Professional Documents
Culture Documents
A
RECENT REPORT BY THE US Data Synthesis Of 394 publications identified, 183 were included for further analy-
Committee for Refugees es- ses of their characteristics; 91 (49.7%) included quantitative data but did not evaluate
timates there are 14.9 mil- measurement properties of instruments used in refugee research, 78 (42.6%) re-
lion refugees and 22 million ported on statistical relationships between measures (presuming validity), and 14 (7.7%)
were only about statistical properties of instruments. In these 183 publications, 125
internally displaced persons in the
different instruments were used; of these, 12 were developed in refugee research. None
world.1 Most have experienced signifi- of these instruments fully met all 5 evaluation criteria, 3 met 4 criteria, and 5 met only
cant trauma, including torture,2-4 as evi- 1 of the criteria. Another 8 standard instruments were designed and developed in non-
denced by prevalence studies in clin- refugee populations but adapted for use in refugee research; of these, 2 met all 5 cri-
ics and nonrepresentative community teria and 6 met 4 criteria.
samples.5-12 Conclusions The majority of articles about refugee trauma or health are either de-
Health problems of refugees have also scriptive or include quantitative data from instruments that have limited or untested
been documented. Clinical research validity and reliability in refugees. Primary limitations to accurate measurement in refu-
demonstrates a high prevalence of post- gee research are the lack of theoretical bases to instruments and inattention to using
traumatic stress and depression symp- and reporting sound measurement principles.
toms,6,10,13-15 and community studies us- JAMA. 2002;288:611-621 www.jama.com
ing self-rated scales2-4,8,10,16 and structured
diagnostic interviews9,17-19 have found types of insults experienced,4,6,20,23,24,27,28 Author Affiliations are listed at the end of this
article.
wide variation in the prevalence of the yet the significance of these symptoms Corresponding Author and Reprints: Michael Holli-
symptoms of posttraumatic stress (4%- is not clear since many are not charac- field, MD, Departments of Psychiatry and Family and
86%) and depression (5%-31%). Refu- teristic of posttraumatic stress disorder Community Medicine, University of New Mexico
Health Sciences Center, 2400 Tucker Ave NE, Albu-
gees experience multiple symp- (PTSD), depression, or other defined dis- querque, NM 87131 (e-mail: mhollifield@salud.unm
toms,4,5,20-26 perhaps due to the many orders.29-35 A few community studies .edu).
©2002 American Medical Association. All rights reserved. (Reprinted) JAMA, August 7, 2002—Vol 288, No. 5 611
RESULTS
Figure. Results of Literature Search and Instrument Evaluation
Review of Publications for Content
Of the 394 publications identified by Literature Search
search criteria for further review, 187
were excluded for not meeting inclu- 394 Articles Identified
sion criteria (153 were not primarily 181 From Online Databases
135 From New Mexico Refugee Project (NMRP) Database
about refugees and 34 were nonempiri- 78 From NMRP Hard Copy Files
cal or were not about trauma or health
211 Articles Excluded From This Review
status measurement) and 24 publica- 187 Articles Did Not Meet Inclusion Criteria
24 Articles Had Insufficient Data for Review
tions had insufficient data for review
(Figure). Thus, 183 articles were fur- 183 Articles Eligible (125 Measurement Instruments)
ther analyzed. Of these, 178 (97%) were
105 Instruments Excluded From This Review
about health status (61% mental health, (Not Developed in or Adapted for
16% physical health, and 20% both), and Refugee Research)
tion about the construct and item de- from their experience with Cambo- by the Pol Pot regime. The WTS full-
velopment is scant. Thus, the 17- dian adolescents, was designed to mea- scale score had an adequate internal
event list may be incomplete or biased, sure stress due to resettlement. In one consistency ("=.74), acceptable inter-
limiting generalizability. For ex- study with 38 adolescents, the RSS score rater reliability (! = 0.88), and ac-
ample, experiences of women, as illus- discriminated between those who had counted for 15.4% of PTSD score vari-
trated in the work by Allotey52 and Bon- psychiatric illness and those who did ance and 6.7% of depression score
nerjea, 53 are not well represented. not using diagnostic interviews, and ac- variance.55 Both the RSS and the WTS
Furthermore, the design of multiple counted for 11.7% of PTSD score vari- demonstrated modest predictive valid-
possible responses may confuse the re- ance but did not account for the de- ity of psychiatric disorder, and the WTS
spondent and limit reliability and there- pression score variance.55 The War demonstrated acceptable reliability.
fore validity. Finally, validity and reli- Trauma Scale (WTS), also developed by They were developed by experienced
ability of the torture item in particular Clarke et al55 from their clinical expe- investigators using rational rather than
have not been reported. rience, consists of 42 items in both an empirical methods. However, it is un-
The 32-item Resettlement Stressor interview and self-report format, mea- clear from the literature how the items
Scale (RSS), developed by Clarke et al55 suring traumatic experiences inflicted for each scale were constructed, devel-
614 JAMA, August 7, 2002—Vol 288, No. 5 (Reprinted) ©2002 American Medical Association. All rights reserved.
oped, or designed, and they were re- No refugee-specific instruments that is clear. Modest reliability and fair va-
ported in only 1 article. assess prewar/conflict or nonwar/ lidity in diagnosing PTSD was demon-
The Post Migration Living Difficul- conflict trauma-related experiences in strated in clinical populations. How-
ties Scale (PMLD), developed by Silove refugees were found. A number of gen- ever, in a community study the sensitivity
et al,9,54 is used to assess current life eral measures of lifetime trauma have and specificity of the “greater than 2.5”
stressors of asylum seekers. Each of the been developed but have not been cutoff score in diagnosing PTSD was 16%
23 items of this administered survey is adapted or used with refugees.56 and 100%, respectively, and the most ef-
rated on a 5-point scale from “no prob- Developed and Described Health Sta- ficient score for diagnosis was 1.17
lem” to “a very serious problem,” with tus Measures. From our review, 2 in- (sensitivity/specificity = 98%/100%).57
a composite score determined.9,54 Its con- struments measuring health status that The HTQ includes a limited range of pos-
struct, development, and design are only have been developed in refugee re- sible symptoms, some are not reliable,
partially described. Principal compo- search are well described in the litera- and their ability to predict impairment
nent analyses yielded 5 factors account- ture. Part 4 of the self-report HTQ, de- has not been shown. Finally, as the au-
ing for 69.8% of the variance of the 23 veloped from clinical experience by thors discuss, generalizability of the HTQ
items: refugee determination process; Mollica et al,7 lists 30 symptom items, 16 and the construct validity of PTSD in
health, welfare, and asylum problems; generated from the Diagnostic and Sta- general and in refugees need further
family concerns; general adaptation tistical Manual of Mental Disorders, Re- study.
stressors; and social and cultural isola- vised Third Edition (DSM-III-R) criteria The VDS, a self-report questionnaire
tion.54 These 5 factors were evaluated for PTSD, and 14 which are “presum- developed by Kinzie et al58 to screen Viet-
among asylum seekers, refugees, and im- ably, culture-specific symptoms associ- namese refugees for depression, was de-
migrants. Asylum seekers scored higher ated with PTSD.” Possible responses are veloped using a well described rational,
than immigrants on all 5 factors, and “not at all,” “a little,” “quite a bit,” or “ex- consensus approach from extensive clini-
higher than refugees on factors 1, 2, and tremely.” In the same convenience cal experience. Culturally appropriate
3. Refugees scored higher than immi- sample of 91 patients described earlier, terms were added to existing Western
grants on factors 2 and 3.54 Thus, the internal consistency was excellent symptoms of depression, and designed
PMLD is valid in discriminating be- ("=.96), the symptom prevalence ranged with items on a 3-point Likert scale. Af-
tween these 3 groups, but no other va- from 44% to 92%, and the 1-week item ter pilot testing, the final 15-item scale
lidity or reliability data are published. test-retest reliability ranged from poor to measures 3 symptom types: physical
The PMLD is an important concept mea- excellent (r=0.32-0.85; median, .59).7 An symptoms associated with depression in
suring life experiences other than war, average item score of greater than 2.5 was the West, Western psychological symp-
but its usefulness is limited because of predictive of a PTSD diagnosis by clini- toms of depression, and symptoms un-
the lack of description about its design, cal interview (78% sensitive, 65% related to Western concepts. The VDS is
development, reliability and validity, and specific). The purpose, construct defi- valid in discriminating between refugee
scoring. nition, and design of part 4 of the HTQ patients with depression and those with
©2002 American Medical Association. All rights reserved. (Reprinted) JAMA, August 7, 2002—Vol 288, No. 5 615
anxiety or schizophrenia, and a cutoff out fully discussing the construction of distinguished between depressed and
score of 13 out of a possible 34 points the actual instrument (eg, type or num- nondepressed refugees in London. De-
demonstrated 91% sensitivity and 96% ber of items, scoring). We include these velopment was not described further and
specificity for diagnosing DSM-III– works because the concepts and meth- to our knowledge reliability has not been
defined major depression in a commu- ods hold promise for refugee research. reported. McCloskey et al65 used simi-
nity sample.58 Reliability has not been re- Ekblad et al62 used a 7-question quali- lar methods to integrate DSM-III-R cri-
ported. tative interview to define and com- teria for PTSD, the Child Behavior
Less-Developed Health Status Mea- pare quality of life (QOL) between 14 Checklist,66 11 items about political vio-
sures. Two additional refugee health sta- Iranian refugees and 8 Swedes at a pri- lence, and 3 items about family conflict
tus measures have been reported, but to mary health care clinic in Sweden. The into an evaluation of trauma and health
our knowledge have not been formally authors found that 3 of their thematic status for Mexican and Central Ameri-
named, used in other research, or pub- domains parallel 3 of 6 domains of the can women and their children.65 The au-
lished. Bolton59 used 3 ethnographic World Health Organization quality of thors report quantitative data but no sta-
qualitative methods to investigate Rwan- life instrument (social relationships, tistical testing of these measures. While
dans’ perceptions of problems follow- level of independence, environment), there are significant limitations to Van
ing the 1994 genocidal conflict and the demonstrating a form of validity.63 Ira- Velsen’s and McCloskey’s work, each
local validity of Western concepts, and nian refugees endorsed more social con- demonstrate the integration of quanti-
to adapt existing measures for local use. cepts of quality of life than Swedes, tative measures into a qualitative clini-
Two of the 18 identified problems about demonstrating discriminant validity. cal interview, which can be used to en-
mental health were further developed. While this measure is incomplete, it is hance the validity of measures.
“Guhahamuka” (mental trauma), a con- important since QOL is understudied Cunningham and Cunningham5 de-
cept that emerged after 1994, has 36 in refugees45 and is an important com- veloped 3 checklists to gather data about
symptom items while “agahinda” (deep ponent of overall health and welfare.63 symptoms, trauma, and resettlement
sadness or grief), an older concept, con- Further development of QOL mea- problems from case records at a treat-
sists of 16 items. Guhahamuka and aga- sures in refugees is needed. ment program for multinational refu-
hinda include all symptoms required for Weine et al64 report on an interview gees in Australia. The checklists were de-
DSM-IV major depression and PTSD di- to investigate important concepts of pro- veloped from literature about symptoms
agnoses, which was interpreted as sup- vider (primary care professionals, so- and trauma in refugees, and from the au-
porting content validity of both syn- cial service workers, or refugee mental thors’ experience with resettlement prob-
dromes in Rwandans. Agahinda was 95% health professionals) knowledge, as well lems. Principal component analyses
sensitive and 38% specific for depres- as attitudes toward, and service provi- yielded 6 trauma factors and 6 symp-
sion measured by the published cutoff sion patterns for, Bosnian refugees with tom factors, with 2 trauma items ac-
point on the HSCL-25, and its test- PTSD. The instrument was developed by counting for 43% of the PTSD score vari-
retest reliability was modest but accept- the authors using rational, consensus ance. This research demonstrates the
able (r=0.67).60 This instrument is new methods. In their study of 30 ran- concept of using factor analyses to de-
and its design is not well described. domly selected providers, primary care fine potentially relevant trauma and
Beiser and Fleming61 used principal professionals had less knowledge about health constructs for refugees, but the
component analysis to identify 4 men- and provided less service to refugees checklists are limited by not being em-
tal health factors (panic, depression, so- with PTSD than did mental health or so- pirically developed or tested for their sta-
matization, and well-being) in South- cial service workers, demonstrating a tistical properties.
east Asian refugees and Euro-Canadians form of discriminant validity.64 No fur-
from interviews using 6 existing scales ther design, development, or metric Nonrefugee Instruments Adapted
as sources. The final 52-item adminis- properties were reported. for and Tested in Refugees
tered instrument demonstrated accept- Van Velsen et al14 report the devel- We found 8 instruments developed in
able internal consistency (" for the 4 opment of the Survivor of Torture As- nonrefugee research that had at least 1
scales ranged from .72-.91) and valid- sessment Record (STAR), a semistruc- statistical property tested in refugee re-
ity by discriminating between the 23 psy- tured clinical interview that incorporates search (Table 2). Two instruments, the
chiatrically ill and the 30 well respon- many instruments—such as the HSCL- HSCL-25 and the Beck Depression In-
dents.59 Refugees and Canadians scored 25—and other investigator-chosen items ventory, met all 5 evaluation criteria.
similarly on all 4 scales. The design and to determine 3 scaled scores: trauma The HSCL-25,67-69 a self-adminis-
development of this instrument is not (scored 0-7), loss of health (scored 0-9), tered questionnaire originally de-
well described and reliability data were and social losses (scored 0-6). Validity signed to measure change in 15 anxi-
not reported. was shown by the correlation between ety and 10 depression symptoms in
Underdeveloped Potential Instru- the trauma and loss-of-health scales psychotherapy,70 has been validated in
ments. Five authors report work with- (r=0.59), and by the fact that all 3 scales the general US population68 and used
616 JAMA, August 7, 2002—Vol 288, No. 5 (Reprinted) ©2002 American Medical Association. All rights reserved.
in many refugee studies. The content refugees who have experienced tor- validity (vs a standard) or reliability
and design on a 4-point severity scale ture, nontorture trauma, and mi- tested in refugee samples. The Posttrau-
is acceptable to Indochinese popula- grants who have not experienced war matic Symptom Scale-10, a 10-
tions, and reviews in the cultural psy- trauma.10 Neither scale nor item test- question survey measuring symptoms of
chiatry literature consider the mea- retest reliability was reported. PTSD, was found to have excellent in-
sure valid.71,72 An average-item score The Symptom Checklist-90 (SCL- ternal consistency (" = .92) and test-
greater than 1.75 indicates “clinically 90),77,78 developed to measure change retest reliability (r = 0.89) in Bosnian
significant distress.”7 Mollica et al69 in psychological symptoms with treat- refugees, but has not been tested for va-
tested the HSCL-25 in 3 Indochinese ment, consists of 10 scales and has been lidity in refugees.88 The Beck Depres-
groups, showing excellent test-retest re- used in 4 of the 183 studies in this re- sion Inventory89 demonstrated excel-
liability (r=0.89 for total scale; r=0.82 view. The SCL-90 depression scale was lent internal consistency ("=.93) and
for each scale), good validity in pre- valid in differentiating depressed from excellent test-retest reliability (r=0.92),
dicting diagnosed depression (88% sen- nondepressed Hmong refugees who and distinguished depressed vs nonde-
sitivity, 73% specificity) or the pres- were either patients in a psychiatric pressed Hmong refugees against a cli-
ence of any major DSM-III-R–defined clinic or who were from a community nician interview (94% sensitivity, 78%
Axis I disorder with either scale or the sample, and the depression scale cor- specificity), demonstrating validity.79
total score (93% sensitive, 76% related well with the Zung Depression The Norbeck Social Support Question-
specific). The greater than 1.75 average- Scale (r=0.67), demonstrating concur- naire (NSSQ), which measures dimen-
item score used as a diagnostic proxy rent validity.79,80 Further, the somati- sions of support that demonstrate ex-
for anxiety and depression is consis- zation scale correlated well (r = 0.40- cellent test-retest reliability and moderate
tent with community data in general US 0.52) with the somatic concern item of concurrent validity in Western stud-
populations.67,68 The HSCL-25 has good the Brief Psychiatric Rating Scale and ies,90 was adapted to study the relation-
reliability and validity in clinical refu- the somatic anxiety subscale of the ship of 3 kinds of support (social net-
gee samples, but is limited to symp- Hamilton Anxiety Scales.81 The depres- work size, emotional support, esteem
toms of anxiety and depression, may not sion scale of a translated Vietnamese support) to health in Namibian refu-
be a valid indicator of the full range of version of the SCL-90 correlated well gees. The authors found that support and
symptoms in refugees, and its ability to with the VDS (r=0.81).3 However, we coping style moderated the relation-
predict impairment has not been well found no reliability testing of the ship between chronic stress (years in ex-
studied in refugees. SCL-90 in refugee research. ile) and health status (anxiety, physi-
The Impact of Events Scale (IES)73 Other adapted instruments were cal symptoms, physical signs, and
has been used in 8 of the 183 studies tested in single studies identified in this hospitalization in the previous year),
in this review. The 15-item measure has review. The anxiety and depression demonstrating a form of predictive va-
7 intrusion and 8 avoidance items on scales from the Health Opinion Survey lidity. 91 The NSSQ showed good inter-
3-point descriptive scales measuring in- (HOS),82 an instrument derived from the nal consistency (" = .83), but no fur-
trusive thoughts and body sensations general Cornell Medical Index to mea- ther adaptations or reliability testing
and avoidance behaviors after trauma. sure psychophysiological symptoms, have been conducted among refugees.
It is valid and reliable in general popu- were administered at time of interview
lations,74 and its development is well de- to a community sample of 2180 South- COMMENT
scribed. The 2 scales had satisfactory in- east Asian refugees from 3 countries. A Half (n=91) of the 183 articles about
ternal consistency (" = .82 and .74, factor analysis demonstrated that anxi- measurement of refugee trauma and
respectively) and accounted for 41% of ety and depression were common and health evaluated for this review re-
the variance in IES scores in a study of had the same meaning for all 3 groups.83 ported quantitative data but did not re-
1787 Croatian and Bosnian chil- Further analyses reported in a subse- port evaluation of association be-
dren,75 confirming the 2-scale con- quent article demonstrated that a single tween or statistical properties of the
struct, although individual items fit dif- factor resembling the concept of neur- instruments used. Forty-three percent
ferently than in the original 20-item asthenia accounted for 40% of the dis- (n = 78) reported associations be-
version of the scales. Principal compo- tress scores on the HOS.84 A commu- tween measures, assuming that valid-
nent analysis suggests a third scale, nity sample of Vietnamese refugees ity had already been determined. In
named “numbing,” which requires fur- demonstrated high and persistent lev- these 183 articles, 125 different instru-
ther validation.75,76 Higher intrusion and els of physical and psychological symp- ments were used. However, only 12 in-
total scores in children with more toms on the Cornell Medical Index struments have been developed and
trauma events demonstrated validity, al- (CMI), a general health-status question- tested specifically in refugee research;
though trauma events did not predict naire, compared with normative data 3 of these met 4 of 5 evaluation crite-
avoidance scores. The IES scores also from the United States and Britain.85-87 ria, but none fully met all 5 criteria rec-
distinguish between 3 groups of adult Neither the HOS nor the CMI have had ommended for a developed instru-
©2002 American Medical Association. All rights reserved. (Reprinted) JAMA, August 7, 2002—Vol 288, No. 5 617
ment, and none have been fully cution that is explicitly or implicitly than if their relationship to impairment
published in the literature evaluated in sanctioned by the state). Operational- is established. Only a few community
this review. Only 41 (22%) of the 183 izing this definition into a measure prevalence studies of psychiatric disor-
articles used these 12 instruments. An- would focus on persons who have been ders using diagnostic instruments have
other 8 well-described instruments displaced because of a threat to their been reported,9,17-19 and these did not as-
adapted for use in refugee research were safety. However, the optimal construct sess the relationship of disorders to im-
identified and evaluated; 2 of these in- definition of “refugee” is an empirical pairment. Symptoms, disorders, and even
struments met all 5 criteria, 6 met 4 cri- question that requires further study. objective evidence of disease do not nec-
teria (but these were all designed and Likewise, there is need for further essarily imply impairment, as demon-
developed in nonrefugee popula- study of what constitutes refugee strated in people with renal disease, panic
tions), and these 6 have not been tested trauma. No empirically developed in- disorder, and heart disease.110-113 Ill-
for either validity or reliability among struments assess the complete range of ness, on the other hand, is defined by loss
refugees. Of the 183 articles, 19% (35) trauma experiences in refugees. It is dif- of functioning, and impairs a person in
used the 8 adapted instruments. ficult to define all relevant events and a highly contextual manner and is not
The primary limitations to accurate types of events that influence health sta- necessarily defined in current medical
measurement of trauma and health sta- tus,* including the understudied ef- nosology.114 Research has focused on in-
tus in refugees are the lack of theory- fects of non–conflict-related events,17,54 fectious diseases, PTSD, anxiety, and de-
based construct definitions to guide the since trauma may precede and post- pression, with some research focusing on
development and design of instru- date experiences related to war and con- physical injury, nutrition, and preven-
ments specifically in refugee popula- flict, genocide, disaster, or oppression tive health. Consideration must be given
tions and inattention to use and report- and because subjective experiences are to other symptom complexes that are
ing of sound measurement principles. highly variable.3,4,19,21,54,93,94 Thus, fur- more strongly associated with impair-
These shortcomings may account for ther community-based empirical re- ment.7,20,35,97,115-121 For example, it is not
conflicting data between studies— search to better define the range and clear that “PTSD” is the most appropri-
different phenomena given the same type of events that are associated with ate construct for traumatized refugees
name are being used to evaluate and adverse health status is needed. Refu- with symptoms currently defined as
compare populations. Improving mea- gee researchers might consider adopt- PTSD. Culture and language compli-
surement may help to clarify events that ing methodologies from life events re- cate diagnosis,122 and polytrauma is
are traumatic and predictive of poor search to better define how and what pathogenic for disorders that are differ-
health, enable clinicians to better di- events are weighted as traumatic and ent compared with those found in non-
agnose and care for patients, assist pub- predictive of poor health.107,108 The con- refugee populations.17,18,123 The work of
lic health officials to develop better pre- cept of “polytrauma” might be devel- Bolton60 is important in this regard as it
vention models, allow scientists to oped and used in research, since refu- demonstrates how community-based,
conduct more useful research, and pro- gees experience multiple events in empirical qualitative data validate refu-
vide more accurate documentation of multiple contexts over time. This con- gees’ illness experiences. However, this
human rights abuses. cept, applicable to other popula- work has yet to demonstrate how local
tions,109 is especially important for refu- illness constructs are related to impair-
Improving Theory and gee research to remind investigators of ment. In addition to improving measure-
Construct Definitions the multiple events to consider as mod- ment about the full range of symptoms
What instruments often lack is good, erators of health status. and valid illness constructs, measures of
theory-based construct definitions that Measurement constructs of health sta- self-rated impairment for refugees are
guide the design and development of tus are better developed than are those needed, since negative self-perceptions
measures, as shown in Table 1. For ex- for trauma. Extant studies evaluate medi- of health may predict future physical ill-
ample, legal definitions that distin- cal and psychological symptoms, disor- ness, mortality, and quality of life, inde-
guish “refugee” from “asylee” from “in- ders, diseases, and impairment. How- pendent of objective health status.124-127
ternally displaced person” are not ever, no community-based empirically
necessarily predictors of trauma expe- developed instruments assess the full Improvements in
riences or health status.92 It might be range of symptoms in refugees, and valid Measurement Principles
that, for research purposes, a “refugee” illness constructs associated with im- Design and Development. Instru-
or “displaced person” is best defined as pairment are underdeveloped and re- ments developed in community refu-
a person who has fled his/her social liv- quire further study. For example, psy- gee populations using empirical ap-
ing context because of threat to the safety chiatric symptom counts are less proaches combining qualitative and
or integrity of themselves or family meaningful indicators of adverse health quantitative methods may create mea-
members due to any cause (eg, war, civil sures that are more valid in represent-
conflict, disaster, oppression, or perse- *References 7, 10, 17, 18, 20, 95-106. ing the experiences of refugees than
618 JAMA, August 7, 2002—Vol 288, No. 5 (Reprinted) ©2002 American Medical Association. All rights reserved.
methods where data are only obtained of 13 on the VDS demonstrated excel- indicates the need for improvements in
rationally via expert and consensus ap- lent validity to DSM-III-R major de- the development, use, and reporting of
proaches. 1 2 8 , 1 2 9 Qualitative tech- pression diagnosis in a community instruments used to measure trauma and
niques, such as in-depth interviews and sample, and the HSCL-25 anxiety and health status in refugee research.
focus groups, help identify the range, depression scales demonstrate excel-
Author Affiliations: Departments of Family and Com-
depth, and meaning of possible re- lent test-retest reliability in Southeast munity Medicine (Dr Hollifield) and Psychiatry (Drs Hol-
sponses in a population,130-133 and al- Asian refugees. However, proper in- lifield, Warner, Lian, and Stevenson), University of New
Mexico Health Sciences Center, Albuquerque; Sleep
low for development of culturally in- strument development should demon- and Human Health Institute, Albuquerque, NM (Dr
formed quantitative measures. These strate internal consistency (ie, item in- Krakow); Departments of Anthropology and Psychia-
try, Case Western Reserve University, Cleveland, Ohio
new instruments must then be vali- tercorrelation), stability (ie, consistent (Dr Jenkins); Northern Colorado Family Practice Cen-
dated using iterative statistical and field scores over time, such as test-retest re- ter, Greeley (Dr Kesler); University of Minnesota and
testing methods. Further, culturally in- liability), and validity of the construct the Minneapolis Veterans Affairs Center, Minneapo-
lis, Minn (Dr Westermeyer).
formed quantitative instruments must (ie, correlation with a standard). In- Author Contributions: Study concept and design:
be designed to be linguistically and vi- strument validity for psychiatric disor- Hollifield, Warner, Krakow, Jenkins, Westermeyer.
Acquisition of data: Hollifield, Lian, Kesler, Stevenson.
sually acceptable and understandable ders is difficult to establish, since psy- Analysis and interpretation of data: Hollifield, Warner,
to various refugee groups. chiatric diagnostic interviews as Lian, Jenkins.
Drafting of the manuscript: Hollifield, Warner, Lian,
Testing Statistical Properties: Va- standards may be insufficiently valid.136 Stevenson.
lidity and Reliability. There are many Thus, determining what standards to Critical revision of the manuscript for important in-
tellectual content: Hollifield, Warner, Lian, Krakow,
kinds of validity and reliability that use for validity testing can be a diffi- Jenkins, Kesler, Westermeyer.
must be demonstrated for a measure to cult methodological problem. Per- Statistical expertise: Hollifield, Warner, Jenkins.
be accurate across groups. For ex- haps measures of impairment that are Obtained funding: Hollifield, Warner, Kesler.
Administrative, technical, or material support: Lian,
ample, the HTQ reports the validity of valid predictors of future health out- Kesler, Westermeyer, Stevenson.
the “greater than 2.5” average item score comes may be the best standards against Study supervision: Hollifield, Warner, Krakow, Jenkins,
Westermeyer.
in predicting PTSD, but this was in out- which developed mental health mea- Qualitative methods: Jenkins.
patient psychiatric patients, and this sures should be tested. Funding/Support: This work was supported by Na-
tional Institute of Mental Health grant R01 MH 59574.
score was not corroborated in a com- Acknowledgment: We thank James Ruiz and Valo-
munity based sample. Nevertheless, this Limitations rie Eckert, MPH, for helping to assemble material for
cutoff score has been used in other com- There are limitations to this review. First, this work.
©2002 American Medical Association. All rights reserved. (Reprinted) JAMA, August 7, 2002—Vol 288, No. 5 619
post-migration stressors. Br J Psychiatry. 1997;170: 32. Foy D. Treating PTSD: Cognitive-Behavioral Strat- 55. Clarke G, Sack WH, Goff B. Three forms of stress
351-357. egies. New York, NY: Guilford Press; 1992. in Cambodian adolescent refugees. J Abnorm Child
10. Thompson M, McGorry P. Psychological se- 33. Green BL, Lindy JD, Grace MC. Posttraumatic stress Psychol. 1993;21:65-77.
quelae of torture and trauma in Chilean and Salva- disorder: toward DSM-IV. J Neurol Ment Disord. 1985; 56. Norris F, Riad J. Standardized self-report mea-
dorean migrants: a pilot study. Aust N Z J Psychiatry. 173:406-411. sures of civilian trauma and posttraumatic stress dis-
1995;29:84-95. 34. Horowitz M, Wilner N, Kaltreider N, Alverez W. order. In: Wilson J, Keane T, eds. Assessing Psycho-
11. Tan ES, Kripinski J, Chiu E. Stresses Experienced Signs and symptoms of post-traumatic stress disor- logical Trauma and PTSD. New York, NY: Guilford
Prior to Arrival in Australia. Sydney, Australia: Per- der. Arch Gen Psychiatry. 1980;37:85-92. Press; 1997:7-42.
gammon Press; 1986. 35. Van der Kolk BA, Pelcovitz D, Roth S, Mandel FS, 57. Smith Fawzi MC, Murphy E, Pham T, Lin L, Poole
12. Gong-Guy E. California Southeast Asian Mental McFarlane A, Herman JL. Dissociation, somatization, C, Mollica RF. The validity of screening for post-
Health Needs Assessment: Contract 85-76282A-2. and affect dysregulation: the complexity of adapta- traumatic stress disorder and major depression among
Oakland, Calif: Asian Community Mental Health Ser- tion of trauma. Am J Psychiatry. 1996;153:83-93. Vietnamese former political prisoners. Acta Psychiatr
vices: California State Dept of Mental Health; 1987. 36. Cheung P. Somatisation as a presentation in de- Scand. 1997;95:87-93.
13. Mollica RF, Wyshak G, Lavelle J, Truong T, Tor pression and post-traumatic stress disorder among 58. Kinzie JD, Manson SM, Vinh DT, Tolan NT, Anh
S, Yang T. Assessing symptom change in Southeast Cambodian refugees. Aust N Z J Psychiatry. 1993; B, Pho TN. Development and validation of a Vietnam-
Asian refugee survivors of mass violence and torture. 27:422-428. ese-language depression rating scale. Am J Psychia-
Am J Psychiatry. 1990;147:83-88. 37. Westermeyer J, Bouafuely M, Neider J, Callies A. try. 1982;139:1276-1281.
14. Van Velsen C, Gorst-Unsworth C, Turner S. Sur- Somatization among refugees: an epidemiologic study. 59. Bolton P. Local perceptions of the mental health
vivors of torture and organized violence: demogra- Psychosomatics. 1989;30:34-43. effects of the Rwandan genocide. J Nerv Ment Dis.
phy and diagnosis. J Trauma Stress. 1996;9:181- 38. Gavagan T, Brodyaga L. Medical care for immi- 2001;189:243-248.
193. grants and refugees. Am Fam Physician. 1998;57: 60. Bolton P. Cross-cultural validity and reliability test-
15. Weine S, Laub D. Narrative constructions of his- 1061-1068. ing of a standard psychiatric assessment instrument
torical realities in testimony with Bosnian survivors of 39. Mollica RF, McInnes K, Sarajlic N, Lavelle J, Sara- without a gold standard. J Nerv Ment Dis. 2001;189:
“ethnic cleansing.” Psychiatry. 1995;58:246-260. jlic I, Massagli MP. Disability associated with psychi- 238-242.
16. Cervantes RC, Salgado de Snyder VN, Padilla AM. atric comorbidity and health status in Bosnian refu- 61. Beiser M, Fleming JAE. Measuring psychiatric dis-
Posttraumatic stress in immigrants from Central gees living in Croatia. JAMA. 1999;282:433-439. order among Southeast Asian refugees. Psychol Med.
America and Mexico. Hosp Community Psychiatry. 40. Eitinger L. Concentration Camp Survivors. Lon- 1986;16:627-639.
1989;40:615-619. don, England: Oslo University Press; 1964. 62. Ekblad S, Abazari A, Eriksson N-G. Migration stress-
17. Basoglu M, Paker M, Ozmen E, Tasdemir O, Sa- 41. Segal J, Hunter EJ, Segal Z. Universal conse- related challenges asociated with perceived quality of
hin D. Factors related to long-term traumatic stress quences of captivity: stress reactions among diver- life: a qualitative analysis of Iranian refugees and Swed-
responses in survivors of torture in Turkey. JAMA. gent populations of prisoners of war and their fami- ish patients. Transcultural Psychiatry. 1999;36:329-
1994;272:357-363. lies. Int Soc Sci J. 1976;28:593-609. 345.
18. Basoglu M, Paker M, Paker O, et al. Psychologi- 42. Ursano R, Boydstein J, Wheatley R. Psychiatric ill- 63. Sartorius N, Janca A. World Health Organiza-
cal effects of torture: a comparison of tortured with ness in US Air Force Viet Nam prisoners of war: a five- tion Quality of Life Assessment Instrument. In:
nontortured political activists in Turkey. Am J Psy- year follow-up. Am J Psychiatry. 1981;138:310- Thornicroft G, Tansella M, eds. Mental Health Out-
chiatry. 1994;151:76-81. 314. comes Measures. New York, NY: Springer; 1996:
19. Cheung P. Posttraumatic stress disorder among 43. Friedman MJ, Schnurr PP. The relationship be- 153-177.
Cambodian refugees in New Zealand. Int J Soc Psy- tween trauma, posttraumatic stress disorder, and physi- 64. Weine SM, Kuc G, Dzudza E, Razzano L, Pavk-
chiatry. 1994;40:17-26. cal health. In: Friedman MJ, Charney DS, Deutch AY, ovic I. PTSD among Bosnian refugees: a survey of pro-
20. Allodi FA, Cowgill G. Ethical and psychiatric as- eds. Neurobiological and Clinical Consequences of viders’ knowledge, attitudes and service patterns. Com-
pects of torture: a Canadian study. Can J Psychiatry. Stress: From Normal Adaptation to PTSD. Philadel- munity Ment Health J. 2001;37:261-271.
1982;27:98-102. phia, Pa: Lippincott-Raven; 1995:507-524. 65. McCloskey LA, Fernandez-Esquer ME, South-
21. Castillo R, Waitzkin H, Ramirez Y, Escobar JI. So- 44. Schnurr PP. Trauma, PTSD, and physical health. wick K, Locke C. The psychological effects of political
matization in primary care, with a focus on immi- PTSD Res Q. 1996;7:1-6. and domestic violence on Central American and Mexi-
grants and refugees. Arch Fam Med. 1995;4:637- 45. De Vries J, Van Heck GL. Quality of life and refu- can immigrant mothers and children. J Community Psy-
646. gees. Int J Ment Health. 1994;23:57-75. chol. 1995;23:95-116.
22. Eitinger L. The symptomatology of mental dis- 46. Westermeyer J, Janca A. Language, culture and 66. Achenbach T, Edelbrock C. Manual for the Child
ease among refugees in Norway. J Ment Sci. 1960; psychopathology: conceptual and methodological is- Behavior Checklist and Revised Child Behavior Pro-
106:947-966. sues. Transcultural Psychiatry. 1997;34:291-311. file. Burlington, Vt: University Associates in Psychia-
23. Hougen HP. Physical and psychological se- 47. Weathers F, Keane T, King L, King D. Psycho- try; 1983.
quelae to torture: a controlled clinical study of exiled metric theory in the development of posttraumatic 67. Derogatis LR, Lipman RS, Rickels K, Uhlenhuth
asylum applicants. Forensic Sci Int. 1988;39:5-11. stress disorder assessment tools. In: Wilson J, Keane EH, Covi L. The Hopkins Symptom Checklist (HSCL):
24. Hougen HP, Kelstrup J, Petersen HD, Rasmus- T, eds. Assessing Psychological Trauma and PTSD. New a measure of primary symptom dimensions. In: Basel
sen OV. Sequelae to torture: a controlled study of tor- York, NY: Guilford Press; 1997:98-135. P, ed. Modern Problems in Pharmacopsychiatry. Basel,
ture victims living in exile. Forensic Sci Int. 1988;36: 48. McDowell I, Newell C. The Quality of Well- Switzerland: Karger; 1974.
153-160. Being Scale (formerly the Index of Well-Being). In: 68. Winokur A, Winokur DF, Rickels K, Cox D. Symp-
25. Pederson S. Psychopathological reactions to ex- Measuring Health. New York, NY: Oxford University toms of emotional distress in a family planning ser-
treme social displacements (refugee neuroses). Psy- Press; 1996:483-492. vice: stability over a four-week period. Br J Psychia-
choanal Rev. 1949;36:344-354. 49. Wilson JP, Keane TM. Assessing Psychological try. 1984;144:395-399.
26. Weinstein HM, Dansky L, Iacopino V. Torture and Trauma and PTSD. New York, NY: Guilford Press; 1997. 69. Mollica RF, Wyshak G, de-Marneffe D, Khuon F,
war trauma survivors in primary care practice. West J 50. Mollica R. The Harvard Trauma Questionnaire Lavelle J. Indochinese versions of the Hopkins Symp-
Med. 1996;165:112-118. Manual: Indochinese Versions. Cambridge, Mass: Har- tom Checklist-25: a screening instrument for the psy-
27. Chester B, Holtan N. working with refugee sur- vard University; undated. chiatric care of refugees. Am J Psychiatry. 1987;144:
vivors of torture. West J Med. 1992;157:301-304. 51. Wyshak G. The relationship between change in 497-500.
28. Petersen HD, Christensen ME, Kastrup M, Thom- reports of traumatic events and symptoms of psychi- 70. Parloff M, Kelman H, Frank J. Comfort, effective-
sen JL, Foldspang A. General health assessment in refu- atric distress. Gen Hosp Psychiatry. 1994;16:290- ness, and self-awareness as criteria of improvement
gees claiming to have been tortured. Forensic Sci Int. 297. in psychotherapy. Am J Psychiatry. 1954;111:343-
1994;67:9-16. 52. Allotey P. Travelling with “excess baggage”: health 351.
29. American Psychiatric Association. Diagnostic and problems of refugee women in Western Australia. 71. Butcher JN. Psychological evaluation. In: Wes-
Statistical Manual of Mental Disorders, Fourth Edi- Women Health. 1998;28:63-81. termyer J, Williams C, Nguyen A, eds. Mental Health
tion. Washington, DC: American Psychiatric Associa- 53. Bonnerjea L. Shaming the World: The Needs of Services for Refugees. Washington, DC: Govern-
tion; 1994. Women Refugees. United Kingdom: Change Inter- ment Printing Office; 1991:111-122.
30. Brett E, Ostroff R. Imagery and posttraumatic tress national Reports and World University Service; 1985. 72. Kinzie JD, Manson SM. The use of self-rating scales
disorder: an overview. Am J Psychiatry. 1985;142: 54. Silove D, Steel Z, McGorry P, Mohan P. Trauma in cross-cultural psychiatry. Hosp Community Psy-
417-424. exposure, postmigration stressors, and symptoms of chiatry. 1987;38:190-196.
31. Brett E, Spitzer R, William J. DSM-III-R criteria for anxiety, depression and post-traumatic stress in Ta- 73. Horowitz NJ, Wilmer N, Alvarez N. Impact of
post-traumatic stress disorder. Am J Psychiatry. 1988; mil asylum-seekers: comparison with refugees and im- Events Scale: a measure of subjective stress. Psycho-
145:1232-1236. migrants. Acta Psychiatr Scand. 1998;97:175-181. som Med. 1979;41:209-218.
620 JAMA, August 7, 2002—Vol 288, No. 5 (Reprinted) ©2002 American Medical Association. All rights reserved.
74. Paton D. Assessing the impact of disasters on help- 96. Turner S, Landau T, Hinshelwood J, Bamber H. vivors of prolonged and repeated trauma. J Trauma
ers. Couns Psychol. 1990;3:149-152. Torture of Turkish Kurds. Lancet. 1989;1:1319. Stress. 1992;5:377-391.
75. Dyregrov A, Kuterovac G, Barath A. Factor analy- 97. Turner SW, Gorst-Unsworth C. Psychological se- 118. Jenkins JH. Cultural comments on adjustment
sis of the Impact of Event Scale with children in war. quelae of torture: a descriptive model. Br J Psychia- and stress disorders. In: Mezzich JE, Kleinman AMD,
Scand J Psychol. 1996;37:339-350. try. 1990;157:475-480. Horacio F Jr, Parron DL, eds. Culture and Psychiatric
76. Schwartzwald J, Solomon Z, Weisenberg M, Miku- 98. Mollica RF, Caspi-Yavin Y. Measuring torture and Diagnosis: A DSM-IV Perspective. Washington, DC:
lincer M. Validation of the Impact of Event Scale for torture-related symptoms. Psychol Assess. 1991;3: American Psychiatric Association; 1996.
the psychological sequelae of combat. J Consult Clin 581-587. 119. Kalucy RS. The health needs of victims of tor-
Psychol. 1987;55:251-256. 99. Shrestha NM, Sharma B, Van Ommeren M, et al. ture. Med J Aust. 1988;148:321-323.
77. Derogatis L, Cleary P. Confirmation of the di- Impact of torture on refugees displaced within the de- 120. Roth S, Newman E, Pelcovitz F, Van der Kolk B,
mensional structure of the SCL-90: a study in con- veloping world. JAMA. 1998;280:443-448. Mandel FS. Complex PTSD in victims exposed to sexual
struct validation. J Clin Psychol. 1977;33:981-989. 100. Van Willigan LHM, Hondius AJK, Van der Ploeg and physical abuse: results from the DSM-IV Field Trial
78. Lipman RS, Covi L, Shapiro AK. The Hopkins HM. Health problems of refugees in The Nether- for Posttraumatic Stress Disorder. J Trauma Stress.
Symptom Checklist (HSCL): factors derived from the lands. Trop Geogr Med. 1995;47:118-124. 1997;10:539-555.
HSCL-90. J Affect Disord. 1979;1:9-24. 101. National Institute of Mental Health. Mental 121. Turner SW, Gorst-Unsworth C. Psychological se-
79. Westermeyer J, Vang T, Neider J. A comparison Health Research for Survivors of Torture and Related quelae of torture. In: Wilson J, Raphael B, eds. Inter-
of refugees using and not using a psychiatric service: Trauma. Rockville, Md: National Institute of Mental national Handbook of Traumatic Stress Syndromes.
an analysis of DSM-III criteria and self-rating scales Health; 1998. New York, NY: Plenum Press; 1993.
in cross-cultural context. J Operational Psychiatry. 102. Green BL. Traumatic stress and disaster: Men- 122. deGirolamo GD, McFarlane AC. The epidemi-
1983;14:36-41. tal health effects and factors influencing adaptation. ology of PTSD: A comprehensive review of the inter-
80. Westermeyer J. Two self-rating scales for depres- In: Mak F, Nadelson C, eds. International Review of national literature. In: Marsella A, Friedman M, Ger-
sion in Hmong refugees: assessment in clinical and non- Psychiatry. Washington, DC: American Psychiatric rity E, Scurfield R, eds. Ethnocultural Aspects of
clinical samples. J Psychiatr Res. 1986;20:103-113. Press; 1996:177-210. Posttraumatic Stress Disorder. Washington, DC: Ameri-
81. Westermeyer J, Neider J, Callies A. Psychosocial 103. Wilson JP. The need for an integrative theory can Psychological Association; 1996.
adjustment of Hmong refugees during their first de- of posttraumatic stress disorder. In: Williams MB, Som- 123. Punakami RI. Political violence and mental health.
cade in the United States. J Nerv Ment Dis. 1989; mer JF, eds. Handbook of Posttraumatic Therapy. Int J Ment Health. 1989;17:3-15.
177:132-139. Westport, Conn: Greenwood Press; 1994:3-18. 124. Idler EL, Kasl S. Health perceptions and sur-
82. Leighton D, Harding J, Macklin D, Macmillan A, 104. Litz BT, Orsillo SM, Friedman M, Ehlich P, Ba- vival: do global evaluations of health status really pre-
Leighton A. The Character of Danger. New York, NY: tres A. Posttraumatic stress disorder associated with dict mortality? J Gerontol Soc Sci. 1991;46:55-65.
Basic Books; 1963. peacekeeping duty in Somalia for U.S military per- 125. Mossey JA, Shapiro E. Self-rated health: a pre-
83. Chung RC, Kagawa-Singer M. Predictors of psy- sonnel. Am J Psychiatry. 1997;154:178-184. dictor of mortality among the elderly. Am J Public
chological distress among southeast Asian refugees. 105. Soloman Z. The effect of prior stressful experi- Health. 1982;72:800-808.
Soc Sci Med. 1993;36:631-639. ence in coping with war trauma and captivity. Psy- 126. Singer E, Garfinkel R, Cohen SM, Strole L. Mor-
84. Chung RC-Y, Singer M. Interpretation of symp- chol Med. 1995;25:1289-1294. tality and mental health: evidence from the midtown
tom presentation and distress: a Southeast Asian refu- 106. Yehuda R, McFarlane AC. Conflict between cur- Manhattan restudy. Soc Sci Med. 1976;10:517-525.
gee example. J Nerv Ment Dis. 1995;183:639-648. rent knowledge about posttraumatic stress disorder 127. Zola IK. Culture and symptoms: an analysis of
85. Brodman K, Erdmann AJ Jr, Wolff HG. Manual and its original conceptual basis. Am J Psychiatry. 1995; patients’ presenting complaints. Am Sociol Rev. 1966;
of Cornell Medical Index-Health Questionnaire. New 152:1705-1713. 31:615-630.
York: New York Hospital and the Departments of 107. Brown GW. Life-events and psychiatric illness: 128. Weiss MG, Doongaji DR, Siddhartha S, et al. The
Medicine (Neurology) and Psychiatry, Cornell Uni- some thoughts on methodology and causality. J Psy- Explanatory Model Interview Catalogue (EMIC): con-
versity Medical College; 1956. chosom Res. 1972;16:311-320. tribution to cross-cultural research methods from a
86. Brown AC, Fry J. The Cornell Medical Index- 108. Holmes TH, Rahe RH. The Social Readjustment study of leprosy and mental health. Br J Psychiatry.
Health Questionnaire in the identification of neurotic Rating Scale. J Psychosom Res. 1967;11:213-218. 1992;160:819-830.
patients in general practice. J Psychosom Res. 1962; 109. Krakow B, Hollifield M, Schrader R, et al. A con- 129. Weiss M. Explanatory Model Interview Cata-
6:185. trolled study of imagery rehearsal for chronic night- logue (EMIC): framework for comparative study of ill-
87. Lin M, Tazuma L, Masuda M. Adaptational prob- mares in sexual assault survivors with PTSD: a pre- ness. Transcultural Psychiatry. 1997;34:235-263.
lems of Vietnamese refugees. Arch Gen Psychiatry. liminary report. J Trauma Stress. 2000;13:589-609. 130. Pelto PJ, Pelto GH. Anthropological Research:
1979;36:955-961. 110. Harris LE, Luft FC, Rudy DW, Tierney WM. Clini- The Structure of Inquiry. 2nd ed.Cambridge, Mass:
88. Thulesius H, Hakansson A. Screening for post- cal correlates of functional status in patients with Cambridge University Press; 1984.
traumatic stress disorder symptoms among Bosnian chronic renal insufficiency. Am J Kidney Dis. 1993; 131. Spradley JP. The Ethnographic Interview. New
refugees. J Trauma Stress. 1999;12:167-174. 21:161-166. York, NY: Holt Rinehart & Winston; 1977.
89. Beck A, Ward C, Mendelson M, Mock J, Er- 111. Hollifield M, Katon W, Skipper B, Chapman T, 132. Bernard HR. Research Methods in Anthropol-
baugh J. An inventory for measuring depression. Arch Ballenger JC, Fyer A. Panic disorder and quality of life: ogy: Qualitative and Quantitative Approaches. 2nd
Gen Psychiatry. 1961;4:561-571. variables predictive of functional impairment. Am J Psy- ed.Thousand Oaks, Calif: Sage Publications; 1994.
90. Norbeck JS, Lindsey AM, Carrieri VL. The devel- chiatry. 1997;154:766-772. 133. Helitzer-Allen DL, Makhambera M, Wangel AM.
opment of an instrument to measure social support. 112. Julius M, Hawthorne VM, Carpenter-Alting P, Obtaining sensitive information: the need for more than
Nurs Res 1981;30:264-249. Kneisley J, Wolfe RA, Port FK. Independence in ac- focus groups. Reproductive Health Matters. 1994;3:
91. Shisana O, Celentano DD. Relationship of chronic tivities of daily living for end-stage renal disease pa- 75-82.
stress, social support, and coping style to health among tients: biomedical and demographic correlates. Am J 134. Cook TD, Campbell DT. Quasi-experimenta-
Namibian refugees. Soc Sci Med. 1987;24:145-157. Kidney Dis. 1989;13:61-69. tion: Design and Analysis Issues for Field Settings. Bos-
92. Summerfield D, Toser L. ‘Low intensity’ war and 113. Sullivan MD, LaCroix AZ, Baum C, Grothaus LC, ton, Mass: Houghton Mifflin; 1979.
mental trauma in Nicaragua: a study in a rural com- Katon WJ. Functional status in coronary artery dis- 135. Cohen P, Cohen J. The clinician’s illusion. Arch
munity. Med War. 1991;7:84-99. ease: a one-year prospective study of the role of anxi- Gen Psychiatry. 1984;41:1178-1182.
93. Solomon Z, Mikulincer M, Flum H. Negative life ety and depression. Am J Med. 1997;103:348-356. 136. Brugha TS, Bebbington PE, Jenkins R, et al. Cross
events, coping responses, and combat-related psy- 114. Kleinman A. Rethinking Psychiatry: From Cul- validation of a general population survey diagnostic
chopathology: a prospective study. J Abnorm Psy- tural Category to Personal Experience. 7th ed. New interview: a comparison of CIS-R with SCAN ICD-10
chol. 1988;97:302-307. York, NY: The Free Press; 1988. diagnostic categories. Psychol Med. 1999;29:1029-
94. Steel Z, Silove D, Bird K, McGorry P, Mohan P. 115. Allodi FA. Assessment and treatment of torture 1042.
Pathways from war trauma to posttraumatic stress victims: a critical review. J Nerv Ment Dis. 1991;179: 137. Porter M, Haslam N. Forced displacement in Yu-
symptoms among Tamil asylum seekers, refugees, and 4-11. goslavia: a meta-analysis of psychological conse-
immigrants. J Trauma Stress. 1999;12:421-435. 116. Goldfeld AE, Mollica RF, Pesavento BH, Fara- quences and their moderators. J Trauma Stress. 2001;
95. Rasmussen OV, Dam AM, Nielsen IL. Torture: A one SV. The physical and psychological sequelae of 14:817-834.
study of Chilean and Greek victims. In: Evidence of torture: symptomatology and diagnosis. JAMA. 1988; 138. Keyes EF. Mental health status in refugees: an
Torture. London, England: Amnesty International Pub- 259:2725-2729. integrative review of current research. Issues Ment
lications; 1977. 117. Herman JL. Complex PTSD: a syndrome in sur- Health Nurs. 2000;21:397-410.
©2002 American Medical Association. All rights reserved. (Reprinted) JAMA, August 7, 2002—Vol 288, No. 5 621