AuthorDirectorate-General for Migration and Home Affairs (European Commission)
1 Introduction
The Appendic es provide detailed i nformation used in th e research to prod uce the main repor t of the Study. The
Appendices mat ch the primar y sections tha t they suppor t.
2 Methodology
2.1 Victims of Trafficking: number and distribution
Tables A2.1.1 and A2.1. 2 offer a more detailed summa ry of the numb er and distributio n of victims derived fro m
the European Co mmission Data Repor t provided in Table 2.5.1, incl uding missing data.
It is the predicte d distribution by age and sex wit hin exploitation group that is us ed throughout for the purp oses of costing.
Table A2.1.1 Number and distribution of victims, with and without missing data, EU-28
EU-2 8
Actual distribution Predicted distribution a
N % N %
Tota l 11 832 3 700 524 433 11 832 2 708 804 8 38
Sexual exploitation 5 374 45 6 626 56
Female child 1 034 19 1 295 20
Male child 132 2165 2
Female adult 3 931 73 4 921 74
Male adult 196 4245 4
Unknown/other 81 2 0 0
Labour exploitatio n 2 459 21 2 958 25
Female child 75 396 3
Male child 475 19 611 21
Female adult 341 14 439 15
Male adult 1 40 9 57 1 812 61
Unknown/other 159 6 0 0
Other exploitation 1 789 15 2 248 19
Female child 267 15 372 17
Male child 143 8199 9
Female adult 760 42 1 059 47
Male adult 444 25 618 28
Unknown/other 175 10 0 0
Unknown exploitatio n 823 700
Missing b1 387 12 0 0
a Unknown/oth er cases recoded to the known prof‌ile.
b Includes 447 vi ctims registered in Bulgari a who were not in the Data Report .
Table A2.1.2 Number and distribution of victims, with and without missing data, EU-27
Actual distribution Predicted distribution a
N % N %
Tota l 8 027 3 70 0 524 433 8 027 2 708 80 4 838
Sexual exploitation 4 06 6 51 5 137 64
Female child 713 18 919 18
Male child 91 2117 2
Female adult 3 043 75 3 923 76
Male adult 138 3178 3
Unknown/other 81 2 0 0
Labour exploitatio n 884 11 1 124 14
Female child 7 1 11 1
Male child 75 8116 2
Female adult 159 18 246 22
Male adult 484 55 750 67
Unknown/other 159 18 0 0
Other exploitation 1 359 17 1 766 22
Female child 200 15 298 17
Male child 107 8160 9
Female adult 500 37 746 42
Male adult 377 28 562 32
Unknown/other 175 13 0 0
Unknown exploitatio n 336 400
Missing b1 382 17 0 0
a Unknown/oth er cases recoded to the known prof‌ile.
b Includes 447 vi ctims registered in Bulgari a who were not in the Data Report .
2.2 Systematic Review of Scientific Literature
We conducted a sys tematic review of the scientific literat ure of the harms of trafficking in hu man beings, drawing
on a previous review in whic h Oram was involved (Ottisova et al. 2016). The revie w established: the prevalence of
physical violence, se xual violence and t hreats among those who are vict ims of trafficking . It offered disaggrega-
tions, where possible , by type of exploitation (sexual , labour, other). The systematic review pr otocol was registered
with PROSPE RO (the Prospec tive Register of Systemat ic Reviews). The registr ation number is CRD 42019141929
and the protocol c an be accessed via th e following link: http s:// /PROSPERO/display_record.
Inclusion criteria re quired that studies (i) reporte d empirical research based on on e of the following study designs:
cross-sect ional survey; ca se control stu dy; cohort st udy; case ser ies analysis; experim ental study w ith baseline
measures for the out comes of interest ; or secondar y analysis of organisatio nal records; (ii) were publish ed as
a scientific journ al article, doctoral th esis or dissertation, or research r eport; (iii) reported on phy sical violence while
trafficked; se xual violence while tr afficked; sex wit hout consent while trafficked; thr eat while trafficked; physical
health harms; ment al health harms ; or reproduct ive health harms; (iv) inclu ded adult or child males or female s
who self-identified or were believe d by the research t eam to have been trafficked according to t he definition laid
down by the United N ations Palermo Prot ocol. No langu age restrict ions were used. A lower date li mit of the 17th
April 2015 was used, as t he review updated a p revious review (Ott isova et al. 2016) which used an up per date
limit of 17th April 2015.
In order to identify s tudies, we searched electron ic bibliographic databases Psyc INFO (Online database of psycho -
logical literature); M EDLINE (Online dat abase of health a nd medical journal s and other news so urces); EMBAS E
(Online database of h ealth and medical journals); CIN AHL (Online database of nursing a nd allied health literature);
The Cochrane C entral Register of Controlled Trials (CENT RAL); and PILOTS (Onl ine database, Published Int ernational
Literature On Traumatic Stre ss157). Returns were de- duplicated to r emove any duplicate r eferences, resulting in
a total of 1 713 references for scre ening.
Screening followed a t wo-stage process: in the f irst stage title and abstra cts were screened against the i nclusion
criteria for likely relevance, in t he second sta ge the full-text of p otentially relevan t studies were read in full a nd
again considered agains t the inclusion crit eria. If it was not clear fro m the title and a bstract whe ther the paper
met the inclusion cri teria for the review, it was inclu ded for full text scre ening. One reviewer (M P) screened all
studies across t hese two stages; a second revie wer (PB) reviewed 10% of studies included at t he title and abstract
screening stag e and 20% of stud ies included at the fu ll-text screeni ng stage. Disagree ments were resolved by
discussion or with reference t o a third reviewer (SO). If studies coll ected data on ou tcomes of interest a mong
trafficked indi viduals but did not report it, st udy authors were contact ed to request information. Studies i ncluded
after full-text scr eening underwent reference list scre ening and citation tracking usin g Web of Science and Google
Scholar. Finally, hand search es of key websites were conduc ted158.
Of the 1 718 article s identified during electronic se arches and subjec ted to title and abstract s creening, 94 were
taken through to full-text screening. Reference list screening and citation tracking identified 19 further potentially
relevant paper s while website searche s identified a fur ther 7 potential ly relevant papers . Thus a total of 120
papers under went full-text sc reening; 53 of these m et the inclusion crit eria. A furth er 37 papers were ident ified
in the previous sys tematic reviews (Ott isova et al. 2016), resulting in a tot al of 90 papers at that stage.
In order to ensure align ment with other Stu dy component s (i.e. output s from analysis of the A PMS and CSE W),
additional inclusio n criteria were then ap plied. Specif ically, studies were requ ired to report o n at least one of the
following high-level cate gories: physical viole nce; sexual violen ce; threats. Stu dies were excluded if t hey did not
report on at leas t one of these categories of interes t, for example because they prov ided aggregated data on any
physical/sex ual violence, did not disaggr egate between v iolence and other eleme nts (e.g. physical v iolence and
material neglect ), or focused only on a very sp ecific type of vio lence or coercion (e.g. thr eatened with gu n/knife,
rather than thr eatened more widely). Papers were also exc luded because data were more compr ehensively reported
in another art icle that was already included . Details of the excluded pap ers are provided at the end of this sec tion.
Data extrac tion was conducted by one re viewer using a standardised data ex traction sheet. Ex tracted data included
bibliographic informat ion, study design a nd setting , sample size and char acteristics, a nd outcomes. O utcomes
were extrac ted based on t he following high-level cat egories: physica l violence; sexual v iolence159 and threats .
Although the A PMS and the C SEW allow for finer gr ained disaggregat ions of type of violence , limitations of the
data available from st udies included in t he review meant th is was not possible here. For ex ample, the majorit y
of the studies did not dis aggregate betwe en types of physic ally violent act s or violence with injur y and without
injury. As far as possible, dat a has been ext racted disaggr egated by type of e xploitation (sexu al, labour, other).
Where exploitat ion type was not specified, i t was categorised as unspecif ied and was excluded from final a nalysis
of pooled prevalence e stimates. Informat ion was also ex tracted on d uration of traffic king; either median (wit h
first and thir d quartiles) or mean (with standar d deviation). Study authors were cont acted for data in cases where
studies include d duration data that were not disaggregat ed by type of exploitation an d age of the victims, or where
data were not provided in t he necessary format required by th e software used for con ducting the meta-analysis.
Duration data were conver ted to mont hs using the ratios of 12 mon ths to 1 year; 4. 35 weeks to 1 month; a nd
30 days to 1 month.
(157) ( PROSPE RO – Prospect ive Register of Systemat ic Reviews; EMB ASE - Excer pta Medica Data base; MEDL INE - Medical L iterature Analy sis and Retrieval Sys tem
Online; Ps ycINFO – Psyc hological Inform ation Database ; CINAHL - C umulative Inde x to Nursing and A llied Health Liter ature; PILOTS - P ublished Inter national
Literature on Traum atic Stress.
Searches wer e conducted of r elevant websites , including: Eur opean Commissio n https://ec.europa. eu/commission/index _en; World Healt h Organisation https://; United Nation s Office on Drugs an d Crime
For the purp oses of the Stud y, all victims of sexual e xploitation are a ssumed to have exp erienced sexua l violence, as th ey would not have been a ble to consent
to these exp eriences due to being v ictims of traffickin g. Taking this approa ch means that prevale nce estimates for victi ms of sexual exploitatio n will differ
from prevalenc e estimates report ed in the identified pa pers. It is possible th at some of the victims of se xual exploitation di d not experience sex ual contact
or sex of this form an d were instead exploite d into other areas of sexua l exploitation. Ho wever, this information was n ot available in the studie s and hence
it is suggeste d that victims of sexu al exploitation are a ssumed to experien ce a 100% victimisat ion rate for sexual violen ce.

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT