Abstract
Background: Since the September 11, 2001, World Trade Center
terrorist attack, airports worldwide have heightened their security standards
in efforts to discourage terrorist attacks. Patients have become increasingly
concerned about whether their metallic implants will set off airport metal
detectors. The purpose of this study was to assess rates of detection of
various orthopaedic implants by airport detectors with the new security
sensitivities.
Methods: One hundred and twenty-nine volunteers with a total of 149
implants were asked to walk through an M-Scope three-zone metal detector at
two sensitivity settings. Low sensitivity was equivalent to the United States
Transportation Security Administration setting for regular security, and high
sensitivity was equivalent to its standard for high security.
Results: Of the 149 implants in 129 patients who were screened,
eighty-four (56%) were trauma hardware, including intramedullary nails,
plates, screws, and Kirschner wires, and sixty-five (44%) were arthroplasty
implants. Seventy-seven (52%) of the 149 implants were detected by the metal
detector at one or both settings. Multivariate analysis revealed that the type
(p < 0.001), material (p < 0.001), and location (p < 0.001) of the
implant were independent predictors of detection. The overall rate of
detection was 88% for prosthetic replacements compared with 32% for plates,
with the likelihood of detection being fifteen times greater (odds ratio =
15.0, 95% confidence interval = 5.9 to 39.1) for the prosthetic replacements.
All total hip replacements and 90% of the total knee replacements were
detected at the low-sensitivity setting. Intramedullary nails and Kirschner
wires were not detected. The overall detection rate was 67% for implants in
the lower extremity, 17% for those in the upper extremity, and 14% for those
in the spine. The detection rate for implants in the lower extremity was ten
times higher than that for implants in the upper extremity and eleven times
higher than that for implants in the spine.
Conclusions: More than half of all orthopaedic implants may be
detected by metal detectors used at commercial airports. Total joint
prostheses will routinely set off the detector, whereas nails, plates, screws,
and wires are rarely detected. Cobalt-chromium and titanium implants are more
likely to be detected than stainless-steel implants.
Since September 11, 2001, airports in the United States and across the
world have heightened their security standards in efforts to discourage
terrorist attacks1.
Patients with metallic orthopaedic implants have become increasingly concerned
about these implants setting off detectors. They often ask if they require a
physician's note for air travel. Orthopaedic surgeons have had limited data
available to identify which patients must be warned about their implants
causing delays during air travel.
In response to this concern, the American Academy of Orthopaedic Surgeons
issued a statement in 2001 informing physicians that, as a result of higher
levels of security, orthopaedic surgeons should consider writing notes for
patients with metal
implants2. The
Transportation Safety Administration's official statement regarding medical
implants is that all individuals with such implants that set off the detector
will be patted down as an extra screening procedure. Individuals who carry an
identification card signed by a physician can bypass the metal detector and
move directly to the individual
screening3.
At the current time, no trials have been performed in the United States to
analyze the detection of orthopaedic implants under the new national security
guidelines. The goal of this study was to assess the ability of metal
detectors to detect various common orthopaedic implants. The results will not
only aid surgeons in counseling patients regarding their implants but also aid
security agencies by identifying which medical devices commonly set off metal
detectors.
We carried out this study during the clinic hours at our department of
orthopaedic surgery over a one-month period. The study was approved by our
institutional review board. Patients with all types of orthopaedic implants
were invited to participate. Prior to enrollment, informed consent was
obtained, and a brief history was elicited. All patients with a cardiac
pacemaker, those with other metallic implants, and those unable to walk
without assistive devices were excluded. Approximately 55% of patients who
were eligible for this study decided to participate.
Enrolled patients were escorted to the metal detector set up in the
physical therapy area at our clinic. After being asked to remove any metallic
objects from their body or clothing, including watches, earrings, belts, and
shoes, they walked through the metal detector twice at each of two settings.
If the alarm sounded during either of the passes at either of the two
settings, the result was considered positive for that setting.
All patients walked through an M-Scope three-zone metal detector (Fisher
Labs, Los Banos, California), which is one of the metal detectors currently
being used by the Transportation Security Administration (TSA) at airports all
across the United States. The detector was set at two sensitivity settings;
one was equivalent to TSA low-security standards, and the other was equivalent
to TSA increased-security standards. The detector settings were obtained from
manufacturer engineers who are familiar with detector settings at TSA
laboratories, as the TSA does not post their detector settings for security
reasons.
Operative notes for each patient were reviewed in order to determine the
location and type of implant, which were cross-referenced with the patient's
radiographs. No new radiographs were made as part of this study. The
metallurgic composition of the implant was obtained directly from published
specifications and/or direct contact with the manufacturer.
Statistical Methods
Implants were grouped according to type, location, and material
composition. Data were analyzed within groups with use of chi-square and
Fisher exact tests. In addition, stepwise multiple logistic regression
analysis was applied to adjust for covariates among the groups. Odds ratios
and 95% confidence intervals were calculated, and two-tailed values of p <
0.05 were considered significant. Statistical analysis was performed with SPSS
version-12.0 software (SPSS, Chicago, Illinois).
One hundred and twenty-nine patients with a total of 149 implants were
screened. Eighty-four implants (56%) were trauma hardware, including
intramedullary nails, plates, screws, and Kirschner wires, and sixty-five
(44%) were arthroplasty implants (Fig.
1).
Fifty-seven (38%) of the 149 implants were detected at the low-sensitivity
setting, and seventy-seven (52%) were detected at the high-sensitivity
setting. Multivariate analysis revealed that the type of implant, its material
composition, and the location of the implant in the body were all independent
predictors of detection (Table
I).
Of the sixty-five arthroplasty implants, fifty (77%) were detected at the
low-sensitivity setting and fifty-seven (88%), at the high-sensitivity
setting. All twenty-eight total hip prostheses and twenty-seven (90%) of the
thirty total knee replacements were detected. Only two of the four hip
hemiprostheses were detected, and only at the high-sensitivity setting.
Upper-extremity prostheses, such as total shoulder replacements, total wrist
replacements, and radial head replacements, were not detected.
The detection rate was considerably higher for the prostheses than for the
plates. Only seven (14%) of the fifty plates were detected at the
low-sensitivity setting, and sixteen (32%) of the fifty were detected at the
high-sensitivity setting (odds ratio = 15.0, 95% confidence interval = 5.9 to
39.1). Single 3.5-mm stainless-steel fracture-fixation plates were largely
undetected. Of the twenty-two patients with screws, none set off the alarm at
the low-sensitivity setting and four (18%) did so at the high-sensitivity
setting. None of the four Kirschner wires or eight intramedullary nails were
detected at either setting.
There was a significant difference in the detection rates among the
different materials (Fig. 2).
Cobalt-chromium implants were detected most often, with twenty-five (86%) of
the twenty-nine implants detected at the low-sensitivity setting and all
twenty-nine detected at the high-sensitivity setting. Of the fifty-three
titanium implants, twenty-seven (51%) were detected at the low-sensitivity
setting and thirty-two (60%) were detected at the high-sensitivity setting
compared with five (7%) and sixteen (24%) of the sixty-seven stainless-steel
implants, respectively (p < 0.001). Compared with stainless steel,
cobalt-chromium was seventy-three times more likely to be detected (95%
confidence interval = 9.2 to 586) and titanium was five times more likely to
be detected (95% confidence interval = 2.3 to 10.9).
Overall, the detection rates varied according to the extremity in which the
implant was situated. The total detection rate at the high-sensitivity setting
was seventy (67%) of the 105 lower-extremity implants compared with five (17%)
of the upper-extremity implants and two (14%) of the fourteen spine implants.
The likelihood of detection was ten times higher for the lower-extremity
implants than for the upper-extremity implants (odds ratio = 10.0, 95%
confidence interval = 3.4 to 28.4) and eleven times higher than for the spine
implants (odds ratio = 11.0, 95% confidence interval = 2.4 to 52.4).
Since September 11, 2001, patients have become increasingly worried about
their orthopaedic implants potentially causing inconveniences at airport
security checkpoints. Only a few investigators have provided data with which
to counsel patients regarding these concerns. Previous studies of this issue
were published outside of the United States and were performed before
September 11, 2001.
Initial studies demonstrated a general insensitivity of airport detectors
to metal implants. In 1992, Pearson and
Matthews4 found that
most orthopaedic implants, such as plates and screws as well as total hip and
knee replacements, were not identified by metal detectors. Only the
Austin-Moore straight fenestrated endoprosthesis set off a detector. In 1994,
van Rhijn and
Veraart5 concluded
that airport detectors, as a rule, did not detect metal implants.
More recent studies have documented that airport detectors can be set off
by specific orthopaedic implants. In 1997, Grohs and
Gottsauner-Wolf6
found that the detectors identified all implants heavier than 195 g. Basu et
al.7 studied the
ability of an implant to set off metal detectors at low and high-security
settings both in vivo and when strapped to a healthy volunteer. They concluded
that only cannulated hip screws, Austin-Moore prostheses, and more than three
joint replacements in one patient set off metal detectors. In a study from
London, Kamineni et
al.8 found that in
vivo total knee and hip replacements were readily detected, while shoulder and
ankle prostheses were not detected. They found no correlation between body
mass index and the likelihood of detection.
Our study documents that orthopaedic implants, as a whole, are more likely
to be detected than has previously been reported. Ninety percent of total knee
replacements and 100% of total hip replacements in this study were detected,
regardless of whether they were unilateral or bilateral.
We observed that the metallic composition of the implant was an independent
predictor of detection. Cobalt-chromium and titanium implants were detected
more often than were those made of stainless steel. This pattern was
consistent among the different types of implants, with titanium plates being
detected more often than stainless-steel plates and titanium prostheses being
detected more often than stainless-steel prostheses. Cobalt-chromium seems to
be the most commonly detected material, but it was found only in total knee
prostheses. Since none of the plates or screws analyzed in this study were
made of cobalt-chromium, it is not possible to comment on its ability to set
off detectors when it is situated in other parts of the body.
The location of the implant is another independent predictor of detection.
Upper-extremity and spine implants were less likely to be detected than were
lower-extremity implants, regardless of the type or material composition. The
greater size and weight of implants in the lower extremities may account for
this difference. Upper extremities were more likely to have a single implant,
such as a distal radial plate, and those single implants were smaller and of
lower profile than those in the lower extremities.
Increasing the sensitivity setting of the detector increased the rate of
implant detection as a whole from 38% to 52%. Most remarkably, raising the
sensitivity setting doubled the detection rate of plates from 14% to 32%,
increased the detection rate of screws from 0% to 18%, and tripled the
detection of stainless-steel implants from 7% to 24%. Raising the sensitivity
setting resulted in only a modest increase in the detection of prostheses
(from 77% to 88%) and no increase in the detection of nails or wires (0%).
This increase in detector sensitivity is comparable with what is expected when
security at airports is changed from standard to high, as may occur during
holidays, busy travel seasons, times of war, or high terrorist threat.
There were several limitations to this study. Although our sample size was
adequate to provide a diverse group of implants and our numbers allowed us to
make generalizations among different types of implants, a much larger cohort
of patients is required to critically evaluate each implant individually. The
study was also limited to an assessment of the sensitivity of one particular
detector. Despite the fact that the sensitivity settings were comparable with
airport settings, different detectors may have different detection rates.
Finally, the sensitivity of a metal detector, according to the manufacturer,
can be influenced by local magnetic interference, such as that coming from
fluorescent lighting or medical imaging devices. Repeating this study in
different locations may show a difference in detection rates.
On the basis of data from this study, we can make the following
observations regarding orthopaedic implants and airport metal detectors: (1)
total hip and knee prostheses can be expected to be identified by airport
detectors, (2) intramedullary nails and Kirschner wires are unlikely to be
detected, (3) lower-extremity implants are much more likely than
upper-extremity and spine implants to be detected, and (4) cobalt-chromium and
titanium implants are much more likely to be detected than stainless steel.
These observations can aid surgeons in advising patients with orthopaedic
implants who are concerned about metal detection at airport security points.
?
Note: The authors thank Douglas Ayers for his help with the
execution of this project.
What we've learned. Five years after 9/11, travel is tougher than
ever. Here are 46 ways to cope. The Washington Post.
2006Sept10;Travel:
1.1
2006
Pearson WG, Matthews LS. Airport
detection of modern orthopedic implant metals. Clin Orthop Relat
Res. 1992;280:
261-2.280261
1992
[PubMed]
van Rhijn LW, Veraart BE. [Metal
detectors for security checks mostly insensitive for metal implants].
Ned Tijdschr Geneeskd.
1994;138: 825-7.
Dutch.138825
1994
[PubMed]
Grohs JG, Gottsauner-Wolf F. Detection
of orthopaedic prostheses at airport security checks. J Bone Joint Surg
Br. 1997;79:
385-7.79385
1997
[CrossRef]
Basu P, Packer GJ, Himstedt J. Detection
of orthopaedic implants by airport metal detectors. J Bone Joint Surg
Br. 1997;79:
388-9.79388
1997
[CrossRef]
Kamineni S, Legge S, Ware H. Metallic
orthopaedic implants and airport metal detectors. J
Arthroplasty. 2002;17:
62-5.1762
2002
[CrossRef]