You are on page 1of 9

Langenbeck's Archives of Surgery

https://doi.org/10.1007/s00423-020-01884-1

SYSTEMATIC REVIEWS AND META-ANALYSES

Routine versus selective computed tomography in non-traumatic


acute abdominal pain: meta-analysis of randomised trials
Shahab Hajibandeh 1,2 & Mohamed Loutfi 2 & Shahin Hajibandeh 3 & Adel Abulkhir 4 & Sheik Rehman 1 &
Moustafa Mansour 2 & Reza Arsalani Zadeh 4

Received: 8 February 2020 / Accepted: 15 April 2020


# Springer-Verlag GmbH Germany, part of Springer Nature 2020

Abstract
Purpose To compare the routine vs. selective use of computed tomography (CT) in patients presenting with non-traumatic acute
abdominal pain (AAP) to a surgical service.
Methods We conducted a systematic review of literature and meta-analysis of outcomes according to PRISMA statement
standards to compare the routine vs. selective use of CT in adult patients presenting with non-traumatic AAP.
Results Analysis of 722 patients from 4 randomised controlled trials showed no difference between the routine CT and selective
CT groups in terms of proportion of correct diagnoses (OR 1.36,95% CI 0.89, 2.07, P = 0.15), mortality (RD 0.03, 95% CI −
0.08, 0.02, P = 0.27] and length of hospital stay (LOS) [MD − 0.26, 95% CI − 2.07, 1.55, P = 0.78).
Conclusions The routine use of CT does not improve the proportion of correct diagnoses and mortality compared to selective use
of CT in adult patients with non-traumatic AAP. The available evidence regarding the influence of routine CT on LOS may be
subject to type 2 error. These findings, however, may not apply to the elderly patient with AAP and further studies are required.

Keywords Abdominal pain . Emergencies . Tomography . X-ray computed

Introduction younger patients [3–5], contributing to higher morbidity and


mortality in these group of patients [6]. This highlights the
Acute abdominal pain (AAP) is a common surgical emergen- importance of early accurate and thorough investigation of
cy requiring admission to hospital [1, 2]. The initial clinical AAP in all patients, especially the elderly, to reduce morbidity
assessment of patients with AAP allow the surgeon to estab- and mortality.
lish relevant differential diagnoses; however, making an accu- The use of computed tomography (CT) to investigate AAP
rate early diagnosis is often very difficult, necessitating further has increased significantly over the past few years. In some
laboratory and radiological investigations. The accuracy of the countries, a routine early CT is commonly used in the triage of
preliminary diagnosis of the underlying cause for acute abdo- patients presenting with AAP while in other countries, triage
men is less reliable in elderly patients (age > 65) compared to of such patients is based mainly on clinical findings supported
by plain radiography and CT if indicated on clinical grounds.
CT can diagnose a wide range of acute abdominal conditions
* Shahab Hajibandeh [7–12], facilitating early diagnosis and management, which
shahab_hajibandeh@yahoo.com can potentially reduce the risk of morbidity and mortality
and shorten the length of hospital stay. The benefits of routine
1
Department of General Surgery, Glan Clwyd Hospital, use of CT in patients presenting with AAP to reduce hospital
Rhyl, Denbighshire, UK admissions and shorten hospital length of stay (LOS) and its
2
Department of General Surgery, North Manchester General Hospital, impact on patients’ morbidity and mortality have been
Northern Care Alliance NHS Group, Manchester, UK controversial.
3
Department of General Surgery, Sandwell and West Birmingham We performed a systematic review of literature and meta-
Hospitals NHS Trust, Birmingham, UK analysis of outcomes to compare the routine and selective use
4
Department of General Surgery, Royal Bolton Hospital, Bolton, UK of CT in patients presenting with non-traumatic AAP.
Langenbecks Arch Surg

Methods in each of the above databases were adapted accordingly. The


literature search strategy is outlined in Table 1. The biblio-
This systematic review was performed according to an graphic lists of relevant articles and reviews were interrogated
agreed predefined protocol. The review was conducted for further potentially eligible studies. No language restric-
and presented according to Preferred Reporting Items for tions were applied in our search strategies.
Systematic Reviews and Meta-Analyses (PRISMA) state-
ment standards [13]. Study selection

Eligibility criteria The title and abstract of articles identified from the literature
searches were assessed independently by two authors (Shahab
We included RCTs comparing routine vs. selective use of CT in H, ML). The full texts of relevant reports were retrieved, and
adult patients of any gender presenting to hospital due to non- those articles that met the eligibility criteria of our review were
traumatic AAP. All patients who presented with AAP and selected. Any discrepancies in study selection were resolved
underwent CT scan or other imaging modalities while they by discussion between the authors. An independent third au-
were in emergency department or any other wards were con- thor (Shahin H) was consulted in the event of disagreement.
sidered eligible. The patients who had CT scan in emergency
department and were discharged due to normal findings were Data collection
also considered eligible. The exclusion criteria were age <
18 years, pregnancy, acute abdominal trauma, contraindication We created an electronic data extraction spreadsheet which
to iodinated contrast media and need for immediate surgery. was pilot-tested in randomly selected articles and adjusted
accordingly. This included study-related data (first author,
Intervention of interest year of publication, country of origin of the corresponding
author, journal, study design and sample size, and clinical
Routine CT was defined as CT within 24 h of presentation to condition of the study participants), baseline demographic
hospital with AAP regardless of clinical assessment. and clinical information of the study populations (age, gender,
Presentation to hospital was defined as either registration in diagnosis) and primary (correct diagnosis) and secondary
emergency department in case of direct presentation to emer- (mortality, LOS) outcomes. Data collection was performed
gency department or registration in surgical assessment unit in independently by two authors (Shahab H, ML), and disagree-
case of referral from community. ments were resolved by discussion. If no agreement could be
reached, a third author (Shahin H) was consulted.
Comparison of interest
Methodological quality and risk of bias assessment
Selective CT was the comparison of interest, defined as CT at
the discretion of physician based on clinical assessment. Any Two authors (Shahab H, ML) independently assessed the meth-
patient with abdominal pain allocated to this group would odological quality and risk of bias of the included articles using the
have initial plain radiographs or ultrasound, and, if necessary Cochrane tool for assessing the risk of bias of randomized con-
on clinical grounds as recognised by the on-call general sur- trolled trials [14]. The Cochrane tool assesses domains including
geon, CT within or after 24 h, or not at all. selection, performance, detection, attrition, reporting and other
sources of bias and, for each individual domain, classifies studies
Outcome measures into low, unclear and high risk of bias. Disagreements were re-
solved by discussion between the reviewers. If no agreement could
The proportion of correct diagnoses was the primary outcome be reached, a third author acted as an adjudicator.
measure. The final diagnosis established at maximum follow
up was used as reference for correct diagnosis. The secondary Data synthesis and statistical analyses
outcome measures included mortality and LOS.
For dichotomous outcome variables, we calculated the odds ratio
Literature search strategy (OR) as the summary measure. We calculated the risk difference
(RD) when more than half of the studies had zero events in both
Two authors independently (Shahab H, ML) searched the fol- groups. For continuous parameters, we calculated the mean differ-
lowing electronic databases: MEDLINE, EMBASE, ence (MD) between the two groups. Individual patient was used as
CINAHL and the Cochrane Central Register of Controlled the unit of analysis. We used intention-to-treat data from the indi-
Trials (CENTRAL). The last search was run on 06 vidual clinical studies. We used the Review Manager 5.3 software
May 2019. Thesaurus headings, search operators and limits for data synthesis [14]. Random or fixed-effects modelling was
Langenbecks Arch Surg

Table 1 The literature search


strategy Search no. Search strategya

1 Routine: TI,AB,KW
2 Routine near 2 ct: TI,AB,KW
3 Routine near 2 computed tomography: TI,AB,KW
4 Mandatory: TI,AB,KW
5 No. 1 OR no. 2 OR no. 3 OR no. 4
6 Selective: TI,AB,KW
7 Selective near 2 ct: TI,AB,KW
8 Selective near 2 computed tomography: TI,AB,KW
9 No. 6 OR no. 7 OR no. 8
10 MeSH descriptor: [computed tomography scanner, x-ray] explode all trees
11 Computed tomography: TI,AB,KW
12 ct: TI,AB,KW
13 No. 10 OR no. 11 OR no. 12
14 abdom* near 2 pain
15 MeSH descriptor: [acute abdomen] explode all trees
16 MeSH descriptor: [abdominal pain] explode all trees
17 No. 14 OR no. 15 OR no .16
18 No. 5 AND no. 9 AND no. 13 AND no. 17
a
This search strategy was adopted for following databases: MEDLINE, EMBASE, CINAHL and the Cochrane
Central Register of Controlled Trials (CENTRAL)

used as appropriate for analysis. We report the results in a Forest databases, 69 articles were not relevant to the topic of this
plot with 95% confidence intervals (CIs). Heterogeneity among study and were excluded directly. The remaining 7 studies
the studies was assessed using the Cochran Q test (χ2). We quan- were relevant to the topic of this study and, after assessing
tified inconsistency by calculating I2 and interpreted it using the their full texts, 3 more were excluded; two were additional
following guide: 0 to 50% may represent low heterogeneity, 50 to analyses of a previously published RCT, and one was exclud-
75% may represent moderate heterogeneity and 75 to 100% may ed, as the comparison group in the study did not undergo
represent high heterogeneity. selective CT. Therefore, 4 RCT studies [15–18] enrolling
722 patients were selected for this review. Overall, 369 pa-
Sensitivity analyses tients were included in the routine CT group and 353 patients
were included in the selective CT group. The included popu-
To explore potential sources of heterogeneity and assess the ro- lations in the individual studies were comparable by baseline
bustness of our results, additional analyses were conducted for characteristics. The literature search flow chart (PRISMA for-
outcomes that were reported by at least four studies. For each mat), baseline characteristics of the included studies and base-
outcome, we repeated the primary analysis using the random- line characteristics and final diagnoses of the included popu-
effects and fixed-effect model. In addition, we calculated the risk lation are shown in Fig. 1; Tables 2 and 3, respectively.
ratio (RR) and risk difference (RD) for each dichotomous vari-
able. We assessed the effect of each study on the overall effect Description of included studies
size and heterogeneity by repeating the analysis after removing
one study at a time. In order assess the consistency of our results, Lehtimäki 2013 [15]: This was an RCT from Finland that
we removed the studies including patients with right lower quad- included 254 patients presenting to the surgical emergency
rant pain or non-specific abdominal pain only. department with acute abdominal pain of unknown cause.
Patients were randomised to routine CT or selective CT.
Patients in routine CT group underwent abdominopelvic CT
Results imaging within 24 h of presentation to the emergency depart-
ment. Patients in selective CT group underwent initial diag-
Results of the search nostic imaging such as plain radiographs or ultrasound scans
and if necessary on clinical grounds CT scan.
Searches of electronic databases identified 76 articles. Among Sala 2007 [16]: This was an RCT from UK that included
the studies that were identified through search of electronic 198 patients presenting to the on-call general surgeons with a
Langenbecks Arch Surg

abdominal pain who were reviewed by the on-call general


surgeons following self-referral to the emergency department.
Patients were randomised to early routine CT or selective CT.
Patients in routine CT group underwent early abdominal and
pelvic CT scan with oral and intravenous contrast. Patients in
selective CT group underwent initial diagnostic imaging such
as plain radiographs or ultrasound scans and if necessary on
clinical grounds CT scan.
Ng 2002 [18]: This was an RCT from UK that included 118
patients presenting with acute abdominal pain who were ad-
mitted under the care of the surgical team. Patients were
randomised to early routine CT or standard practice including
selective CT. Patients in routine CT group underwent early
abdominopelvic CT imaging within 24 h of presentation to
the emergency department. Patients in selective CT group
underwent initial diagnostic imaging such as plain radio-
graphs or ultrasound scans and if necessary on clinical
grounds CT scan.

Risk of bias in included studies

The summary and results of methodological quality assess-


ment of the 4 RCTs [15–18] are demonstrated graphically in
Fig. 2. All of the included studies were judged to be of low
risk of bias in terms of random sequence generation and allo-
cation concealment. The risk of performance bias was consid-
ered to be high or unclear due to lack of information about
blinding of participants and personnel in all of the included
studies. Three of the included studies [15–17] were judged to
be of unclear risk of detection bias due to lack of information
about blinding of outcome assessment. All of the included
studies were judged to be of low risk of reporting, attrition
or other sources of bias.

Outcome synthesis

Proportion of correct diagnoses

OR analysis Proportion of correct diagnoses was reported in 4


Fig. 1 Study flow diagram
studies [15–18], enrolling 652 patients (Fig. 3). There was no
difference in proportion of correct diagnoses between the two
diagnosis of non-specific acute abdominal pain either follow- groups (OR 1.36, 95% CI 0.89, 2.07, P = 0.15). A low level of
ing self-referral to the emergency department or following heterogeneity among the studies existed (I2 = 21%, P = 0.28).
referral by their general practitioner. Patients were randomised
to early routine CT or standard practice including selective Sensitivity analysis Removing one study at a time did not
CT. Patients in routine CT group underwent early change the direction of the effect size and the overall hetero-
abdominopelvic CT imaging within 24 h of presentation to geneity. The direction of the effect size remained unchanged
the emergency department. Patients in selective CT group when RR or RD were calculated. The use of random-effects or
underwent initial diagnostic imaging such as plain radio- fixed-effect models did not affect the direction of the effect
graphs or ultrasound scans and if necessary on clinical size. Removing the studies that included patients with right
grounds CT scan. lower quadrant pain or non-specific abdominal pain only did
Lee 2007 [17]: This was an RCT from USA that included not change the direction of effect size and overall
152 patients presenting with acute right lower quadrant heterogeneity.
Langenbecks Arch Surg

Table 2 Baseline characteristics of the included studies

Study Journal Design Clinical Intervention Comparison Definition Definition of Sample size Methodological
presentation of routine selective CT quality
CT
Total Routine Selective
CT CT

Ng 2002 BMJ RCT Patients with Routine CT Standard Early CT CT requested 118 55 63 Moderate
acute scan practice within at
abdominal with 24 h of discretion
pain selective admis- of
presenting CT sion physician
to hospital imaging based on
clinical
assess-
ment
Lee 2007 Academic RCT Patients with Routine CT Standard Early CT CT requested 152 72 80 Moderate
emer- acute right scan practice within at
gency lower with 24 h of discretion
medi- quadrant selective admis- of
cine pain CT sion physician
presenting imaging based on
to hospital clinical
assess-
ment
Lehtmiki European RCT Patients with Routine CT Standard Early CT CT requested 254 143 111 Moderate
2013 radiolo- acute scan practice within at
gy abdominal with 24 h of discretion
pain selective admis- of
presenting CT sion physician
to hospital imaging based on
clinical
assess-
ment
Sala 2007 Clinical RCT Patients with Routine CT Standard Early CT CT requested 198 99 99 Moderate
radiolo- acute scan practice within at
gy abdominal with 24 h of discretion
pain selective admis- of
presenting CT sion physician
to hospital imaging based on
clinical
assess-
ment

RCT randomised controlled trial, CT computed tomography

Mortality lower quadrant pain or non-specific abdominal pain only did


not change the direction of effect size and overall
RD analysis Mortality was reported in 4 studies [15–18], en- heterogeneity.
rolling 722 patients (Fig. 3). There was no difference in the
risk of mortality between the two groups (RD − 0.03, 95% CI LOS
− 0.08, 0.02, P = 0.27). A high level of heterogeneity among
the studies existed (I2 = 88%, P < 0.00001). MD analysis LOS was reported in 3 studies [15, 16, 18], en-
rolling 570 patients (Fig. 3). There was no difference in LOS
Sensitivity analysis Removing one study at a time did not between the two groups (MD − 0.26, 95% CI -2.07, 1.55, P =
change the direction of the effect size and the overall hetero- 0.78). A moderate level of heterogeneity among the studies
geneity. The direction of the effect size remained unchanged existed (I2 = 71%, P = 0.03).
when RR or OR were calculated. The use of fixed-effect mod-
el changed the direction of the effect size in favour of routine Sensitivity analysis We did not perform sensitivity analysis for
CT. Removing the studies that included patients with right this outcome because it was reported by fewer than 4 studies.
Langenbecks Arch Surg

Table 3 Baseline demographics and final diagnoses of the included population

Lehtmiki 2013 Sala 2007 Lee 2007 Ng 2002

Routine CT Selective CT Routine CT Selective CT Routine CT Selective CT Routine CT Selective CT

Mean/median age (SD/range) 53 (17) 54 (17) 59 (42–73) 57 (43–75) 35.1 (14.5) 33.3 (12.6) NR NR
Male gender 68/143 53/111 42/99 30/99 24/72 31/80 NR NR
Female gender 75/143 58/111 57/99 69/99 48/72 49/80 NR NR
Final diagnosis:
Acute appendicitis 30/143 20/111 10/99 7/99 39/72 38/80 6/55 3/63
Acute diverticulitis 22/143 14/111 12/99 12/99 NR NR 10/55 12/63
Acute pancreatitis 13/143 10/111 4/99 6/99 NR NR 4/55 4/63
Biliary disease 9/143 10/111 5/99 1/99 NR NR 4/55 7/63
Bowel obstruction 10/143 4/111 8/99 5/99 NR NR 5/55 4/63
Urological disease 6/143 7/111 3/99 6/99 NR NR 1/55 1/63
Non-specific abdominal pain 27/143 33/111 20/99 18/99 NR NR 10/55 8/63
Inflammatory bowel disease 3/143 3/111 NR NR NR NR NR NR
Colitis NR NR 4/99 4/99 NR NR NR NR
Gastroenteritis NR NR 1/99 2/99 NR NR 1/55 2/63
Malignancy 3/143 1/111 8/99 14/99 NR NR 1/55 1/63
Perforated viscus NR NR 0/99 2/99 NR NR 1/55 3/63
Vascular disease 4/143 0/111 1/99 0/99 NR NR 0/55 1/63
Thoracic disease 2/143 3/111 1/99 0/99 NR NR 2/55 2/63
Gynaecological disease NR NR 4/99 2/99 NR NR 1/55 1/63

CT computed tomography, NR not reported

Discussion traumatic AAP, we found no benefits of routine use of CT


compared to selective approach. It has been argued that selec-
We performed a systematic review of literature and meta- tive use of CT, with ultrasonography in all patients and CT
analysis of outcomes to compare routine use of CT vs. selective only after negative or inconclusive ultrasonography, gives the
use of CT in patients presenting with non-traumatic AAP. We highest sensitivity for detecting urgent conditions. With this
identified 4 RCTs enrolling 722 patients. The results of analyses selective CT strategy, it has been shown that only 50% of
suggested that there was no difference in the proportion of patients would require CT [24].
correct diagnoses, mortality and LOS between the routine use The use of CT scan for patients presenting to emergency
and selective use of CT in patients presenting with AAP. department with AAP is variable among different coun-
The use of CT scan in patients with AAP with clinical tries. In some countries such as USA, early CT is com-
suspicion of diverticulitis or appendicitis has been investigat- monly used in the triage of patients presenting with AAP
ed by some authors. In patients with suspected diverticulitis, because admission to General Surgery setting prior to CT
use of CT has been shown to be helpful in detecting compli- scan is not a common practice. On the other hand, in many
cations to assist in therapeutic decisions [19]. In patients with European countries including UK, triage of patients is
suspected acute appendicitis, use of CT has been shown to based mainly on clinical findings supported by plain radi-
reduce rates of negative appendicectomy and lower overall ography and CT scan will be done if indicated on clinical
hospital costs by preventing unnecessary admissions and sur- grounds. The findings of this study may have some impli-
gical explorations [20–22]. We know from available evidence cations in facilitating efficient and safe referral of patients
that CT increases the accuracy of the clinical diagnosis and from emergency department to General Surgery unit with-
diagnostic certainty, facilitating management decisions in pa- out performing early CT in emergency department. Many
tients with AAP. However, whether CT should be routinely causes of AAP such as acute cholecystitis, peptic ulcer
used as the first imaging modality for patients with AAP is disease or acute appendicitis may be diagnosed without a
controversial. need for CT scan; therefore, not performing CT scan in
Routine use of CT in trauma has been shown to reduce emergency department should not necessarily delay refer-
missed injury and mortality rates [23]. In terms of non- ral of a patient to General Surgery team.
Langenbecks Arch Surg

Fig. 2 Risk of bias summary and


graph showing authors’
judgements about each risk of
bias item for randomised trials

The reported outcomes of our review should be viewed and conclusions for these subgroups. The number of eligible stud-
interpreted in the context of inherent limitations. The sample ies for this review was less than 10, not allowing for formal
size of this study was adequate for robust conclusions regard- assessment of publication bias as planned in our protocol. The
ing the correct diagnosis rate and mortality; however, the sam- available data was not adequate to synthesise some clinically
ple size was not adequate for LOS, subjecting our results to important outcomes such as non-therapeutic surgery and
type 2 error for this outcome. One of the included studies missed diagnoses. Moreover, the available data from the in-
included patients with acute right lower quadrant pain and cluded studies was not adequate to determine what proportion
excluded patients with other forms of acute abdominal pain, of patients in the selective group already had a correct diag-
subjecting our results to potential selection bias. The included nosis by other imaging. Finally, the available data did not
studies had some differences in their design in terms of includ- allow for subgroup analysis based on the age of the patients;
ed population (one study included patients with right-sided therefore, we were not able to define a subgroup of patients
abdominal pain and one study included patients with non- who may benefit from routine use of CT.
specific abdominal pain) and time frame for performing CT
scan (one study performed CT scan within 1 h of presentation
and other 3 RCTs within 24 h of presentation). All of these can Conclusions
potentially subject the results of this study to bias. Moreover,
we excluded cases that were pregnant and cases with acute The best available evidence suggests that routine use of CT
abdominal trauma, contraindication to iodinated contrast me- does not improve the proportion of correct diagnoses and
dia and need for immediate surgery without delay; therefore, mortality compared to selective use of CT in adult patients
the results of current study cannot be used for definite with non-traumatic AAP. The available evidence regarding
Langenbecks Arch Surg

a) Correct diagnosis

b) Mortality

c) Length of hospital stay

Fig. 3 Forest plots of the comparisons of outcomes between routine CT death, M-H Mantel-Haenszel methods, random random-effects model,
and selective CT groups. a Correct diagnosis: event = correct diagnosis, weight: the amount of information that each study contributes. c Length
M-H Mantel-Haenszel methods, fixed fixed-effect model, weight: the of hospital stay: IV Inverse variance; random random-effects model;
amount of information that each study contributes. b Mortality: event = weight: the amount of information that each study contributes

the influence of routine CT on LOS may be subject to type 2 References


error. These findings, however, may not apply to the elderly
patient with AAP and further studies are required. 1. Nawar EW, Niska RW, Xu J (2007) National hospital ambulatory
medical care survey: 2005 emergency department summary. Adv
Data 386(386):1–32
Authors’ contributions Study conception and design: Shahab H,
2. Hastings RS, Powers RD (2011) Abdominal pain in the ED: a 35
Acquisition of data: ML, AA, SR, Shahin H, Shahab H. Analysis and
year retrospective. Am J Emerg Med 29(7):711–716
interpretation of data: SR, Shahin H, Shahab H. Drafting of manuscript:
all the authors. Critical revision of manuscript: all the authors. 3. Lewis LM, Banet GA, Blanda M, Hustey FM, Meldon SW, Gerson
LW (2005) Etiology and clinical course of abdominal pain in senior
patients: a prospective, multicenter study. J Gerontol A Biol Sci
Compliance with ethical standards Med Sci 60(8):1071–1076
4. Laurell H, Hansson LE, Gunnarsson U (2006) Acute abdominal
Conflict of interest The authors declare that they have no conflict of pain among elderly patients. Gerontology 52(6):339–344
interest. 5. Kizer KW, Vassar MJ (1998) Emergency department diagnosis of
abdominal disorders in the elderly. Am J Emerg Med 16(4):357–362
Ethical approval Considering the nature of this study, ethical approval 6. Marco CA, Schoenfeld CN, Keyl PM, Menkes ED, Doehring MC
was not required. (1998) Abdominal pain in geriatric emergency patients: variables
associated with adverse outcomes. Acad Emerg Med 5(12):1163–
Human and animal rights This study is a systematic review with meta- 1168
analysis of outcomes which does not include research directly involving 7. Malone AJ, Wolf CR, Malmed AS, Melliere BF (1993) Diagnosis
human or animal participation. of acute appendicitis: value of unenhanced CT. Am J Roentgenol
160:763–766
Informed consent Considering the nature of this study, informed con- 8. Rao PM, Rhea JT, Novelline RA, Dobbins JM, Lawrason JN,
sent was not required. Sacknoff R, Stuk JL (1998) Helical CT with only colonic contrast
material for diagnosing diverticulitis: prospective evaluation of 150
patients. Am J Roentgenol 170:1445–1449
Langenbecks Arch Surg

9. Smith RC, Rosenfield AT, Choe KA, Essenmacher KR, Verga M, 17. Lee CC, Golub R, Singer AJ, Cantu R Jr, Levinson H (2007)
Glickman MG, Lange RC (1995) Acute flank pain: comparison of Routine versus selective abdominal computed tomography scan in
noncontrastenhanced CT and intravenous urography. Radiology the evaluation of right lower quadrant pain: a randomized con-
194:789–794 trolled trial. Acad Emerg Med 14(2):117–122
10. Siewert B, Raptopoulos V (1994) CT of the acute abdomen: find- 18. Ng CS, Watson CJ, Palmer CR, See TC, Beharry NA, Housden BA,
ings and impact on diagnosis and treatment. Am J Roentgenol 163: Bradley JA, Dixon AK (2002) Evaluation of early abdominopelvic
1317–1324 computed tomography in patients with acute abdominal pain of
11. Fukuya T, Hawes DR, Lu CC, Chang PJ, Barloon TJ (1992) CT unknown cause: prospective randomised study. BMJ. 325(7377):
diagnosis of smallbowel obstruction: efficacy in 60 patients. Am J 1387
Roentgenol 158:765–769 19. Pradel JA, Adell J-F, Taourel P, Djafari M, Monnin-Delhom E,
12. Maglinte DDT, Reyes BL, Harmon BH, Kelvin FM, Turner WW Jr, Bruel J (1997) Acute colonic diverticulitis: prospective comparative
Hage JE et al (1996) Reliability and role of plain film radiography evaluation with US and CT. Radiology 205:503–512
and CT in diagnosis of smallbowel obstruction. Am J Roentgenol 20. Walker S, Haun W, Clark J, McMillin K, Zeren F, Gilliland T
167:1451–1455 (2000) The value of limited computed tomography with rectal con-
13. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, trast in the diagnosis of acute appendicitis. Am J Surg 180:450–454
Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D 21. Raman SS, Osuagwu FC, Kadell B, Cryer H, Sayre J, Lu DS (2008)
(2009) The PRISMA statement for reporting systematic reviews Effect of CT on false positive diagnosis of appendicitis and perfo-
and meta-analyses of studies that evaluate healthcare interventions: ration. N Engl J Med 358:972–973
explanation and elaboration. BMJ 339:b2700
22. Krajewski S, Brown J, Phang PT, Raval M, Brown CJ (2011)
14. Higgins JP, Altman DG, editors. Chapter 8: assessing risk of bias in
Impact of computed tomography of the abdomen on clinical out-
included studies. In: Higgins JP, Green S, editors. Cochrane hand-
comes in patients with acute right lower quadrant pain: a meta-
book for systematic reviews of interventions. Version 5.0.1 [up-
analysis. Can J Surg 54(1):43–53
dated September 2008]. Available at: http://hiv.cochrane.org/sites/
23. Hajibandeh S, Hajibandeh S (2015) Systematic review: effect of
hiv. cochrane.org/files/uploads/Ch08_Bias.pdf. Accessed April 15,
whole-body computed tomography on mortality in trauma patients.
2017
J Inj Violence Res 7(2):64–74
15. Lehtimäki T, Juvonen P, Valtonen H, Miettinen P, Paajanen H,
Vanninen R (2013) Impact of routine contrast-enhanced CT on 24. Laméris W, van Randen A, van Es HW, van Heesewijk JP, van
costs and use of hospital resources in patients with acute abdomen. Ramshorst B, Bouma WH et al (2009) Imaging strategies for de-
Results of a randomised clinical trial. Eur Radiol 23(9):2538–2545 tection of urgent conditions in patients with acute abdominal pain:
16. Sala E, Watson CJ, Beadsmoore C, Groot-Wassink T, Fanshawe diagnostic accuracy study. BMJ. 338:b2431
TR, Smith JC et al (2007) A randomized, controlled trial of routine
early abdominal computed tomography in patients presenting with Publisher’s note Springer Nature remains neutral with regard to jurisdic-
non-specific acute abdominal pain. Clin Radiol 62(10):961–969 tional claims in published maps and institutional affiliations.

You might also like