Our search yielded 1714 discrete studies, of which 311 required full text review (Fig. 1). Twenty-nine articles met inclusion criteria and were included in our review (Table 2) [32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60]. Two additional articles and reports were identified by hand-searching the citation lists of included articles and by expert consultations [61, 62]. Quality assessment of the articles yielded 21 “high quality studies” that fulfilled 7 or greater of the nine criteria and 10 “moderate quality studies” that fulfilled between 4 and 6 of the criteria. (Table 2 is categorized by study quality).
Demographic analysis
Of the total of 31 studies included in the analysis, 16 were retrospective cohort studies, seven were prospective cohort studies, six were case series and two were non-peer reviewed reports from reputable human rights organizations. The number of subjects ranged from two to 3697 (median 31) (one study met the inclusion criteria of five subjects, and although several of the subjects sustained injuries from another weapon, the study was nonetheless included). In studies in which gender was reported, 57% of subjects were male and 43% were female. In studies in which age was reported, the age ranged from 3 months to 94 years, with an mean age of 25.7 years. The injury context included protests (10), arrests (five), military or police training exercises (five), accidental exposures (five), and a detention center riot (one); some of the studies included injuries in more than one context. The eligible studies included data from 11 countries and were published between 1993 and 2015 (Fig. 2).
Twenty-six studies included data on injuries caused by Agent CS and 14 included data on injuries caused by Agent OC. Sixteen studies evaluated dry aerosolized forms, such as grenade-type deployment of the chemical agent and 15 included sprays formulated with solvents. Seven of the studies recorded injuries that resulted from the projectile munition containing chemical irritant causing direct trauma to subjects.
Analysis of injuries and deaths
In the included articles, a total of 5910 people were exposed to chemical irritants and sought medical attention, of whom 5131 (87%) suffered injuries or died as a result of the exposure. Of those who suffered injuries, two people died and 67 (1.3%) suffered permanent disability. The majority fully recovered from their injuries (98.7%).
Deaths
Two deaths were documented in the selected articles. A report from Bahrain documented the case of a man who died of respiratory arrest after agent CS aerosol was fired inside his home. In another case, the chemical irritant projectile munition contributed to one death from traumatic brain injury after protests in Nepal. There were no deaths associated with agent OC.
Permanent injuries and disabilities
Fifty-eight people experienced permanent disability (Fig. 3). Eighteen of the disabilities were secondary to traumatic injuries from the projectile munitions. These included globe ruptures and blindness (four people), traumatic brain injury resulting in a persistent vegetative state (one person), limb amputations (three people), and functional loss of limbs (10 people). Persistent psychiatric symptoms were documented in 14 people and persistent symptoms of asthma and other respiratory complaints were reported in 23 people. Chronic dermatological conditions such as hypersensitivity reactions were documented by skin testing in three people. In one study of 297 individuals seeking care and/or evaluation of injuries following the 2013 Gezi Park protests in Turkey, 117 psychiatric evaluations were conducted. Of those, 50 (43%) met diagnostic criteria for acute stress disorder, 27 (23%) met diagnostic criteria for post-traumatic stress disorder (PTSD), and nine (8 %) met diagnostic criteria for major depressive disorder [62].
Injuries
There were 9261 documented injuries, with multiple injuries occurring in each individual. In total, 6878 (74.2%) of the injuries were categorized as mild, 1582 (17%) were moderate injuries, and 865 (8.7%) were severe injuries (Fig. 4). While many body systems were affected, the majority of injuries were to the skin, eyes, and cardiopulmonary system. Injury severity varied depending on body system affected (Fig. 5). Severity followed a pattern of fewer injuries among the higher severity categories for dermal, ocular, neurological, cardiopulmonary, and intra-abdominal injuries. For instance, while the majority of skin injuries (2539) were mild, 351 of the injuries were moderate and 180 were severe. Similarly, for cardiopulmonary injuries, there were 1220 mild injuries, 328 moderate injuries and 131 severe injuries. On the other hand, there were more severe neurological injuries (12%) than moderate neurologic injuries (1 %). Injuries to the musculoskeletal system, as well as psychological injuries, were all categorized as severe, based on the definition employed.
Chemical agent
The nature of chemical irritant exposure and injuries is also related to the chemical agent used, mechanism of deployment, environmental conditions, and context of use. Fourteen studies dealt exclusively with agent CS and 10 studies exclusively studied agent OC. Three studies included injury data from both chemical agents or did not differentiate between the two. Four studies reported exclusively on traumatic injuries from the projectile munition, while three other studies reported some injuries from the projectile munition among other injuries from the chemical agents themselves. Among 7156 documented injuries specifically from agent OC, only 6 % were categorized as severe. In contrast, 27.9% of 1148 injuries from agent CS were categorized as severe (Fig. 6).
Other factors that may impact injury severity
To assess for other factors that may impact injury severity, we utilized a qualitative approach by perusing the articles for data that may not fit into categorical variables but would regardless be relevant. Several of these factors were noted in the article text and are highlighted.
Deployment mechanism
The included studies documented injuries secondary to both dry aerosolized and soluble spray forms of both chemicals. Proximity to the area where the chemical was released and the force of the propellant affected outcomes [38, 40, 47, 53, 60]. We identified 5366 mild injuries, 884 moderate injuries, and 483 severe injuries from spray forms of agent CS and OC. Among the injuries from spray forms of chemical irritants, 7.2% were severe. We documented 1512 mild injuries, 676 moderate injuries, and 281 severe injuries from aerosolized forms of agent CS and agent OC. Among injuries from aerosolized forms, 11.6% were severe. Comparative analysis of the deployment mechanisms using pooled data was not conducted, given the concern for confounding factors.
The use of projectile munitions was documented to cause 231 injuries, of which 63 (27%) were severe. There were 73 traumatic injuries to the head and neck, including at least four people who lost vision in an eye due to projectile munition trauma. We documented 45 injuries to the torso (chest, abdomen, back, and genitalia). There were 61 upper extremity injuries and 34 lower extremity injuries (including at least three people requiring amputations and 10 with severe functional loss of a limb due to neurovascular injuries). Eighteen dermal injuries (8 %) included bruises, lacerations, and heat burns.
Other factors
Several other factors were documented as exacerbating the potential for injury, but they lacked detailed data for analysis, such as documentation of specific injuries. Utilizing the weapons in confined spaces and in areas where people could not easily escape potentially increased the exposure to the irritant either in quantity or over time [54]. One study in a detention center suggested that the excessive number of injuries may have been caused by the crowded and enclosed setting, which offered no opportunity for people to escape [33]. Use of chemical irritants in areas with high heat or humidity potentially exacerbated skin irritation, and windy conditions risked the contamination of law enforcement officers, bystanders, or nearby residences and businesses [33, 43, 53]. One study noted that the use of agent CS for military training on a particularly humid day, followed by strenuous exercise by trainees, may have caused severe respiratory injuries which resulted in several people requiring ICU-level care [53]. Direct targeting of the face and eyes by hand-held spray has been noted to cause trauma and toxicity to the cornea and conjunctiva of the eye [40, 41, 45, 50].