Original article

Prehospital ultrasound imaging improves management of

abdominal trauma
F. Walcher1 , M. Weinlich1,3 , G. Conrad2 , U. Schweigkofler4 , R. Breitkreutz5 , T. Kirschning5 and
I. Marzi1
1
Department of Trauma Surgery, Johann Wolfgang Goethe University, Frankfurt/Main, 2 German Air Rescue, Filderstadt, Germany, 3 International
SOS, London, UK and Frankfurt, Germany, 4 Berufsgenossenschaftliche Unfallklinik, Frankfurt/Main and 5 Department of Anesthesiology, Johann
Wolfgang Goethe University, Frankfurt/Main, Germany
Correspondence to: Dr F. Walcher, Department of Trauma Surgery, Johann Wolfgang Goethe University, Theodor-Stern-Kai 7, 60590 Frankfurt/Main,
Germany (e-mail: walcher@trauma.uni-frankfurt.de)

Background: Blunt abdominal trauma with intra-abdominal bleeding is often underdiagnosed or even
overlooked at trauma scenes. The purpose of this prospective, multicentre study was to compare the
accuracy of physical examination and prehospital focused abdominal sonography for trauma (PFAST) to
detect abdominal bleeding.
Methods: Six rescue centres took part in the study from December 2002 to December 2003, including
230 patients with suspected abdominal injury. The accuracy of physical examination at the scene and
PFAST were compared. Later examinations in the emergency department (ultrasonography and/or
computed tomography) were used as the reference standard.
Results: The complete protocol and follow-up was obtained in 202 patients. The sensitivity, specificity
and accuracy of PFAST were 93 per cent, 99 per cent and 99 per cent, respectively, compared with
93 per cent, 52 per cent and 57 per cent for physical examination at the scene. Scanning with PFAST
occurred a mean(s.d.) 35(13) min earlier than ultrasound in the emergency department. Abdominal
bleeding was detected in 14 per cent of patients. Using PFAST led to a change in either prehospital
therapy or management in 30 per cent of patients, and a change to admitting hospital in 22 per cent.
Conclusion: In this study, PFAST was a useful and reliable diagnostic tool when used as part of surgical
triage at the trauma scene.

Paper accepted 8 September 2005

Published online 2 December 2005 in Wiley InterScience (www.bjs.co.uk). DOI: 10.1002/bjs.5213

Introduction of uncontrolled haemorrhage, immediate diagnosis and

urgent laparotomy offers the only chance of survival. This
Over the last few decades, mortality rates and the incidence
raises the question of how to obtain the diagnosis as early
of multiple organ failure among multiply injured patients
as possible in order to make a decision regarding surgical
have decreased owing to improvements in rescue systems,
treatment. In the case of blunt abdominal trauma, no
trauma management and intensive care. However, the
outcome in severely injured patients remains heavily relevant physical signs can be trusted to provide adequate
influenced by initial life support and early surgical care. information regarding the need for surgery. Even patients
Within these factors, time plays a major role, especially with completely normal clinical examination and vital signs
with respect to early management of major blunt abdominal may have abdominal pathology3 . It has been shown that
trauma and perforating truncal injuries1,2 . the assessment of response to resuscitation by sequential
Abdominal and pelvic injuries are major causes of early physiology scores could help to determine the need for
death after severe trauma, hence it is important to focus urgent laparotomy in an abdominal emergency4 .
on their initial assessment and management. In the case Until recently, focused abdominal sonography for
trauma (FAST) performed immediately after admission
The Editors have satisfied themselves that all authors have contributed to the receiving trauma centre has been the standard
significantly to this publication procedure for diagnosing bleeding due to abdominal

Copyright  2005 British Journal of Surgery Society Ltd British Journal of Surgery 2006; 93: 238–242
Published by John Wiley & Sons Ltd
Prehospital ultrasound for abdominal trauma 239

injury5,6 . For more than a decade, it has been known that life support. Following physical examination, preliminary
the presence of free abdominal fluid detected by FAST, diagnosis made by the emergency doctor with respect
in combination with haemodynamic instability, requires to abdominal trauma and the blood pressure and pulse
urgent laparotomy7 – 11 . rate were recorded. The duration and findings of PFAST
Before the use of FAST, management of blunt were also recorded. Changes in prehospital therapy and
abdominal trauma was very challenging during the patient management resulting from the findings of PFAST
prehospital interval, and crucial time may have been lost in were documented on structured questionnaires, including
patients with undiagnosed intra-abdominal bleeding1 . free form answers. The study endpoint was the diagnosis
Recently, it has been demonstrated that the detection of or exclusion of a haemoperitoneum by ultrasound or
abdominal bleeding can be achieved before hospital admis- CT10,11,17,19 , performed once the patient arrived in the
sion using prehospital focused abdominal sonography for emergency department. In addition, the trauma leaders
trauma (PFAST)12,13 . That study was conducted among at the receiving hospital were interviewed by structured
the ground ambulance service in Frankfurt, Germany, questionnaire retrospectively to find out whether they had
where PFAST was shown to improve overall trauma man- modified their preparation for the patient as a result of the
agement. Ultrasound evaluation was exclusively performed PFAST findings, which were communicated to them from
by a single investigator. the scene.
The purpose of the present multicentre study was to All investigators, including the emergency doctors who
evaluate the feasibility of PFAST performed routinely performed PFAST and the radiologists who interpreted
by different emergency doctors and paramedics at the the CTs, were independent and blinded to the results of
trauma scene. Additionally, the study aimed to compare the other tests. Statistical analyses were performed using
the accuracy of PFAST with physical examination, using BiAS for Windows (Epsilon, Frankfort).
ultrasonography and computed tomography (CT) in the
emergency department as the ‘gold standard’.
Results

Materials and methods A total of 230 patients were included in the study with
suspected abdominal trauma; the trial profile is given
Between December 2002 and December 2003, a prospec- in Fig. 1. The causes of injury were as follows: motor
tive, multicentre study was performed involving five air vehicle accidents 37 per cent, pedestrians struck by a
rescue centres in South West Germany (Frankfurt/Main, vehicle 17 per cent, motorcycle accidents 16 per cent, fall
Freiburg, Leonberg, Mannheim, Ochsenfurt) and one from a height 10 per cent, cycling 4 per cent, gunshot
ground ambulance team in Frankfurt/Main. The research or stab wounds 2 per cent and others 14 per cent. The
protocol was approved by the ethics committee of the demographic data are listed in Table 1.
university hospital. On 219 occasions (95 per cent) the rescue team stated
All patients with suspected abdominal trauma were that there was enough time for PFAST to be performed
eligible for inclusion in the study. Prehospital focused without exceeding the intervals of prehospital care. In the
abdominal sonography for trauma was performed prospec- remaining 11 patients (5 per cent), the prehospital time
tively at the scene using standard techniques, as described at the scene was prolonged by up to 4 min, in order
elsewhere3,6,7,9,11 . The main focus was on the detection of to complete the protocol. Some 214 (93 per cent) of the
haemoperitoneum; no specific attempt was made to eval- investigations performed on the scene provided good or
uate individual parenchymal organ pathologies14,15 , bowel acceptable images for establishing a definitive diagnosis.
or mesenteric injury16 . The participating emergency doc- In the remaining 16 (7 per cent), PFAST investigation
tors from the rescue centres included surgeons, internists was incomplete owing to unfavourable circumstances for
and anaesthetists. Doctors and paramedics who were not conducting ultrasound. These included failure to obtain a
familiar with ultrasound imaging received training in the clear image or technical failure, because of bright sunlight
use of PFAST in a 1-day course17,18 . The hand-held ultra- in one case, artifacts due to air emphysema in two patients
TM
sound device used in all cases was the PRIMEDIC with severe thoracic trauma, or gross obesity of four
HandyScan (Metrax GmbH, Rottweil, Germany). This patients. The entire protocol and follow-up was obtained
device with a 3·5 MHz curved array transducer was in 202 patients.
designed specifically for prehospital care. The mean(s.d.) investigation time was 2·4(0·8) min. All
At the trauma scene a primary survey with stabilization scans were performed either at the scene, in the ground
was performed, according to the principles of advanced ambulance or in the helicopter before transport of the

Copyright  2005 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery 2006; 93: 238–242
Published by John Wiley & Sons Ltd
240 F. Walcher, M. Weinlich, G. Conrad, U. Schweigkofler, R. Breitkreutz, T. Kirschning and I. Marzi

Emergency calls
(n = 6823)

Patients with trauma

(blunt and penetrating)
(n = 1573)

Failure of protocol:
PFAST Lost to follow-up
(n = 230) Insufficient documentation
(n = 28)
Entire protocol and follow-up
(n = 202)

Positive Negative
PFAST PFAST
(n = 27) (n = 175)

True positive False positive True negative False negative

(n = 26) (n = 1) (n =173) (n = 2)

Fig. 1 Trial profile of the multicentre study

Table 1 Demographic data and the severity of trauma according patients were managed by emergency physicians and
to the injury severity score20 of all 230 patients paramedics trained in the use of PFAST, and were included
in the study. Primary survey was performed before PFAST.
Age (years)* 35·5(17·9) 3–70
Sex ratio: M : F (%) 152 : 78 (66 : 33)
Twenty-eight patients were excluded because they were
ISS of all patients† 16·2(16·3) lost to follow-up.
ISS of patients with negative FAST† 14·0(13·9) Overall, the incidence of free abdominal blood on
ISS of patients with positive FAST† 30·6(22·8)
ultrasound was 28 of 202 (14 per cent) including 26 true
positives and two false negatives. Therefore PFAST had a
*Values are mean(s.d.) range; †values are mean(s.d.). ISS, injury severity
score; FAST, focused abdominal ultrasonography for trauma. sensitivity, specificity and accuracy of 93 (95 per cent c.i.
76 to 99) per cent, 99 (95 per cent c.i. 97 to 100) per cent
patient. On average, PFAST was performed 35(13) min and 99 (95 per cent c.i. 96 to 100) per cent, respectively.
earlier than ultrasound or CT in the receiving hospital. In 17 patients, positive findings were found on PFAST
As a result of the physical examination at the trauma in only one of three probe locations (right upper quadrant,
scene, the emergency doctors suspected that 110 patients left upper quadrant, retrovesical space); nine patients had
had abdominal trauma with bleeding. In only 26 patients free abdominal fluid in two or three anatomical sites. Free
was this bleeding confirmed and validated by ultrasound abdominal fluid was found mainly in the pouch of Douglas
or CT in the admitting hospital. Thus in 84 patients, the (18 patients), followed by the right upper quadrant (12
suspicion of abdominal bleeding based on clinical signs, patients) and the left upper quadrant (eight patients).
haemodynamic parameters and the mechanism of injury There was no correlation between the location of free
was not confirmed. Therefore, the sensitivity of physical fluid detected by PFAST and the need for subsequent
examination alone for intra-abdominal bleeding was 93 laparotomy.
(95 per cent confidence interval (c.i.) 76 to 99) per cent, In 42 patients (21 per cent), prehospital care at the
the specificity 52 (95 per cent c.i. 44 to 59) per cent and trauma scene was changed because of the findings of
the accuracy 57 (95 per cent c.i. 50 to 64) per cent. PFAST. Mostly fluid resuscitation was modified by
reducing volume replacement in order to reduce blood loss
by permissive hypotension, once significant head injury
Using PFAST at the trauma scene
was excluded. In 61 patients (30 per cent), prehospital
Some 6823 emergency calls were recorded; 1573 patients management was also changed. When intra-abdominal
had blunt or penetrating trauma (Fig. 1). A total of 230 bleeding was found on PFAST, the overall management at

Copyright  2005 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery 2006; 93: 238–242
Published by John Wiley & Sons Ltd
Prehospital ultrasound for abdominal trauma 241

the scene was influenced to avoid any therapy beyond In 95 per cent of the patients, there was enough time to
advanced life support. Additional therapy was either complete the PFAST investigation within the prehospital
omitted or performed more quickly in order to shorten phase; the procedure itself took on average less than
the time to surgery. If PFAST was negative, deliberated 2·5 min. Other studies have reported that the mean(s.d.)
rescue of patients who were trapped could go ahead. time taken to perform FAST in the emergency department
The report from the scene to the receiving trauma centre was 154(13) s10 and 2·6(1·2) min7 .
was supplemented by the findings of PFAST in 105 patients In only 7 per cent of patients was PFAST not completed
(52 per cent), and in 44 patients (22 per cent) the choice owing to suboptimal conditions for imaging. However, in
of receiving hospital was changed. As a result of additional these rare cases it is recommended that the patients are
information, all trauma teams modified their preparations treated as if they had a positive PFAST, because physical
by including an abdominal surgeon and preparing theatre examination alone would have resulted in a correct diagno-
for urgent laparotomy. sis in only half of the patients. The major prerequisites for
Early follow-up of the 28 patients who had intra- success with PFAST are an intensive training programme
abdominal bleeding showed that laparotomy was needed and good equipment. Several hand-held ultrasound units
in 12 patients, including seven who had splenectomy. have been designed for mobile use25,26 .
Fourteen patients had no surgical intervention and two In approximately one third of the patients in this
died before surgical control of the haemorrhage. One of study, the findings of PFAST had an influence on trauma
the two with a false negative PFAST required laparotomy; management at the scene. In the event of intra-abdominal
the other was managed conservatively. bleeding, the prehospital phase was minimized to allow
immediate transport of the patient to hospital; helicopter
Discussion transport was considered27 . In contrast, if the PFAST was
negative, the routine algorithm for trauma care at the scene
In contrast to the low accuracy of physical examination
was followed, including primary and secondary survey in
and haemodynamic measurement, PFAST was highly
accordance with advanced life support.
reliable in the detection of a haemoperitoneum. In this
In the event of a positive PFAST, patients should ideally
study, only two false negative findings and one false
go to an appropriate trauma centre. Clarke et al. showed
positive finding were noted. Intra-abdominal bleeding is
that, for patients with abdominal bleeding, the probability
a dynamic situation and therefore one explanation for
of death increased by approximately 1 per cent for every
the false negative findings might be that the ultrasound
3 min delay in the emergency department1 . Therefore,
examination was performed so soon after the trauma that
the closest appropriate hospital should be chosen and the
haemorrhage due to splenic laceration was not yet apparent
trauma team should be informed. A change in the choice of
enough to give a positive result11 . Therefore, it is suggested
admitting hospital was made in approximately 20 per cent
that PFAST should be repeated every 15 min during the
prehospital period interval if the initial PFAST findings of patients in this study. This may be only a minor
are negative but physical examination is suspicious. advantage in an urban setting, but in rural hospitals with
The location of free abdominal fluid detected during less experience in treating life-threatening injuries, PFAST
PFAST in this study differed from the data of Rozycki21 and gives the surgeons more information and more time to
Hahn22 , who showed a correlation between parenchymal prepare28 . Helicopter transport may be required to bring
organ injury and the appearance of free blood, primarily a patient to an appropriate trauma centre if the regional
located in Morison’s pouch. In this study, most pathological hospital is unable to arrange immediate laparotomy.
findings were found in the pelvis, as also described by The air rescue centres that took part in this multicentre
Nance in children23 . The results may differ because PFAST study now include the information from PFAST in their
was performed on patients found in different positions at standard reports from a trauma scene. Following the results
the scene of the trauma. of this study, one major air rescue provider in Germany (24
The accuracy of PFAST was high; the data correlated helicopters and four fixed wing aircrafts) has incorporated
favourably with studies concerning the accuracy of and established PFAST into its algorithm for trauma
ultrasound imaging under optimal conditions in the management.
emergency department11,18,19 . Even teams who received
1-day training in PFAST reported accurate results. Data Acknowledgements
from the literature suggest that training programmes
provide competence in FAST and are associated with a The authors would like to thank all of our collaborators, in
steep learning curve17,18,24 . particular the emergency physicians and paramedics who

Copyright  2005 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery 2006; 93: 238–242
Published by John Wiley & Sons Ltd
242 F. Walcher, M. Weinlich, G. Conrad, U. Schweigkofler, R. Breitkreutz, T. Kirschning and I. Marzi

took part in this study. Special thanks go to the head of the 14 Brown MA, Casola G, Sirlin CB, Hoyt DB. Importance of
German Air Rescue, Dr Braun. His help and high level of evaluating organ parenchyma during screening abdominal
professionalism was crucial for organizing the multicentre ultrasonography after blunt trauma. J Ultrasound Med 2001;
study. The authors also thank Mrs Weihgold and Dr Peitz 20: 577–583.
15 Poletti PA, Kinkel K, Vermeulen B, Irmay F, Unger PF,
for collection of the relevant data for this study.
Terrier F. Blunt abdominal trauma: should US be used to
detect both free fluid and organ injuries? Radiology 2003; 227:
References 95–103.
16 Stassen NA, Lukan JK, Carillo EH, Spain DA, Richardson
1 Clarke JR, Trooskin SZ, Doshi PJ, Greenwald L, Mode CJ.
JD. Abdominal seat belt marks in the era of focused abdomi-
Time to laparotomy for intra-abdominal bleeding from
nal sonography for trauma. Arch Surg 2002; 137: 718–722.
trauma does affect survival for delays up to 90 minutes. J
17 Shackford SR, Rogers FB, Osler TM, Trabulsy ME,
Trauma 2002; 52: 420–425.
Clauss DW, Vane DW. Focused abdominal sonogram for
2 Bickell WH, Wall MJ Jr, Pepe PE, Martin RR, Ginger VF,
trauma: the learning curve of nonradiologist clinicans in
Allen MK et al. Immediate versus delayed fluid resuscitation
detecting hemoperitoneum. J Trauma 1999; 46: 553–562.
for hypotensive patients with penetrating torso injuries. N
18 Frezza EE, Solis RL, Silich RJ, Spence RK, Martin M.
Engl J Med 1994; 331: 1105–1109.
3 Blaivas M, Sierzenski P, Theodoro D. Significant Competency-based instruction to improve the surgical
hemoperitoneum in blunt trauma victims with normal signs technique and accuracy of the trauma ultrasound. Am
and clinical examination. Am J Emerg Med 2002; 20: Surgeon 1999; 65: 884–888.
218–221. 19 McCarter FD, Luchette FA, Molloy M, Hurst JM, Davis K
4 O’Dair GN, Leaper DJ. Sequential physiology scoring Jr, Johannigman JA et al. Institutional and individual learning
facilitates objective assessment of resuscitation in patients curves for focused abdominal ultrasound for trauma:
with an intraabdominal emergency. Br J Surg 2003; 90: cumulative sum analysis. Ann Surg 2000; 231: 689–700.
1445–1450. 20 Baker SP, O’Neill B, Haddon W Jr, Long WB. The injury
5 McKenney MG, Martin L, Lentz K, Lopez C, Sleemann D, severity score: a method for describing patients with multiple
Aristide G et al. 1,000 consecutive ultrasounds for blunt injuries and evaluating emergency care. J Trauma 1974; 14:
abdominal trauma. J Trauma 1996; 40: 607–610. 187–196.
6 Rozycki GS, Shackford SR. Ultrasound, what every trauma 21 Rozycki GS, Ochsner MG, Feliciano DV, Thomas B,
surgeon should know. J Trauma 1996; 40: 1–4. Boulanger BR, Davis FE et al. Early detection of
7 Boulanger BR, McLellan BA, Brenneman FD, Wherrett L, hemoperitoneum by ultrasound examination of the right
Rizoli SB, Culhane J et al. Emergent abdominal sonography upper quadrant: a multicenter study. J Trauma 1998; 45:
as a screening test in a new diagnostic algorithm for blunt 878–883.
trauma. J Trauma 1996; 40: 867–874. 22 Hahn DD, Offermann SR, Holmes JF. Clinical importance
8 Hoffmann R, Nerlich M, Muggia-Sullam M, Pohlemann T, of intraperitoneal fluid in patients with blunt intra-abdominal
Wippermann B, Regel G et al. Blunt abdominal trauma in injury. Am J Emerg Med 2002; 20: 595–600.
cases of multiple trauma evaluated by ultrasonography: a 23 Nance ML, Mahboubi S, Wickstrom M, Prendergast F,
prospective analysis of 291 patients. J Trauma 1992; 32: Stafford PW. Pattern of abdominal free fluid following
452–458. isolated blunt spleen or liver injury in the pediatric patient. J
9 Kimura A, Otsuka T. Emergency center ultrasonography in Trauma 2002; 52: 85–87.
the evaluation of hemoperitoneum: a prospective study. J 24 Smith RS, Kern SJ, Fry WR, Helmer SD. Institutional
Trauma 1991; 31: 20–23. learning curve of surgeon-performed trauma ultrasound.
10 Wherrett LJ, Boulanger BR, McLellan BA, Brenneman FD, Arch Surg 1998; 133: 530–535.
Rizoli SB, Culhane J et al. Hypotension after blunt 25 Strode CA, Rubal BJ, Gerhardt RT, Bulgrin JR, Boyd SY.
abdominal trauma: the role of emergent abdominal Wireless and satellite transmission of prehospital focused
sonography in surgical triage. J Trauma 1996; 41: 815–820. abdominal sonography for trauma. Prehosp Emerg Care 2003;
11 Rozycki GS, Ballard RB, Feliciano DV, Schmidt JA, 7: 375–379.
Pennington SD. Surgeon-performed ultrasound for the 26 Brooks A, Davies B, Connolly J. Prospective evaluation of
assessment of truncal injuries; lessons learned from 1540 handheld ultrasound in the diagnosis of blunt abdominal
patients. Ann Surg 1998; 228: 557–567. trauma. J R Army Med Corps 2002; 148: 19–21.
12 Walcher F, Kortüm S, Kirschning T, Weihgold N, Marzi I. 27 Frankema SP, Ringburg AN, Steyerberg EW, Edwards MJ,
[Optimized management of polytraumatized patients Schipper IB, van Vugt AB. Beneficial effect of helicopter
by prehospital ultrasound.] Unfallchirurg 2002; 105: medical services on survival of severely injured patients. Br J
986–994. Surg 2004; 91: 1520–1526.
13 Heegaard W, Plummer D, Dries D, Fracscone R, Pippert G, 28 Brammer RD, Bramhall SR, Mirza DF, Mayer AD,
Steel D et al. Ultrasound for the air medical clinician. Air McMaster P, Buckels JA. A 10-year experience of complex
Medical Journal 2004; 23: 20–23. liver trauma. Br J Surg 2002; 89: 1532–1537.

Copyright  2005 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery 2006; 93: 238–242
Published by John Wiley & Sons Ltd