Emerg Radiol (2005) 12: 3033

DOI 10.1007/s10140-005-0437-9
ORIGINAL ARTI CLE
Jonathan H. Valente
.
Thomas Lemke
.
Mark Ridlen
.
Dale Ritter
.
Brian Clyne
.
Steven E. Reinert
Aluminum foreign bodies: do they show up on x-ray?
Received: 1 March 2005 / Accepted: 15 June 2005 / Published online: 2 December 2005
# Am Soc Emergency Radiol 2005
Abstract The objective of this study is to evaluate the
utility of radiographs in the detection of aluminum foreign
bodies (FB). Aluminum can tabs were placed at the upper
esophagus/posterior pharyngeal area in ten randomly
selected cadavers. Anteriorposterior (AP) and lateral
(LAT) radiographies were performed before and after
placement. Twenty sets of randomly ordered radiographs
were assessed by two blinded radiologists for the presence
of radio-opaque FB. For any positive reading on an AP or a
LAT radiograph, the sensitivity, specificity, positive pre-
dictive value (PPV), and negative predictive value (NPV)
for radiologist Awere 80, 90, 89, and 82%, respectively, and
for radiologist B were 90, 100, 100, and 91%, respectively.
These values were also calculated using only AP and LAT
views. AluminumFBcan often be visualized on radiographs.
The sensitivity of this method, however, is not adequate to
completely rule out their presence. Additional testing in
these cases is warranted. Conversely, a high PPV suggests
that therapy based on this finding alone is a logical choice.
Keywords Radiograph
.
Aluminum
.
Foreign body
Introduction
Numerous reports have shown poor visibility of aluminum
foreign bodies (FB) located in the posterior pharynx or in
the esophagus [19]. Many of these are case reports or
reviews of case reports. Some have evaluated the use of
metal detectors to identify metallic FB and have shown this
modality to be very useful [6, 7]. One paper described
aluminum as nonradio-opaque [10]. Aluminum has low
radiodensity, which makes visualization on radiographs
difficult [8]. The atomic numbers of soft tissue and alu-
minum (7.5 and 13, respectively) are similar [1]. Other
investigators, however, have reported adequate visibility of
aluminum on plain radiographs [11].
We sought to evaluate the utility of standard radiography
in the evaluation of aluminum FB located in the upper
esophageal/posterior pharyngeal area using a cadaver
model to simulate typical patient encounters. To our
knowledge, this is the first controlled trial to investigate
this question in a human model.
Materials and methods
Ten cadaver models at the Brown University Anatomy
Laboratory were randomly selected for use in this study.
Prior to placement of FB, standard anteriorposterior (AP)
and lateral (LAT) radiographies of the upper chest and
lower neck were performed on each cadaver using standard
radiographic techniques.
Aluminum can tabs (FB) were placed at the junction of
the upper esophagus/posterior pharyngeal area. Placement
was verified with a fiber-optic scope (Olympus). Repeat
J. H. Valente
.
T. Lemke
.
B. Clyne
Department of Emergency Medicine, Brown Medical School,
Providence, RI, USA
J. H. Valente
Department of Pediatrics, Brown Medical School,
Providence, RI, USA
J. H. Valente (*)
Department of Emergency Medicine and Pediatrics,
Brown Medical School, Rhode Island Hospital,
593 Eddy Street, 2nd Floor, Potter Building,
Providence, RI 02903, USA
e-mail: JValente@Lifespan.org
Tel.: +1-401-4447745
Fax: +1-401-4444307
M. Ridlen
Department of Radiology, Brown Medical School,
Providence, RI, USA
D. Ritter
Department of Ecology and Evolutionary Biology,
Brown Medical School,
Providence, RI, USA
S. E. Reinert
Rhode Island Hospital,
593 Eddy Street,
Providence, RI 02903, USA
standard AP and LAT radiographs, with the FB in place on
each cadaver, were performed. Examples of the radiographs
have been provided (Figs. 1, 2, 3, and 4).
All radiographs were coded, and identification numbers
were concealed. Twenty sets of radiographs (ten with FB
and ten without FB) were arranged in random order after
which two blinded, board-certified radiologists were asked
to assess each set for the presence or absence of radio-
opaque FB. If FB were thought to be present, the
radiologist was asked on which view(s) they could be
visualized (AP, LAT, or both). The radiologists were not
provided with the number of FB placed or the number of
cadavers used in the study.
To measure the detectability of FB, we calculated
sensitivity, specificity, positive predictive value (PPV),
and negative predictive value (NPV) in the assessment of
each of the two readers. We used statistics to gauge the
degree of interrater agreement in the assessment of
positivity or negativity for FB. Statistics were analyzed
with Stata v.8 (Stata Corp., College Station, TX).
Fig. 1 LAT view demonstrates FB (arrows)
Fig. 2 LAT view without FB
Fig. 3 FB not visible on AP view
Fig. 4 AP view without FB
31
Results
There were five male and five female cadavers used in this
study, and the mean age was 79.8 years. There were a total
of 20 sets of radiographs (ten sets with FB and ten sets
without FB). Sensitivity and NPVfor both radiologists
were substantially higher in LAT view vs AP view. PPV
was greater in LAT view vs AP view for radiologist A, but
not for radiologist B, while specificity was similarly
assessed in both views by both interpreters (Table 1).
For any positive reading on an AP or a LAT radiograph,
the sensitivity, specificity, PPV, and NPV for radiologist A
were 80, 90, 88.89, and 81.82%, respectively. For any
positive reading on an AP or a LAT radiograph, the
sensitivity, specificity, PPV, and NPV for radiologist B
were 90, 100, 100, and 90.91% respectively. Sensitivity,
specificity, PPV, and NPV for both radiologists using only
the AP view or only the LAT view were also determined
(Table 1).
Interrater reliability was poor for the AP views (=
0.0714, P=0.6338), near-perfect for the LAT views
(=0.898, P=0.0000), and substantial when both views
were used by the raters (=0.798, P=0.0002).
Discussion
To our knowledge, this is the first randomized human study
designed to assess the ability of plain radiographs to detect
aluminum FB in the posterior pharyngeal/upper esophageal
region. The fact that aluminum has low radiodensity may
not be widely known to practitioners, as Bradburn et al. [9]
reported that this fact was not widely known in a straw
poll among colleagues. As a historical note, in 1973, the
US Mint considered replacing bronze pennies with cheaper
aluminum pennies. This legislation was defeated after
concerns surfaced that it may be difficult to see aluminum
on radiographs [12]. This fact has also been our anecdotal
experience in surveys of other practitioners. We feel that
this paper will fill this gap in the medical literature.
One limitation of our study was the use of cadavers.
Cadavers can have less soft tissue fluid and less air in the
lungs, potentially altering the visualization of FB. Howev-
er, the cadavers used in this study were fresh cadavers
scheduled for use in the medical school anatomy course for
the upcoming year and were relatively normal in terms of
lung air volume and soft tissue fluid density. As expected,
the cadavers were elderly patients. The elderly typically
have decreased bone density. Both of these factors could
have facilitated the radiological location of FB. In addition,
none of the cadavers was extremely obese, as obesity
disqualifies people from being donors. This, too, may have
improved the ability of our radiologists to interpret these
radiographs. Assessments by both radiologists were far
superior when using LAT view vs AP view. Inability to
visualize aluminum FB on the AP view is likely due to both
the low radiodensity of aluminum and the superimposition
of FB over underlying vertebral bodies.
While needing prospective validation, this study sug-
gests that the first step in the evaluation of possible
aluminum FB should be plain film radiographs. If positive,
the clinician can move directly to removal. If negative,
additional studies may need to be performed. An alter-
native to this approach is the use of a metal detector to
assess the presence of metallic FB in medical centers with
access to this modality.
Conclusion
Clearly, aluminum FB can often be visualized on routine
radiographs. The sensitivity of this method, however, is not
adequate to completely rule out the presence of aluminum
FB. Additional testing in these cases is warranted.
Conversely, a high PPV does make this a valuable test,
with therapy based on this finding alone being a rational
choice.
Acknowledgements An Olympus fiber-optic scope was provided for
use by the company for this study. There was no financial
involvement by the company. Brown Medical School and the
Department of Radiology allowed the use of the cadaver laboratory
and radiographic equipment.
References
1. Stewart GD, Lakshmi MV, Jackson A (1994) Aluminum ring
pulls: an invisible foreign body. J Accid Emerg Med 11(3):201
203
2. Eggli KD, Potter BM, Garcia V, Altman RP, Breckbill DL
(1986) Delayed diagnosis of esophageal perforation by alumi-
num foreign bodies. Pediatr Radiol 16(6):511513
Table 1 Diagnostic testing measures (sensitivity, specificity, PPV, and NPV) based on interpretations of the presence or absence (positive or
negative) of aluminum FB in 20 radiograph sets (ten with FB, ten with no FB) of the upper chest and lower neck (95% confidence intervals
in parentheses)
View Reader Sensitivity Specificity PPV NPV
AP A 0 (00) 90 (77100) 0 (00) 47 (2569)
B 20 (238) 100 (100100) 100 (100100) 56 (3477)
LAT A 80 (6298) 100 (100100) 100 (100100) 83 (67100)
B 90 (77100) 100 (100100) 100 (100100) 91 (78100)
AP or LAT A 80 (6298) 90 (77100) 89 (75100) 82 (6599)
B 90 (77100) 100 (100100) 100 (100100) 91 (78100)
32
3. Conradi SE (1982) New aluminum-jacketed ammunition: the
case of the invisible jacket. Am J Forensic Med Pathol 3
(2):153155
4. Burington JD (1976) Aluminum pop tops. A hazard to child
health. JAMA 235(24):26142617
5. Rogers LF, Igini JP (1975) Beverage can pull-tabs. Inadvertent
ingestion or aspiration. JAMA 233(4):345348
6. Ryan J, Perz-Avila CA, Cherukuri A, Tidey B (1995) Using a
metal detector to locate a swallowed ring pull. J Accid Emerg
Med 12(1):6465
7. Arena L, Baker SR (1990) Use of a metal detector to identify
ingested metallic foreign bodies. Am J Roentgenol 155(4):803
804
8. Conners GP (2000) Finding aluminum foreign bodies. Pediatr
Rev 21(5):172
9. Bradburn DM, Carr HF, Renwick I (1994) Radiographs and
aluminum: a pitfall for the unwary. BMJ 308:1226
10. Stack LB, Munter DW (1996) Foreign bodies in the gastroin-
testinal tract. Emerg Med Clin North Am 14(3):493522
11. Ellis GL (1993) Are aluminum foreign bodies detectable
radiographically? Am J Emerg Med 11(1):1213
12. Dorst JP, Reichelderfer TE, Sanders RC (1982) Radiodensity of
the proposed new penny. Pediatrics 69(2):224225
33
Reproducedwith permission of thecopyright owner. Further reproductionprohibited without permission.