswas24].

Cited factors responsible for the increase in MDR organisms include delay in appropriate management, wide use of antimicrobial drugs without prescription in the community, and the presence of metallic or organic fragments in the wound [Reference Teicher10, Reference Al-Faham, Habboub and Takriti27]. It is recommended that in patients undergoing surgery such fragments must be removed or monitored for a long period of time [Reference Aras9]. However, several studies have shown that this does not apply for the central nervous system (CNS) injuries as no relationship was found between the presence of retained fragments and the development of either a seizure disorder or a CNS infection warranting the conclusion that it is unnecessary to re-operate for retained bone fragments, as they do not increase immediate or late complications [Reference Carey28–Reference Barhoum31]. The hospital setting is the major source of MDR organisms associated with blast injuries and nosocomial patient-patient transmission as well as faecal colonization with ESBL-producing Gram-negative bacteria often being responsible for acquisition of infection [Reference Peretz25, Reference Murphy32]. Turton et al. showed the ABC strains responsible for wound infections in the United States and UK in American and British soldiers, respectively, who had shared exposure to various medical facilities, were genetically identical [Reference Turton33]. Nosocomial transmission has been reported to be a greater contributing factor to wound infection over environmental contamination at the time of injury [Reference Turton33–Reference Vento35].

This increase in MDR organisms has also translated to novel resistance at the molecular level. Carbapenem resistance in 66% of ABC isolates in Syria has been reported [Reference Hamzeh, Al Najjar and Mahfoud36] and this was ascribed to the bla _NDM−1 gene in isolates from four Syrian patients, wounded in Syria, while being treated in Lebanon, which was the first time it was described in that country [Reference Rafei37]. The bla _NDM−1 gene was first described in New Delhi in 2008 in Enterobacteriaceae [Reference Yong38] and differs from the most common acquired OXA-type resistance mechanism against carbapenemases [Reference Rafei37, Reference Baroud39]. Data from Syrian hospitals on carbapenem resistance are lacking possibly due to the severe damage to healthcare infrastructure and the fact that Syrian doctors and health professionals are overwhelmed and understaffed dealing with the huge patient-care burden [Reference Stone-Brown40, Reference Dewachi41].

Israel

Israel has witnessed numerous wars with bordering Arab countries. Medical military authorities during the Yom Kippur war recommended prophylactic penicillin administration to all wounded soldiers within 30–60 min after injury. In addition to penicillin prophylaxis, many hospitals added 1 g/day of streptomycin followed by oral 3–4 g/day of each of cloxacillin or ampicillin [Reference Klein, Berger and Yekutiel42, Reference Romanoff43]. During that war, infection rates ranging from 4·9% to 58·3% were reported with P. aeruginosa being the most common pathogen [Reference Romanoff43–Reference Sidi45], and this species was again the most prevalent in the ensuing 1982 war with Lebanon [Reference Simchen46]. In both theatres the increased incidence of infection was associated with penetrating abdominal wounds involving the colon, extensive soft tissue loss, burns involving >25% of the body surface, multiple operations, open drains inserted in the first operation, and wounds located below the diaphragm [Reference Klein, Berger and Yekutiel42, Reference Simchen and Sacks44, Reference Simchen46]. Infections due to penetrating abdominal wounds involving the colon can be due to heavy colonic bacterial flora mainly Bacteroides spp. and Enterobacteriaceae. For this reason, Simchen & Sacks [Reference Simchen and Sacks44] recommended the prompt administration of clindamycin and gentamicin in combination for all abdominal injuries and treatment was continued for 48–72 h if colonic involvement was shown at laparotomy.

Despite its widespread use on the battlefield during the Yom Kippur war, the administration of prophylactic antibiotics did not have a significant impact in preventing wound infection [Reference Klein, Berger and Yekutiel42]. Furthermore, this practice possibly contributed significantly to an increase in infections due to Gram-negative rods [Reference Finland47–Reference Tillotson and Finland51]. An increase in carbenicillin-resistant P. aeruginosa and gentamicin-resistant Klebsiella strains was noted in burn patients [Reference Sidi45].

Candidaemia among immunocompetent patients following a bomb blast at an Israeli market place, which contained a significantly high concentration of airborne Candida organisms, was reported by Wolf et al. [Reference Wolf52]. This infection was recorded in seven (30%) of 21 patients between 4 and 16 days after the injury and Candida spp. was the most frequent cause of bloodstream infections, with inhalation injury appearing to be the best predictor of infection. Mortality in such patients was 43% and underlined the need to monitor patients for fungal infection where injuries were sustained in a similar setting and highlighted the wide range of organisms responsible for infected wounds through environmental exposure at the time of injury [Reference Wolf52].

Lebanon

Lebanon has been plagued by wars and military conflicts since 1975 and the country has experienced, on several occasions, violent armed battles ranging from internal civil war to full-blown war with the Israeli Army. Over 40 years, the nature and intensity of military confrontations has varied, but the results remain the same – deaths and casualties among civilians and fighters [Reference Winslow53]. Relatively few reports have been published describing battle injuries in Lebanon during the civil war (1975–1978) and the war with Israel (1982–2006) [Reference Nohra29, Reference Sfeir54–Reference Nassoura57]. In the civil war, a study of the management of 1021 Lebanese patients over a 10-year period (1975–1984) at a tertiary hospital reported an infection rate of 12% with S. aureus, P. aeruginosa, and E. coli being the most common organisms, respectively [Reference Zaytoun, Shikhani and Salman55]. Another report from the same centre on the outcome of 1500 patients, described a 2·1% incidence rate of sepsis as the second cause of death after haemorrhage (3·7%) [Reference Nassoura57].

In more recent times Lebanese hospitals have acted as referral centres for war casualties from Syria and Iraq as well as their own national soldiers and civilians. The variability in the source of injuries and the various approaches for early management at the point of injury adds further to the complexity of the treatment of such injuries [Reference Bakhach, Abu-Sitta and Dibo58] and risking the possibility of importation of MDR strains. In 2010 two NDM-1-producing K. pneumoniae imported from Iraq were detected in Lebanon [Reference El-Herte59] and between 2011 and 2013 a 60% prevalence rate of carbapenem-resistant ABC was recorded in Tripoli, the largest Lebanese city bordering Syria [Reference Rafei60], thus highlighting the need for effective measures to control the spread of such pathogens in the country.

Intracranial infections, including brain abscesses, following missile injuries to the brain were reported by the neurosurgery division at the American University of Beirut Medical Centre in patients treated between 1981 and 1988 [Reference Taha, Haddad and Brown56]. More recently, the 2006 conflict with Israel, resulted in the release of more than four million sub-munitions over Lebanese soil including one million unexploded duds [Reference Beehner61]. A study of 350 injured causalities by Fares et al. in 2013 revealed an infection rate overall of 19·4% with bacterial infections accounting for 86·8% and fungi for the remainder. P. aeruginosa was the most common single isolate (30·5%) followed by E. coli [Reference Fares62]. This underlines the common finding that Gram-negative infections are generally more prevalent among war-related injuries compared to Gram-positive organisms and infected wounds are most often located below the diaphragm. Environmental contamination of penetrating pieces of clothes and metallic foreign objects increases the risk of streptococcal, staphylococcal and enterobacterial infections which are commonly found on skin or clothing covering the wound [Reference Murray12, Reference Fares62, Reference Korzeniewski63].

Treatment and prevention

The use of prophylactic antibiotics for war-related trauma is controversial. Current guidelines of the Infectious Disease Society of America and the Surgical Infection Society for the prevention of infections associated with combat-related injuries strongly suggest that the use of broad-spectrum antibiotics at time of injury should be discouraged due to their potential to select resistant organisms. Furthermore, they do not support the use of aminoglycosides or fluoroquinolones for enhanced Gram-negative antimicrobial coverage as well as the administration of penicillin to prevent clostridial gangrene or streptococcal infection [Reference Hospenthal64]. Velmahos et al. [Reference Velmahos65, Reference Velmahos66] considered that in severe trauma patients the use of broad spectrum antibiotic coverage that includes MDR pathogens such as ABC is not needed at the time of injury, and administration of antibiotics for >24 h does not offer additional protection against sepsis, organ failure, and death, but rather increases the probability of infection with MDR pathogens. However, the guidelines do advocate the administration of systemic antibiotics within 3 h post-injury to prevent infectious complications and sepsis. The choice of antibiotics depends on the location of the wound, with cefazolin preferred for extremity, central nervous system, and thorax wounds, and metronidazole for abdominal wounds; topical silver sulfadiazine and mafenide acetate are suggested for burn trauma [Reference Hospenthal64].

ABC is a common pathogen in wound infections across the Middle Eastern region and the reported most active antimicrobial agents against these organisms are colistin, polymyxin B, and minocycline [Reference Griffith67–Reference Scott69]. Imipenem was originally recommended for prophylaxis against ABC organisms in war wounds but its empirical use is now discouraged, and should be reserved for management of a proven or suspected ABC infection [Reference Hawley70]. Injuries sustained in certain environmental settings may predispose to fungal colonization thus warranting antifungal prophylaxis with fluconazole [Reference Wolf52, Reference Fares62].

Wound debridement and delayed primary closure including the removal of possible foreign debris, first introduced in World War II [Reference Churchill71], have been shown to significantly decrease the incidence of infection on admission [Reference Klein, Berger and Yekutiel42, Reference Tintle72]. Primary wound closure after debridement creates tension at the wound edges, which consequently compromises blood supply, increases the risk of infection and wound dehiscence [Reference Bakhach, Abu-Sitta and Dibo58]. Early drainage of haemothoraces, avoidance of thoracotomy as primary treatment, and the separate treatment of abdominal and thoracic injuries have also contributed to lower infection rates in patients with chest injuries [Reference Romanoff43, Reference Senanayake73]. In countries such as Syria, where bacterial resistance is likely as a result of wide over-the-counter use of antibiotics [Reference Al-Assil, Mahfoud and Hamzeh26, Reference Otoom and Sequeira74], hospitals in bordering countries receiving casualties are taking extra measures to ensure appropriate infection control practices including routine microbiological screening on admission, the use of a specific trauma room for immediate resuscitation, and the introduction of isolation bays on the wards [Reference Biswas24].

A system for the treatment and prevention of war trauma-related wounds has been an evolving discipline since World War I. Currently, the gold standard approach used by the US military consists of four levels of care divided according to proximity to the battlefield and type of management. The first level consists of initial care in the battlefield and includes stabilizing fractures, bandaging wounds with sterile dressings, and administration of single dose antimicrobials if evacuation is delayed. Level-2 care takes place in a field hospital and consists of tetanus vaccinations and immunoglobulin administration, wound irrigation with saline to remove gross environmental contamination, and topical antimicrobials for burns. Combat support hospitals provide level-3 treatment with inpatient care, intensive care units, and operating rooms; this also includes surgical wound management with external fixation of open fractures, in addition to treatment provided at level 2. Finally, level-4 treatment is provided by regional hospitals or hospital ships, located outside the battlefield area, and covers general and specialized inpatient medical and surgical care [Reference Hospenthal64]. Despite the success of this system in reducing morbidity and mortality, its implementation requires large amounts of financial and logistical resources, which are beyond the capability of most national armies. Wars in the Middle East including the US-led ‘War on Terror’ increasingly involve smaller nation armies and paramilitary groups that do not have such healthcare systems in place and as a consequence, civilian medical centres and hospitals in countries such as Iraq, Lebanon, and Syria bear the major burden of these conflicts.