G Model

JINJ-6288; No. of Pages 7

Injury, Int. J. Care Injured xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Injury
journal homepage: www.elsevier.com/locate/injury

The risk of PTSD and depression after an airplane crash and its
potential association with physical injury: A longitudinal study
Juul Gouweloos a,b,*, Ingri L.E. Postma c, Hans te Brake a, Marit Sijbrandij d,e,
Rolf J. Kleber b, J. Carel Goslings c
a
Impact – National Knowledge and Advice Centre for Psychosocial Care Concerning Critical Incidents, Partner in Arq Psychotrauma Expert Group, Nienoord 5,
1112XE Diemen, The Netherlands1
b
Utrecht University, Department of Clinical and Health Psychology, Utrecht, The Netherlands
c
Academic Medical Centre, Trauma Unit Department of Surgery, Amsterdam, The Netherlands
d
VU University Amsterdam, Department of Clinical Psychology, Amsterdam, The Netherlands
e
EMGO Institute for Health and Care Research, Amsterdam, The Netherlands

A R T I C L E I N F O A B S T R A C T

Article history: In 2009, a commercial airplane crashed near Amsterdam. This longitudinal study aims to investigate (1)
Accepted 1 July 2015 the proportion of survivors of the airplane crash showing a probable posttraumatic stress disorders
(PTSD) or depressive disorder, and (2) whether symptoms of PTSD and depression were predicted by
Keywords: trauma characteristics. Identifying these trauma characteristics is crucial for early detection and
Post-traumatic stress disorder (PTSD) treatment. Of the 121 adult survivors, 82 participated in this study 2 months after the crash and
Depression 76 participated 9 months after the crash. Risk for PTSD and depression was measured with the self-report
Injury severity
instruments Trauma Screening Questionnaire and Patient Health Questionnaire-2. Trauma character-
Hospitalisation
istics assessed were Injury Severity Score (ISS), hospitalisation, length of hospital stay, and seating
Airplane crash
Longitudinal study position in the plane. Two months after the crash, 32 participants (of N = 70, 46%) were at risk for PTSD
and 28 (of N = 80, 32%) were at risk for depression. Nine months after the crash, 35 participants (of N = 75,
47%) were at risk for PTSD and 24 (of N = 76, 35%) were at risk for depression. There was a moderate
correlation between length of hospital stay and symptoms of PTSD and depression 9 months after the
crash (r = .33 and r = .45, respectively). There were no differences in seating position between
participants at high risk vs. participants at low risk for PTSD or depression. Mixed design ANOVAs
showed also no association between the course of symptoms of PTSD and depression 2 and 9 months
after the crash and ISS or hospitalisation. This suggests that health care providers need to be aware that
survivors may be at risk for PTSD or depression, regardless of the objective severity of their physical
injuries.
ß 2015 Elsevier Ltd. All rights reserved.

Introduction suffering and functional impairment [2,3]. Studies on PTSD and

depression among air crash survivors are rare. In 1995, Gregg et al.
On 25 February 2009, a Boeing 737-800 crashed near Schiphol [4] found PTSD prevalence of 40% and depression prevalence of
airport, Amsterdam. Most occupants (93%) survived the crash. 33%. PTSD is characterised by involuntary intrusive thoughts of the
Ninety-five percent of survivors were injured [1]. Following such event, avoidance, negative alterations in cognition and mood and
an event, survivors are at risk for developing posttraumatic mental heightened arousal [5]. Acute PTSD may be diagnosed one month
disorders, particularly Posttraumatic Stress Disorder (PTSD), major after the traumatic event; chronic PTSD is diagnosed when
depression and other anxiety disorders, that may cause significant symptoms persist for over 3 months. Depression may be diagnosed
when symptoms of depressed mood and/or loss of interest in life
activities last longer than 2 weeks [5].
Early identification of symptoms of PTSD and depression is
* Corresponding author at: Utrecht University, Department of Clinical and Health
important to prevent a chronic course of PTSD; acute PTSD may be
Psychology, Utrecht, The Netherlands. Tel.: +31 6 26252372.
E-mail address: [email protected] (J. Gouweloos). treated effectively with brief psychotherapy [6]. Identifying the
1
http://www.linkedin.com/company/2851054?trk=prof-0-ovw-curr_pos. risk factors in the acute phase following trauma that predict PTSD

http://dx.doi.org/10.1016/j.injury.2015.07.005
0020–1383/ß 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Gouweloos J, et al. The risk of PTSD and depression after an airplane crash and its potential association
with physical injury: A longitudinal study. Injury (2015), http://dx.doi.org/10.1016/j.injury.2015.07.005
G Model
JINJ-6288; No. of Pages 7

2 J. Gouweloos et al. / Injury, Int. J. Care Injured xxx (2015) xxx–xxx

and depression is crucial to facilitating early identification. and crew) of 12 different nationalities, 9 were fatally wounded
Characteristics of the traumatic event may affect the development [1]. All 126 survivors (including 5 children) were screened and
of symptoms of PTSD and depression following trauma [7,8]. For treated for injuries at the emergency departments of several
instance, trauma severity and proximity to the stressor are hospitals. Demographic data (age, gender and nationality) and
associated with an increased risk for symptoms [7,8]. Both extensive medical data on all survivors were gathered. The regional
concepts refer to the degree to which someone is directly exposed Community Health Service (CHS) conducted a survey to identify
to the traumatic event, for instance by measuring perceived life symptoms of PTSD and depression 2 and 9 months after the crash
threat or physical danger. Physical injury is often also considered a using self-report instruments, administered by telephone. Survi-
possible risk factor [7], however, research on the relationship vors were invited to participate by letter or phone call. Interviews
between physical injury and mental health problems following were conducted in Turkish, Dutch or English.
trauma demonstrates conflicting results. Most studies carried out Fig. 1 provides a flow diagram showing survivors of the crash and
in injured trauma patients investigated survivors of motor vehicle the participants of this study. The inclusion criterion was age above
accidents and the majority of these studies report no significant 14 years. Response rates were 68% at timepoint 1 (at 2 months;
relationship between injury severity and incidence of PTSD and/or n = 82, total adult survivors N = 121) and 63% at timepoint 2 (at
depression [9–11]. However, since most of these studies were 9 months; n = 76). The main reasons given for refusal to participate
conducted in severely injured trauma patients, it remains possible were that the individual had moved on with his or her life, had
that patients with severe injuries may be at higher risk for PTSD already received psychological treatment or did not want to talk
than patients with no or very mild injuries. about their complaints. These reasons also explain why, especially
This study examined two research questions: (1) what was the at timepoint 1, some participants chose to complete only a brief part
proportion of survivors of the February 2009 airplane crash of the study protocol that consisted of 2 items (PHQ-2).
showing a probable posttraumatic stress disorder (PTSD) or
depressive disorder 2 months and 9 months after the crash? Outcome measures
and (2) to what extent were symptoms of PTSD and depression
associated with trauma characteristics (injury severity, hospita- To address our first research question, symptoms of PTSD and
lisation, length of stay in hospital and seating position) among depression were measured. Symptoms of PTSD were measured
survivors of this airplane crash? using the Trauma Screening Questionnaire (TSQ), a ten-item
The study population offered several advantages for research questionnaire developed to enable early identification of individ-
into this relationship. The population was homogeneous with uals at risk for PTSD [12,13]. The TSQ uses a yes/no response format
respect to the type of trauma, as the index trauma was shared by all and asks about symptoms during the past week. It consists of five
participants, and the survivors varied in terms of severity of injury items about re-experiencing and five items about arousal taken
– from not injured to severely injured – and length of stay in a from the DSM IV (Diagnostic and Statistical Manual of Mental
hospital after the crash. Also, we were able to include medical and Disorders, 4th ed.) PTSD criteria [14]; scores range from 0
psychological data of victims. (asymptomatic) to 10. A score 6 was considered to indicate that
The Medical Research Ethics Committee (MREC) of the the individual was at risk for PTSD [15]. The TSQ is considered to
Academic Medical Centre Amsterdam and the regional MREC of identify accurately individuals at risk for a PTSD diagnosis using
Noord Holland gave approval for this study. this threshold, when compared with a ‘‘gold standard’’, clinician-
administered interview; sensitivities of .76–.86 and specificities of
Methods .93–.97 have been reported [12].
Symptoms of depression were measured by the Patient Health
Study population Questionnaire 2 (PHQ-2), a two-item measure that inquires about
the frequency of depressed mood and anhedonia over the past two
On 25 February 2009, a commercial airplane crashed near weeks [16,17]. The PHQ-2 uses a four-option response format (not
Amsterdam in the Netherlands. Of the 135 occupants (passengers at all; several days; more than half the days; nearly every day).

Fig. 1. Flow diagram of survivors and participants.

Please cite this article in press as: Gouweloos J, et al. The risk of PTSD and depression after an airplane crash and its potential association
with physical injury: A longitudinal study. Injury (2015), http://dx.doi.org/10.1016/j.injury.2015.07.005
G Model
JINJ-6288; No. of Pages 7

J. Gouweloos et al. / Injury, Int. J. Care Injured xxx (2015) xxx–xxx 3

Total score ranges from 0 to 6. A cut-off score of 3 was used to the squares of the highest AIS scores for the three most injured
identify those at risk for depression [18]. The PHQ-2 score 3 has body regions (head or neck, face, chest, abdominal or pelvic
been found to agree well with formal diagnosis, sensitivities of .83– contents, extremities or pelvic girdle, and external). The ISS scale
.87 and specificities of .78–.92 have been reported [16,18]. ranges from 1 to 75. To compare survivors with no or minor
To address our second research question, hospitalisation, length injuries and moderate to severe injuries, we used a threshold of ISS
of stay in hospital (LOS), Injury Severity Score (ISS) and seating scores greater than 8 [21]. Higher thresholds may exclude a
position in the airplane were measured. Hospitalisation was substantial number of participants with severe trauma [22].
measured dichotomously and indicated whether a participant had Information about seating position in the plane was provided
been admitted to a hospital after being treated in the emergency by the Dutch Safety Board. We used seating position as a means to
department. Sixty-four victims were hospitalised. LOS was gauge the degree of difficulty for victims to reach safety after the
measured in days. crash, calculated by the distance to the nearest exit. Number of
The ISS is based on the Abbreviated Injury Score (AIS) and has seats and rows survivors had to pass before reaching the nearest
been arguably the most used injury severity measure since its exit were counted. Survivors used the following exits (Fig. 2): two
development in 1974 [19,20]. The ISS is calculated as the sum of emergency exits above the right wing, one emergency exit above

Fig. 2. Seating position with respect to injury severity, risk for depression and risk for PTSD. The left figure shows risk for depression at timepoint 1 or 2. The right figure shows
risk for PTSD at timepoint 1 or 2. The middle figure shows injury severity directly after the crash.

Please cite this article in press as: Gouweloos J, et al. The risk of PTSD and depression after an airplane crash and its potential association
with physical injury: A longitudinal study. Injury (2015), http://dx.doi.org/10.1016/j.injury.2015.07.005
G Model
JINJ-6288; No. of Pages 7

4 J. Gouweloos et al. / Injury, Int. J. Care Injured xxx (2015) xxx–xxx

the left wing, tear on the right sight between rows 7 and 8 difference between the study sample and non-responders in terms
(caused by the crash), and an opening at the rear after row of mean scores on ISS and LOS and group distributions on ISS and
28. The rear section of the plane (from row 29) had broken off and hospitalisation. After the crash 3 adult survivors were not injured
become separated from the main fuselage during impact, thereby (ISS = 0) and 118 were injured (ISS > 0). Of the 121 adult survivors,
creating an additional means of escape for passengers. A score of 45% (n = 54) had an ISS score of 1 (e.g. bruises, lacerations) and 30%
1 was given to each seat a survivor had to pass in his/her row, and (n = 36) were moderately to severely injured e.g. fractures,
to each additional row he/she had to pass before reaching the multiple trauma (ISS score > 8). Of those hospitalised (n = 64),
exit. These scores were then tallied and ranged from 0 (next to 21 stayed at the hospital for longer than 1 week and 3 participants
exit) to 10. stayed more than 1 month.

Analysis Research question 1: Participants showing a probable PTSD or

depressive disorder
Statistics were computed in SPSS Statistics 20, with p-levels of
<.05 taken to indicate statistical significance. To examine the At timepoints 1 and 2, respectively, 32 (of N = 70) and 35 (of
characteristics of the participants and investigate our first research N = 75) participants were considered to be at risk for PTSD
question, we conducted descriptive statistics, independent t-tests (indicated by a score of 6). Mean TSQ scores were 5.2 at timepoint
and Chi-square tests. To test our second research question, 1 (SD = 3.5) and 4.8 at timepoint 2 (SD = 3.5). Of those participants
bivariate correlations (Pearson), mixed design ANOVAs and who completed the TSQ at both timepoints (N = 64), 21 showed a
independent t-tests were conducted. Bivariate correlations were probable PTSD at both moments in time. Furthermore, 28 (of
computed to examine the association between demographic N = 80) participants at timepoint 1 and 24 participants (of N = 76)
variables (age, gender), trauma characteristics (ISS as continuous at timepoint 2 were at risk for depression (indicated by a PHQ-2
variables, LOS, hospitalisation) and symptoms of PTSD and score of 3). Mean PHQ-2 scores were 2.1 at timepoint 1 (SD = 2.0)
depression. This was done separately for timepoint 1 (2 months and 1.8 at timepoint 2 (SD = 1.9). A minority of 9 participants (of
after the crash) and timepoint 2 (9 months after the crash). Among N = 66) showed a probable depression at both timepoints.
those participants who completed the TSQ or PHQ-2 at both Risk for PTSD and depression co-occurred: 18 participants (of
timepoints, mixed design ANOVAs were conducted to examine N = 68) showed both a probable PTSD and probable depression at
whether ISS and hospitalisation (as dichotomous, between- timepoint 1. At timepoint 2 this was the case for 22 participants (of
subjects factors) were related to the course of PTSD symptoms N = 75).
and depressive symptoms at 2 and 9 months after the crash
(within-subject factor with two levels). Independent t-tests were Research question 2: Association of PTSD and depression with trauma
conducted to examine whether number of seats to nearest exit characteristics
differed between participants at high risk of PTSD or depression vs.
participants at low risk. ISS, hospitalisation (being hospitalised or not), age and gender
were not associated with PTSD or depression at either timepoint.
Results At timepoint 2, among those hospitalised, longer LOS correlated
with a higher score on PTSD symptoms (n = 40, r = .33, p = .04) and
Characteristics of participants depressive symptoms (n = 41, r = .45, p = .001). At timepoint 1, LOS
did not significantly correlate with PTSD symptoms (n = 36, r = .12,
Table 1 presents demographic data (gender, age, nationality) for p = .48) or depressive symptoms (n = 41, r = .14, p = .39).
adult survivors and participants who completed the TSQ and/or Tables 2 and 3 show mean TSQ and PHQ-2 scores for
PHQ-2. There were no significant differences in the distribution of participants grouped by injury severity and hospitalisation.
gender, age and nationality between the participants who Regarding TSQ score and PHQ-2 score there was no significant
completed the study protocol and non-responders. interaction between time and ISS (F(1, 62) = 1.14, p = .29 and F(1,
Table 1 includes ISS, hospitalisation and LOS for the population 64) = .47, p = .50, respectively), indicating that the course of PTSD
of adult survivors and the samples at timepoints 1 and 2. At symptoms and depressive symptoms did not differ significantly
timepoint 1 (N = 82) and timepoint 2 (N = 76) there was no between participants with high and low injury severity. We

Table 1
Demographics and physical injury of the participants.

Adult survivors Timepoint 1 Non-responders Timepoint 2 Non-responders

(N = 121) (N = 82) timepoint 1 (N = 76) timepoint 2
(N = 39) (N = 45)

N (%) M (SD) N (%) M (SD) N (%) M (SD) N (%) M (SD) N (%) M (SD)

Male 85 (70) 55 (67) 29 (74) 49 (65) 35 (78)

Female 36 (30) 27 (33) 10 (26) 27 (35) 10 (22)
Age 40.2 (13.2) 40.4 (13.7) 39.7 (12.4) 41.7 (14.1) 37.6 (11.4)
Dutch 60 (50) 47 (57) 13 (33) 45 (59) 15 (33)
Turkisha 46 (38) 28 (34) 18 (46) 26 (34) 20 (45)
Otherb 15 (12) 7 (9) 8 (21) 5 (7) 10 (22)
ISS 0–8 85 (70) 60 (73) 25 (64) 53 (70) 32 (71)
ISS > 8 36 (30) 22 (27) 14 (36) 23 (30) 13 (29)
ISS 6.6 (9.3) 5.4 (6.5) 9.0 (13.1) 6.1 (7.0) 7.4 (12.2)
Hospitalised 64 (53) 42 (50) 22 (56) 41 (54) 23 (51)
LOS 9.9 (15.7) 8.1 (9.8) 13.2 (23.0) 8.3 (9.9) 12.6 (22.6)
Distance to nearest exit 4.7 (2.5) 4.6 (2.5) 4.7 (2.7) 4.6 (2.4) 4.7 (2.8)
a
We compared the distribution of Turkish and Dutch participants vs. non-responders. For other nationalities, groups were too small for chi square analysis.
b
Other nationalities were mostly Iranian, American, English and Syrian.

Please cite this article in press as: Gouweloos J, et al. The risk of PTSD and depression after an airplane crash and its potential association
with physical injury: A longitudinal study. Injury (2015), http://dx.doi.org/10.1016/j.injury.2015.07.005
G Model
JINJ-6288; No. of Pages 7

J. Gouweloos et al. / Injury, Int. J. Care Injured xxx (2015) xxx–xxx 5

Table 2 2 months after the crash. This was the case for 22 (29%)
Mean TSQ score with respect to ISS and hospitalisation.
participants 9 months after the crash. These rates are relatively
N (64)a TSQ Timepoint 1 TSQ Timepoint 2 high, compared to previously reported prevalence of 10%
M (SD) M (SD) [2,23]. Possible explanations are that, firstly, all survivors were
ISS = 0–8 46 5.0 (3.4) 4.3 (3.5)
in close proximity to the event and were unable to escape;
ISS  9 18 5.7 (3.7) 5.8 (3.5) proximity is an important risk factor for mental health problems
Not hospitalised 30 5.0 (3.3) 3.7 (3.5) [7,24]. Although close proximity varies between events, it is
Hospitalised 34 5.4 (3.6) 5.5 (3.3) common in accidents such as an airplane crash. Close proximity
a
Number of participants that completed the TSQ at both timepoints. might explain why Gregg at al. [5] also found prevalence rates of
40% for PTSD and 33% for major depression among survivors of an
air crash in England (in which 47 people died and most of the
Table 3 79 survivors were injured) in the year after the crash. In 1988 Sloan
Mean PHQ-2 score with respect to ISS and hospitalisation.
[25] followed up 32 survivors of a non-fatal charter flight crash
N (66)a PHQ-2 Timepoint 1 PHQ-2 Timepoint 2 and also found initially intense stress that subsided over the
M (SD) M (SD) following months. However, results of other types of accidents,
ISS = 0–8 47 1.8 (2.0) 1.6 (1.8)
such as motor vehicle accidents, contradict this explanation:
ISS  9 19 2.1 (1.6) 2.2 (2.0) although some studies find a high risk of mental health problems
Not hospitalised 31 1.7 (2.0) 1.3 (1.6) [26–29], other do not find elevated rates [30,31].
Hospitalised 35 2.1 (1.8) 2.1 (2.0) A second explanation for the rather high percentage of
a
Number of participants that completed the PHQ-2 at both timepoints. participants showing a probable PTSD or depression relates to
the use of self-report screening instruments. These are known to
overestimate mental health problems compared to structured
also found no significant main effect of time on PTSD symptoms clinical interviews [32]. This explanation cannot in itself explain
(F(1, 62) = .64, p = .43) or depressive symptoms (F(1, 64) = .09, the higher prevalence, as many studies of mental health
p = .76), indicating that participants’ symptoms of PTSD and problems in disaster survivors have used self-report question-
depression did not change between 2 and 9 months after the naires and reported lower prevalence [3,33,34]. It is important
crash. We also found no significant main effect of low and high to note that the TSQ and PHQ-2 questionnaires are considered
injury severity groups on PTSD symptoms (F(1, 62) = 1.53, p = .22) accurate for the early identification of PTSD and depression.
or depressive symptoms (F(1, 64) = .94, p = .34), which means that A third explanation relates to cultural differences. Drogendijk
participants with low and high injury severity did not differ in their et al. [35,36] found that Turkish migrant victims of a disaster
level of PTSD symptoms and depressive symptoms. scored considerably higher than native Dutch victims on instru-
In case of hospitalisation as an independent (between group) ments assessing mental health problems and posttraumatic stress.
variable, there was no significant interaction effect between time To test this explanation we compared Turkish and Dutch
and hospitalisation regarding TSQ and PHQ-2 score (F(1, 62) = 3.83, participants in our sample, but found no group differences in
p = .06 and F(1, 64) = .21, p = .65, respectively). This means that the either TSQ or PHQ-2 score.
course of PTSD symptoms and depressive symptoms did not differ A fourth explanation might be that some survivors have not
between hospitalised participants and not-hospitalised partici- received the mental health care they needed. Survivors can be
pants. There was also no significant main effect of time on PTSD dissatisfied with the support provided after an airplane crash
symptoms (F(1, 62) = 2.43, p = .12) or depressive symptoms (F(1, [37]. The CHS actively sought to identify all survivors with mental
64) = .47, p = .50), and no significant main effect of hospitalisation health problems to help them find local psychosocial care.
on PTSD symptoms (F(1, 62) = 2.14, p = .15) or depressive Nevertheless, this explanation cannot be ruled out.
symptoms (F(1, 64) = 2.34, p = .13), indicating that both time The second research question focused on whether symptoms of
and hospitalisation had no effect on the symptom level of PTSD and PTSD and depression were associated with trauma characteristics.
depression. Injury severity and hospitalisation were not associated with the
The seating distribution of participants at risk for PTSD or course of symptoms of PTSD and depression. Previous studies also
depression is shown in Fig. 2 and Table 1. Visual inspection of Fig. 2 did not find any relations between physical injuries and mental
suggests no relationship between seating position and later being problems [9–11]. A possible explanation is that the subjective
at risk for PTSD or depression at both timepoints. Survivors later experience of the severity of an event may be more important than
assessed as at risk were spread throughout the plane. With respect objective indicators of trauma severity (such as ISS or hospitalisa-
to the number of seats and rows survivors had to pass before tion) [9,29,38]. Interestingly, seating position seemed also not
reaching the nearest exit, independent t-tests showed no differ- related. Those showing a probable PTSD or depression were not
ence at both timepoints between participants at high risk vs. nearer to an exit, where they might have been exposed to the crash
participants at low risk for PTSD (timepoint 1: t(68) = 1.02, for a shorter period of time or might have been less afraid of not
p = .31; timepoint 2: t(73) = 1.40, p = .17) or depression (time- being able to exit. Evidently, they were also not overly represented
point 1: t(78) = 1.74, p = .09; timepoint 2: t(74) = .73, p = .47). at the front of the plane, where the severe and critical injuries
occurred.
Discussion Among those hospitalised, length of stay in a hospital was
significantly correlated with symptoms of PTSD and symptoms of
The first research question focused on the proportion of depression 9 months after the crash. This result is consistent with
survivors of a commercial airplane crash near Amsterdam, in the the findings of Sijbrandij et al. [9], who reported that injury tends
Netherlands, showing a probable PTSD or depressive disorder. We to be associated with late-onset symptoms rather than early
found that 2 months after the crash 32 survivors (46%) were at risk symptoms. In the long term survivors may become functionally
for PTSD and 28 (32%) for depression. Nine months after the crash, impaired and have work or relationship difficulties that may
still 35 survivors (47%) were at risk for PTSD and 24 (35%) for contribute to symptoms of depression and PTSD [39]. They suggest
depression. Risk for PTSD and depression also co-occurred: 18 that survivors may focus on physical recovery first and become
(27%) participants showed both a probable PTSD and depression aware of psychological distress later.

Please cite this article in press as: Gouweloos J, et al. The risk of PTSD and depression after an airplane crash and its potential association
with physical injury: A longitudinal study. Injury (2015), http://dx.doi.org/10.1016/j.injury.2015.07.005
G Model
JINJ-6288; No. of Pages 7

6 J. Gouweloos et al. / Injury, Int. J. Care Injured xxx (2015) xxx–xxx

Limitations References

[1] Postma IL, Winkelhagen J, Bloemers FW, Heetveld MJ, Bijlsma TS, Goslings JC.
This study has a number of limitations. Obviously, sample size February 2009 airplane crash at Amsterdam Schiphol airport: an overview of
was limited, which reduced the statistical power of the study. The injuries and patient distribution. Prehosp Disaster Med 2011;26:299–304.
TSQ and PHQ-2 do not measure the whole spectrum of symptoms [2] Breslau N, Kessler RC, Chilcoat HD, Schultz LR, Davis GC, Andreski P. Trauma
and posttraumatic stress disorder in the community: the Detroit area survey of
of PTSD and depression (e.g. the TSQ specifically does not address trauma. Arch Gen Psychiatry 1998;55:626–32.
avoidance symptoms). This could have resulted in misclassifica- [3] Norris FH, Friedman MJ, Watson PJ, Byrne CM, Diaz E, Kaniasty K. 60,000
tion of individuals in our sample with undetected symptoms of disaster victims speak: Part I. An empirical review of the empirical literature,
1981–2001. Psychiatry 2002;65:207–39.
PTSD or depression. Although the PHQ-2 has proven to accurately [4] Gregg W, Medley I, Fowler-Dixon R, Curran P, Loughrey G, Bell P, et al.
detect and monitor depression over time [18], the TSQ has mainly Psychological consequences of the Kegworth air disaster. Br J Psychiatry
proven accurate for the early identification of PTSD and more 1995;167:812–7.
[5] American Psychiatric Association. Diagnostic and statistical manual of mental
research is needed to assess its accuracy months after the event.
disorders. 5th ed. Arlington, VA: American Psychiatric Publishing; 2013.
There was no information available on pre-existing symptoms of [6] Roberts NP, Kitchiner NJ, Kenardy J, Bisson J. Multiple session early psycho-
PTSD and depression in the population under study, so one cannot logical interventions for the prevention of post-traumatic stress disorder.
assume that PTSD and depression symptoms are ‘new’. Because we Cochrane Database Syst Rev 2009;3:1–43.
[7] Bonanno GA, Brewin CR, Kaniasty K, La Greca AM. Weighing the costs of
started measuring symptoms at 2 months after the crash, we could disaster: consequences, risks, and resilience in individuals, families, and
not identify survivors who suffered from symptoms within the first communities. Psychol Sci Publ Interest 2010;11:1–49.
2 months but recovered naturally before the study started. There [8] Brewin CR, Andrews B, Valentine JD. Meta-analysis of risk factors for post-
traumatic stress disorder in trauma-exposed adults. J Consult Clin Psychol
were also no longer term (>2 years) measurements in order to 2000;68:748–66.
investigate delayed onset PTSD and/or depression. Finally, note [9] Sijbrandij M, Engelhard IM, De Vries GJ, Luitse JS, Carlier IVE, Gersons BP, et al.
that there are many other possible predictors of PTSD and The role of injury and trauma-related variables in the onset and course of
symptoms of posttraumatic stress disorder. J Clin Psychol Med Settings
depression that were not tested in this study (e.g. history of 2013;20:449–55.
mental illness, childhood trauma, sense of death threat, social [10] Fujita G, Nishida Y. Association of objective measures of trauma exposure from
support) [7,8]. motor vehicle accidents and posttraumatic stress symptoms. J Trauma Stress
2008;21:425–9.
Given these limitations we strongly recommend future
[11] Delahanty DL, Raimonde AJ, Spoonster E, Cullado M. Injury severity, prior
research to confirm our findings, using different and larger trauma history, urinary cortisol levels, and acute PTSD in motor vehicle
samples with varying severity of physical injury, to improve our accident victims. J Anxiety Disord 2003;17:149–64.
[12] Brewin CR, Rose S, Andrews B, Green J, Tata P, McEvedy C, et al. Brief screening
understanding of the relationship between proximity to a stressor
instrument for post-traumatic stress disorder. Br J Psychiatry 2002;181:158–
and subjective and objective injury, and the possible influence of 62.
length of hospital stay on mental health. [13] Dekkers AMM, Olff M, Näring GWB. Identifying persons at risk for PTSD after
trauma with TSQ in the Netherlands. Community Ment Health J 2010;46:20–5.
[14] American Psychiatric Association. Diagnostic and statistical manual of mental
Conclusions disorders. 4th ed., text rev. Washington, DC: American Psychiatric Association;
2000.
Mental health risks of airplane accidents have rarely been [15] Brewin CR, Fuchkan N, Huntley Z, Scragg P. Diagnostic accuracy of the Trauma
Screening Questionnaire after the 2005 London bombings. J Trauma Stress
studied. This study showed a risk for PTSD and depression among 2010;23:393–8.
survivors of an airplane crash. Objectively measured physical [16] Kroenke K, Spitzer RL, Williams JB. The Patient Health Questionnaire-2:
injuries and hospitalisation had no association with course of the validity of a two-item depression screener. Med Care 2003;41:1284–92.
[17] Pedersen SS, Denollet J, De Jonge J, Simsek C, Serruys PW, Van Domburg RT.
symptoms of either PTSD or depression. Raising awareness of these Brief depression screening with the PHQ-2 associated with prognosis follow-
results among health care providers is important. Victims’ need for ing percutaneous coronary intervention with paclitaxel-eluting stenting. J Gen
mental health care cannot be related to their often much more Intern Med 2009;24:1037–42.
[18] Löwe B, Kroenke K, Gräfe K. Detecting and monitoring depression with a two-
visible physical needs, so monitoring mental health needs is item questionnaire (PHQ-2). J Psychosom Res 2005;58:163–71.
particularly important, not only during the first days after an [19] Baker SP, O’Neill B, Haddon Jr W, Long WB. The Injury Severity Score: a method
incident, but also over the following weeks and months. Survivors for describing patients with multiple injuries and evaluating emergency care. J
Trauma 1974;14:187–96.
without severe injuries may nevertheless suffer from mental
[20] Rutledge R, Hoyt DB, Eastman AB, Sise MJ, Velky T. Comparison of the Injury
health problems; communication and cooperation between the Severity Score and ICD-9 diagnosis codes as predictors of outcome in injury:
medical health care system and community health services is analysis of 44,032 patients. J Trauma 1997;42:477–87.
[21] Hannan EL, Waller CH, Farrell LS, Rosati C. Elderly trauma inpatients in New
therefore essential to deliver optimal long term care.
York state: 1994–1998. J Trauma 2004;56:1297–304.
[22] Palmer C. Major trauma and the injury severity score – where should we set
Conflict of interest the bar? Annu Proc Assoc Adv Automot Med 2007;51:13–29.
[23] Person C, Tracy M, Galea S. Risk factors for depression after a disaster. J Nerv
Ment Dis 2006;194:659–66.
There a no conflicts of interest. [24] Van den Berg B, Wong A, Van der Velden PG, Boshuizen HC, Grievink L. Disaster
exposure as a risk factor for mental health problems, eighteen months, four
Acknowledgements and ten years post-disaster – a longitudinal study. BMC Psychiatry 2012;12:
147–60.
[25] Sloan P. Post-traumatic stress in survivors of an airplane crash-landing: a
This study was made possible by a collaboration between clinical and exploratory research intervention. J Trauma Stress
Academic Medical Centre in Amsterdam (on behalf of the Medical 1988;1:211–29.
[26] Haagsma JA, Ringburg AN, Van Lieshout EMM, Van Beeck EF, Patka P, Schipper
Research Group Turkish Airlines Crash), VU University Amsterdam, IB, et al. Prevalence rate, predictors and long-term course of probable post-
EMGO Institute for Health and Care Research, Utrecht University traumatic stress disorder after major trauma: a prospective cohort study. BMC
and Impact – National Knowledge and Advice Centre for Psychiatry 2012;12:236–45.
[27] Shih RA, Schell TL, Hambarsoomian K, Belzberg H, Marshall GN. Prevalence of
Psychosocial Care Concerning Critical Incidents. We would like
posttraumatic stress disorder and major depression after trauma center
to thank the participants, who made this research possible. We hospitalization. J Trauma 2010;69:1560–6.
would also like to thank Jolanda ten Brinke and Ietje Boele of the [28] Beck JG, Coffey SF. Assessment and treatment of PTSD after a motor vehicle
regional public health service of Kennemerland and the inter- collision: empirical findings and clinical observations. Prof Psychol Res Pract
2007;38:629–39.
viewers for their great work regarding acquisition of data and care [29] Schnyder U, Moergeli H, Klaghofer R, Buddeberg C. Incidence and prediction of
about the participants. We also thank Niels van der Aa for his posttraumatic stress disorder in severely injured accident victims. Am J
advice about methods and statistical analyses. Psychiatry 2001;158:594–9.

Please cite this article in press as: Gouweloos J, et al. The risk of PTSD and depression after an airplane crash and its potential association
with physical injury: A longitudinal study. Injury (2015), http://dx.doi.org/10.1016/j.injury.2015.07.005
G Model
JINJ-6288; No. of Pages 7

J. Gouweloos et al. / Injury, Int. J. Care Injured xxx (2015) xxx–xxx 7

[30] Malt U. The long-term psychiatric consequences of accidental injury: [35] Drogendijk AN, Van der Velden PG, Gersons BPR, Kleber RJ. Lack of social
a longitudinal study of 107 adults. Br J Psychiatry 1999;156: support among ethnic minorities after a disaster: comparative study. Br J
589–95. Psychiatry 2011;198:317–22.
[31] Zatzick DF, Rivara FP, Nathens AB, Jurkovich GJ, Wang J, Fan MY, et al. A [36] Drogendijk AN, Van der Velden PG, Kleber RJ. Acculturation and post-disaster
nationwide US study of post-traumatic stress after hospitalization for physical mental health problems among affected and non-affected immigrants: a
injury. Psychol Med 2007;37:1469–80. comparative study. J Affect Disord 2012;138:485–9.
[32] Engelhard IM, Van den Hout MA, Weerts J, Arntz A, Hox JJ, McNally RJ. [37] Klerx van Mierlo F. Nasleep van een vliegtuigramp: resultaten van de derde
Deployment-related stress and trauma in Dutch soldiers returning from Iraq enquête onder slachtoffers en nabestaanden van de vliegtuigramp in Faro,
Prospective Study. Br J Psychiatry 2007;191:140–5. Portugal 1992. Aftermath of an airplane crash: results of a third survey among
[33] Drogendijk AN, Dirkzwager AJ, Grievink L, Van Der Velden PG, Marcelissen FG, victims and asurvivors of the airplane crash in Faro, Portugal 1992 Tilburg:
Kleber RJ. The correspondence between persistent self-reported post-trau- INTERVICT, University of Tilburg; 2009.
matic problems and general practitioners’ reports after a major disaster. [38] Brasel KJ, deRoon-Cassini T, Bradley CT. Injury severity and quality of life:
Psychol Med 2007;37:193–202. whose perspective is important? J Trauma 2010;68:263–8.
[34] Van der Velden PG, Yzermans CJ, Grievink L. The Enschede fireworks disaster. [39] Hours M, Chossegros L, Charnay P, Tardy H, Nhac-Vu H, Boisson D, et al.
In: Neria Y, Geleo S, Norris FH, editors. Mental health and disasters. New York: Outcomes one year after a road traffic accident: results from the ESPARR
Cambridge University Press; 2009. p. 94–115. cohort. Accid Anal Prev 2013;50:92–102.

Please cite this article in press as: Gouweloos J, et al. The risk of PTSD and depression after an airplane crash and its potential association
with physical injury: A longitudinal study. Injury (2015), http://dx.doi.org/10.1016/j.injury.2015.07.005