402 by suchenfz


									Posttraumatic Stress Among Hospitalized
and Nonhospitalized Survivors of Serious
Car Crashes: A Population-Based Study
Shanthi Ameratunga, M.B.Ch.B., Ph.D.
Sandar Tin Tin, M.B.B.S., M.P.H.
John Coverdale, M.D., M.Ed.
Jennie Connor, M.B.Ch.B., Ph.D.
Robyn Norton, M.P.H., Ph.D.

Objective: This study investigated          with posttraumatic stress disor-              Methods
the prevalence of posttraumatic             der. At 18 months, 23% of hospi-              This population-based prospective
stress among survivors of serious           talized passengers, 11% of hospi-             cohort study was designed to recruit
injury-producing car crashes.               talized drivers, and 7% of nonhos-            all hospitalized car occupants (pas-
Methods: This population-based              pitalized drivers reported signifi-           sengers and drivers) as well as non-
prospective cohort study, con-              cant levels of stress. Conclusions:           hospitalized drivers involved in crash-
ducted in New Zealand, recruited            Strategies to prevent disabling se-           es that resulted in hospital admission
hospitalized car occupants (pas-            quelae of crashes must address                of at least one occupant. The study
sengers and drivers) as well as             the needs of hospitalized and non-            covered all such crashes in the Auck-
nonhospitalized drivers after a             hospitalized survivors. (Psychi-              land region (population approximate-
crash in which at least one occu-           atric Services 60:402–404, 2009)              ly one million) between October 1998
pant was hospitalized. Fifty-nine                                                         and July 1999. As described in detail
hospitalized passengers (62%)
and 209 drivers (72%) completed
five- and 18-month interviews.
                                            R     ecent reviews have highlighted
                                                  the increasing global burden of
                                            road traffic injuries while acknowl-
                                                                                          elsewhere (4), we excluded individu-
                                                                                          als aged less than 16 years, those un-
                                                                                          able to provide informed consent (for
The Impact of Event Scale as-               edging methodological difficulties in         example, those with significant cogni-
sessed symptoms of posttraumatic            estimating the incidence of disabling         tive problems), and those who sur-
stress. Results: At five months             nonfatal sequelae related to these in-        vived fatal crashes or experienced fur-
28% of hospitalized passengers,             juries (1). Researchers have attrib-          ther crashes during the follow-up pe-
24% of hospitalized drivers, and            uted the widely divergent estimates           riod (because of the anticipated re-
24% of nonhospitalized drivers              of posttraumatic stress among sur-            sponse burden and changes in expo-
reported symptoms consistent                vivors of motor vehicle crashes to sev-       sure status). The Regional Ethics
                                            eral factors, including selected clinic-      Committee approved the study. After
Dr. Ameratunga and Dr. Tin Tin are affil-
                                            based study populations, varying out-         complete description of the study,
iated with the Section of Epidemiology      come measures and follow-up peri-             participants gave informed consent.
and Biostatistics, School of Population     ods, and idiosyncratic litigation and            Of the 299 drivers and 96 passen-
Health, Faculty of Medical and Health       compensation schemes (2,3).                   gers who met the study eligibility cri-
Sciences, University of Auckland, Private      The primary aim of this study was          teria, 209 (70%) and 59 (61%), re-
Bag 92019, Auckland Mail Centre, Auck-      to investigate the prevalence of post-        spectively, completed both five- and
land 1142, New Zealand (e-mail: s.amer-     traumatic stress among survivors of           18-month interviews by telephone or,
atunga@auckland.ac.nz). Dr. Coverdale is    serious injury-producing car crashes          if necessary, in person. Compared
with the Menninger Department of Psy-       in Auckland, New Zealand. Sub-                with participants retained in the
chiatry and Center for Medical Ethics,      groups of interest included hospital-         study, those lost to follow-up were
Baylor College of Medicine, Houston,
                                            ized drivers and passengers as well as        younger and more likely to be male,
Texas. Dr. Connor is with the Department
of Social and Preventive Medicine,
                                            nonhospitalized drivers. The state-           to be single, to report hazardous
Dunedin School of Medicine, University      funded “no fault” injury compensa-            drinking habits at recruitment, and to
of Otago, Dunedin, New Zealand. Dr.         tion system covers all residents re-          move residence frequently.
Norton is with The George Institute for     gardless of their age, occupation, lia-          Information collected at recruit-
International Health, University of Syd-    bility, and socioeconomic or insur-           ment (usually within 48 hours of the
ney, New South Wales, Australia.            ance status.                                  crash) included self-reported so-
402                                                     PSYCHIATRIC SERVICES   ♦ ps.psychiatryonline.org ♦ March 2009 Vol. 60 No. 3
ciodemographic data; presence of               Table 1
hazardous drinking patterns, which             Survivors of motor vehicle crashes who had symptoms suggestive of
was defined as a score ≥8 on the Al-           posttraumatic stress disorder (PTSD) at two follow-up points after the crash,
cohol Use Disorders Inventory Test             by survivor characteristica
(5); marijuana or other recreational
drug use; and a history of psychiatric                                                                  Reported                Reported
treatment. Because of diagnostic dif-                                                 Total             symptoms                symptoms
                                                                                      sample            at 5 months             at 18 months
ficulties related to posttraumatic                                                    (N=268)           (N=64)                  (N=33)b
stress disorder (PTSD) and particu-
lar challenges with telephone inter-           Characteristic                         N       %         N       %        p      N       %      p
views, we obtained self-reported in-
formation about the presence or ab-            Status after crash                                                        .8                    .03
                                                 Driver admitted to hospital          161      60      37       24              17      11
sence of posttraumatic stress using              Driver not hospitalized               48      18      11       24               3       7
the Impact of Event Scale (IES) (6)              Hospitalized passenger                59      22      16       28              13      23
at five and 18 months after the crash.         Age                                                                       .02                   .6
The 15 items were scored by using                15–24                                 79      29      18       23               9      12
standard criteria. Possible scores on            25–44                                 91      34      30       35              14      16
                                                 ≥45                                   98      37      16       17              10      11
the IES range from 0 to 75, with               Gender                                                                    .3                    .006
higher scores indicating greater lev-            Male                                 150      56      31       22              11       8
els of intrusive thoughts or avoidance           Female                               118      44      33       28              22      20
symptoms. We used a cutoff score of            Ethnicity                                                                 .9                    .7
27 to define significant posttraumatic           Maori                                 51      19      13       26               7      15
                                                 Non-Maori                            217      81      51       24              26      13
stress based on the findings of a re-          Highest education level                                                   .7                    .8
cent validation study (7). Other vari-           <4-year high school                  112      42      28       26              13      12
ables included in this analysis were             Ηigh school graduate
scores on the Short Form–36 (SF-36)                 or some college                   156      58      36       24              20      13
(8), a two-item case-finding instru-           Living situation                                                          .08                   .7
                                                 Single or unmarried                   97      36      30       33              14      15
ment for depression (9), and a global            Married or living with
health transition indicator that ascer-             a partner                         140      52      29       21              16      12
tained whether participants consid-            Separated, divorced, or
ered their overall health at 18 months           widowed                               31      12        5      17               3      10
to be worse, the same, or better than          Income at baselinec                                                       .6                    .5
                                                 ≥40,000                               85      32      19       23               9      11
before the crash.                                <40,000                              113      42      25       23              17      16
   We used chi square tests to com-              Declined to answer or
pare the prevalence of significant                  don’t know                         69      26      19       29               6      10
stress by sociodemographic charac-             Hazardous drinking                                                        .1                    .4
teristics, alcohol and drug use, and             Yes                                   63      24      20       32               6      10
                                                 No                                   205      76      44       22              27      14
psychiatric history. Pearson product-          Recreational drug use                                                     .2                    .6
moment correlations determined                   Yes                                   50      19      16       32               5      11
correlations between posttraumatic               No                                   218      81      48       23              28      14
stress at 18 months and the presence           Psychiatric history                                                       .1                    .4
of depressive symptoms, the SF-36                Yes                                   66      25      21       32              10      16
                                                 No                                   202      75      43       22              23      12
mental health subscale score, and
participants’ impression of their              a   Data were incomplete for eight and 14 participants at five and 18 months, respectively. A score of
overall health since the crash. All                ≥27 on the Impact of Events Scale was deemed consistent with PTSD (7).
                                               b   Includes 17 patients who had symptoms indicative of PTSD at five months
analyses were conducted using SAS              c   In New Zealand dollars
release 9.1.

The mean±SD overall IES score                  in Table 1, about a fourth of hospi-                  female survivors were significantly
among participants was 16.2±15.9 at            talized passengers and drivers and a                  more likely than male survivors to
five months and 8.7±13.7 at 18                 similar proportion of drivers who                     report an IES score ≥27 at 18
months. Levels of posttraumatic                were not hospitalized after the crash                 months (p=.006).
stress suggestive of PTSD (IES score           reported an IES score ≥27 at five                       At 18 months, an IES score ≥27
≥27) were reported by 64 partici-              months. At 18 months the propor-                      was positively correlated with respon-
pants (25%) at five months and by 33           tion of hospitalized passengers who                   dents’ perception that their overall
(13%) at 18 months (Table 1). Sev-             reported significant levels of stress                 health was worse than their health be-
enteen participants (6%) experi-               remained much the same, whereas                       fore the crash (r=.15, p<.01), with
enced significant levels of stress at          the proportions of drivers with sig-                  lower scores on the SF-36 mental
both five and 18 months. As shown              nificant stress were smaller. Overall,                health subscale (r=.32, p<.001), and
PSYCHIATRIC SERVICES   ♦ ps.psychiatryonline.org ♦ March 2009 Vol. 60 No. 3                                                                        403
with symptoms suggestive of depres-        plore predictors of PTSD. However,              References
sion (r=.19, p=.003). Twelve survivors     being female increased the risk of
                                                                                            1. Ameratunga S, Hijar M, Norton R: Road-
(36%) who had an IES score ≥27             posttraumatic stress, a finding consis-             traffic injuries: confronting disparities to
screened positive at 18 months, and        tent with previous research (10). Dif-              address a global-health problem. Lancet
12 (27%) who screened positive for         ferences by age group were apparent                 367:1533–1540, 2006
depression had an IES score ≥27.           but did not persist at 18 months. Fac-           2. Blaszczynski A, Gordon K, Silove D, et al:
                                           tors that predicted persistent symp-                Psychiatric morbidity following motor vehi-
                                                                                               cle accidents: a review of methodological is-
Discussion and conclusions                 toms of PTSD for a year or more                     sues. Comprehensive Psychiatry 39:111–
This is one of a few studies that used     identified in previous research in-                 121, 1998
a population-based sample of sur-          clude unresolved medical, financial,
                                                                                            3. Kuch K, Cox BJ, Evans RJ: Posttraumatic
vivors to estimate the prevalence of       litigation, or compensation problems;               stress disorder and motor vehicle accidents:
symptoms consistent with PTSD after        peritraumatic dissociation and thought              a multidisciplinary overview. Canadian
motor vehicle crashes. To our knowl-       suppression; a history of psychologi-               Journal of Psychiatry 41:429–434, 1996
edge, it is also the only study from a     cal problems before the crash; alco-             4. Ameratunga S, Norton R, Connor J, et al: A
country with a “no fault” injury com-      hol abuse; and increased levels of                  population-based cohort study of longer-
                                                                                               term changes in health of car drivers in-
pensation scheme, which would have         threat and vulnerability at the time of             volved in serious crashes. Annals of Emer-
reduced the influence of claim-relat-      the crash (10,13,14). The latter may                gency Medicine 48:729–736, 2006
ed concerns (10).                          explain the higher proportion of pas-
                                                                                            5. Babor T, Higgins-Biddle J, Saunders J, et
  The main limitation of our study is      sengers (compared with drivers) re-                 al: AUDIT: The Alcohol Use Disorders
the likelihood that the prevalence of      porting stress symptoms in our study,               Identification Test: Guidelines for Use in
posttraumatic stress was underesti-        although our data were insufficient to              Primary Care. Geneva, World Health Orga-
                                                                                               nization, Department of Mental Health
mated because of the exclusion of          investigate this. Because the accident              and Substance Dependence, 2001
survivors of fatal crashes and passen-     compensation scheme in New
                                                                                            6. Horowitz M, Wilner N, Alvarez W: Impact
gers who were not admitted to the          Zealand covers personal injury for all              of Event Scale: a measure of subjective
hospital when the driver was. Despite      residents, the influence of litigation              stress. Psychosomatic Medicine 41:209–
the low attrition rate compared with       could not be investigated in this                   218, 1979
other studies in this field, psychologi-   study. The differential impact of liti-          7. Coffey S, Gudmundsdottir B, Beck JG, et
cal morbidity experienced by partici-      gation and compensation schemes in                  al: Screening for PTSD in motor vehicle ac-
pants who were interviewed may have        contexts where these operate re-                    cident survivors using the PSS-SR and IES.
                                                                                               Journal of Traumatic Stress 19:119–128,
differed from that experienced by          quires further investigation.                       2006
survivors who were not interviewed.           In a study of 233 survivors fol-
                                                                                            8. Gandek B, Ware J: Health Survey Manual
We assessed symptoms indicative of         lowed up one to two years after a                   and Interpretation Guide. Boston, New
PTSD by use of a standardized self-        motor vehicle crash (13), those with                England Medical Center, Health Institute,
report instrument rather than by a di-     chronic PTSD commonly reported                      1993
agnostic interview. Although the IES       symptoms consistent with major de-               9. Whooley M, Avins A, Miranda J, et al:
does not identify hyperarousal symp-       pression (53%), a mood disorder                     Case-finding instruments for depression:
toms, the psychometric properties of       (62%–68%), or generalized anxiety                   two questions are as good as many. Journal
                                                                                               of General Internal Medicine 12:439–445,
the IES compare well with DSM-             disorder (26%)—levels of comorbid-                  1997
IV–based instruments (11,12).              ity that may reflect higher levels of
                                                                                           10. Mayou R, Ehlers A, Bryant B: Posttraumat-
  A cohort study of 546 participants       psychopathology among treatment-
                                                                                               ic stress disorder after motor vehicle acci-
from the Oxford region found the           seeking survivors (10,13). In our                   dents: 3-year follow-up of a prospective
prevalence of PTSD, as determined          study, which was not restricted to                  longitudinal study. Behaviour Research and
by the Posttraumatic Stress Symp-          survivors seeking follow-up health                  Therapy 40:665–675, 2002
tom Scale, to be 23% at three              care, a third of survivors with signifi-        11. Brewin C: Systematic review of screening
months, 17% at one year, and 11% at        cant levels of posttraumatic stress                 instruments for adults at risk of PTSD.
                                                                                               Journal of Traumatic Stress 18:53–62, 2005
three years (10). Although the Ox-         also reported symptoms suggestive
ford study involved emergency de-          of depression and an appreciable de-            12. Joseph S: Psychometric evaluation of
                                                                                               Horowitz’s Impact of Event Scale: a review.
partment attendees, our study found        terioration in their overall health
                                                                                               Journal of Traumatic Stress 13:101–113,
similar levels of PTSD symptoms            since the crash. Effective strategies               2000
among drivers who were and were            must address potentially preventable
                                                                                           13. Blanchard EB, Hickling EJ, Freidenberg
not hospitalized immediately after         psychological sequelae among hospi-                 BM, et al: Two studies of psychiatric mor-
the crash. This suggests that signifi-     talized and nonhospitalized crash                   bidity among motor vehicle accident sur-
cant levels of psychopathology may         survivors.                                          vivors 1 year after the crash. Behaviour Re-
                                                                                               search and Therapy 42:569–583, 2004
occur among crash survivors who do
not have injuries that require med-        Acknowledgments and disclosures                 14. Beck J, Grant D, Read J, et al: The Impact
ical attention.                            This research was funded by the Health Re-          of Event Scale–Revised: psychometric
                                           search Council of New Zealand.                      properties in a sample of motor vehicle ac-
  Our study was not designed—and                                                               cident survivors. Journal of Anxiety Disor-
did not have sufficient power—to ex-       The authors report no competing interests.          der 22:187–198, 2008

404                                                     PSYCHIATRIC SERVICES    ♦ ps.psychiatryonline.org ♦ March 2009 Vol. 60 No. 3

To top