VOL. 79-B, NO. 3, MAY 1997 385

J. G. Grohs, MD, ResidentF. Gottsauner-Wolf, MD, Associate ProfessorDepartment of Orthopaedics, University of Vienna, Wahringer Gurtel18-20, A-1090 Wien, Austria.

Correspondence should be sent to Dr F. Gottsauner-Wolf.

©1997 British Editorial Society of Bone and Joint Surgery0301-620X/97/37604 $2.00

DETECTION OF ORTHOPAEDIC PROSTHESES AT AIRPORTSECURITY CHECKS

JOSEF G. GROHS, FLORIAN GOTTSAUNER-WOLF

From the University of Vienna, Austria

We studied the detection of joint replacements atairport security checks in relation to their weight, usingtwo types of detector arch. A single-source, unilateraldetector showed different sensitivities for implants ondifferent sides of a test subject.

All implants weighing more than 145 g were detectedby one of the arches. The degree of detection wasdirectly related to the logarithm of the weight of theprosthesis in patients, with a linear correlation(r2 = 0.61). A bilateral arch detected all prosthesesweighing over 195 g.

With their usual sensitivity settings many jointreplacements were detectable; an identification passcontaining the site and weight of such prostheses wouldhelp to avoid the need for body-search procedures.

J Bone Joint Surg [Br] 1997;79-B:385-7.Received 16 January 1997; Accepted 12 February 1997

Patients with implanted joint prostheses often ask whetherthese will produce problems with security checks at air-ports. Each year at Vienna International Airport, manythousands of passengers have re-examinations by hand-helddetectors and body checks because of positive results atdetection arches. Body checks on patients with orthopaedicimplants are often necessary because electronic devicescannot distinguish between these and other metallicsources.

There have been previous reports on the detection ofbraces1 and implants used in the management of trauma.1-4

These suggest that screws, plates, intramedullary nails andsingle joint replacements go undetected, but that large or

bilateral joint implants of stainless steel or weighing over400 g give positive results.

In our department, most hip replacements are of a titani-um (Ti) or cobalt-chromium-molybdenum (Co-Cr-Mo)alloy, and weigh much less than 400 g. Despite this manyof our patients experienced activation of metal detectorarches and the need for individual body searches. Mostproblems were reported from the airports of Vienna, Parisand Antalya (Turkey).

We have therefore studied the detection of orthopaedicjoint replacements in relation to their weight at an inter-national airport with modern security equipment.

PATIENTS AND METHODS

Metal detection. At Vienna International Airport passen-gers walk through metal detector arches after handing overloose metal items such as keys, etc. A positive result leadsto the location of the metal by hand-held detectors, andwhen necessary body searches.

Metals are detected by changes in a magnetic fieldcaused by horizontal movement. Two different types ofdetector arch were tested.

Detector one (Heimann MDT8900, Wiesbaden, Ger-many, 1989 model) had one sender and one receiver panel;levels of detection were indicated by 24 luminous diodes.Passengers activating more than five of these require indi-vidual examination.

Detector two (CEIA 02PN8A, Costruzioni ElettronicheIndustriali Automatismi, Arezzo, Italy, 1994 model) hasseveral magnetic coils giving a uniform magnetic field, andalso has a software filter which can discriminate betweendifferent metals by Fourier transformation. Every positivealarm leads to individual searches.

Both detectors are calibrated according to the directionsof the Austrian Ministry of the Interior: no changes areallowed. This sensitivity setting exceeds the minimum lim-its of the Federal Aviation Administration (USA).Study design. In the first part of the study differentimplants were taped to a volunteer who walked through thearch at constant speed and in both directions. Femoralstems of titanium-aluminium-niobium (Ti-Al-Nb) alloy(Alloclassic; Allopro, Winterthur, Switzerland) in sizes 3 to11 and weighing 125 to 300 g, and implants of vitallium(Co-Cr-Mo alloy) (Howmedica Modular Resection System-

endoprostheses; Howmedica Inc, Limerick, Ireland) weigh-ing 250 to 2060 g, were fixed consecutively at the shoulder,the hip and the knee (Fig. 1).

In the second part of the study 15 patients with jointprostheses walked three times in each direction througheach of the detector arches. Eight had unilateral and fourbilateral cementless hip implants (Alloclassic, Allopro).Each single implant weighed 145 to 215 g (Ti-Al-Nb), or270 g (Ti-Al-Nb with a head of Co-Cr-Mo). The bilateralimplants had individual weights of 255 to 435 g (Ti-Al-Nb)or 535 g (Ti-Al-Nb with a head of Co-Cr-Mo). Two patientshad Duracon knee implants (Howmedica) weighing 430 to570 g and one had a proximal femoral HMRS implant(Howmedica) weighing 1170 g.

The weight of the metal in the prostheses was deter-mined to the nearest 10 g using an electronic balance andthe volume calculated from the specific weight of Ti-Al-Nb(4.6 g/ml) and Co-Cr-Mo (8.5 g/ml). The volume of Co-Cr-Mo alloy implants ranged from 29.4 to 242.4 ml and that ofTi-Al-Nb implants only from 27.2 to 65.2 ml.

Correlations between the level of detection and theweight of the prostheses were calculated for the side andlevel of the implants and the direction of passing throughthe detector arches. Correlations between the level of detec-tion and the volume of the implants were calculated foreach alloy. All calculations were performed using Fig Psoftware (Biosoft, Cambridge, UK).

RESULTS

Implants fixed to a test subject showed different results atthe two detector arches. At arch one (Heimann MDT8900)

detection depended on the weight of the implants, but alsovaried with the side of the body. We found a linear correla-tion between the degree of detection and the logarithm ofthe weight both for prostheses nearer the receiver panel(y = 14.3(logx) – 24.9; r2 = 0.96, n = 30) and for thosenearer the sender panel (y = 13.7(logx) – 29.2; r2 = 0.89,n = 27). Detection began at weights of 125 g and 230 g,respectively, which made a difference of six diodes on thedetector panel for the same weight on opposite sides. Wefound no difference between the implants at shoulder, hipor knee level.

There was also a linear correlation between detection andvolume for implants of Ti-Al-Nb (y = 0.2�–4.5; r2 = 0.98,n = 30) and of Co-Cr-Mo (y = 0.1�+7.1; r2 = 0.94,n = 30). This difference was caused by a higher sensitivityfor small implants of Co-Cr-Mo alloy with volumesbeneath 40 ml.

At detector two (CEIA 02PN8A) all implants weighingmore than 200 g were detected independently of side andlocation. The degree of detection depended only onweight.

Patients with joint replacements also showed weight-dependent detection of implants at detector one (Fig. 2,upper panel); there was a linear correlation with the loga-rithm of the weight (y = 6.4(logx)–8.2; r2 = 0.61, n = 90).All implants of over 145 g were detected, on either side ofthe body. Only the lightest implants failed to trigger thealarm (Fig. 3). The threshold of detection intensity whichwould lead to further examination was calculated to be165 g.

At detector two all implants with weights higher than195 g were detected (Fig. 2, lower panel). Only kneeimplants (two patients 430 and 570 g) and small bilateral

386 J. G. GROHS, F. GOTTSAUNER-WOLF

THE JOURNAL OF BONE AND JOINT SURGERY

Fig. 1

Weights of different cementless prostheses: from left to right, Ti-Al-Nb hip (stem + cup: 135 to 380 g); Co-Cr-Mo-alloy knee (350to 570 g); and Co-Cr-Mo-alloy proximal femoral replacement(with Ti-cup: 780 to 1500 g).

Fig. 2

Detection of prostheses by detector arches related to their weight. Detectorone showed a linear correlation (solid line) between detection and thelogarithm of weight. For detector two the stepwise hatched line is derivedfrom the results for test subjects. Personal search would be performed atimplant weight of over 195 g ( ● Ti-Al-Nb, O Co-Cr-Mo (Duracon, knee),O Co-Cr-Mo (HMRS, proximal femur)).

hip implants (two patients, together 255 to 310 g) failed totrigger the alarm in 61% of attempts.

DISCUSSION

Our study was designed to relate the detection of ortho-paedic prostheses to their weight and volume.

At Vienna International Airport, many small- and medi-um-sized joint replacements were being regularly detected,contrary to previous reports.1-3 Implants such as plates andgamma nails with weights up to 250 g had not been detect-able.4 The differences are partly due to the side of fixationof the implants in the non-uniform magnetic field of detec-tor one.2 Another factor is probably a different sensitivity

setting for non-iron metals; this can vary between differentairports. Modern joint implants contain little iron or aremade of stainless steel and are thus non-magnetic.5 Thedetection of the large HMRS implant6 is in accordance withthat of all other studies.

Prostheses fixed to a test subject were detected moreeasily than implanted prostheses, possibly due to the mask-ing effect of tissue coverage.

The detection of knee implants and bilateral hip implantswas lower (39%) in the newer detector two than in detectorone, possibly because of the filter technique used in detec-tor two, or the relatively small volume of the kneeimplants.

Modern orthopaedic replacements for the hip, knee andshoulder have weights from 133 g up to nearly 3000 g forsegmental prostheses. All except the lightest will be detec-ted at the sensitivity settings at Vienna airport; this isapparently necessary to detect some modern weaponswhich may contain only a small amount of iron. A securitycheck using a detector arch and a hand-held detector forlocalisation would distinguish orthopaedic implants andother metal sources, if an identification pass giving site andweight were provided. This would avoid the need for bodysearching.

The authors gratefully acknowledge the support of the security service ofVienna International Airport.

No benefits in any form have been received or will be received from acommercial party related directly or indirectly to the subject of thisarticle.

REFERENCES1. Suelzer JG. Orthopedic devices and airport metal detectors. J Indiana

State Med Assoc 1973;66:380-2.

2. Pearson WG, Matthews LS. Airport detection of modern orthopaedicimplant metals. Clin Orthop 1992;280:261-2.

3. Evans SC, Ferris BD. Airport metal detectors and orthopedicimplants. Acta Orthop Scand 1993;64:643-4.

4. Van Rhijn LW, Veraart BE. Airport metal detectors and orthopedicimplants: the responses of arch and hand-held devices. Ned TijdschrGeneeskd 1994;138:825-7.

5. Beaupre GS. Airport detection of modern orthopedic metal implants.Clin Orthop 1994;303:291-2.

6. Kotz R, Ritschl P, Trachtenbrodt J. A modular femur tibia recon-struction system. Orthopaedics 1986;9:1639-52.

387DETECTION OF ORTHOPAEDIC PROSTHESES AT AIRPORT SECURITY CHECKS

VOL. 79-B, NO. 3, MAY 1997

Fig. 3

Percentage of prostheses causing an alarm for each type of detector,related to weight: 145 to 190 g represents small hip prostheses(Allopro up to stem size 5), 191 to 400 g large hip or small kneeprostheses and 401 to 1100 g large knee or tumour resection pros-theses (Duracon knee large, HMRS). Individual search would beneeded for hatched groups.