BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

Posttraumatic stress disorder in adult victims of cluster munitions: A 10-year longitudinal study

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-017214

Article Type:

Date Submitted by the Author: 07-Apr-2017

Complete List of Authors: Fares, Jawad; Lebanese University, Research Center, Faculty of Medical ; American University of Faculty of Medicine Gebeily, Souheil; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences; Lebanese University , Department of Neurology, Faculty of Medical Sciences Saad, Mohamad; University of Washington School of Medicine Harati, Hayat; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences Nabha, Sanaa; Lebanese University, Neuroscience Research Center, Faculty of Medical Sciences Said, Najwane; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences Kanso, Mohamad; American University of Beirut Medical Center, Department of Emergency Medicine; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences http://bmjopen.bmj.com/ Abdel Rassoul, Ronza; INSERM U1141, Hôpital Robert Debré; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences Fares, Youssef; Lebanese University, Neuroscience Research Center, Faculty of Medical Sciences; Lebanese University, Department of , Faculty of Medical Sciences

Primary Subject Neurology Heading: on September 27, 2021 by guest. Protected copyright. Secondary Subject Heading: , Public health,

Posttraumatic stress disorder, PTSD, cluster munitions, cluster bombs, Keywords: , Adults

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 4 Original Article 5 6 7 Posttraumatic stress disorder in adult victims of 8 9 cluster munitions: A 10-year longitudinal study 10 11 12 13 Jawad Fares1,2*, Souheil Gebeily1,3, Mohamad Saad1,4, Hayat Harati1, Sanaa Nabha1, 14 1 1,5 1,6 1,7 15 NajwaneFor Said , Mohamad peer Kanso review, Ronza Abdel Rassoul only, Youssef Fares * 16 17 18 19 1. Neuroscience Research Center, Faculty of Medical Sciences, Lebanese University, 20 21 Beirut, Lebanon 22 2. Faculty of Medicine, American University of Beirut, Beirut, Lebanon 23 3. Department of Neurology, Faculty of Medical Sciences, Lebanese University, Beirut, 24 Lebanon 25 26 4. Division of Statistical Genetics, Department of Biostatistics, University of Washington, 27 Seattle, WA 98195, USA 28 5. Department of Emergency Medicine, American University of Beirut Medical Center, 29 Beirut, Lebanon 30 31 6. UMR 1141, Hôpital Robert Debré, Institut National de la Santé et de la Recherche 32 Médicale, 75019 Paris, France 33 7. Department of Neurosurgery, Faculty of Medical Sciences, Lebanese University, Beirut, http://bmjopen.bmj.com/ 34 Lebanon 35 36 37 *Corresponding Authors 38 39 Jawad Fares 40 Email: [email protected] 41 42 on September 27, 2021 by guest. Protected copyright. 43 Youssef Fares 44 Email: [email protected] 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 4 ABSTRACT 5 6 7 Objective: This study aims to explore the shortterm and longterm prevalence and effects of 8 9 Posttraumatic Stress Disorder (PTSD) among victims of cluster munitions. 10 11 12 Design and Setting: A prospective 10year longitudinal study that took place in Lebanon. 13 14 15 Participants:For The 244 Lebanese peer civilian victimsreview of submunition onlyblasts, who were injured in 16 17 2006 and over 18 years old, were interviewed. Included were 239 participants (N=239) who had 18 19 20 been diagnosed with PTSD according to the DSM5 and the PTSD Checklist Civilian Version 21 22 (PCL) in 2006; participants were present for the 10year follow up. 23 24 25 Main Outcome Measures: The PTSD prevalence rates of participants in 2006 and 2016 were 26 27 compared. The analysis of the demographical data pertained to the association of longterm PTSD 28 29 with other variables was performed. Pvalues <0.05 were considered statistically significant for 30 31 all analyses (95% CI). 32 33 34 http://bmjopen.bmj.com/ 35 Results: All the 244 civilians injured by cluster munitions in 2006 responded, and were present 36 37 for longterm follow up in 2016. The prevalence of PTSD decreased significantly from 98% to 38 39 43% after ten years (p<0.001). A lower longterm prevalence was significantly associated with 40 41 the male sex (p<0.001), family support (p<0.001), and religion (p<0.001). Hospitalization 42 on September 27, 2021 by guest. Protected copyright. 43 (p=0.005) and severe functional impairment (p<0.001) posttrauma were significantly associated 44 45 46 with increased prevalence of longterm PTSD. Symptoms of negative cognition and mood were 47 48 more common in the long run. In addition, job instability was the most frequent socioeconomic 49 50 repercussion among the participants (88%). 51 52 53 Conclusions: Psychological symptoms, especially PTSD, remain high in waraffected 54 55 populations many years after the war, and this is particularly evident for Lebanese civilians that 56 57 were victimized by cluster munitions. Screening programs and psychological interventions need 58 59 60 1

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 to be implemented in vulnerable populations exposed to war traumas. Officials and public health 4 5 6 advocates should consider the socioeconomic implications, and help raise awareness against the 7 8 harm induced by cluster munitions and similar weaponry. 9 10 11 12 13 14 STRENGTHS AND LIMITATIONS OF THE STUDY 15 For peer review only 16 - This is the first longitudinal study to analyze PTSD in victims of cluster munition 17 18 explosions, which helps in better understanding the course and prognosis of PTSD in 19 20 21 these individuals. 22 23 - The DSM5 criteria for PTSD was adopted, and higher cutoff scores of the PCL were 24 25 used to aid in the diagnosis and minimize false positives. 26 27 - Functional impairment was measured using the Fares Scale of injuries due to cluster 28 29 munitions, which makes it easier to determine the true effects of cluster munitions and 30 31 project on the mental health service needs. 32 33 34 - More demographical data on health and socioeconomic outcomes could have been http://bmjopen.bmj.com/ 35 36 collected. 37 38 - The female sample size of our study may have contributed to the lack of significant 39 40 differences in some of the outcomes. 41 42 on September 27, 2021 by guest. Protected copyright. 43 KEYWORDS: 44 45 Posttraumatic stress disorder; PTSD; Cluster munitions; Cluster bombs; Adults; Lebanon 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 INTRODUCTION 4 5 6 As long as human beings have engaged in combat, there have often been extraordinarily 7 8 9 damaging psychiatric injuries among those who survive. The 2006 IsraeliLebanese conflict have 10 11 resulted in a large cohort of Lebanese civilian personnel exposed to combat related psychological 12 13 trauma as well as biomechanical trauma, including proximity to blast events. Cluster munitions 14 15 are weapons thatFor scatter smallerpeer submunitions review intended to kill oronly mutilate on impact (Figure 1). 16 17 They have been used by the Israeli forces in the south of Lebanon and are now scattered over 18 19 wide rural areas affecting its inhabitants (Figure 2). 20 21 22 Many states perceive cluster munitions as a main military equity that increases the efficiency of 23 24 25 suppressing, killing or destroying multiple targets within a specified area. However, often, when 26 27 they have been used near populated areas, civilians have died or got injured, either as a direct 28 29 result of the attack and its area effect, or as a result of postconflict unexploded submunitions. 30 31 While all types of ordnances fail to function at some rate, the failure rate for cluster munitions is 32 33

striking. http://bmjopen.bmj.com/ 34 35 36 Due to their easily "pickable" nature, submunitions can inflict various injuries. Biomechanical 37 38 injuries resulting from cluster munitions have been discussed in a series of research articles;18 39 40 41 psychological tribulations remain to be explored. Posttraumatic stress disorder (PTSD) is a 42 on September 27, 2021 by guest. Protected copyright. 43 psychiatric disorder that affects 7–8% of the general population at some point during their 44 45 lifetime;9 however, the prevalence is much higher among certain subgroups, including active duty 46 47 military personnel, and veterans and civilians exposed to blasts, warrelated injuries and, in this 48 49 case, injuries due to cluster munitions. 50 51 52 War experiences can affect mental health; however, largescale studies that focus on the short 53 54 term and longterm impact are rare. Such information may help screening programs in targeting 55 56 57 highrisk populations and raise awareness against the harm induced by cluster munitions and 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 similar weaponry. Therefore, this study aims to explore the shortterm and longterm prevalence 4 5 6 and effects of PTSD among adult victims of cluster munition explosions in Lebanon. 7 8 9 MATERIALS AND METHODS 10 11 12 Study Design 13 14 15 This was a prospectiveFor 10year peer longitudinal review study of victims of clusteronly munitions who got injured 16 17 after the 2006 LebaneseIsraeli conflict. The study took place over two phases: the first phase, 18 19 studying the shortterm prevalence of PTSD, was in 2006, and the second phase, exploring the 20 21 longterm outcome of PTSD, took place in 2016. The participants were first contacted by 22 23 telephone and asked if they would be willing to be interviewed about their mental state following 24 25 their injury and whether they preferred to be interviewed over the telephone or in the clinic. 26 27 28 Almost all the participants preferred to come to the clinic for the interview when they were 29 30 contacted the first time. Ten years after the first interview, the patients were contacted by 31 32 telephone and again asked whether they preferred to be interviewed over the telephone or in the 33 34 clinic. Almost half of the participants were not interested in coming to the clinic, and so, many of http://bmjopen.bmj.com/ 35 36 these followup interviews were held over the telephone. The Institutional Review Board at the 37 38 Lebanese University approved the study and deemed that verbal consent was sufficient for this 39 40 41 study. 42 on September 27, 2021 by guest. Protected copyright. 43 44 Participants 45 46 47 All of the 244 Lebanese civilian victims, who were aged 18 years and above, and injured by 48 49 cluster munitions were interviewed. To be included, participants must be mentally competent, 50 51 have met the DSM5 criteria for PTSD, and present with symptoms that lasted more than one 52 53 month (to rule out Acute Stress Disorder). It is worth mentioning that data from interviews 54 55 56 collected in 2006 were adapted to the new DSM5 after its release in 2013. Military personnel, 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 casualties presenting with injuries caused by other than submunitions, or those who had another 4 5 6 recent traumatic event within a month period were excluded. In addition, those who had PTSD 7 8 symptoms resulting from other traumatic events that occurred in their life were also excluded 9 10 from the study. The participants must be present for followup as well. 11 12 13 Assessment 14 15 For peer review only 16 All casualties were interviewed for the first time at least 1 month after the injury, in order to fall 17 18 within the DSM5 definition of PTSD, and no more than 9 months after the injury, in order to 19 20 avoid a possible aggravated clinical expression as the time grew near the anniversary. The 21 22 disturbance must have caused significant distress or impairment in the individual’s social 23 24 25 interactions, capacity to work or other important areas of functioning; functional impairment was 26 4 27 measured by the Fares Scale of injuries due to cluster munition explosions (Appendix A). In 28 29 addition, PTSD must not be the physiological result of another medical condition, medication, 30 31 drugs or alcohol. Followup interviews were conducted again after 10 years. To enhance 32 33

reliability, interviews were conducted by a single trained neuropsychologist. The interviewer was http://bmjopen.bmj.com/ 34 35 blind to the type of interventions the injured survivors had received. 36 37 38 The PTSD Checklist (PCL) Civilian Version for DSMIV was used to aid in the diagnosis of 39 40 41 PTSD among the participants. This questionnaire is a selfreport measure that can be read by 42 on September 27, 2021 by guest. Protected copyright. 43 respondents themselves or read to them either in person or over the telephone. It can be 44 45 completed in approximately 510 minutes. A total symptom severity score (range = 1785) can be 46 47 obtained by summing the scores from each of the 17 items that have response options ranging 48 49 from 1 "Not at all" to 5 "Extremely". The PCL can be scored to provide a presumptive diagnosis 50 51 by determining whether the total severity score exceeds a given normative threshold. A higher 52 53 54 cutpoint was considered to minimize false positives as it is recommended that in settings with 55 56 expected high rates of PTSD, such as specialty mental health clinics, to consider a higher cut 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 point.1012 The Veterans Affairs National Center for PTSD suggested a PCL cutpoint score of 4 5 13 6 45. Although there are no civilian version corresponding to the PTSD Checklist for DSM5 7 8 (PCL5), preliminary validation studies by the Veterans Affairs National Center for PTSD 9 10 suggests that optimal PCL5 cutpoints appear to be 1114 points lower than for PCL for DSMIV 11 12 cutpoints, with closer to an 11point difference for more stringent cutoffs.13 Still, our cutoff was 13 14 higher, which would help in achieving better screening results. Response categories 3–5 15 For peer review only 16 (Moderately or above) were treated as symptomatic, and responses 1–2 (below Moderately) were 17 18 19 considered as nonsymptomatic. The questions were administered by the interviewer in Arabic 20 21 after being translated by a certified translator. 22 23 24 Demographic and Outcome Variables 25 26 27 The data collected were as follows: (1) Demographics: sex; mean age; age range; mean time since 28 29 injury; hospitalization data; and data on discharge from the emergency room; (2) PTSD related 30 31 variables: Reexperiencing (recollection of the event through thoughts or perceptions, images, 32 33

dreams, illusions or hallucinations, dissociative flashback episodes or intense psychological http://bmjopen.bmj.com/ 34 35 distress or reactivity to cues that symbolize some aspect of the event); Avoidance (evasion of 36 37 38 thoughts, feelings, or conversations associated with the event and /or avoidance of people, places, 39 40 or activities that may trigger recollections of the event); Negative cognitions and mood (inability 41 42 to remember an important aspect of the event, persistent negative emotional state, and persistent on September 27, 2021 by guest. Protected copyright. 43 44 inability to experience positive emotions); Arousal (irritable behavior and angry outbursts, 45 46 reckless or selfdestructive behavior, hypervigilance, exaggerated startle response, concentration 47 48 problems, and/or sleep disturbance); (3) Injury related variables: severity of injury; and functional 49 50 51 impairment; and (4) Personal and social repercussions: marital and family problems; job 52 53 instability; and legal difficulties. 54 55 56 Statistical Analysis 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Data collected from the interviews were processed, and scores of the PCL were calculated for 4 5 6 each individual; scores ≥45 indicated the presence of PTSD. The PTSD prevalence rate of 7 t 8 participants according to sex in 2006 and 2016 were compared by means of paired twotailed 9 10 tests. This test was used to account for the dependent (before and after) observations. It calculates 11 12 the difference within each beforeandafter pair of measurements, determines the mean of these 13 14 differences and reports whether this mean of differences is statistically different than zero. The 15 For peer review only 16 analysis of the demographical data pertained to the association of longterm PTSD with other 17 18 19 variables such as sex, religion, family support and hospitalization, was performed using the Chi 20 21 square test, as it tests the significance of association between two categorical variables. All 22 23 statistical analyses were performed using the Statistical Package for Social Sciences, version 23.0 24 25 (SPSS Inc., Chicago, IL, USA). 26 27 28 RESULTS 29 30 31 All of the 244 civilians, aged 18 years and above, and injured by cluster munitions in 2006 32 33

responded, and were present for longterm followup in 2016. Of 244, 239 individuals (98%) http://bmjopen.bmj.com/ 34 35 were diagnosed with PTSD in 2006 and were the subject of our study (N=239). Participants 36 37 38 served as their own controls because matching a cohort of civilians injured by cluster munitions 39 40 and having PTSD with a cohort of civilians injured by cluster munitions who do not have PTSD 41 42 would have been difficult. on September 27, 2021 by guest. Protected copyright. 43 44 45 The mean age (±SD) of the participants was 27 (±4.3) years, ranging from 18 to 67 years. The 46 47 mean time since injury (±SD) was 2.2 (±1.2) months. The complete profile of the participants is 48 49 displayed in table 1. 50 51 52 The distribution of answers to the PCL checklist in the shortterm and the longterm are presented 53 54 in Appendices B and C respectively. Interestingly, the percentage of participants with PTSD was 55 56 57 98% in 2006 and it decreased to 43% in 2016 (105 of 244). The five individuals who did not 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 develop PTSD in the short run, did not develop PTSD in the long run as well. The most common 4 5 6 symptomatic response that persisted in the long run were: (1) feeling very upset when something 7 8 reminded of the stressful experience (65%); (2) loss of interest in things that were previously 9 10 enjoyable (63%); and (3) feeling distant or cut off from other people (63%). The least common 11 12 symptomatic responses that were present after 10 years were: (1) trouble falling or staying asleep 13 14 (17%); (2) feeling irritable or having angry outbursts (20%); and (3) suddenly acting or feeling as 15 For peer review only 16 if a stressful experience were happening again (20%). 17 18 19 20 Among the participants, more males (94%) than females (6%) were injured and subsequently 21 22 diagnosed with PTSD in the short run. In the long run, the prevalence of PTSD among males 23 24 decreased significantly from 224 to 91 (p<0.001). In females, however, the prevalence of PTSD 25 26 did not change much in the long run (5.8%). In total, the proportion of participants with PTSD 27 28 decreased significantly from 98% (239 of 244) to 43% (105 of 244) after ten years (p<0.001) 29 30 (table 2). Figure 3 depicts the changes in PTSD prevalence between 2006 and 2016. 31 32 33 34 In the long run, family support (p<0.001) and religion (p<0.001) were significantly associated http://bmjopen.bmj.com/ 35 36 with decreased prevalence of PTSD. Hospitalization postincident, opposed to being discharged 37 38 from the emergency room, was significantly associated with increased prevalence of longterm 39 40 PTSD (p=0.005). Severe functional impairment was also significantly correlated with increased 41 42 prevalence of longterm PTSD (p<0.001) (table 3). on September 27, 2021 by guest. Protected copyright. 43 44 45 Personal and socioeconomic repercussions of participants diagnosed with PTSD in the long run 46 47 48 are presented in table 4. 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 DISCUSSION 4 5 6 This is the first longitudinal study to analyze PTSD in victims of cluster munition explosions. 7 8 Overall, the prevalence of PTSD decreased significantly after 10 years. A lower longterm 9 10 prevalence was significantly associated with the male sex, family support and religion. 11 12 Hospitalization posttrauma and severe functional impairment were significantly associated with 13 14 increased prevalence of longterm PTSD. Symptoms of negative cognition and mood were more 15 For peer review only 16 common in the long run. In contrast, symptoms of arousal and reexperiencing were less 17 18 19 common. Job instability was the most frequent socioeconomic repercussion among the 20 21 participants in the long run. 22 23 24 The prevalence of PTSD in the Lebanese victims of cluster munitions was 98% in 2006. A 25 26 previous study on PTSD in civilian populations from Lebanon showed a prevalence of 29.3% in 27 28 the same year post war.14 After 10 years, the prevalence of PTSD among our participants was 29 30 found to be 43%. A recent study have shown a 23.4% prevalence of PTSD in a crosssectional 31

32 15 33 survey of southern Lebanese civilian sample. The striking difference in PTSD prevalence 34 http://bmjopen.bmj.com/ 35 between studies can be attributed to the cluster munition injury that the participants of our study 36 37 endured. Other studies have shown that the difference in trauma exposure cause alterations in 38 39 PTSD prevalence.16 Besides, the instrument of measure used differs from one study to another 40 41 and can lead to varying degrees of PTSD reporting. 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 Previous studies that examine the prevalence of PTSD in countries that had experienced war

46 1720 47 atrocities have shown lower prevalence of PTSD than our study. To examine the mental 48 17 49 health and cognitive effects of war trauma on war survivors in Yugoslavia, Baoğlu et al. found 50 51 that 33% of survivors suffered from PTSD. De Jong et al.19 attempted to study the impact of 52 53 trauma in postconflict lowincome countries where people have survived multiple traumatic 54 55 experiences. They found the prevalence rate of assessed PTSD to be 37.4% in Algeria, 28.4% in 56 57 19 20 58 Cambodia, 15.8% in Ethiopia, and 17.8% in Gaza. Priebe et al. studied mental health in five 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 countries of exYugoslavia and found prevalence rates of 15.6% to 41.8% for anxiety disorders. 4 5 6 The PTSD prevalence rate among victims of cluster munitions in our study is higher than those 7 8 reported in different samples from different countries, even after 10 years. 9 10 11 As there might be a difference in the psychobiology of PTSD between victims of cluster 12 13 munitions and civilians who were survivors of war, we compared our results to studies on 14 15 civilians who Forsustained a traumaticpeer injury. reviewIn a study by Ohry et al.only21 on traumatic brain injury 16 17 (TBI) patients, 33% met criteria for PTSD diagnosis. Shalev et al.22found that 25.5% of injured 18 19 23 20 trauma survivors met PTSD diagnostic criteria at followup. Also, Mora et al. reported a 32% 21 22 prevalence of PTSD in patients with explosionrelated burns. Furthermore, a prospective study 23 24 following traumatic events reported that 29.9% of survivors met criteria for PTSD at 1 month, 25 26 and 17.5% had PTSD at 4 months.24 Prevalence of PTSD among cluster munition victims was 27 28 still much higher. We believe that the injuries resulting from cluster munition blasts are unique 29 30 and different from other blast injuries; the high prevalence of PTSD and the functional 31 32 16 33 impairment sustained in our sample supports this claim. Gaylord et al. compared PTSD 34 http://bmjopen.bmj.com/ 35 prevalence in civilians with burn injuries and combat casualties treated in the same burn unit, and 36 37 found no significant difference in PTSD prevalence. Similarly, the group used the PCL as their 38 39 measure and the cutoff score was defined as ≥44.16They concluded that PTSD is related to the 40 41 type of trauma (burn trauma in their case) and not to the circumstances surrounding the injury 42 on September 27, 2021 by guest. Protected copyright. 43 (being a civilian or combat casualty).16In contrast, we agree with other studies that believe that 44 45 25 26 46 the difference in trauma exposure will cause alterations in PTSD prevalence. As can be 47 48 conferred from our results, the prevalence rate of PTSD resulting from submunition blast injuries 49 50 is much higher than that resulting from other blast related injuries. 51 52 53 Our study showed a significant drop in PTSD prevalence after 10 years of the trauma incident. 54 55 Other studies have also shown that, with time, PTSD decreases in prevalence and severity.27 28 56 57 Farhood29reassessed the prevalence and predictors of psychiatric disorders in a general population 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 from southern Lebanon conducted one year after the July war in 2007, and was thereafter 4 5 6 compared to an assessment conducted one year prewar in 2005 amid political turmoil on the 7 8 same population. Findings revealed a drop in PTSD symptoms in the 2007 sample at a rate of 9 29 10 17.9%, from 24.1% in 2005. 11 12 13 The most common longterm symptoms of PTSD were related to low mood and cognition. 14 15 Several studiesFor have shown peer the cooccurrence review of depressive symptoms only and PTSD early on after 16 17 trauma.21 23 26 3036 Furthermore, research have shown that the occurrence of depression during the 18 19 25 26 34 20 months that follow a traumatic event is an important mediator of chronicity in PTSD. 21 22 23 In our study, the number of males injured by cluster munitions was much higher than that of 24 25 females. The male predominance could be explained by the societal norms of males performing 26 27 farming, grazing and other rural activities where submunitions could be scattered.3 Furthermore, 28 29 the evacuation of females from the area during the war decreased the female casualties. 30 31 32 The female sex was significantly associated with having PTSD after 10 years. This coincides with 33 http://bmjopen.bmj.com/ 34 21 23 26 3036 35 previous research studies that assert that the female sex is a risk factor for PTSD. 36 37 38 Our results indicate that although females were less likely to experience a submunition blast, 39 40 they were more likely to meet the diagnosis of PTSD. Tolin and Foa37 expressed similar findings 41 42 in their cohort when studying sexspecific risk of potentially traumatic events and PTSD. Thus, on September 27, 2021 by guest. Protected copyright. 43 44 sex differences in risk of exposure to cluster munition trauma can only partially account for the 45 46 differential PTSD risk in male and female participants. 47 48 49 50 Our results show that individuals hospitalized after biomechanical trauma were at heightened risk 51 3840 52 for PTSD. Other studies have shown similar findings. Often, patients with PTSD present with 53 54 both medical and psychiatric symptoms. Several studies have suggested that physical injury in the 55 56 context of a psychologically traumatic event is a risk factor for PTSD.41 42Mollica et al.43found 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 that psychological trauma associated with injury in a civilian population was associated with 4 5 44 6 higher rates of PTSD and depression than other types of injury. Also, Hoge et al. found higher 7 8 rates of PTSD and depression in military personnel who may have had a TBI, particularly those 9 45 10 who reported loss of consciousness. Moreover, Schneiderman et al. conferred that a probable 11 12 TBI almost doubled the risk for PTSD. It is worth mentioning that 59% of the participants in our 13 14 study sustained injuries with severe functional impairment. Therefore, the type of injuries, the 15 For peer review only 16 severity of functional impairment and the personal and psychosocial stressors that come with the 17 18 19 situation may play a role in strengthening the relationship between injury and PTSD. 20 21 22 Similar to other studies in the literature, religion and social and familial support were negatively 23 4650 24 correlated to PTSD. It has been shown that a higher index of religious moral beliefs enables 25 26 better control of distress, and provides better mental health stability.46 Hasanović & Pajević46 27 28 assert that religion enables posttraumatic conflicts typical for combatants’ survivors to be more 29 30 easily overcome. CopelandLinder47 found that prayer buffered the effects of stress on physical 31 32 33 health and reduced the deleterious effects of experiencing depressive symptomatology. In another 34 http://bmjopen.bmj.com/ 35 study, social and family support, and religious beliefs were all found to be protective against 36 37 PTSD following war trauma and torture; higher perceived social support was associated with 38 39 higher positive emotions.48 We add that family support and religion help in overcoming post 40 41 trauma personal and psychosocial tribulations and lead to an improvement in mental health. 42 on September 27, 2021 by guest. Protected copyright. 43 44 Our results showed that personal and socioeconomic impairment was present in the participants 45 46 47 diagnosed with PTSD after 10 years. Studies have shown that effects of PTSD on the survivor's 48 49 immediate family and society at large are substantial. In addition to the emotional and cognitive 50 51 symptoms of PTSD, individuals with PTSD are more likely to experience marital and family 52 53 problems,51 52 job instability,53 legal difficulties,51 54 and physical health problems.55 56 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 In the long run, job instability was the most common socioeconomic repercussion (88%). Other 4 5 6 studies have shown that unemployment and low functionality to be associated with lower mental

7 15 8 health conditions, namely PTSD. We believe that the economic costs of PTSD are attributable 9 10 to lost work productivity. The severity of the injury and the associated functional impairment also 11 12 play a role. 13 14 15 Health wise, HogeFor et al.57 addpeer that individuals review with a history of PTSD only have a higher risk of 16 17 cardiovascular, respiratory, gastrointestinal, infectious, nervous system, and autoimmune 18 19 20 disorders, and are more likely to experience anxiety, depression, substance use, and other

21 26 46 51 54 58 59 22 psychiatric disorders. In addition, studies have reported a higher risk of suicidal 23 24 ideation among veterans with PTSD.60 61 25 26 27 In our study, we aimed to identify the prevalence of PTSD and associated factors following 28 29 cluster munitionrelated injuries by adopting a longitudinal design and paying close attention to 30 31 optimizing the research methodology. The longitudinal nature of the study helped understand 32 33 34 better the course and prognosis of PTSD in these individuals. This form of design can also http://bmjopen.bmj.com/ 35 36 elucidate the most effective screening and treatment regimens and factors that influence recovery. 37 38 The study can be considered representative as we were capable of reaching all of the 244 39 40 individuals who met our inclusion criteria. In addition, all of them were reachable for followup. 41 42 Due to the lack of consistent definitions of PTSD, we adopted the DSM5 criteria for PTSD and on September 27, 2021 by guest. Protected copyright. 43 44 used the PCL as a tool to aid in the diagnosis. We used a higher cutoff score to minimize false 45 46 47 positives. We measured functional impairment using the Fares Scale of injuries due to cluster 48 49 munitions, as to our knowledge, functional impairment measures have not been incorporated into 50 51 case definitions, which makes it difficult to determine the true effects of cluster munitions and 52 53 project on the mental health service needs. Furthermore, we studied the nature and severity of the 54 55 impairment along multiple dimensions to include work, family and social relationships. However, 56 57 the instrument of measure used differs from one study to another and can lead to varying degrees 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 of PTSD reporting. Besides, more demographical data on health and socioeconomic outcomes 4 5 6 could have been collected. In addition, information on comorbid mental health conditions like 7 8 depression, anxiety, sleep disorders, eating disorders, substance use disorders, and suicide could 9 10 have been collected. 11 12 13 More research is needed to study the interaction between trauma exposure, preexisting 14 15 psychologicalFor and biological peer vulnerabilities, review and the posttrauma environment.only Research is also 16 17 needed to better quantify the effect of comorbid factors such as aggression, risk taking behaviors, 18 19 20 and physical symptoms to understand the full extent of the effects of warrelated trauma. As the 21 22 prevalence of PTSD and other mental health conditions is high in the Lebanese population, 23 24 advanced neurobiological and neuropsychological studies are needed to better understand the 25 26 pathophysiology of PTSD and how genetic and environmental factors tally up to induce it. 27 28 29 IMPLICATIONS &CONCLUSIONS 30 31 32 This is the first longitudinal study to analyze PTSD in victims of cluster munition explosions. 33 http://bmjopen.bmj.com/ 34 Shortly posttrauma, 98% of those injured by cluster munitions were diagnosed with PTSD. In a 35 36 37 10year followup, 43% still had the disorder. Overall, the prevalence of PTSD decreased 38 39 significantly after 10 years. The male sex, family support and religion are possible protective 40 41 factors against PTSD. Whereas, hospitalization posttrauma and severe functional impairment 42 on September 27, 2021 by guest. Protected copyright. 43 were significantly correlated with increased prevalence of longterm PTSD. Symptoms of 44 45 negative cognition and mood were more common in the long run. In contrast, symptoms of 46 47 arousal and reexperiencing decreased. PTSD is associated with numerous deleterious outcomes, 48 49 50 and its costs on the victims, their immediate family, and society at large are substantial. 51 52 53 Intervention and screening programs should target vulnerable populations in Lebanon, and public 54 55 health advocates and governmental officials should run campaigns to raise awareness against 56 57 cluster munitions and to educate people on their shapes, characteristics and dangers. Moreover, 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 municipal efforts and regional projects that aim to demine unexploded submunitions should be 4 5 6 funded and supported by governmental and nongovernmental organizations. 7 8 9 Laws should be passed and enforced to ban the use of these detrimental weapons that have 10 11 negative effects on human, societal and ecological levels. More countries should join the 12 13 Convention on Cluster Munitions, as these inhumane weapons indiscriminately and 14 15 disproportionatelyFor harm innocentpeer civilians, review thereby violating the wellestablishedonly international 16 17 principles governing conflict and war today. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 FOOTNOTES 4 5 6 7 Contributors: JF and YF designed the study and collected the data. JF carried out the statistical 8 9 analyses and drafted the manuscript. All authors contributed to the analysis of the results. All 10 11 authors critically revised the manuscript for important intellectual content. JF and YF are the 12 13 primary investigators and guarantors of the study. 14 15 For peer review only 16 Funding: This research was funded by a grant from the Central Administration of the Lebanese 17 18 University, Beirut, Lebanon. 19 20 21 22 Competing interests: None declared. 23 24 25 Ethics Approval: The study was approved by the Institutional Review Board at the Faculty of 26 27 Medical Sciences, Lebanese University, Beirut, Lebanon. 28 29 30 Data Sharing Statement: Supplementary material are provided. 31 32 33

http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 REFERENCES 4 5 6 1. Bedrossian N, Haidar M, Fares J, et al. Inflammation and elevation of interleukin-12p40 in 7 9 8 patients with schizophrenia. Frontiers in molecular neuroscience 2016; . 9 2. Fares Y, Ayoub F, Fares J, et al. Pain and neurological sequelae of cluster munitions on 10 children and adolescents in South Lebanon. Neurological Sciences 2013;34(11):1971-76. 11 3. Fares Y, El-Zaatari M, Fares J, et al. Trauma-related infections due to cluster munitions. 12 Journal of infection and public health 2013;6(6):482-86. 13 4. Fares Y, Fares J. Anatomical and neuropsychological effects of cluster munitions. Neurological 14 Sciences 2013;34(12):2095-100. 15 For peer review only 16 5. Fares Y, Fares J, Gebeily S. Head and facial injuries due to cluster munitions. Neurological 17 Sciences 2014;35(6):905-10. 18 6. Hoteit A, Fares J. Psycho-environmental tribulations arising from cluster munitions in South 19 Lebanon. Sci-Afric Journal of Scientific Issues, Research and Essays 2014;2(10):469-73. 20 7. Abou-Abbass H, Bahmad H, Ghandour H, et al. Epidemiology and clinical characteristics of 21 traumatic brain injury in Lebanon: A systematic review. Medicine 2016;95(47). 22 8. Fares Y, Fares J. Neurosurgery in Lebanon: history, development, and future challenges. 23 24 World Neurosurgery 2016. 25 9. Kessler RC, Berglund P, Demler O, et al. Lifetime prevalence and age-of-onset distributions of 26 DSM-IV disorders in the National Comorbidity Survey Replication. Archives of general 27 psychiatry 2005;62(6):593-602. 28 10. Yeager DE, Magruder KM, Knapp RG, et al. Performance characteristics of the posttraumatic 29 stress disorder checklist and SPAN in Veterans Affairs primary care settings. General 30 29 31 hospital psychiatry 2007; (4):294-301. 32 11. Bliese PD, Wright KM, Adler AB, et al. Validating the primary care posttraumatic stress 33 disorder screen and the posttraumatic stress disorder checklist with soldiers returning 34 from combat. Journal of consulting and clinical psychology 2008;76(2):272. http://bmjopen.bmj.com/ 35 12. Freedy JR, Steenkamp MM, Magruder KM, et al. Post-traumatic stress disorder screening 36 test performance in civilian primary care. Family Practice 2010;27(6):615-24. 37 13. PTSD VNCf. Using the PTSD Checklist for DSM-IV (PCL). 2014. 38 39 https://sph.umd.edu/sites/default/files/files/PTSDChecklistScoring.pdf. 40 14. Farhood L, Dimassi H, Lehtinen T. Exposure to war-related traumatic events, prevalence of 41 PTSD, and general psychiatric morbidity in a civilian population from Southern Lebanon. 42 Journal of transcultural nursing 2006;17(4):333-40. on September 27, 2021 by guest. Protected copyright. 43 15. Farhood LF, Fares S, Sabbagh R, et al. PTSD and depression construct: prevalence and 44 predictors of co-occurrence in a South Lebanese civilian sample. European Journal of 45 Psychotraumatology 2016;7. 46 47 16. Gaylord KM, Holcomb JB, Zolezzi ME. A comparison of posttraumatic stress disorder 48 between combat casualties and civilians treated at a military burn center. Journal of 49 Trauma and Acute Care Surgery 2009;66(4):S191-S95. 50 17. Başoğlu M, Livanou M, Crnobarić C, et al. Psychiatric and cognitive effects of war in former 51 Yugoslavia: association of lack of redress for trauma and posttraumatic stress reactions. 52 Jama 2005;294(5):580-90. 53 54 18. Hall BJ, Hobfoll SE, Palmieri PA, et al. The psychological impact of impending forced settler 55 disengagement in Gaza: Trauma and posttraumatic growth. Journal of traumatic stress 56 2008;21(1):22-29. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 19. De Jong JT, Komproe IH, Van Ommeren M, et al. Lifetime events and posttraumatic stress 4 disorder in 4 postconflict settings. Jama 2001;286(5):555-62. 5 6 20. Priebe S, Bogic M, Ajdukovic D, et al. Mental disorders following war in the Balkans: a study 7 in 5 countries. Archives of general psychiatry 2010;67(5):518-28. 8 21. Ohry A, Rattok J, Solomon Z. Post-traumatic stress disorder in brain injury patients. Brain 9 Injury 1996;10(9):687-96. 10 22. Shalev AY, Peri T, Canetti L, et al. Predictors of PTSD in injured trauma survivors: a 11 prospective study. American Journal of Psychiatry 1996;153(2):219-25. 12 13 23. Mora AG, Ritenour AE, Wade CE, et al. Posttraumatic stress disorder in combat casualties 14 with burns sustaining primary blast and concussive injuries. Journal of Trauma and 15 AcuteFor Care Surgery peer 2009;66(4):S178-S85. review only 16 24. Shalev AY, Freedman S, Peri T, et al. Prospective study of posttraumatic stress disorder and 17 depression following trauma. American Journal of Psychiatry 1998. 18 25. DiGrande L, Neria Y, Brackbill RM, et al. Long-term posttraumatic stress symptoms among 19 3,271 civilian survivors of the September 11, 2001, terrorist attacks on the World Trade 20 173 21 Center. American journal of epidemiology 2011; (3):271-81. 22 26. Holbrook TL, Hoyt DB, Coimbra R, et al. Long-term posttraumatic stress disorder persists 23 after major trauma in adolescents: new data on risk factors and functional outcome. 24 Journal of Trauma and Acute Care Surgery 2005;58(4):764-71. 25 27. Greist J. Posttraumatic Stress Disorder (PTSD). Merck Manual 2014. 26 https://www.merckmanuals.com/professional/psychiatric-disorders/anxiety-and- 27 28 stressor-related-disorders/posttraumatic-stress-disorder-ptsd?qt=PTSD&alt=sh. 29 28. F Farhood L. Patterns of Psychiatric Morbidity Before and After a War in Lebanon at Twelve 30 Months Following Cessation of Hostilities. The Open Psychiatry Journal 2014;8(1). 31 29. Farhood L. Patterns of Psychiatric Morbidity Before and After a War in Lebanon at Twelve 32 Months Following Cessation of Hostilities. The Open Psychiatry Journal 2014;8(1). 33

30. Holbrook TL, Hoyt DB, Stein MB, et al. Gender differences in long-term posttraumatic stress http://bmjopen.bmj.com/ 34 disorder outcomes after major trauma: women are at higher risk of adverse outcomes 35 53 36 than men. Journal of Trauma and Acute Care Surgery 2002; (5):882-88. 37 31. Verger P, Dab W, Lamping DL, et al. The psychological impact of terrorism: an epidemiologic 38 study of posttraumatic stress disorder and associated factors in victims of the 1995- 39 1996 bombings in France. American Journal of Psychiatry 2004;161(8):1384-89. 40 32. Brewin CR, Andrews B, Valentine JD. Meta-analysis of risk factors for posttraumatic stress 41 disorder in trauma-exposed adults. Journal of consulting and clinical psychology 42 on September 27, 2021 by guest. Protected copyright. 2000;68(5):748. 43 44 33. Predictors of posttraumatic stress disorder and symptoms in adults: a meta-analysis. Annual 45 Meeting of the International Society for Traumatic Stress Studies, 14th, Nov, 1998, 46 Washington, DC, US; This article is based on a paper presented at the aforementioned 47 meeting; 2008. Educational Publishing Foundation. 48 34. Freedman SA, Brandes D, Peri T, et al. Predictors of chronic post-traumatic stress disorder. A 49 prospective study. The British journal of psychiatry 1999;174(4):353-59. 50 51 35. Javidi H, Yadollahie M. Post-traumatic stress disorder. The international journal of 52 occupational and environmental medicine 2011;3(1 January). 53 36. Farhood LF, Dimassi H. Prevalence and predictors for post-traumatic stress disorder, 54 depression and general health in a population from six villages in South Lebanon. Social 55 psychiatry and psychiatric epidemiology 2012;47(4):639-49. 56 37. Tolin DF, Foa EB. Sex differences in trauma and posttraumatic stress disorder: a quantitative 57 review of 25 years of research. Psychological bulletin 2006;132(6):959. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 38. Abenhaim L, Dab W, Salmi LR. Study of civilian victims of terrorist attacks (France 1982– 4 1987). Journal of Clinical Epidemiology 1992;45(2):103-09. 5 6 39. O'donnell ML, Creamer M, Pattison P, et al. Psychiatric morbidity following injury. American 7 Journal of Psychiatry 2004;161(3):507-14. 8 40. Shih RA, Schell TL, Hambarsoomian K, et al. Prevalence of PTSD and major depression 9 following trauma-center hospitalization. The Journal of trauma 2010;69(6):1560. 10 41. Vasterling JJ, Verfaellie M, Sullivan KD. Mild traumatic brain injury and posttraumatic stress 11 disorder in returning veterans: perspectives from cognitive neuroscience. Clinical 12 29 13 psychology review 2009; (8):674-84. 14 42. Koren D, Norman D, Cohen A, et al. Increased PTSD risk with combat-related injury: a 15 matchedFor comparison peer study of injured review and uninjured soldiers only experiencing the same 16 combat events. American Journal of Psychiatry 2005;162(2):276-28. 17 43. Mollica RF, Henderson DC, Tor S. Psychiatric effects of traumatic brain injury events in 18 Cambodian survivors of mass violence. The British Journal of Psychiatry 19 2002;181(4):339-47. 20 21 44. Hoge CW, McGurk D, Thomas JL, et al. Mild traumatic brain injury in US soldiers returning 22 from Iraq. New England journal of medicine 2008;358(5):453-63. 23 45. Schneiderman AI, Braver ER, Kang HK. Understanding sequelae of injury mechanisms and 24 mild traumatic brain injury incurred during the conflicts in Iraq and Afghanistan: 25 persistent postconcussive symptoms and posttraumatic stress disorder. American 26 journal of epidemiology 2008;167(12):1446-52. 27 28 46. Hasanović M, Pajević I. Religious moral beliefs as mental health protective factor of war 29 veterans suffering from PTSD, depressiveness, anxiety, tobacco and alcohol abuse in 30 comorbidity. Psychiatria Danubina 2010;22(2):203-10. 31 47. Copeland-Linder N. Stress among black women in a South African township: The protective 32 role of religion. Journal of Community Psychology 2006;34(5):577-99. 33

48. Feder A, Ahmad S, Lee EJ, et al. Coping and PTSD symptoms in Pakistani earthquake http://bmjopen.bmj.com/ 34 survivors: Purpose in life, religious coping and social support. Journal of Affective 35 147 36 Disorders 2013; (1):156-63. 37 49. Khamis V. Impact of war, religiosity and ideology on PTSD and psychiatric disorders in 38 adolescents from Gaza Strip and South Lebanon. Social & Medicine 39 2012;74(12):2005-11. 40 50. Johnson H, Thompson A. The development and maintenance of post-traumatic stress 41 disorder (PTSD) in civilian adult survivors of war trauma and torture: A review. Clinical 42 on September 27, 2021 by guest. Protected copyright. psychology review 2008;28(1):36-47. 43 44 51. Gates MA, Holowka DW, Vasterling JJ, et al. Posttraumatic stress disorder in veterans and 45 military personnel: Epidemiology, screening, and case recognition. Psychological 46 services 2012;9(4):361. 47 52. Jordan BK, Marmar CR, Fairbank JA, et al. Problems in families of male Vietnam veterans 48 with posttraumatic stress disorder. Journal of consulting and clinical psychology 49 1992;60(6):916. 50 51 53. Smith MW, Schnurr PP, Rosenheck RA. Employment outcomes and PTSD symptom severity. 52 Mental health services research 2005;7(2):89-101. 53 54. Kulka RA, Schlenger WE, Fairbank JA, et al. Trauma and the Vietnam war generation: Report 54 of findings from the National Vietnam Veterans Readjustment Study: Brunner/Mazel, 55 1990. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 55. Boscarino JA. Posttraumatic stress disorder and physical illness: results from clinical and 4 epidemiologic studies. Annals of the New York Academy of Sciences 2004;1032(1):141- 5 6 53. 7 56. O'Toole BI, Catts SV, Outram S, et al. The physical and mental health of Australian Vietnam 8 veterans 3 decades after the war and its relation to military service, combat, and post- 9 traumatic stress disorder. American Journal of Epidemiology 2009;170(3):318-30. 10 57. Hoge CW, Terhakopian A, Castro CA, et al. Association of posttraumatic stress disorder with 11 somatic symptoms, visits, and absenteeism among Iraq war veterans. 12 13 American Journal of Psychiatry 2007. 14 58. Long N, MacDonald C, Chamberlain K. Prevalence of posttraumatic stress disorder, 15 depressionFor and anxiety peer in a community review sample of New Zealandonly Vietnam War veterans. 16 Australian and New Zealand Journal of Psychiatry 1996;30(2):253-56. 17 59. Karam EG, Fayyad J, Karam AN, et al. Outcome of depression and anxiety after war: a 18 prospective epidemiologic study of children and adolescents. Journal of traumatic stress 19 2014;27(2):192-99. 20 21 60. Jakupcak M, Cook J, Imel Z, et al. Posttraumatic stress disorder as a risk factor for suicidal 22 ideation in Iraq and Afghanistan war veterans. Journal of traumatic stress 23 2009;22(4):303-06. 24 61. Pietrzak RH, Johnson DC, Goldstein MB, et al. Psychosocial buffers of traumatic stress, 25 depressive symptoms, and psychosocial difficulties in veterans of Operations Enduring 26 Freedom and Iraqi Freedom: the role of resilience, unit support, and postdeployment 27 120 28 social support. Journal of affective disorders 2010; (1):188-92. 29 30 31

32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Table 1. Demographics of the Lebanese participants injured with cluster munitions and 4 5 diagnosed with PTSD (N=239) in 2006. SD: Standard Deviation. *Severe functional 6 impairment was measured using the Fares Scale for injuries due to cluster munition 7 explosions (Grades II-IV were considered severe). 8 9 10 Demographic Trait Number (%) 11 12 Male 224 (94) 13 14 Female 15 (6) 15 For peer review only 16 Mean Age (years) ± SD 27 ± 4.3 17 18 Age Range (years) [18, 67] 19 20 Mean time since injury (months) ± SD 2.2 ± 1.2 21 22 23 Hospitalized 230 (96) 24 25 Discharged from emergency room 9 (4) 26 27 Severe functional impairment* 141 (59) 28 29 Family Support 228 (95) 30 31 Religious 220 (92) 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Table 2. Prevalence of PTSD among Lebanese victims of cluster munitions in 2006 and 2016 4 5 across the two sexes. *P value < 0.05 indicates significance. 6 Diagnosed with PTSD Diagnosed with PTSD 7 P value* 8 in 2006 in 2016 9 Male 224 91 <0.001 10 11 Female 15 14 0.334 12 13 Total 239 105 <0.001 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Table 3. Association of long-term PTSD (after 10 years) with variables such as family 4 5 support, religion, hospitalization, and severity of injury in Lebanese victims of cluster 6 munitions (N=239). SE: Standard Error of Mean. *P value < 0.05 indicates significance. 7 †Functional impairment was measured using the Fares Scale for injuries due to cluster 8 munition explosions (Grades II-IV were considered severe). 9 10 11 12 N PTSD rate SE P value* 13 14 Family Support 15 Yes 228 0.41 0.03 For peer review only <0.001 16 No 11 1.00 0 17 Religious 18 Yes 220 0.39 0.03 19 <0.001 20 No 19 1.00 0 21 Hospitalized Yes 230 0.46 0.03 22 0.005 23 No 9 0 0 24 Functional

25 Impairment† 26 Yes 141 0.74 0.04 <0.001 27 No 98 0 0 28 29 30 31 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Table 4. Personal and socioeconomic repercussions in Lebanese victims of cluster munitions 4 5 diagnosed with PTSD, 10 years post injury (N=105). 6 7 Social Repercussions Number (%) 8 9 Marital and Family Problems 21 (20) 10 11 Job Instability 92 (88) 12 13 Legal Difficulties 9 (9) 14 15 For peer review only 16 17 18

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 on September 27, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 1. A cluster munition containing numerous undetonated sub-munitions in southern Lebanon. 33 34 135x102mm (96 x 96 DPI) http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 2. An adopted map showing the distribution of the cluster munition strikes in southern Lebanon as of 31 August 20, 2006. Sources: U.N. Interim Force in Lebanon; U.N. Humanitarian Information Center. 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 Figure 3. Prevalence of PTSD among victims of cluster munitions in 2006 and 2016, in Lebanon. *The 28 change in PTSD prevalence between 2006 and 2016 is significant with p<0.05 29 30 31 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 SUPPLEMENTATY MATERIAL 4 5 6 Appendix A 7 8 Fares Scale of injuries due to cluster munitions explosions4. 9 10 11 Grade Level Percentage of Functional Disability Anatomical Lesion 12 13 14 15 For peer review Amputation only of three fingers: middle finger, 16 - 17 ring finger & little finger 18 Grade I ≤ %25 - Partial loss of a hand or foot 19 - Superficial wounds to the body and face 20 - Loss of skin and/or muscles in the 21 extremities 22 23 24 25 26 - Amputation of two fingers: one of the two is 27 the thumb 28 - Loss of an eye, leg and/or hand 29 Grade II %50 30 - Wound infection 31 - Recuperable lesion in abdomen and/or 32 thorax 33

- Psychological effects http://bmjopen.bmj.com/ 34 35 36 37 38 - Affection of more than two of the four 39 extremities 40 - Partial loss of vision and/or impairment of 41 Grade III %75 42 the auditory system on September 27, 2021 by guest. Protected copyright. 43 - Mutilation and partial loss of the digestive 44 system and /or lungs 45 - Lesion of the spinal cord 46 47 48 49 - Amputation of three extremities 50 51 Grade IV >%75 - Total loss of vision and/or auditory capacity 52 - Tetraplegia 53 - Intellectual incapacity 54 55 56 57 58 59 60 1

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 33 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Appendix B 4 5 6 Distribution of answers of the 244 Lebanese adult injuries of cluster munitions to the 17itemPTSD checklist in late 2006 7 8 and early 2007. 9 10 Not at all A little bit Moderately Quite a bit Extremely No. Response 11 (1) (2) (3) (4) (5) 12 Repeated, disturbing memories, thoughts, or 13 01 5 15 36 62 126 14 images of a stressful experience from the past? 15 Repeated, disturbing Fordreams of a peerstressful review only 02 6 14 41 59 124 16 experience from the past? 17 18 Suddenly acting or feeling as if a stressful 19 03 experience were happening again (as if you 4 13 38 64 125 20 were reliving it)? 21 Feeling very upset when something reminded 22 04 8 13 32 79 112 23 you of a stressful experience from the past? 24 Having physical reactions (e.g., heart 25 pounding, trouble breathing, or sweating) 26 05 8 10 31 59 136 27 when somethingreminded you of a stressful 28 experience from the past? 29 Avoid thinking about or talking about a 30 31 06 stressful experience from the past or avoid 8 13 44 57 122 32 having feelings related to it? 33

Avoid activities or situations because they http://bmjopen.bmj.com/ 34 07 remind you of a stressful experience from the 10 13 45 55 121 35 past? 36 37 Trouble remembering important parts of a 08 12 13 46 61 112 38 stressful experience from the past? 39 Loss of interest in things that you used to 40 09 9 12 47 57 119 41 enjoy? 42 10 Feeling distant or cut off from other people? 11 28 51 47 107 on September 27, 2021 by guest. Protected copyright. 43 Feeling emotionally numb or being unable to 44 11 9 25 53 47 110 45 have loving feelings for those close to you? 46 Feeling as if your future will somehow be cut 12 9 22 50 49 114 47 short? 48 Trouble falling or staying asleep? 12 24 48 55 105 49 13 50 14 Feeling irritable or having angry outbursts? 3 22 42 42 135 51 15 Having difficulty concentrating? 5 17 43 47 132 52 16 Being “super alert” or watchful on guard? 13 28 45 56 102 53 54 17 Feeling jumpy or easily startled? 14 21 48 60 101 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 33 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 Appendix C 3 4 Distribution of answers of the 244 Lebanese adult injuries of cluster munitions to the 17itemPTSD checklist in 2016. 5 6 7 8 Not at all A little bit Moderately Quite a bit Extremely No. Response 9 (1) (2) (3) (4) (5) 10 Repeated, disturbing memories, thoughts, or 11 01 95 59 50 26 14 12 images of a stressful experience from the past? 13 Repeated, disturbing dreams of a stressful 02 80 70 46 36 12 14 experience from the past? 15 For peer review only 16 Suddenly acting or feeling as if a stressful 17 03 experience were happening again (as if you 110 85 24 20 5 18 were reliving it)? 19 20 Feeling very upset when something reminded 04 45 41 56 60 42 21 you of a stressful experience from the past? 22 23 Having physical reactions (e.g., heart pounding, 24 trouble breathing, or sweating) when 05 111 67 30 22 14 25 somethingreminded you of a stressful 26 experience from the past? 27 28 Avoid thinking about or talking about a 29 06 stressful experience from the past or avoid 113 73 30 18 10 30 having feelings related to it? 31 32 Avoid activities or situations because they 33

07 remind you of a stressful experience from the 101 40 32 24 47 http://bmjopen.bmj.com/ 34 past? 35 36 Trouble remembering important parts of a 08 114 48 27 37 18 37 stressful experience from the past? 38 39 Loss of interest in things that you used to 09 54 36 47 61 46 40 enjoy? 41

10 Feeling distant or cut off from other people? 50 40 41 59 54 on September 27, 2021 by guest. Protected copyright. 42 43 Feeling emotionally numb or being unable to 11 73 69 36 38 28 44 have loving feelings for those close to you? 45 Feeling as if your future will somehow be cut 46 12 69 67 43 42 23 47 short? 48 13 Trouble falling or staying asleep? 133 69 12 18 12 49 14 Feeling irritable or having angry outbursts? 127 68 15 20 14 50 51 15 Having difficulty concentrating? 121 61 20 24 18 52 16 Being “super alert” or watchful on guard? 80 52 39 46 27 53 54 17 Feeling jumpy or easily startled? 84 54 35 38 33 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from Page 32 of 33 7 7 5 5 7 7 5,6 5 5

4 4 4 4

7 7 7 7 NA 7 7 7 7 7 7 7 7 NA 1,2 8 8 Reported Reported onpage # 2 2

cohort studies cohort

BMJ Open http://bmjopen.bmj.com/ on September 27, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For matchedFor studies, give matching criteria numberand of exposedand unexposed

) If applicable, how explain loss to was follow-up addressed ) Describe methodsany used tosubgroups examine and interactions ) ) in the Provide abstract informativean balanced summaryand of what was donewhat was and found ) statistical Describe all methods, including thosecontrol confoundingusedfor to ) eligibilitythe Give criteria, theand sourcesand methods of selectionof participants. Describe methods follow-up of 5,6 ) Indicate study’sthe design commonlya with used term in ortitle the abstractthe ) Describe) sensitivity any analyses For peer review only ) Explain missinghow data were addressed e d c b b b a a a For each variable of interest,For variable sourceseach give of data and details of methodsassessment of (measurement). Describe comparability comparability of assessment methods if isthanthere more groupone

applicable collection ( ( ( ( ( ( Recommendation why STROBE 2007 (v4) Statement—Checklist of items that should be included in reports of reports in included be should that items of Statement—Checklist (v4) 2007 STROBE 3 3 objectives, State specific including prespecified any hypotheses 1 1 ( 5 5 Describe setting,the locations, and relevantdates, including periods exposure,of recruitment, follow-up, and data 9 9 Describe any addressefforts to sourcespotential of bias 6 6 ( 7 7 Clearly define outcomes, all exposures, predictors, potential and confounders, effect modifiers.diagnostic Give criteria, if 4 4 Present key elements of study in paperthedesign early 8* 10 10 Explain thehow study wassize arrived at # Item Item

Title andTitle abstract Results Statistical Statistical methods 12 ( Quantitative variables 11 Explain quantitativehow variableshandled were inthe analyses. If describeapplicable, which groupingschosen were and Study size Bias Data sources/Data measurement Variables Participants Setting Study design Methods Objectives Background/rationale 2 Explain scientific the background investigation thefor and rationale being reported Introduction

Section/Topic

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 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 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from 16 16 15,16 11-14 11-14 9,10,11 9-15 9-15 8 8 NA 8 8 8 - Table18 8 8 15 15 9 9 NA 8,9 16 16 8,9 NA NA

BMJ Open http://bmjopen.bmj.com/ on September 27, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml ) Report category boundariescontinuous when variables categorized were ) unadjustedGive estimates ifand, applicable, confounder-adjusted estimates and their precision (eg, confidence95% For peerrelevant, If ) consider translating estimates of relative risk into absolute for risk meaningful periodtime a review only c b a similar studies, other relevantand evidence confounders eligible, includedeligible, in study,the completing follow-up, analysedand interval). Makeinterval). clear confounderswhich andwere foradjusted they why were included which thewhich article presentis based

( ( Summarise (c) follow-up time (eg, average total and amount) (b)number Indicate of participants missingwith data eachof for interest variable Consider (c) diagram flow use of a (b)reasons Give non-participationfor each at stage 18 18 Summarise resultskey to reference with study objectives 20 20 Give cautiousa interpretationoverall of results considering objectives, limitations, of multiplicity resultsanalyses, from 16 16 ( 22 22 the Give of fundingsource and role the of fundersthe presentthe for if study theand, forapplicable, original study on 13* (a) numbersReport of individuals each at of stage study—eg numbers potentially eligible, eligibility,examined for confirmed

An Explanation and Elaboration article discusses itemeach checklist and gives methodological backgroundand published examples of transparent reporting. The STROBE checklist is best used in thiswith conjunction (freely available article onsites the Medicine of PLoS Web http://www.plosmedicine.org/, at Annals of Internal Medicine at http://www.annals.org/, Epidemiologyhttp://www.epidem.com/). at and Information on InitiativeSTROBE the is www.strobe-statement.org.available at information*Give separately cases for and controls studiesin case-controlfor applicable, if and, exposedunexposed and groups in cohort and cross-sectional studies. Note: Funding Other information Generalisability 21 Discuss (external generalisability the of validity)the study results Interpretation Limitations Key Key results Discussion Other analyses 17 Report other analyses done—eg analyses of subgroupsinteractions, sensitivity and and analyses

Main resultsMain Outcome Outcome data 15* Report numbers of outcome events or summary timemeasures over

Descriptive data 14* characteristics (a) Give participantsof study (eg clinical,demographic, social) informationand exposureson and potential

Participants Page 33 of 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 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 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

Posttraumatic stress disorder in adult victims of cluster munitions in Lebanon: A 10-year longitudinal study

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-017214.R1

Article Type: Research

Date Submitted by the Author: 24-May-2017

Complete List of Authors: Fares, Jawad; Lebanese University, Neuroscience Research Center, Faculty of Medical Sciences; American University of Beirut Faculty of Medicine Gebeily, Souheil; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences; Lebanese University , Department of Neurology, Faculty of Medical Sciences Saad, Mohamad; University of Washington School of Medicine Harati, Hayat; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences Nabha, Sanaa; Lebanese University, Neuroscience Research Center, Faculty of Medical Sciences Said, Najwane; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences Kanso, Mohamad; American University of Beirut Medical Center, Department of Emergency Medicine; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences http://bmjopen.bmj.com/ Abdel Rassoul, Ronza; INSERM U1141, Hôpital Robert Debré; Lebanese University , Neuroscience Research Center, Faculty of Medical Sciences Fares, Youssef; Lebanese University, Neuroscience Research Center, Faculty of Medical Sciences; Lebanese University, Department of Neurosurgery, Faculty of Medical Sciences

Primary Subject Neurology Heading: on September 27, 2021 by guest. Protected copyright. Secondary Subject Heading: Mental health, Public health, Global health

Posttraumatic stress disorder, PTSD, cluster munitions, cluster bombs, Keywords: Lebanon, Adults

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 4 Original Article 5 6 Posttraumatic stress disorder in adult victims of cluster 7 8 9 munitions in Lebanon: A 10-year longitudinal study 10 11 12 13 Jawad Fares1,2*, Souheil Gebeily1,3, Mohamad Saad1,4, Hayat Harati1, Sanaa Nabha1, 14 Najwane Said1, Mohamad Kanso1,5, Ronza Abdel Rassoul1,6, Youssef Fares1,7* 15 For peer review only 16 17 18 19 1. Neuroscience Research Center, Faculty of Medical Sciences, Lebanese University, 20 Beirut, Lebanon 21 2. Faculty of Medicine, American University of Beirut, Beirut, Lebanon 22 23 3. Department of Neurology, Faculty of Medical Sciences, Lebanese University, Beirut, 24 Lebanon 25 4. Division of Statistical Genetics, Department of Biostatistics, University of Washington, 26 Seattle, WA 98195, USA 27 28 5. Department of Emergency Medicine, American University of Beirut Medical Center, 29 Beirut, Lebanon 30 6. UMR 1141, Hôpital Robert Debré, Institut National de la Santé et de la Recherche 31 Médicale, 75019 Paris, France 32 33 7. Department of Neurosurgery, Faculty of Medical Sciences, Lebanese University, Beirut, 34 Lebanon http://bmjopen.bmj.com/ 35 36 *Corresponding Authors 37 38 39 Jawad Fares 40 Email: [email protected] 41 42 Youssef Fares on September 27, 2021 by guest. Protected copyright. 43 Email: [email protected] 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 4 ABSTRACT 5 6 7 Objective: This study aims to explore the shortterm and longterm prevalence and effects of 8 9 Posttraumatic Stress Disorder (PTSD) among victims of cluster munitions. 10 11 12 Design and Setting: A prospective 10year longitudinal study that took place in Lebanon. 13 14 15 Participants:For The 244 Lebanese peer civilian victimsreview of submunition onlyblasts, who were injured in 16 17 2006 and over 18 years old, were interviewed. Included were participants who had been 18 19 20 diagnosed with PTSD according to the DSM5 and the PTSD Checklist Civilian Version (PCL) 21 22 in 2006. Interviewees were present for the 10year follow up. 23 24 25 Main Outcome Measures: The PTSD prevalence rates of participants in 2006 and 2016 were 26 27 compared. The analysis of the demographical data pertained to the association of longterm PTSD 28 29 with other variables was performed. Pvalues <0.05 were considered statistically significant for 30 31 all analyses (95% CI). 32 33 34 http://bmjopen.bmj.com/ 35 Results: All the 244 civilians injured by cluster munitions in 2006 responded, and were present 36 37 for longterm follow up in 2016. The prevalence of PTSD decreased significantly from 98% to 38 39 43% after ten years (p<0.001). A lower longterm prevalence was significantly associated with 40 41 the male sex (p<0.001), family support (p<0.001), and religion (p<0.001). Hospitalization 42 on September 27, 2021 by guest. Protected copyright. 43 (p=0.005) and severe functional impairment (p<0.001) posttrauma were significantly associated 44 45 46 with increased prevalence of longterm PTSD. Symptoms of negative cognition and mood were 47 48 more common in the long run. In addition, job instability was the most frequent socioeconomic 49 50 repercussion among the participants (88%). 51 52 53 Conclusions: Psychological symptoms, especially PTSD, remain high in waraffected 54 55 populations many years after the war; this is particularly evident for Lebanese civilians that were 56 57 victimized by cluster munitions. Screening programs and psychological interventions need to be 58 59 60 1

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 implemented in vulnerable populations exposed to war traumas. Officials and public health 4 5 6 advocates should consider the socioeconomic implications, and help raise awareness against the 7 8 harm induced by cluster munitions and similar weaponry. 9 10 11 12 13 14 STRENGTHS AND LIMITATIONS OF THE STUDY 15 For peer review only 16 - This is the first longitudinal study to analyze PTSD in victims of cluster munition 17 18 explosions, which helps in better understanding the course and prognosis of PTSD in 19 20 21 these individuals. 22 23 - The DSM5 criteria for PTSD was adopted, and higher cutoff scores of the PCL were 24 25 used to aid in the diagnosis and minimize false positives. 26 27 - Functional impairment was measured using the Fares Scale of injuries due to cluster 28 29 munitions, which makes it easier to determine the true effects of cluster munitions and 30 31 project on the mental health service needs. 32 33 34 - More demographical data on health and socioeconomic outcomes could have been http://bmjopen.bmj.com/ 35 36 collected. 37 38 - The female sample size of our study may have contributed to the lack of significant 39 40 differences in some of the outcomes. 41 42 on September 27, 2021 by guest. Protected copyright. 43 KEYWORDS: 44 45 Posttraumatic stress disorder; PTSD; Cluster munitions; Cluster bombs; Adults; Lebanon 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 INTRODUCTION 4 5 6 There have often been severe damaging psychiatric injuries among those who survive combat. 7 8 9 The 2006 Israeli–Lebanese conflict have resulted in a large cohort of Lebanese civilian personnel 10 11 exposed to combatrelated psychological trauma, as well as biomechanical trauma resulting from 12 13 proximity to blast events. 14 15 For peer review only 16 Cluster munitions are weapons that scatter smaller submunitions intended to kill or mutilate on 17 18 impact (Figure 1). They have been extensively used by the Israeli forces in the south of Lebanon 19 20 and are now scattered over wide rural areas affecting its inhabitants.1 2 21 22 23 Many states perceive cluster munitions as a main military equity that increases the efficiency of 24 25 suppressing, killing or destroying multiple targets within a specified area. However, often, when 26 27 28 they have been used near populated areas, civilians have died or got injured, either as a direct 29 111 30 result of the attack and its area effect, or as a result of postconflict unexploded submunitions. 31 32 While all types of ordnances fail to function at some rate, the failure rate for cluster munitions is 33 34 striking. In 2006, 4 million submunitions were dropped over Lebanese soil; more than one http://bmjopen.bmj.com/ 35 36 million of which remained unexploded.12 37 38 39 Due to their easily "pickable" nature, submunitions can inflict various injuries. Biomechanical 40 41 18

injuries resulting from cluster munitions have been discussed in a series of research articles; on September 27, 2021 by guest. Protected copyright. 42 43 44 however, psychological tribulations remain to be explored. Posttraumatic stress disorder (PTSD) 45 46 is a psychiatric disorder that affects 7–8% of the general population at some point during their 47 48 lifetime;13 however, the prevalence is much higher among certain subgroups, including active 49 50 duty military personnel, and veterans and civilians exposed to blasts and warrelated injuries. 51 52 53 War experiences can affect mental health, yet largescale studies that focus on the shortterm and 54 55 longterm impact are rare. Such information may help screening programs in targeting highrisk 56 57 populations and raise awareness against the harm induced by cluster munitions and similar 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 weaponry. Therefore, this study aims to explore the shortterm and longterm prevalence and 4 5 6 effects of PTSD among adult victims of cluster munition explosions in Lebanon. 7 8 9 MATERIALS AND METHODS 10 11 12 Study Design 13 14 15 This was a prospectiveFor 10year peer longitudinal review study of victims of clusteronly munitions who got injured 16 17 after the 2006 IsraeliLebanese conflict. The study took place over two phases: the first phase, 18 19 studying the shortterm prevalence of PTSD, was in 2006, and the second phase, exploring the 20 21 longterm outcome of PTSD, took place in 2016. The participants were first contacted by 22 23 telephone and asked if they would be willing to be interviewed about their mental state following 24 25 their injury, and whether they preferred to be interviewed over the telephone or in the clinic. 26 27 28 Almost all the participants preferred to come to the clinic for the interview when they were 29 30 contacted the first time. Ten years after the first interview, the patients were contacted by 31 32 telephone and again asked whether they preferred to be interviewed over the telephone or in the 33 34 clinic. Almost half of the participants were not interested in coming to the clinic, and so, many of http://bmjopen.bmj.com/ 35 36 these followup interviews were held over the telephone. The Institutional Review Board at the 37 38 Lebanese University approved the study and deemed that verbal consent was sufficient. 39 40 41 42 Participants on September 27, 2021 by guest. Protected copyright. 43 44 45 All the 244 Lebanese civilian victims, who were aged 18 years and above, and injured by cluster 46 47 munitions were interviewed. To be included, participants must be mentally competent, have met 48 49 the DSM5 criteria for PTSD, and present with symptoms that lasted more than one month (to 50 51 rule out Acute Stress Disorder). It is worth mentioning that data from interviews collected in 2006 52 53 were adapted to the new DSM5 after its release in 2013. Military personnel, casualties presenting 54 55 56 with injuries caused by other than submunitions, or those who had another recent traumatic event 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 within a month period were excluded. In addition, those who had PTSD symptoms resulting from 4 5 6 other traumatic events that occurred in their life were also excluded from the study. The 7 8 participants must be present for followup as well. 9 10 11 Assessment 12 13 14 All casualties were interviewed for the first time at least 1 month after the injury, in order to fall 15 For peer review only 16 within the DSM5 definition of PTSD, and no more than 9 months after the injury, in order to 17 18 avoid a possible aggravated clinical expression as the time grew near the anniversary. The 19 20 disturbance must have caused significant distress or impairment in the individual’s social 21 22 interactions, capacity to work or other important areas of functioning. Functional impairment was 23 24 2 25 measured by the Fares Scale of injuries due to cluster munition explosions (Appendix A). In 26 27 addition, PTSD must not be the physiological result of another medical condition, medication, 28 29 drugs or alcohol. Followup interviews were conducted again after 10 years. To enhance 30 31 reliability, interviews were conducted by a single trained neuropsychologist. The interviewer was 32 33

blind to the type of interventions the injured survivors had received. After their first interview, http://bmjopen.bmj.com/ 34 35 participants were provided with a list of authorized institutions that provide psychological health 36 37 38 services and hospitals/medical centers that have psychiatry departments to increase the benefits of 39 40 the study. 41 42 on September 27, 2021 by guest. Protected copyright. 43 The PTSD Checklist (PCL) Civilian Version for DSMIV was used to aid in the diagnosis of 44 45 PTSD among the participants. This questionnaire is a selfreport measure that can be read by 46 47 respondents themselves or read to them either in person or over the telephone. It can be 48 49 completed in approximately 510 minutes. A total symptom severity score (range = 1785) can be 50 51 obtained by summing the scores from each of the 17 items that have response options ranging 52 53 54 from 1 "Not at all" to 5 "Extremely". The PCL can be scored to provide a presumptive diagnosis 55 56 by determining whether the total severity score exceeds a given normative threshold. A higher 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 cutpoint was considered to minimize false positives as it is recommended that in settings with 4 5 6 expected high rates of PTSD, such as specialty mental health clinics, to consider a higher cut

7 1416 8 point. The Veterans Affairs National Center for PTSD suggested a PCL cutpoint score of 9 17 10 45. Although there is no civilian version corresponding to the PTSD Checklist for DSM5 11 12 (PCL5), preliminary validation studies by the Veterans Affairs National Center for PTSD 13 14 suggests that optimal PCL5 cutpoints appear to be 1114 points lower than for PCL for DSMIV 15 For peer review only 16 cutpoints, with closer to an 11point difference for more stringent cutoffs.17 Response categories 17 18 19 3–5 (Moderately or above) were treated as symptomatic, and responses 1–2 (below Moderately) 20 21 were considered as nonsymptomatic. The questions were administered by the interviewer in 22 23 Arabic after being translated by a certified translator. The questionnaire was sdeemed valid and 24 25 reliable after being piloted among a sample of 10 patients with PTSD. 26 27 28 Demographic and Outcome Variables 29 30 31 The data collected were as follows: (1) Demographics: sex; mean age; age range; mean time since 32 33

injury; hospitalization data; and data on discharge from the emergency room; (2) PTSD related http://bmjopen.bmj.com/ 34 35 variables: Reexperiencing (recollection of the event through thoughts or perceptions, images, 36 37 38 dreams, illusions or hallucinations, dissociative flashback episodes or intense psychological 39 40 distress or reactivity to cues that symbolize some aspect of the event); Avoidance (evasion of 41 42 thoughts, feelings, or conversations associated with the event and /or avoidance of people, places, on September 27, 2021 by guest. Protected copyright. 43 44 or activities that may trigger recollections of the event); Negative cognitions and mood (inability 45 46 to remember an important aspect of the event, persistent negative emotional state, and persistent 47 48 inability to experience positive emotions); Arousal (irritable behavior and angry outbursts, 49 50 51 reckless or selfdestructive behavior, hypervigilance, exaggerated startle response, concentration 52 53 problems, and/or sleep disturbance); (3) Injury related variables: severity of injury; and functional 54 55 impairment; and (4) Personal and social repercussions: marital and family problems; job 56 57 instability; and legal difficulties. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Statistical Analysis 4 5 6 Data collected from the interviews were processed, and scores of the PCL were calculated for 7 8 9 each individual; scores ≥45 supported the diagnosis of PTSD. The PTSD prevalence rate of 10 11 participants according to sex in 2006 and 2016 were compared by means of paired twotailed t 12 13 tests. This test was used to account for the dependent (before and after) observations. It calculates 14 15 the differenceFor within each peerbeforeandafter review pair of measurements, only determines the mean of these 16 17 differences and reports whether this mean of differences is statistically different than zero. The 18 19 analysis of the demographical data pertained to the association of longterm PTSD with other 20 21 22 variables such as sex, religion, family support and hospitalization, was performed using the Chi 23 24 square test, as it tests the significance of association between two categorical variables. All 25 26 statistical analyses were performed using the Statistical Package for Social Sciences, version 23.0 27 28 (SPSS Inc., Chicago, IL, USA). 29 30 31 RESULTS 32 33 34 All the 244 civilians, aged 18 years and above, and injured by cluster munitions in 2006 http://bmjopen.bmj.com/ 35 36 responded, and were present for longterm followup in 2016. Of 244, 239 individuals (98%) 37 38 were diagnosed with PTSD in 2006 and were the focus of our study (N=239). Participants served 39 40 41 as their own controls because matching a cohort of civilians injured by cluster munitions and 42 on September 27, 2021 by guest. Protected copyright. 43 having PTSD with a cohort of civilians injured by cluster munitions who do not have PTSD 44 45 would have been difficult. 46 47 48 The mean age (±SD) of the participants was 27 (±4.3) years, ranging from 18 to 67 years. The 49 50 mean time since injury (±SD) was 2.2 (±1.2) months. The complete profile of the participants is 51 52 displayed in Table 1. 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Interestingly, the percentage of participants with PTSD was 98% in 2006 (239 of 244) and it 4 5 6 significantly decreased to 43% in 2016 (105 of 244). The 5 individuals who were not diagnosed 7 8 with PTSD in 2006, did not develop it in 2016. 9 10 11 The distribution of answers to the PCL checklist in the shortterm and longterm are presented in 12 13 Appendices B and C, respectively. The most common symptomatic responses that persisted in the 14 15 long run were:For (1) feeling verypeer upset when review something reminded ofonly the stressful experience (65%); 16 17 (2) loss of interest in things that were previously enjoyable (63%); and (3) feeling distant or cut 18 19 off from other people (63%). The least common symptomatic responses that were present after 10 20 21 22 years were: (1) trouble falling or staying asleep (17%); (2) feeling irritable or having angry 23 24 outbursts (20%); and (3) suddenly acting or feeling as if a stressful experience were happening 25 26 again (20%). 27 28 29 Among the participants, more males (94%) than females (6%) were diagnosed with PTSD in 30 31 2006. The number of males diagnosed with PTSD decreased significantly from 224 in 2006, to 91 32 33

in 2016 (p<0.001). However, the number of females diagnosed with PTSD did not change http://bmjopen.bmj.com/ 34 35 36 significantly in the long run (Table 2). 37 38 39 In the long run, family support (p<0.001) and religion (p<0.001) were significantly associated 40 41 with decreased prevalence of PTSD. Hospitalization postincident, opposed to being discharged 42 on September 27, 2021 by guest. Protected copyright. 43 from the emergency room, was significantly associated with increased prevalence of longterm 44 45 PTSD (p=0.005). Severe functional impairment was also significantly correlated with increased 46 47 prevalence of longterm PTSD (p<0.001) (Table 3). 48 49 50 51 Personal and socioeconomic repercussions of participants diagnosed with PTSD in the long run 52 53 are presented in Table 4. 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 DISCUSSION 4 5 6 This is the first longitudinal study to analyze PTSD in victims of cluster munition explosions. 7 8 Overall, the prevalence of PTSD decreased significantly after 10 years. A lower longterm 9 10 prevalence was significantly associated with the male sex, family support and religion. 11 12 Hospitalization posttrauma and severe functional impairment were significantly associated with 13 14 increased prevalence of longterm PTSD. Symptoms of negative cognition and mood were more 15 For peer review only 16 common in the long run. In contrast, symptoms of arousal and reexperiencing were less 17 18 19 common. Job instability was the most frequent socioeconomic repercussion among the 20 21 participants in the long run. 22 23 24 The prevalence of PTSD in the Lebanese victims of cluster munitions was 98% in 2006. A 25 26 previous study on PTSD in civilian populations from Lebanon showed a prevalence of 29.3% in 27 28 the same year post war.18 After 10 years, the prevalence of PTSD among our participants was 29 30 found to be 43%. A recent study have shown a PTSD prevalence of 23.4% in a crosssectional 31

32 19 33 survey of Southern Lebanese civilian sample. The striking difference in PTSD prevalence 34 http://bmjopen.bmj.com/ 35 between studies within the same time frame can be attributed to the cluster munition injury that 36 37 the participants of our study endured. 38 39 40 Previous studies that examine the prevalence of PTSD in countries that had experienced war 41 on September 27, 2021 by guest. Protected copyright. 42 atrocities have shown lower prevalence of PTSD than our study.2023 To examine the mental 43 44 20 45 health and cognitive effects of war trauma on war survivors in Yugoslavia, Baoğlu et al. found

46 22 47 that 33% of survivors suffered from PTSD. De Jong et al. attempted to study the impact of 48 49 trauma in postconflict lowincome countries where people have survived multiple traumatic 50 51 experiences. They found the prevalence rate of assessed PTSD to be 37.4% in Algeria, 28.4% in 52 53 Cambodia, 15.8% in Ethiopia, and 17.8% in Gaza.22 Priebe et al.23studied mental health in five 54 55 countries of exYugoslavia and found prevalence rates ranging from 15.6% to 41.8% for anxiety 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 disorders. The PTSD prevalence rate among victims of cluster munitions in our study is still 4 5 6 higher than those reported in different samples from different countries, even after 10 years. 7 8 9 As there might be a difference in the psychobiology of PTSD between victims of cluster 10 11 munitions and civilians who were survivors of war, we compared our results to studies on 12 13 civilians who sustained a traumatic injury. In a study by Ohry et al.24 on traumatic brain injury 14 15 (TBI) patients,For 33% met criteriapeer for PTSD reviewdiagnosis. Shalev et al. only25found that 25.5% of injured 16 17 trauma survivors met PTSD diagnostic criteria at the 6month followup. Also, Mora et al.26 18 19 20 reported a PTSD prevalence of 32% in patients with explosionrelated burns. Furthermore, a 21 22 prospective study following traumatic events reported that 29.9% of survivors met criteria for 23 27 24 PTSD at 1 month, and 17.5% had PTSD at 4 months. Prevalence of PTSD among cluster 25 26 munition victims was still much higher. We believe that the injuries resulting from cluster 27 28 munition blasts are unique and different from other blast injuries. The high prevalence of PTSD 29 30 and the functional impairment sustained in our sample may support this claim. 31 32 33 34 Our study showed a significant drop in PTSD prevalence after 10 years of the trauma incident. http://bmjopen.bmj.com/ 35 28 29 36 Other studies have also shown that, with time, PTSD decreases in prevalence and severity. 37 38 Farhood30 reassessed the prevalence and predictors of psychiatric disorders in a general 39 40 population from Southern Lebanon conducted one year after the July war in 2007, and was 41 42 thereafter compared to an assessment conducted one year prewar in 2005 amid political turmoil on September 27, 2021 by guest. Protected copyright. 43 44 on the same population. Findings revealed a drop in PTSD symptoms in the 2007 sample at a rate 45 46 30 47 of 17.9%, from 24.1% in 2005. We believe that the ability to cope and social support may lessen 48 49 the impact of trauma events over time. 50 51 52 The most common longterm symptoms of PTSD were related to low mood and cognition. 53 54 Several studies have shown the cooccurrence of depressive symptoms and PTSD early on after 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 trauma.24 26 3138 Furthermore, research have shown that the occurrence of depression during the 4 5 31 36 39 6 months that follow a traumatic event is an important mediator of chronicity in PTSD. 7 8 9 In our study, the number of males injured by cluster munitions was much higher than that of 10 11 females. The male predominance could be explained by the societal norms of males performing 12 13 farming, grazing and other rural activities where submunitions could be scattered.6 Furthermore, 14 15 the evacuationFor of females frompeer the area during review the war decreased onlythe female casualties. 16 17 18 The female sex was significantly associated with having PTSD after 10 years. This coincides with 19 20 24 26 3138 21 previous research studies that assert that the female sex is a risk factor for PTSD. 22 23 24 Our results show that individuals hospitalized after biomechanical trauma were at heightened risk 25 26 for PTSD. Other studies have shown similar findings.4042 Often, patients with PTSD present with 27 28 both medical and psychiatric symptoms. Several studies have suggested that physical injury in the 29 30 context of a psychologically traumatic event is a risk factor for PTSD.43 44 Mollica et al.45 found 31 32 that psychological trauma associated with injury in a civilian population was associated with 33 http://bmjopen.bmj.com/ 34 46 35 higher rates of PTSD and depression than other types of injury. Also, Hoge et al. found higher 36 37 rates of PTSD and depression in military personnel who may have had a TBI, particularly those 38 39 who reported loss of consciousness. Moreover, Schneiderman et al.47conferred that a probable 40 41 TBI almost doubled the risk for PTSD. It is worth mentioning that 59% of the participants in our 42 on September 27, 2021 by guest. Protected copyright. 43 study sustained injuries with severe functional impairment. Therefore, the type of injuries, the 44 45 severity of functional impairment, and the personal and psychosocial stressors that come with the 46 47 48 situation may play a role in strengthening the relationship between injury and PTSD. 49 50 51 Similar to other studies in the literature, religion and social and familial support were negatively 52 53 correlated to PTSD.4852 It has been shown that a higher index of religious moral beliefs enables 54 55 better control of distress, and provides better mental health stability.48 Hasanović & Pajević48 56 57 assert that religion enables posttraumatic conflicts typical for combatants’ survivors to be more 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 easily overcome. CopelandLinder49 found that prayer buffered the effects of stress on physical 4 5 6 health and reduced the deleterious effects of experiencing depressive symptomatology. In another 7 8 study, social and family support, and religious beliefs were all found to be protective against 9 10 PTSD following war trauma and torture; higher perceived social support was associated with 11 12 higher positive emotions.50 We affirm that family support and religion help in overcoming post 13 14 trauma personal and psychosocial tribulations and lead to an improvement in mental health. 15 For peer review only 16 17 Our results showed that personal and socioeconomic impairment was present in the participants 18 19 20 diagnosed with PTSD after 10 years. Studies have shown that effects of PTSD on the survivor's 21 22 immediate family and society at large are substantial. In addition to the emotional and cognitive 23 24 symptoms of PTSD, individuals with PTSD are more likely to experience marital and family 25 26 problems,53 54 job instability,55 legal difficulties,53 56 and physical health problems.57 58 27 28 29 In the long run, job instability was the most common socioeconomic repercussion (88%). Other 30 31 studies have shown that unemployment and low functionality to be associated with lower mental 32 33 19 34 health conditions, namely PTSD. We believe that the economic costs of PTSD are attributable http://bmjopen.bmj.com/ 35 36 to lost work productivity. The severity of the injury and the associated functional impairment also 37 38 play a role. 39 40 41 Health wise, Hoge et al.59 add that individuals with a history of PTSD have a higher risk of 42 on September 27, 2021 by guest. Protected copyright. 43 cardiovascular, respiratory, gastrointestinal, infectious, nervous system, and autoimmune 44 45 disorders, and are more likely to experience anxiety, depression, substance use, and other 46 47 31 48 53 56 60 61 48 psychiatric disorders. In addition, studies have reported a higher risk of suicidal 49 62 63 50 ideation among veterans with PTSD. 51 52 53 In our study, we aimed to identify the prevalence of PTSD and associated factors following 54 55 cluster munitionrelated injuries by adopting a longitudinal design and paying close attention to 56 57 optimizing the research methodology. The longitudinal nature of the study helped in better 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 understanding the course and prognosis of PTSD in these individuals. This form of design can 4 5 6 also elucidate the most effective screening and treatment regimens and factors that influence 7 8 recovery. The study can be considered representative as we were capable of reaching all of the 9 10 individuals who met our inclusion criteria. In addition, all individuals were reachable for follow 11 12 up. Due to the lack of consistent definitions of PTSD, we adopted the DSM5 criteria for PTSD 13 14 and used the PCL as a tool to aid in the diagnosis. We used a higher cutoff score to minimize 15 For peer review only 16 false positives. We measured functional impairment using the Fares Scale of injuries due to 17 18 19 cluster munitions. To our knowledge, functional impairment measures have not been incorporated 20 21 into previous case definitions, which makes it difficult to determine the true effects of cluster 22 23 munitions and project on the mental health service needs. Furthermore, we studied the nature and 24 25 severity of the impairment along multiple dimensions to include work, family and social 26 27 relationships. However, the instrument of measure used differs from one study to another and can 28 29 lead to varying degrees of PTSD reporting. As many of the followup interviews were held over 30 31 32 the telephone, we were limited by time. More demographical data on health, treatment, and 33 34 socioeconomic outcomes could have been gathered. In addition, information on comorbid mental http://bmjopen.bmj.com/ 35 36 health conditions like depression, anxiety, sleep disorders, eating disorders, substance use 37 38 disorders, and suicide could have been collected. 39 40 41 More research is needed to study the interaction between trauma exposure, preexisting 42 on September 27, 2021 by guest. Protected copyright. 43 psychological and biological vulnerabilities, and the posttrauma environment. Research is also 44 45 46 needed to better quantify the effect of comorbid factors such as aggression, risk taking behaviors, 47 48 and physical symptoms to understand the full extent of the effects of warrelated trauma. As the 49 50 prevalence of PTSD and other mental health conditions is high in the Lebanese population, 51 52 advanced neurobiological and neuropsychological studies are needed to better understand the 53 54 pathophysiology of PTSD and how genetic and environmental factors tally up to induce it. 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 IMPLICATIONS & CONCLUSIONS 4 5 6 7 This is the first longitudinal study to analyze PTSD in victims of cluster munition explosions. 8 9 Shortly posttrauma, 98% of those injured by cluster munitions were diagnosed with PTSD. In a 10 11 10year followup, 43% still had the disorder. Overall, the prevalence of PTSD decreased 12 13 significantly after 10 years. The male sex, family support and religion are possible protective 14 15 factors againstFor PTSD. Whereas, peer hospitalization review posttrauma and severeonly functional impairment 16 17 were significantly correlated with increased prevalence of longterm PTSD. Symptoms of 18 19 20 negative cognition and mood were more common in the long run. In contrast, symptoms of 21 22 arousal and reexperiencing decreased in prevalence. PTSD is associated with numerous 23 24 deleterious outcomes, and its costs on the victims, their immediate family, and society at large are 25 26 substantial. 27 28 29 Intervention and screening programs should target vulnerable populations in Lebanon. Public 30 31 health advocates and governmental officials should run campaigns to raise awareness against 32 33 34 cluster munitions, and educate people on their shapes, characteristics and dangers. Moreover, http://bmjopen.bmj.com/ 35 36 municipal efforts and regional projects that aim to demine unexploded submunitions should be 37 38 funded and supported by governmental and nongovernmental organizations. 39 40 41 Laws should be passed and enforced to ban the use of these detrimental weapons that have 42 on September 27, 2021 by guest. Protected copyright. 43 negative effects on human, societal and ecological levels. More countries should join the 44 45 Convention on Cluster Munitions, as these inhumane weapons indiscriminately and 46 47 48 disproportionately harm innocent civilians, thereby violating the wellestablished international 49 50 principles governing conflict and war today. 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 FOOTNOTES 4 5 6 7 Contributors: JF and YF designed the study and collected the data. JF carried out the statistical 8 9 analyses and drafted the manuscript. All authors contributed to the analysis of the results. All 10 11 authors critically revised the manuscript for important intellectual content. JF and YF are the 12 13 primary investigators and guarantors of the study. 14 15 For peer review only 16 Funding: This research was funded by a grant from the Central Administration of the Lebanese 17 18 University, Beirut, Lebanon. 19 20 21 22 Competing interests: None declared. 23 24 25 Ethics Approval: The study was approved by the Institutional Review Board at the Faculty of 26 27 Medical Sciences, Lebanese University, Beirut, Lebanon. 28 29 30 Data Sharing Statement: Supplementary material are provided. 31 32 33

http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 REFERENCES 4 5 6 1. Hoteit A, Fares J. Psycho-environmental tribulations arising from cluster munitions in South 7 8 Lebanon. Sci-Afric Journal of Scientific Issues, Research and Essays 2014;2(10):469-73. 9 2. Fares Y, Fares J. Anatomical and neuropsychological effects of cluster munitions. Neurological 10 Sciences 2013;34(12):2095-100. 11 3. Abou-Abbass H, Bahmad H, Ghandour H, et al. Epidemiology and clinical characteristics of 12 traumatic brain injury in Lebanon: A systematic review. Medicine 2016;95(47) 13 4. Bedrossian N, Haidar M, Fares J, et al. Inflammation and elevation of interleukin-12p40 in 14 patients with schizophrenia. Frontiers in molecular neuroscience 2016;9 15 For peer review only 16 5. Fares Y, Ayoub F, Fares J, et al. Pain and neurological sequelae of cluster munitions on 17 children and adolescents in South Lebanon. Neurological Sciences 2013;34(11):1971-76. 18 6. Fares Y, El-Zaatari M, Fares J, et al. Trauma-related infections due to cluster munitions. 19 Journal of infection and public health 2013;6(6):482-86. 20 7. Fares Y, Fares J. Neurosurgery in Lebanon: history, development, and future challenges. World 21 Neurosurgery 2016 22 8. Fares Y, Fares J, Gebeily S. Head and facial injuries due to cluster munitions. Neurological 23 24 Sciences 2014;35(6):905-10. 25 9. Hoteit A. War against architecture, identity and collective memory. International Journal of 26 Development Research 2015;5(2) 27 10. Hoteit A. Role of the landscape in the preservation of collective memory and the 28 enhancement of national belonging. Canadian Social Science 2015;11(3):42. 29 11. Hoteit A. Architectural Education in the Arab World and Its Role in Facing the Contemporary 30 31 Local and Regional Challenges. Canadian Social Science 2016;12(7):1-7. 32 12. Beehner L. The Campaign to Ban Cluster Bombs. Backgrounder, 2006. 33 13. Kessler RC, Berglund P, Demler O, et al. Lifetime prevalence and age-of-onset distributions of 34 DSM-IV disorders in the National Comorbidity Survey Replication. Archives of general http://bmjopen.bmj.com/ 35 psychiatry 2005;62(6):593-602. 36 14. Yeager DE, Magruder KM, Knapp RG, et al. Performance characteristics of the posttraumatic 37 stress disorder checklist and SPAN in Veterans Affairs primary care settings. General 38 39 hospital psychiatry 2007;29(4):294-301. 40 15. Bliese PD, Wright KM, Adler AB, et al. Validating the primary care posttraumatic stress 41 disorder screen and the posttraumatic stress disorder checklist with soldiers returning 42 from combat. Journal of consulting and clinical psychology 2008;76(2):272. on September 27, 2021 by guest. Protected copyright. 43 16. Freedy JR, Steenkamp MM, Magruder KM, et al. Post-traumatic stress disorder screening 44 test performance in civilian primary care. Family Practice 2010;27(6):615-24. 45 17. PTSD VNCf. Using the PTSD Checklist for DSM-IV (PCL). 2014. 46 47 https://sph.umd.edu/sites/default/files/files/PTSDChecklistScoring.pdf. 48 18. Farhood L, Dimassi H, Lehtinen T. Exposure to war-related traumatic events, prevalence of 49 PTSD, and general psychiatric morbidity in a civilian population from Southern Lebanon. 50 Journal of transcultural nursing 2006;17(4):333-40. 51 19. Farhood LF, Fares S, Sabbagh R, et al. PTSD and depression construct: prevalence and 52 predictors of co-occurrence in a South Lebanese civilian sample. European Journal of 53 54 Psychotraumatology 2016;7 55 20. Başoğlu M, Livanou M, Crnobarić C, et al. Psychiatric and cognitive effects of war in former 56 Yugoslavia: association of lack of redress for trauma and posttraumatic stress reactions. 57 Jama 2005;294(5):580-90. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 21. Hall BJ, Hobfoll SE, Palmieri PA, et al. The psychological impact of impending forced settler 4 disengagement in Gaza: Trauma and posttraumatic growth. Journal of traumatic stress 5 6 2008;21(1):22-29. 7 22. De Jong JT, Komproe IH, Van Ommeren M, et al. Lifetime events and posttraumatic stress 8 disorder in 4 postconflict settings. Jama 2001;286(5):555-62. 9 23. Priebe S, Bogic M, Ajdukovic D, et al. Mental disorders following war in the Balkans: a study 10 in 5 countries. Archives of general psychiatry 2010;67(5):518-28. 11 24. Ohry A, Rattok J, Solomon Z. Post-traumatic stress disorder in brain injury patients. Brain 12 13 Injury 1996;10(9):687-96. 14 25. Shalev AY, Peri T, Canetti L, et al. Predictors of PTSD in injured trauma survivors: a 15 prospectiveFor study. peer American Journal review of Psychiatry 1996;153(2):219-25. only 16 26. Mora AG, Ritenour AE, Wade CE, et al. Posttraumatic stress disorder in combat casualties 17 with burns sustaining primary blast and concussive injuries. Journal of Trauma and 18 Acute Care Surgery 2009;66(4):S178-S85. 19 27. Shalev AY, Freedman S, Peri T, et al. Prospective study of posttraumatic stress disorder and 20 21 depression following trauma. American Journal of Psychiatry 1998 22 28. Greist J. Posttraumatic Stress Disorder (PTSD). Merck Manual 2014. 23 https://www.merckmanuals.com/professional/psychiatric-disorders/anxiety-and- 24 stressor-related-disorders/posttraumatic-stress-disorder-ptsd?qt=PTSD&alt=sh. 25 29. Farhood L. Patterns of Psychiatric Morbidity Before and After a War in Lebanon at Twelve 26 Months Following Cessation of Hostilities. The Open Psychiatry Journal 2014;8(1) 27 28 30. Farhood L. Patterns of Psychiatric Morbidity Before and After a War in Lebanon at Twelve 29 Months Following Cessation of Hostilities. The Open Psychiatry Journal 2014;8(1) 30 31. Holbrook TL, Hoyt DB, Coimbra R, et al. Long-term posttraumatic stress disorder persists 31 after major trauma in adolescents: new data on risk factors and functional outcome. 32 Journal of Trauma and Acute Care Surgery 2005;58(4):764-71. 33

32. Holbrook TL, Hoyt DB, Stein MB, et al. Gender differences in long-term posttraumatic stress http://bmjopen.bmj.com/ 34 disorder outcomes after major trauma: women are at higher risk of adverse outcomes 35 36 than men. Journal of Trauma and Acute Care Surgery 2002;53(5):882-88. 37 33. Verger P, Dab W, Lamping DL, et al. The psychological impact of terrorism: an epidemiologic 38 study of posttraumatic stress disorder and associated factors in victims of the 1995- 39 1996 bombings in France. American Journal of Psychiatry 2004;161(8):1384-89. 40 34. Brewin CR, Andrews B, Valentine JD. Meta-analysis of risk factors for posttraumatic stress 41 disorder in trauma-exposed adults. Journal of consulting and clinical psychology 42 on September 27, 2021 by guest. Protected copyright. 2000;68(5):748. 43 44 35. Predictors of posttraumatic stress disorder and symptoms in adults: a meta-analysis. Annual 45 Meeting of the International Society for Traumatic Stress Studies, 14th, Nov, 1998, 46 Washington, DC, US; This article is based on a paper presented at the aforementioned 47 meeting; 2008. Educational Publishing Foundation. 48 36. Freedman SA, Brandes D, Peri T, et al. Predictors of chronic post-traumatic stress disorder. A 49 prospective study. The British journal of psychiatry 1999;174(4):353-59. 50 51 37. Javidi H, Yadollahie M. Post-traumatic stress disorder. The international journal of 52 occupational and environmental medicine 2011;3(1 January) 53 38. Farhood LF, Dimassi H. Prevalence and predictors for post-traumatic stress disorder, 54 depression and general health in a population from six villages in South Lebanon. Social 55 psychiatry and psychiatric epidemiology 2012;47(4):639-49. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 39. DiGrande L, Neria Y, Brackbill RM, et al. Long-term posttraumatic stress symptoms among 4 3,271 civilian survivors of the September 11, 2001, terrorist attacks on the World Trade 5 6 Center. American journal of epidemiology 2011;173(3):271-81. 7 40. Abenhaim L, Dab W, Salmi LR. Study of civilian victims of terrorist attacks (France 1982– 8 1987). Journal of Clinical Epidemiology 1992;45(2):103-09. 9 41. O'donnell ML, Creamer M, Pattison P, et al. Psychiatric morbidity following injury. American 10 Journal of Psychiatry 2004;161(3):507-14. 11 42. Shih RA, Schell TL, Hambarsoomian K, et al. Prevalence of PTSD and major depression 12 13 following trauma-center hospitalization. The Journal of trauma 2010;69(6):1560. 14 43. Vasterling JJ, Verfaellie M, Sullivan KD. Mild traumatic brain injury and posttraumatic stress 15 disorderFor in returning peer veterans: perspectives review from cognitive only neuroscience. Clinical 16 psychology review 2009;29(8):674-84. 17 44. Koren D, Norman D, Cohen A, et al. Increased PTSD risk with combat-related injury: a 18 matched comparison study of injured and uninjured soldiers experiencing the same 19 combat events. American Journal of Psychiatry 2005;162(2):276-28. 20 21 45. Mollica RF, Henderson DC, Tor S. Psychiatric effects of traumatic brain injury events in 22 Cambodian survivors of mass violence. The British Journal of Psychiatry 23 2002;181(4):339-47. 24 46. Hoge CW, McGurk D, Thomas JL, et al. Mild traumatic brain injury in US soldiers returning 25 from Iraq. New England journal of medicine 2008;358(5):453-63. 26 47. Schneiderman AI, Braver ER, Kang HK. Understanding sequelae of injury mechanisms and 27 28 mild traumatic brain injury incurred during the conflicts in Iraq and Afghanistan: 29 persistent postconcussive symptoms and posttraumatic stress disorder. American 30 journal of epidemiology 2008;167(12):1446-52. 31 48. Hasanović M, Pajević I. Religious moral beliefs as mental health protective factor of war 32 veterans suffering from PTSD, depressiveness, anxiety, tobacco and alcohol abuse in 33

comorbidity. Psychiatria Danubina 2010;22(2):203-10. http://bmjopen.bmj.com/ 34 49. Copeland-Linder N. Stress among black women in a South African township: The protective 35 36 role of religion. Journal of Community Psychology 2006;34(5):577-99. 37 50. Feder A, Ahmad S, Lee EJ, et al. Coping and PTSD symptoms in Pakistani earthquake 38 survivors: Purpose in life, religious coping and social support. Journal of Affective 39 Disorders 2013;147(1):156-63. 40 51. Khamis V. Impact of war, religiosity and ideology on PTSD and psychiatric disorders in 41 adolescents from Gaza Strip and South Lebanon. Social Science & Medicine 42 on September 27, 2021 by guest. Protected copyright. 2012;74(12):2005-11. 43 44 52. Johnson H, Thompson A. The development and maintenance of post-traumatic stress 45 disorder (PTSD) in civilian adult survivors of war trauma and torture: A review. Clinical 46 psychology review 2008;28(1):36-47. 47 53. Gates MA, Holowka DW, Vasterling JJ, et al. Posttraumatic stress disorder in veterans and 48 military personnel: Epidemiology, screening, and case recognition. Psychological services 49 2012;9(4):361. 50 51 54. Jordan BK, Marmar CR, Fairbank JA, et al. Problems in families of male Vietnam veterans 52 with posttraumatic stress disorder. Journal of consulting and clinical psychology 53 1992;60(6):916. 54 55. Smith MW, Schnurr PP, Rosenheck RA. Employment outcomes and PTSD symptom severity. 55 Mental health services research 2005;7(2):89-101. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 56. Kulka RA, Schlenger WE, Fairbank JA, et al. Trauma and the Vietnam war generation: Report 4 of findings from the National Vietnam Veterans Readjustment Study: Brunner/Mazel 5 6 1990. 7 57. Boscarino JA. Posttraumatic stress disorder and physical illness: results from clinical and 8 epidemiologic studies. Annals of the New York Academy of Sciences 2004;1032(1):141- 9 53. 10 58. O'Toole BI, Catts SV, Outram S, et al. The physical and mental health of Australian Vietnam 11 veterans 3 decades after the war and its relation to military service, combat, and post- 12 13 traumatic stress disorder. American Journal of Epidemiology 2009;170(3):318-30. 14 59. Hoge CW, Terhakopian A, Castro CA, et al. Association of posttraumatic stress disorder with 15 somaticFor symptoms, peer health care visits, review and absenteeism amongonly Iraq war veterans. 16 American Journal of Psychiatry 2007 17 60. Long N, MacDonald C, Chamberlain K. Prevalence of posttraumatic stress disorder, 18 depression and anxiety in a community sample of New Zealand Vietnam War veterans. 19 Australian and New Zealand Journal of Psychiatry 1996;30(2):253-56. 20 21 61. Karam EG, Fayyad J, Karam AN, et al. Outcome of depression and anxiety after war: a 22 prospective epidemiologic study of children and adolescents. Journal of traumatic stress 23 2014;27(2):192-99. 24 62. Jakupcak M, Cook J, Imel Z, et al. Posttraumatic stress disorder as a risk factor for suicidal 25 ideation in Iraq and Afghanistan war veterans. Journal of traumatic stress 26 2009;22(4):303-06. 27 28 63. Pietrzak RH, Johnson DC, Goldstein MB, et al. Psychosocial buffers of traumatic stress, 29 depressive symptoms, and psychosocial difficulties in veterans of Operations Enduring 30 Freedom and Iraqi Freedom: the role of resilience, unit support, and postdeployment 31 social support. Journal of affective disorders 2010;120(1):188-92. 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Table 1. Demographics of the participants injured by cluster munitions and diagnosed with 4 5 PTSD in 2006 (N=239). SD: Standard Deviation. *Severe functional impairment was measured 6 using the Fares Scale for injuries due to cluster munition explosions (Grades IIIV were 7 considered severe). 8 9 10 Demographic Trait Number (%) 11 12 Male 224 (94) 13 14 Female 15 (6) 15 For peer review only 16 Mean Age (years) ± SD 27 ± 4.3 17 18 Age Range (years) [18, 67] 19 20 Mean time since injury (months) ± SD 2.2 ± 1.2 21 22 23 Hospitalized 230 (96) 24 25 Discharged from emergency room 9 (4) 26 27 Severe functional impairment* 141 (59) 28 29 Family Support 228 (95) 30 31 Religious 220 (92) 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Table 2. Prevalence of PTSD among the 244 adult victims of cluster munitions in 2006 and 4 5 2016 across the two sexes. *P value < 0.05 indicates significance. 6 Diagnosed with PTSD Diagnosed with PTSD 7 P value* 8 in 2006 in 2016 9 Male 224 91 <0.001 10 11 Female 15 14 0.334 12 13 Total 239 105 <0.001 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Table 3. Association of long-term PTSD (after 10 years) with variables such as family 4 5 support, religion, hospitalization, and severity of injury in victims of cluster munitions 6 (N=239). SE: Standard Error of Mean. *P value < 0.05 indicates significance. †Functional 7 impairment was measured using the Fares Scale for injuries due to cluster munition explosions 8 (Grades IIIV were considered severe). 9 10 11 12 N PTSD rate SE P value* 13 14 Family Support 15 Yes 228 0.41 0.03 For peer review only <0.001 16 No 11 1.00 0 17 Religious 18 Yes 220 0.39 0.03 19 <0.001 20 No 19 1.00 0 21 Hospitalized Yes 230 0.46 0.03 22 0.005 23 No 9 0 0 24 Functional

25 Impairment† 26 Yes 141 0.74 0.04 <0.001 27 No 98 0 0 28 29 30 31 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Table 4. Personal and socioeconomic repercussions in the 105 victims of cluster munitions 4 5 diagnosed with PTSD, 10 years post injury. 6 7 Social Repercussions Number (%) 8 9 Marital and Family Problems 21 (20) 10 11 Job Instability 92 (88) 12 13 Legal Difficulties 9 (9) 14 15 For peer review only 16 17 18

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Figure Legends 4 5

6 7 8 Figure 1. A cluster munition containing numerous undetonated submunitions in South Lebanon. 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 33 34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 1. A cluster munition containing numerous undetonated sub-munitions in South Lebanon. 33 34 43x32mm (300 x 300 DPI) http://bmjopen.bmj.com/ 35 36 37 38 39 40 41 42 on September 27, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Appendix A 4 5 6 Fares Scale of injuries due to cluster munitions explosions4 7 8 9 Grade Level Percentage of Functional Disability Anatomical Lesion 10 11 12 13 - Amputation of three fingers: middle finger, 14 ring finger & little finger 15 For peer≤ %25 review only 16 Grade I - Partial loss of a hand or foot 17 - Superficial wounds to the body and face 18 - Loss of skin and/or muscles in the 19 extremities 20 21 22 23 - Amputation of two fingers: one of the two is 24 25 the thumb 26 - Loss of an eye, leg and/or hand 27 Grade II %50 - Wound infection 28 - Recuperable lesion in abdomen and/or 29 30 thorax 31 - Psychological effects 32 33 34 http://bmjopen.bmj.com/ 35 - Affection of more than two of the four 36 37 extremities 38 - Partial loss of vision and/or impairment of 39 Grade III %75 the auditory system 40 - Mutilation and partial loss of the digestive 41 42 system and /or lungs on September 27, 2021 by guest. Protected copyright. 43 - Lesion of the spinal cord 44 45 46 47 - Amputation of three extremities 48 Grade IV >%75 - Total loss of vision and/or auditory capacity 49 - Tetraplegia 50 51 - Intellectual incapacity 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 31 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 3 Appendix B 4 5 6 Distribution of answers of the 244 Lebanese adult injuries of cluster munitions to the 17-item-PTSD checklist in late 2006 7 8 and early 2007. 9 10 Not at all A little bit Moderately Quite a bit Extremely No. Response 11 (1) (2) (3) (4) (5) 12 Repeated, disturbing memories, thoughts, or 13 01 5 15 36 62 126 14 images of a stressful experience from the past? 15 Repeated, disturbingFor dreams of apeer stressful review only 02 6 14 41 59 124 16 experience from the past? 17 18 Suddenly acting or feeling as if a stressful 19 03 experience were happening again (as if you 4 13 38 64 125 20 were reliving it)? 21 Feeling very upset when something reminded 22 04 8 13 32 79 112 23 you of a stressful experience from the past? 24 Having physical reactions (e.g., heart 25 pounding, trouble breathing, or sweating) 26 05 8 10 31 59 136 27 when somethingreminded you of a stressful 28 experience from the past? 29 Avoid thinking about or talking about a 30 31 06 stressful experience from the past or avoid 8 13 44 57 122 32 having feelings related to it? 33

Avoid activities or situations because they http://bmjopen.bmj.com/ 34 07 remind you of a stressful experience from the 10 13 45 55 121 35 past? 36 37 Trouble remembering important parts of a 08 12 13 46 61 112 38 stressful experience from the past? 39 Loss of interest in things that you used to 40 09 9 12 47 57 119 41 enjoy? 42 10 Feeling distant or cut off from other people? 11 28 51 47 107 on September 27, 2021 by guest. Protected copyright. 43 Feeling emotionally numb or being unable to 44 11 9 25 53 47 110 45 have loving feelings for those close to you? 46 Feeling as if your future will somehow be cut 12 9 22 50 49 114 47 short? 48 Trouble falling or staying asleep? 12 24 48 55 105 49 13 50 14 Feeling irritable or having angry outbursts? 3 22 42 42 135 51 15 Having difficulty concentrating? 5 17 43 47 132 52 16 Being “super alert” or watchful on guard? 13 28 45 56 102 53 54 17 Feeling jumpy or easily startled? 14 21 48 60 101 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 31 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from

1 2 Appendix C 3 4 Distribution of answers of the 244 Lebanese adult injuries of cluster munitions to the 17-item-PTSD checklist in 2016. 5 6 7 8 Not at all A little bit Moderately Quite a bit Extremely No. Response 9 (1) (2) (3) (4) (5) 10 Repeated, disturbing memories, thoughts, or 11 01 95 59 50 26 14 12 images of a stressful experience from the past? 13 Repeated, disturbing dreams of a stressful 02 80 70 46 36 12 14 experience from the past? 15 For peer review only 16 Suddenly acting or feeling as if a stressful 17 03 experience were happening again (as if you 110 85 24 20 5 18 were reliving it)? 19 20 Feeling very upset when something reminded 04 45 41 56 60 42 21 you of a stressful experience from the past? 22 23 Having physical reactions (e.g., heart pounding, 24 trouble breathing, or sweating) when 05 111 67 30 22 14 25 somethingreminded you of a stressful 26 experience from the past? 27 28 Avoid thinking about or talking about a 29 06 stressful experience from the past or avoid 113 73 30 18 10 30 having feelings related to it? 31 32 Avoid activities or situations because they 33

07 remind you of a stressful experience from the 101 40 32 24 47 http://bmjopen.bmj.com/ 34 past? 35 36 Trouble remembering important parts of a 08 114 48 27 37 18 37 stressful experience from the past? 38 39 Loss of interest in things that you used to 09 54 36 47 61 46 40 enjoy? 41

10 Feeling distant or cut off from other people? 50 40 41 59 54 on September 27, 2021 by guest. Protected copyright. 42 43 Feeling emotionally numb or being unable to 11 73 69 36 38 28 44 have loving feelings for those close to you? 45 Feeling as if your future will somehow be cut 46 12 69 67 43 42 23 47 short? 48 13 Trouble falling or staying asleep? 133 69 12 18 12 49 14 Feeling irritable or having angry outbursts? 127 68 15 20 14 50 51 15 Having difficulty concentrating? 121 61 20 24 18 52 16 Being “super alert” or watchful on guard? 80 52 39 46 27 53 54 17 Feeling jumpy or easily startled? 84 54 35 38 33 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from Page 30 of 31 7 7 5 5 7 7 5,6 5 5

4 4 4 4

7 7 7 7 NA 7 7 7 7 7 7 7 7 NA 1,2 8 8 Reported Reported onpage # 2 2

cohort studies cohort

BMJ Open http://bmjopen.bmj.com/ on September 27, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For matchedFor studies, give matching criteria numberand of exposedand unexposed

) If applicable, how explain loss to was follow-up addressed ) Describe methodsany used tosubgroups examine and interactions ) ) in the Provide abstract informativean balanced summaryand of what was donewhat was and found ) statistical Describe all methods, including thosecontrol confoundingusedfor to ) eligibilitythe Give criteria, theand sourcesand methods of selectionof participants. Describe methods follow-up of 5,6 ) Indicate study’sthe design commonlya with used term in ortitle the abstractthe ) Describe) sensitivity any analyses For peer review only ) Explain missinghow data were addressed e d c b b b a a a For each variable of interest,For variable sourceseach give of data and details of methodsassessment of (measurement). Describe comparability comparability of assessment methods if isthanthere more groupone

applicable collection ( ( ( ( ( ( Recommendation why STROBE 2007 (v4) Statement—Checklist of items that should be included in reports of reports in included be should that items of Statement—Checklist (v4) 2007 STROBE 3 3 objectives, State specific including prespecified any hypotheses 1 1 ( 5 5 Describe setting,the locations, and relevantdates, including periods exposure,of recruitment, follow-up, and data 9 9 Describe any addressefforts to sourcespotential of bias 6 6 ( 7 7 Clearly define outcomes, all exposures, predictors, potential and confounders, effect modifiers.diagnostic Give criteria, if 4 4 Present key elements of study in paperthedesign early 8* 10 10 Explain thehow study wassize arrived at # Item Item

Title andTitle abstract Results Statistical Statistical methods 12 ( Quantitative variables 11 Explain quantitativehow variableshandled were inthe analyses. If describeapplicable, which groupingschosen were and Study size Bias Data sources/Data measurement Variables Participants Setting Study design Methods Objectives Background/rationale 2 Explain scientific the background investigation thefor and rationale being reported Introduction

Section/Topic

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 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 BMJ Open: first published as 10.1136/bmjopen-2017-017214 on 18 August 2017. Downloaded from 16 16 15,16 11-14 11-14 9,10,11 9-15 9-15 8 8 NA 8 8 8 - Table18 8 8 15 15 9 9 NA 8,9 16 16 8,9 NA NA

BMJ Open http://bmjopen.bmj.com/ on September 27, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml ) Report category boundariescontinuous when variables categorized were ) unadjustedGive estimates ifand, applicable, confounder-adjusted estimates and their precision (eg, confidence95% For peerrelevant, If ) consider translating estimates of relative risk into absolute for risk meaningful periodtime a review only c b a similar studies, other relevantand evidence confounders eligible, includedeligible, in study,the completing follow-up, analysedand interval). Makeinterval). clear confounderswhich andwere foradjusted they why were included which thewhich article presentis based

( ( Summarise (c) follow-up time (eg, average total and amount) (b)number Indicate of participants missingwith data eachof for interest variable Consider (c) diagram flow use of a (b)reasons Give non-participationfor each at stage 18 18 Summarise resultskey to reference with study objectives 20 20 Give cautiousa interpretationoverall of results considering objectives, limitations, of multiplicity resultsanalyses, from 16 16 ( 22 22 the Give of fundingsource and role the of fundersthe presentthe for if study theand, forapplicable, original study on 13* (a) numbersReport of individuals each at of stage study—eg numbers potentially eligible, eligibility,examined for confirmed

An Explanation and Elaboration article discusses itemeach checklist and gives methodological backgroundand published examples of transparent reporting. The STROBE checklist is best used in thiswith conjunction (freely available article onsites the Medicine of PLoS Web http://www.plosmedicine.org/, at Annals of Internal Medicine at http://www.annals.org/, Epidemiologyhttp://www.epidem.com/). at and Information on InitiativeSTROBE the is www.strobe-statement.org.available at information*Give separately cases for and controls studiesin case-controlfor applicable, if and, exposedunexposed and groups in cohort and cross-sectional studies. Note: Funding Other information Generalisability 21 Discuss (external generalisability the of validity)the study results Interpretation Limitations Key Key results Discussion Other analyses 17 Report other analyses done—eg analyses of subgroupsinteractions, sensitivity and and analyses

Main resultsMain Outcome Outcome data 15* Report numbers of outcome events or summary timemeasures over

Descriptive data 14* characteristics (a) Give participantsof study (eg clinical,demographic, social) informationand exposureson and potential

Participants Page 31 of 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 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