Annals of the ICRP

Published on behalf of the lnternational Commission Prevention of Ac

on Radiological Protection
Undergo
Editor: J. VALENTIN, ICRP, SE-l7l l6 Stockholm. Sweden

International Commission on Radiological Protection 200f-2m5

Chairman: Prof. R. H. Clarke, National Radiological Protection Board, Chilton, Didcot, Oxfordshire,
oxL1 qRQ, UK
Vice-Chairman: Dr. L.-E. Holm, Swedish Radiation Protection Institute (SS1), SE-171 16 Stockholm,
Sweden
Scientific Secretary: Dr. J. Valentin, ICRP, SE-171 16 Stockholm, Swefun; fax: -146 8 729 72gB

Members of the Main Commission of the ICRP

R. Alexakhin, Obninsk, Russia Y. Sasaki, Chiba-shi, Japan D. Beninson, Buenos Aires,

J. D. Boice Jr, Rocktille, MD C. Streffer, Essen, Germany Argentina (Emeritus)
R. Cox, Didcot, UK A. Sugier, Fontenay-aux- Roses, H. J. Dunster, Oxford, UK (Emeritus)
G. J. Dicus, Washington, DC France B. Lindell, Stockholm, Sweden (Emeritus)
A. J. GonzAlez, Vienna, Austria Z. Pan, Beijing, China W. K. Sinclair, Escondido, CA (Emeritus)
F. A. Mettler Jr, Albuquerque, NM B. C. Winkler, Centurion, South Africa L. S. Taylor, Mitchellville, MD (Emeritus)

Subscription Information
Annals of the ICRP is published as I volume of 4 issues per annum
Annual Institutional Subscription Rates 2000: Europe, The CIS and Japan, 526 Dutch Guilders. All other countries
US$267. Associated Personal Subscription Rates are available on request lor those whose institutions are library
subscribers. Sterling prices exclude VAT. Non-VAT registered customers in the European Community will bL
charged the appropriate VAT in addition to the price listed. Prices include postage and insurance and are iubject to
change without notice. (Each report will be published as soon as material is received from the ICRP. so that issues
will not necessarily appear at regular intervals.)
Any enquiries relating to subscriptions should be sent to:
The Americas: Elsevier Science, Customer Support Department- PO Box 945, New York, Ny 10010, USA
[Te1.: (+ l) 212-633-3730(+ 1) 888 4ES-INFO. Fax: (+ l) 212-633-3680. E-mail: [email protected]].
Japan: Elsevier Science Customer Support Department. 9-15 Higashi-Azabu l-chome, Minato-ku, Tokyo 106-0044, Fig. 1. Radiation overdosage to the br
Japan [Tel.: (+3) 5561-5033. Fax: (+3) 5561-5047. E-mail: infor@elserier-cojp].
Asia Pacific (excluding Japaa): Elsrvier Science (Singapore) he Ltd- No. I Tempsek Avenue, 17-01 Millenia Tower.
Singapore 039192. [el.: (+ 65) 43+3727. Fa-x: ( - 65) 337-2:30- E-mail: asiainfo@ elsevier.com.sg].
Rest of the World: Elsevier Science Customer Senice Deporm€nt. PO Box 2ll. l00l AE Amsterdam. The
Netherlands. [Tel.: (+31) 2M85-3757. Far: ( -31) 1]48!gla E-mail: [email protected].

Back Issues
Back issues of all previously published volumes are arailablc drccr from Elserier Science offices.

This journal is indexed/abstracted in: App- Heahh Phy- A,bff-, Biosis Dara-- CABS and SSSA/CISA/ECA/ISMEC.
@TMThe paper used in this publication meets thc mrimimm lquirmrs
of American National Standard lor
klformation Sciences Permanence of Paper for Primod Lihul M,ilcrials. ANSI 239.48-1984.
I
Periodicals postage paid at Rahway. New Jerue,v- -dmdr qf dfrs ICIP IXSSN 01,166453) is published four times a year,
in March, June, September and December- in oc rufumc. @ Ehada Scl.nce Lrd- The Boulevard. Langford Lane, International Comr
Kidlington, Oxford OX5 IGB- Ir.K. The ennual sbmriprinn in 6E USA is $267.
Annals of the ICRP is circulated b1'\lercur-v lolrrmlnrl lid.il, 365 Bl,air Road, Avenel, NJ 07001. USA. POST-
MASTER: Please send address corrections to ,lmc& qfofu fff,P- c'o Cuqoner Services, Elsevier Science Inc., 655
Avenue of the Americas, New Yort- N)' tml0- USA-
 NOI^IVCd!Id
,tq
uollceloJd lecrSolorpe5 uo uorssruruoJ l?uorleuJolul eql
UOd CI:IHSI-IgNd
NIINIITVA'f
rollpg
,(dereq; uo]lelp€U EuroE;epufl sluerl€d
o] sernsodxg Ielueplccv Jo uorluenoJd
98 NOIJVJITflNd dUJI
duSl oL,ll +o sleuuv
UK Elsevier Science Ltd, The Boulevard, Langford Lane,
Kidlington, Oxford OX5 1GB, UK

USA Elsevier Science Inc., 660 White Plains Road,

Tarrytown, New York 10591-5153, USA PERGAMON
JAPAN Elsevier Science Japan, Tsunashima Building Annex,
3-20-12 Yushima, Bunkyo-ku, Tokyo 113, Japan

Copyright O 2001 ICRP published by Elsevier Science Ltd

All rights reserved. IT HAS TC
The International Commission on Radiological
Protection encourages the publication of translations
of this report. No part of this publication may be
reproduced, stored in a retrietal syslem or tlansmitted
in any form or by any means electronic, electrostatic,
It's a question that we all face
magnetic tape, mechanical photocopying, recording or fessionally: how can we achieve s

otherwise or republished in any form, v)ithout permission And it's tempting, but rather f
in writing from the copyright owner. late as 'by coercion'.
Of course, a few strict rules arc
Third edition 2000 licensing, inspection, and at time
ISBN 008 0440827 parents, trying to prevent accidel
rssN 0146-6453
alternative to explicit bans couplr
Published quarterly (March, June, September, December) execution of such punishments).
September 2000 issue Unfortunately, this approach i
all sensible parents will want to i
No responsibility is assumed by the Publisher or the dren's minds, such that safety is
ICRP for any injury and/or damage to persons or siderations, not the result of rules
property as a matter ofproducts liability, negligence,
or otherwise, or from any use or operation of any we avoid illegal and hazardous t
methods, products, instructions, or ideas contained in round the corner catching us, bu
the material herein. The recommendations and advice of Applying this parable to radii
the ICRP reflect understanding and evaluation of the
current scientifio evidence as given in this report. If and having 'sumcient numbers of in
when lurther relevant information becomes available, the unreliable operators on a short le
ICRP will review its recommendations. Because of rapid desire by operators to run a safe
advances in the medical sciences, in particular, diagnoses
and drug doses should be independently verified. Although In such a scenario, a primary
all advertising material is expected to conform to ethical viding as much information as 1
(medical) standards, inclusion in this publication does not one of the important messages in
constitute a guarantee or endorsement of the quality or
value of such product or of the claims made by its tem with feedback to operators
manufacturer. dents.
For such a system to work, mu
be able to trust operators. In a st
prove that they have the knowler
an internal audit mechanism) req
issued is an indication of trust frr
way around: operators must be a
One aspect is that all operators
licenses to trustworthy operatorr
disadvantaged. Another aspect, I
Typeset by Variorum Publishing Ltd, Rugby must be confident that the regular
Printed and bound in Great Britain by Polestar Wheatons Ltd,
Exeter
JAICRP 157p Disk used No. pages 1, DTD=4.2.0
Version 7

Prevention of Accidental Exposures to Patients

Undergoing Radiation Therapy

Fig. 1. Radiation overdosage to the brain, as a result of which this child lost his ability to speak and walk
and has permanent epilation.

A publication from the

International Commission on Radiological Protection.Fig. 1

i
 ICRP Publication 86

Editorial
IT HAS TO COME FROM WITHIN

It’s a question that we all face now and then, in our private lives as well as pro-
fessionally: how can we achieve safety?
And it’s tempting, but rather futile, to come up with replies that essentially trans-
late as ‘by coercion’.
Of course, a few strict rules are a necessity; there must be enforcement in terms of
licensing, inspection, and at times penalties for non-compliance. Exasperated young
parents, trying to prevent accidents to their small children, may feel that there is no
alternative to explicit bans coupled with threats of dire punishment (and, if need be,
execution of such punishments).
Unfortunately, this approach is not very efficient, and as the children grow older,
all sensible parents will want to instil a sense of personal responsibility in their chil-
dren’s minds, such that safety is the result of the children’s own wishes and con-
siderations, not the result of rules set by parents or of fear of punishment. As adults,
we avoid illegal and hazardous behaviour, not because there might be a policeman
round the corner catching us, but because it actually makes sense to behave safely.
Applying this parable to radiation safety, we can never achieve true safety by
having ‘sufficient numbers of inspectors’ or ‘sufficiently severe penalties’ to keep
unreliable operators on a short leash. Safety must come from within, from a genuine
desire by operators to run a safe operation.
In such a scenario, a primary task of the regulator’s is to help licensees by pro-
viding as much information as possible about the prevention of accidents. This is
one of the important messages in the present publication: an incident reporting sys-
tem with feedback to operators is vital in order to prevent the recurrence of acci-
dents.
For such a system to work, mutual trust is required. Obviously, regulators need to
be able to trust operators. In a sense, that’s what licensing is about: operators must
prove that they have the knowledge, the organisation, and the resources (including
an internal audit mechanism) required to run a safe operation, and the license once
issued is an indication of trust from the regulator. But it also has to work the other
way around: operators must be able to trust regulators.
One aspect is that all operators need to be sure that the regulator will only award
licenses to trustworthy operators - otherwise, serious operators would be unfairly
disadvantaged. Another aspect, perhaps less obvious at first sight, is that operators
must be confident that the regulator really is there primarily to achieve safety, not to
1
 ICRP Publication 86

find and punish culprits. This is particularly evident in the case of incident reporting.
If the inevitable result of incident reporting is that the operator is punished, then few
operators will be willing to come forward with their problems, and in the end safety
will suffer for lack of information transfer.
Probably the best proof of this comes from aviation, where there is a long history
of incident reporting. In aviation, operators and their staff are trained to volunteer
information, not only about accidents, but also about unsafe situations that could
have developed into an accident but did not go that far. They provide this incident
information, convinced that the authorities will not want to press charges or punish
them other than in cases of blatant and deliberate disregard for rules and proce-
dures. The important point here is that the system of mutual trust has generated an
unsurpassed level of safety that makes aviation the safest mode of transport by far.
Safety experts sometimes speak of four stages of safety awareness. Operators at
the lowest stage simply try to evade authorities and rules. The next stage is when the
operator complies with regulations, but tries to minimise the short-term efforts and
costs. The third stage is when the operator is willing to spend any amount of effort
and money to achieve safety, but requests the regulator to tell them exactly what to
do. The highest level of safety awareness is when the operator is not only willing to
invest whatever it takes, but also actively participates in furthering the safety objec-
tive, and is capable of conducting discussions with peers and regulators on an equal
footing in order to further improve safety.
At least some of the accidents described in the present report would probably not
have happened, or at least could have had less devastating consequences, if all
operators had reached the highest stage of safety awareness. If this report can con-
tribute to increasing safety awareness, then it will have achieved its objective.

Jack Valentin

2
 CONTENTS

PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

MAIN POINTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . ... .. ... .. ... .. ... .. .. 11

Causes of accidental exposure in radiotherapy . ... .. ... .. ... .. ... .. .. 13
Summary of recommendations . . . . . . . . . . . . . ... .. ... .. ... .. ... .. .. 15
The future . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... .. ... .. ... .. ... .. .. 16

1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.1. The potential for an accidental exposure involving radiotherapy patients 17
1.2. Criteria for accidental exposure with radiotherapy patients. . . . . . . . . . . 19
1.3. Clinical detectability of differences in dose prescription and delivery . . . . 21
1.4. Other events of interest to the prevention of accidental exposure . . . . . . 22

2. CASE HISTORIES OF MAJOR ACCIDENTAL EXPOSURES IN

RADIOTHERAPY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1. Incorrect 60Co decay chart and lack of verification (USA, 1974-76) . . . . 23
2.2. Lack of acceptance procedures for a treatment planning system
(UK, 1982-1990) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3. Six accidental exposures involving software problems in several accelerators
of the same type (Canada and USA, 1985-87) . . . . . . . . . . . . . . . . . . . . . . . 24
2.4. Computer file not updated for 60Co source change (USA, 1987-88). . . . . 26
2.5. Incorrect repair followed by lack of communication (Spain, 1990) . . . . . 26
2.6. Malfunction of brachytherapy high dose rate equipment (USA, 1992) . . 27
2.7. Beam miscalibration following the exchange of a 60Co source
(Costa Rica, 1996) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

3. CLINICAL CONSEQUENCES OF ACCIDENTAL EXPOSURE IN

RADIOTHERAPY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.1. Side effects and complications – normal therapy versus accidental exposure 31
3.2. Consequences of accidental exposure in radiotherapy . . . . . . . . . . . . . . . 31
3.3. Impact of individual radiosensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.4. Clinical detection of accidental exposures in radiotherapy . . . . . . . . . . . 35

4. CAUSES OF AND FACTORS CONTRIBUTING TO ACCIDENTAL

EXPOSURES IN RADIOTHERAPY . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.1. External beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.2. Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.3. Accidents involving public exposure and environment contamination . . . 43
4.4. Generic lessons learned . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3
 ICRP Publication 86

5. RECOMMENDATIONS FOR THE PREVENTION OF ACCIDENTAL

EXPOSURES IN RADIOTHERAPY . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.1. Structural organisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.2. Education and training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.3. Acceptance testing and commissioning of equipment . . . . . . . . . . . . . . . 50
5.4. Follow-up of equipment faults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.5. Communication. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.6. Patient identification and patient chart. . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.7. External beam radiotherapy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.8. Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.9. The potential for accidental exposures in the future . . . . . . . . . . . . . . . . 55

APPENDIX A: UNCERTAINTY IN RADIOTHERAPY . . . . . . . . . . . . . . 57

A.1. The influence of dose uncertainty in radiotherapy treatments. . . . . . . . . 57
A.2. Uncertainty in external beam radiotherapy . . . . . . . . . . . . . . . . . . . . . . 59

APPENDIX B: QUALITY AUDITS OF THE CALIBRATION OF

RADIOTHERAPY BEAMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

APPENDIX C: CASE HISTORIES OF ACCIDENTS RELATED TO

DECOMMISSIONING OF RADIOTHERAPY EQUIPMENT AND
SOURCES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
C.1. Illegal import, storage, and disposal of a teletherapy 60C unit
(Mexico, 1984). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
C.2. Abandonment of a teletherapy 137Cs unit (Brazil, 1988) . . . . . . . . . . . . . 66

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

4
 PREFACE

Over the years, the International Commission on Radiological Protection (ICRP),

referred to below as ‘the Commission’, has issued many reports providing advice on
radiological protection and safety in medicine. Its Publication 73 is a general over-
view of this area. These reports summarise the general principles of radiation pro-
tection and provide advice on the application of those principles to the various uses
of ionising radiation in medicine and biomedical research.
Most of these reports are of a general nature, and the Commission wishes to
address some specific situations where difficulties have been observed. It is desirable
that reports on such problem areas be written in a style which is accessible to those
who may be directly concerned in their daily work, and that every effort is taken to
ensure wide circulation of such reports.
A first step in this direction was taken at the Commission’s meeting in Oxford,
United Kingdom, in September 1997. At that time, on the recommendation of ICRP
Committee 3, the Commission established several Task Groups to produce reports
on topical issues in medical radiation protection.
Some such reports have now appeared in print, and the present one is the result of
the work of yet another one of those Task Groups. The terms of reference of this
Task Group on Accident Prevention and Safety in Radiation Therapy were to pre-
pare a document with recommendations to prevent accidents. The Task Group was
requested to base its report on ICRP Publications 60 and 73, giving special emphasis
to recent accidents and to the lessons to be drawn from them. The scope of the
report to be produced was outlined in some detail, and it was requested that
recommendations and check-lists be developed.
The membership of the Task Group was as follows:

P. Ortiz López (Chairman) P. Andreo J.-M. Cosset

A. Dutreix T. Landberg

Corresponding members were:

L.V. Pinillos-Ashton W. Yin

The Commission is grateful to P.J. Biggs for kindly reviewing the manuscript by
correspondence. Numerous helpful comments were received from interested parties
through the Commission’s customary public consultation via the Internet.
The membership of Committee 3 during the period of preparation of this report
was:

F.A. Mettler, Jr. (Chairman) J.-M. Cosset M.J. Guiberteau

L.K. Harding (Secretary) J. Liniecki (Vice-Chairman) S. Mattsson
H. Nakamura P. Ortiz López L.V. Pinillos-Ashton
5
 ICRP Publication 86

M.M. Rehani H. Ringertz M. Rosenstein

Y. Sasaki C. Sharp W. Yin
W.Y. Ussov

This report aims to serve the purposes described above. In order to be as useful as
possible for those purposes, its style differs in a few respects from the usual style of
the Commission’s publications in the Annals of the ICRP. Some colour figures are
also included in order to illustrate more clearly the possible consequences of acci-
dents to patients undergoing radiotherapy.
The report was approved for publication by the Commission in October 2000.

6
 ICRP Publication 86

Prevention of accidental exposures to patients

undergoing radiation therapy

ICRP Publication 86
Approved by the Commission in October 2000

Abstract-This publication aims to assist in the prevention of accidental exposures involving

patients undergoing treatment from external beam or solid brachytherapy sources. It does not
directly deal with therapy involving unsealed sources. The document is addressed to a diverse
audience of professionals directly involved in radiotherapy procedures, hospital adminis-
trators, and health and regulatory authorities. The approach adopted is to describe illustrative
severe accidents, discuss the causes of these events and contributory factors, summarise the
sometimes devastating consequences of these events, and provide recommendations on the
prevention of such events. The measures discussed include institutional arrangements, staff
training, quality assurance programmes, adequate supervision, clear definition of responsi-
bilities, and prompt reporting.
In many of the accidental exposures described in this report, a single cause cannot be
identified. Usually, there was a combination of factors contributing to the accident, e.g.,
deficient staff training, lack of independent checks, lack of quality control procedures, and
absence of overall supervision. Such combinations often point to an overall deficiency in
management, allowing patient treatment in the absence of a comprehensive quality assurance
programme. Factors common to many accidents are identified and discussed in detail. The use
of radiation therapy in the treatment of cancer patients has grown considerably and is likely
to continue to increase. Major accidents are rare, but are likely to continue to happen unless
awareness is increased. In this report, explicit recommendations on measures to prevent
radiotherapy accidents are given with respect to regulations, education, and quality assurance.
# 2001 ICRP. Published by Elsevier Science Ltd. All rights reserved.
Keywords: Radiotherapy, Accidents, Safety, Radiation protection, QA

7
 ICRP Publication 86

MAIN POINTS

. Doses received during radiotherapy are on the upper edge of tolerable doses to
normal tissues. As a result, accidental overdosages have often had devastating
and sometimes fatal consequences.
. Accidental exposures involving a 10% or more overdosage should be detect-
able by a well-trained clinician, based upon an unusually high incidence of
adverse patient reactions.
. Underdosage accidents are difficult to detect clinically and may only be man-
ifest as poor tumour control.
. Radiotherapy is increasing worldwide and accidents may be expected to
increase in frequency, if measures for prevention are not taken.
. While a number of serious and fatal radiotherapy accidents are reported, it is
likely that many more have occurred but were either not recognised or repor-
ted to regulatory authorities or published in the literature.
. The complex equipment and techniques used in radiotherapy mandate that for
accident prevention, there must be sound and risk-informed regulations, man-
agerial commitment at the hospital level, an adequate number of trained staff,
adequate resources, a functional implemented quality assurance programme,
good communication, and continuing education.
. There is a danger in not fully appreciating that modern equipment and new
technologies require more quality assurance and highly qualified maintenance.
. Persons in charge of radiotherapy facilities should ensure that there is proper
commissioning of new equipment and proper decommissioning of old equip-
ment and sources.

9
 SUMMARY

(a) Curative or palliative radiation therapy has three major concerns: efficacy,
quality of life, and safety (Horiot, 1995). From the point of view of radiation safety,
radiotherapy is a very special application of radiation because:

. human beings are directly placed in a very intense radiation beam (external
beam therapy), or radiation sources are placed in direct contact with tissue
(brachytherapy), to deliver intentionally very high doses (20 Gy–80 Gy), and
. overdosage as well as underdosage may have severe consequences.

(b) From the prescription to the delivery of a radiotherapy treatment, a team of

professionals from a number of disciplines is involved in a large number of steps. In
the case of external beam radiotherapy, a treatment is usually delivered with 20 to 40
fractions or sessions, each requiring a large number of machine and patient para-
meters to be set up by the radiotherapy technologist. Based on these considerations,
it becomes apparent that there is a significant potential for mistakes leading to an
accidental exposure in radiotherapy, and that special preventive measures have to be
implemented.
(c) This document aims to assist in the prevention of accidental exposures in
radiotherapy, mainly those involving patients undergoing treatment from external
beam or solid brachytherapy sources. The document does not directly deal with
therapy involving unsealed sources (nuclear medicine radiophamaceuticals). The
approach adopted is to: (i) describe illustrative severe accidental exposures that have
occurred in radiotherapy procedures during the last 25 years, using case histories1;
(ii) discuss why these events occurred, highlighting their causes and any contributory
factors; (iii) summarise the consequences of these events, which were in some cases
devastating; and (iv) provide recommendations on measures to prevent such events,
primarily through institutional arrangements, staff training, quality assurance pro-
grammes, adequate supervision, clear definition of responsibilities, and prompt
reporting.
(d) The document is addressed to professionals directly involved in radiotherapy
procedures (radiation oncologists/radiotherapists2, medical physicists, radiotherapy
technologists3, dosimetrists, radiotherapy nurses, maintenance engineers, etc.) as
well as to hospital administrators and health and regulatory authorities. It is inten-
ded to provide information and recommendations useful to such a diverse audience;
compromises in the scientific detail are, therefore, unavoidable.

1
It should be noted that, in addition to the accidents reported in detail in this document, a consider-
able number of other events have occurred in radiotherapy where information could enhance the body of
knowledge on this topic. Unfortunately most of these events have not been reported. Not all reported
incidents have been included (cf. Järvinen, 1997).
2
The specialist physician concerned is usually referred to as a radiation oncologist in the United States
and as a radiotherapist in Europe. In this report, the term radiation oncologist will be used.
3
In some countries radiotherapy technologists are called radiation therapists, radiographers, radio-
therapy technicians, etc.

11
 ICRP Publication 86

Table 1a. Major reported accidental exposures involving patients undergoing external beam radiotherapy
and brachytherapy treatments

Country Year No. of patients Causes and main contributing factors

affected

60
USA 1974–76 426 Co dose calculations based on erroneous decay curve
(varying overdoses)
No independent verification of dose calculations
More than two years without beam measurements

Canada and 1985–87 3 deaths from Accelerator software for the control of treatment features
USA radiation was transferred from other equipment without sufficient
consideration to safety
60
Germany 1986–87 86 Co dose calculations based on erroneous dose tables
(varying overdoses) No independent determination
of the dose rate

UK 1988 207 Error in the calibration of a 60Co therapy unit

(25% overdose) No independent calibration of the beam

UK 1988–89 22 Error in the identification of 137Cs brachytherapy

sources (dosimetry errors between-20% and +10%)
No independent determination of source strength

USA 1987–88 33 A computer file (of a treatment planning system)

for the use of trimmers was not updated for a new
60
Co source because it was not intended for further use
The computer file was used (overdoses of 75%)
There was no manual verification of the calculated
treatment time
There was no independent verification of the treatment time

Spain 1990 27 (15 deaths directly Error in the maintenance of a clinical linear accelerator;
from radiation; procedures for transferring machine from/to maintenance
two deaths with radiation (notification of physicists) not followed
as major contribution)
Conflicting signals and displays not analysed
Procedures for periodic beam verifications not implemented
or insufficient
Overdosage ranging from 200% to 600%

UK 1982–91 1,045 (492 patients Inappropriate commissioning of a computerised

developed a local radiotherapy Treatment Planning System (5-30% underdosage)
recurrence, possibly
as a result of
the underdosage)
Lack of quality control

USA 1992 1 (death from radiation) High dose rate brachytherapy source left inside the
patient. Source dislodged from equipment
Conflicting monitor signals (display at the machine
and area monitor) ignored
A survey with an available portable instrument was
not conducted
Costa Rica 1996 114 (17 deaths from Error in the calibration of a 60Co therapy unit.
radiation)
Lack of independent beam calibration and quality control.
Recommendations of external audit ignored
Overdosages of about 60%

12
 ICRP Publication 86

Table 1b. Major accidents due to abandoned radiotherapy sources involving public exposure and/or
large-scale contamination

Country Year Consequences on humans Other consequences

Mexico 1984 4,000 people exposed 30,000 table stands and 6,000 tons
(five of them from 2 to 7 Gy) of reinforcing rods contaminated
814 buildings partly or totally demolished
Brazil 1988 Four persons died due to 3,500 m3 of radioactive waste generated
radiation exposure
112 persons monitored for
contamination
Georgia 1994 One person died due to
radiation exposure
Turkey 1999 11 persons severely exposed
to radiation
Thailand 2000 3 persons died due to
radiation exposure

In most of the accidents in Table 1b the sources were waiting long for a disposal or return to the manu-
facturer. A major reason for the long delay in the decision was the lack of a disposal capability, the high cost
of disposal, and the lack of provisions for the end of life of the sources prior to their purchasing and use.

(e) A summary of the major accidental exposures involving patients undergoing

external beam radiotherapy and brachytherapy treatments is given in Table 1a,
which also summarises the main factors that contributed to each event. Detailed
case histories are given in Chapter 2. Factors contributing to these and other
accidental exposures, classified according to the stage in the radiotherapy process
at which they occurred, are in Chapter 4. Given the devastating consequences of
accidents involving radiotherapy sources, for exposure of the public and/or large
scale contamination these are also illustrated in Table 1b (case histories from some
of them are in Appendix C).

Causes of accidental exposures in radiotherapy

(f) In many of the accidental exposures reported in this document a single cause
cannot be identified. Usually there was a combination of factors contributing to the
accident, e.g., deficient staff training, lack of independent checks, lack of quality
control procedures, and/or absence of overall supervision. Such combinations often
point to an overall deficiency in management, which allowed patient treatment in the
absence of essential institutional arrangements, i.e. a comprehensive quality assurance
programme. To prevent these events, appropriate institutional arrangements should
be required by regulation, and compliance with them should be verified (Ortiz, 1998).
The following factors, common to many accidental exposures, can be identified.

Deficiencies in education and training

(g) The lack of accredited professionals with adequate education and training,
mainly in the field of medical physics, is a major factor in accidents caused by the
13
 ICRP Publication 86

erroneous calibration of radiotherapy beams and brachytherapy sources. As illu-

strated in Table 1a, this type of accidental exposures has involved large numbers of
patients. Insufficiently trained radiation oncologists, working without supervision,
caused misinterpretations and consequent incorrect treatment. A lack of trained
brachytherapy nurses contributed to accidents with radiation sources in the bra-
chytherapy ward. Lack of understanding of beam parameters by a maintenance
engineer led to a maladjustment of accelerator energy, causing an accident with
devastating consequences to many patients. In most cases, the training of profes-
sionals had included only the ‘normal’ circumstances likely to occur in radiotherapy,
but did not consider the identification of ‘unusual’ situations. As a result, initiating
events were not detected and the accidental exposure occurred.

Deficiencies in procedures and protocols

(h) Severe accidental exposures have been caused by inadequate commissioning

procedures for radiotherapy equipment (testing and measurement of operational
parameters prior to use), computers, and accessories, or by the lack of periodic
verification. These accidental exposures could have been avoided if widely available
quality control recommendations and protocols had been followed. The use of bra-
chytherapy sources without verification of their activity, or a misunderstanding of
the units used to specify their source strength, also caused accidents. Procedures for
dealing with equipment repair and maintenance were not followed, and equipment
was used for treatment without a previous verification. Deficiencies of procedures
and documentation of treatment prescription and planning, patient identification,
and patient set-up have also contributed to such events.

Equipment faults

(i) This document does not address issues related to the design of radiotherapy
equipment by providing specific recommendations to manufacturers. However, there
have been some major accidental exposures, with fatal consequences, as a result of
inappropriate design and factory testing that could not be detected by hospital
quality controls. In one case, factory tests of accelerator software that had been
transferred from an old design to a new one failed to anticipate operating conditions
that could occur in practice the manufacturer needed a considerable time to identify
the cause, to inform users, and to correct the problem. In another case, equipment
faults that appeared sporadically, and could not be resolved by maintenance engin-
eers, led to a decision to operate the machine with interlocks disabled (‘physical’
mode as opposed to ‘clinical’ mode); the result was a fatal accidental exposure.

Deficient communication and transfer of essential information

(j) A frequent cause of accidents was misunderstanding of the treatment prescrip-

tion, treatment plan, or patient identification data. Changes in personnel, without
transfer of information on calibration procedures and treatment planning, caused
14
 ICRP Publication 86

accidental exposures. Deficient communication of patient reactions, conflicting sig-

nals and displays, as well as of equipment malfunction contributed to missing
opportunities of early detection of accidental exposures.

Lack of independent checks (defence in depth4)

(k) In many accidental exposures unidentified single mistakes remained undiscovered.

These could have been prevented if independent checks had been incorporated into
procedures. For instance, mistakes in beam calibration after source installation or
exchange could have been identified by independent calibration by a second person
or by a consistency check with the manufacturer’s source certificate. A systematic
application of multiple checks would have drastically reduced accident potential.

Inattention and unawareness

(l) Examples of unawareness included cases when it was not realised that after
source change (higher activity), calculated treatment times should be substantially
shorter than before. An error in calibration, treatment planning, or calculation should
become apparent to attentive staff before treatment delivery. In external beam therapy,
treatment delivery involves a large number of steps, utilising substantial data in a
repetitive manner, but varying from patient to patient. Radiotherapy technologists
often have to enter machine parameters onto treatment charts for upwards of one
hundred beams per day5. Inattention can lead to the use of a wrong prescription, the
wrong patient chart or the selection or recording of wrong parameters. Preparation and
delivery of treatment by a radiotherapy technologist requires a continuous and high
level of concentration, which may be difficult to maintain when the workload is high.
Measures to maintain concentration and awareness require a working environment
where distractions are minimised and redundancy checks undertaken (i.e., two indivi-
duals should be involved in patient set-up).

Unsecured long-term storage or abandonment of radiotherapy sources

(m) Several major accidents, involving members of the public and with severe
environmental consequences, were related to the long-term unsecured storage of
radiotherapy sources, either before or after decommissioning.

Summary of recommendations

(n) To prevent accidental exposures in radiotherapy, the following measures are

recommended:

4
The application of more than a single protective measure for a given safety objective such that the
objective is achieved even if one of the protective measures fails, for example, multiple independent
checks.
5
This task is achieved by utilising machines with ‘record-and-verify’ systems.

15
 ICRP Publication 86

. regulations for radiation safety should be in place and enforced, and com-
pliance should be monitored;
. appropriate education of professionals directly involved in radiotherapy pro-
cedures should be undertaken as a high priority, with continuing education on
new treatment techniques;
. a written, comprehensive, and functional quality assurance programme should
be adopted and implemented by every radiotherapy institution. This should
include a clear chain of responsibilities and procedures for purchasing, accep-
tance test and commissioning, calibration, treatment planning and delivery,
maintenance, two-way communication, and incident reporting.

The future

The benefits of radiation therapy in the treatment of cancer patients are widely
acknowledged and as a result, the use of this treatment modality has grown con-
siderably. Usage is likely to continue to increase. Major accidental exposures are
rare, but it is likely that they will continue to happen unless awareness is increased.
Accidents will usually occur as the result of inadequate education and training, lack
of quality assurance, poor infrastructure, equipment failure, and improper decom-
missioning. Unless these issues are properly addressed and dealt with, more acci-
dental exposures are likely to occur, as current and new technology developments
are disseminated.

16
 1. INTRODUCTION

(1) Radiotherapy is concerned primarily with tumour cure or palliation. Modern

radiotherapy has three major concerns: efficacy, quality of life, and safety (Horiot,
1995). It is always necessary to be aware of the potential for an accident, the relative
importance of human factors, and the wider consequences of an accident.
(2) A radiation accident is an unintended event (operator error, equipment failure,
or other mishap) that has or may have adverse consequences (IAEA, 1996).
Depending on the type of persons exposed, accidents can be divided into three major
groups:

(i) members of the general public irradiated as a result of failure of imple-

mentation of radiation protection and safety rules;
(ii) clinical staff irradiated during preparation of radiation sources or patient
treatment and maintenance staff irradiated during installation, repairs,
source change, or other equipment servicing;
(iii) patients injured during treatment.

(3) The use of the term ‘radiation accident’ for events involving the public and
staff, i.e. groups (i) and (ii), has a relatively straightforward interpretation as neither
workers nor public are intentionally irradiated; this makes it easier to distinguish a
normal exposure from an accidental one. For patients undergoing treatment, i.e.
group (iii), the use of the term ‘accident’ deserves special consideration. The major
focus of this report is in preventing injury to patients. The majority of radiation
therapy accidents have occurred in category (iii).

1.1. The potential for an accidental exposure involving radiotherapy patients

(4) The following aspects are relevant to understanding the potential for accidents
involving patients:

. Very high doses (20 Gy to 80 Gy) are intentionally delivered to the ‘clinical
target volume’ (ICRU 1999). As radiation passes through the body, some nor-
mal tissues receive doses often similar to that delivered to the target volume,
thus resulting in an expected incidence of side effects (Chapter 3).
. The radiation beam is focused directly at the patient (external beam radio-
therapy) or sealed radioactive sources are placed in contact with tissues or
organs (brachytherapy). Any significant mistake in beam delivery or in the
placement of sources can have negative, and often serious, consequences.
. Radiotherapy involves many steps between prescription and dose delivery.
Each step may involve a large number of parameters that must be selected,
adjusted, recorded, and communicated between different professionals. For
example, the delivery of 30 fractions by external beam, each with four fields,
requires around 15 parameters to be set for the first field and half this number
changed for the other fields; in total, the requirement is to set about 1,000
17
 ICRP Publication 86

parameters for the entire treatment. The set-up for each patient is similar, but
not identical. The number of parameters is much larger in conformal therapy6
using multi-leaf collimators and intensity modulated beams, although these are
usually computer controlled.
. For treatment units without computerised ‘record-and-verify’ systems, the
radiation technologist has to enter manually parameters in patients’ treatment
charts. This may be required for up to one hundred treatment beams daily.
This is done in a repetitive way, but is different for each patient.
. Sophisticated technology, computer calculations, and data transfer are com-
bined with manual activities e.g. preparation of organ shielding, and immobi-
lisation devices.

(5) The combination of these diverse aspects in radiotherapy requires special and
specific safety measures. The design of a safety plan has to consider likely accidental
exposure scenarios. Two main approaches are available for identifying these sce-
narios: retrospective analysis of past adverse events, and prospective methods (event
trees and fault trees combined with probabilistic assessment) that may identify
additional weaknesses. Situations that have not yet produced an accidental exposure
may be overlooked by retrospective methods. However, there are difficulties with
prospective methods, e.g. dealing with situations where dissimilar staff training may
play a significant role. Both methods have their merits and are complementary.
(6) In this document, retrospective methods are used to analyse several accidents,
which occurred in radiotherapy in the period 1974-2000. Data has been obtained
from papers published on individual accidents, as well as from regular reports such
as those issued regularly by the US Nuclear Regulatory Commission (NRC) and a
safety report of the International Atomic Energy Agency (IAEA, 2000a). The num-
ber of accidents in Table 3 (Chapter 4) is large enough to identify the major initia-
tors and contributing factors and, in most cases, to provide information on their
frequency and consequences. Further weaknesses may be revealed, and additional
scenarios anticipated, by a complementary prospective assessment (ICRP, 1997).
(7) The information provided in this report includes primarily classical equipment and
treatment modalities. Currently, there is insufficient information about possible accidents
involving new technologies and treatment techniques, such as ‘gamma knife’ units,7
multi-leaf collimators, intensity beam modulation, intravascular brachytherapy, and high
dose rate brachytherapy. Although accidents have already been reported for those new
modalities, only prospective methods of analysis may help to prevent accidents.

6
In conformal therapy the cross section of the radiation beam is shaped (conformed) to the target
volume, normally using multi-leaf collimators, rather than rectangular collimators. Conformal therapy
allows a more efficient and effective treatment, by delivering high-dose to the target volume while keeping
doses low to tissues and organs outside the target area.
7
Several accidents involving gamma knives have been reported recently in the US Nuclear Regulatory
Commission Reports to the Congress on Abnormal Occurrences, NUREG 0090, Volumes 17 and 22. One
of the cases related to a failure of the patient’s couch to retract from the treatment psotion and the couch
had to be retracted manually; two other cases related to errors in the preparation of the treatment plan
from the prescription and in its implementation.

18
 ICRP Publication 86

1.2. Criteria for accidental exposures with radiotherapy patients

(8) In radiotherapy, a ‘normal’ radiation exposure is a treatment that closely fol-

lows the plan specified in the treatment prescription. An accidental exposure can
therefore be considered to have occured if there is a substantial deviation from the
prescription.8
(9) Broadly, the prescription includes: identification of the patient; definition of
the target volume; the dose to be delivered to the anatomical site(s); the dose frac-
tionation or treatment schedule; and the final dose distribution. Thus, a departure
from the treatment prescription could be: the wrong patient, the wrong anatomical
site, a substantial difference from the prescribed dose,9 the wrong dose distribution
(including incorrect beam modality or energy), or the wrong fractionation.
(10) The term ‘substantial difference from the prescription’ implies a quantifica-
tion of the difference and its impact on the outcome, i.e., the clinical consequences.
Although there are no international standards that include the quantification of a
substantial difference or a threshold, there are approaches to classify hazards related
to radiotherapy patients. The proposal of the American Association of Physicists in
Medicine Task Group 35 (AAPM, 1993) deserves special mention, despite being
focused on accelerators and the assumption that weekly quality controls are always
performed which would discover errors or equipment malfunctions within one week.
This AAPM proposal is based on the US Food and Drug Administration (FDA)
regulations, which define a Class I hazard as a condition that could cause death or
serious injury and a Class II hazard, where the risk of serious injury is small.
(11) Based on this classification, AAPM (1993) developed a sub-classification of
Class I hazards that is valid for the conditions described above. According to this,
type A hazards can be directly responsible for complications threatening patient life;
type B hazards are those that increase the probability of an unacceptable outcome of
the treatment (increased rate of not life-threatening complications or lack of tumour
control), but usually do not pose a threat to life. These criteria are summarised in
Table 2. Type A hazards are associated with an overdose equivalent to 25% or more
of the total prescribed dose. This includes the situation where a dose is delivered
outside the intended radiation field, since such circumstances may produce severe
clinical complications when a critical organ (e.g., the spinal cord) is near the radia-
tion field. A Type A hazard also includes selection of incorrect beam energy or
modality, e.g., irradiation with an electron beam when a photon beam was intended
or vice versa. A type B hazard corresponds to overdoses in the range 5%-25% and
to most underdosage situations. The interested reader is referred to the original
report (AAPM, 1993), where illustrative examples are given.
(12) It is emphasised that inclusion of most cases of underdose in the type B group
of events is based on the assumption that the error would be discovered within one
week (with weekly quality controls) and corrective actions would be successfully

8
Situations resulting from an incorrect prescription are not the subject of this report.
9
For simplicity, in this document the term ‘prescribed dose’ refers to the dose at different points of the
target volume and other tissues and organs, resulting from a clinically accepted dose distribution.

19
 ICRP Publication 86

Table 2. Summary of the AAPM TG-35 sub-classification of Class I hazards in radiotherapy (AAPM,
1993)a. Remarks added here.

Type Criteria Remarks

Type A 25% overdose or more of ‘The rationale for this choice is related to the
the total prescribed dose observation that a 25% to 50% increase in
total dose will often place the patient in the
range of the LD50/5 (the probability of 50%
lethal complications within five years). . .’.
For a typical treatment of 40-60 Gy, an overdose
of 25% of the prescribed total dose corresponds to
10-15 Gy. This excess in dose can be reached either
with an error on each fraction for several fractions
during the week or with a large error in a single fraction.
Type B 5% (i) to 25% dose excess (i) The value 5% is derived from the TG 35 criteria
over the total dose(ii) where an overdosage of 20% during one week
corresponds approximately to an overdosage of
about 5% over the whole treatment.
and most underdose situations (ii) If the underdosage is not discovered within a
time in which correction to the treatment can be
successfully applied, the hazard should be
considered as type A with similar percentage
as for an overdose as indicated in the text
a
Class I hazards are defined by the USA FDA as a condition that could cause death or serious injury.
TG-35 considers type A hazards as those that can likely be responsible for life-threatening complications.
Type B hazards increase the probability of an unacceptable treatment outcome (complications or lack of
tumour control). The criteria refer to a typical treatment prescription of 40–60 Gy total dose with 2 Gy
per fraction, and is based on the assumption that weekly quality controls are performed that will discover
errors or equipment malfunctions within one week.

implemented during the rest of the treatment. This assumption, however, is not
always met. For an example, see case history number 2 of Chapter 2, where the
underdosage accident involved 1,045 patients during a 10 year period. Such an
accident cannot be considered as being type B. If the underdosage is not discovered
within a time period such that corrective actions can be successfully applied, the
consequences may be fatal, since the progression of the disease reaches advanced
stages and opportunities for remedial actions are not available. Severe underdosage
may thus be more adequately classified using limit values similar to overdosages, i.e.
an underdose of 25% or more of the total prescribed dose is a type A hazard.

1.3. Clinical detectability of differences in dose prescription and delivery

(13) The question of how large a dose prescription-delivery difference can be

detected by observing the clinical outcome of a treatment has relevance in relation to
the consequences of human errors or equipment malfunctions. A review of the
information on clinically detected dose differences for under- and overdosage situa-
tions is therefore relevant:
20
 ICRP Publication 86

(i) Dutreix (1984) has reported that a radiation oncologist observed skin reac-
tions on skin folds and also diarrhoea, which were more severe than usual,
in patients irradiated for gynaecological tumours with a prescribed tumour
dose of 50 Gy to the whole pelvis, given in five fractions per week for five
weeks. The effects observed were caused by an error in the calibration of a
treatment unit that led to a systematic overdosage to patients. Estimated
overdosages were 7% in 67 patients and 10% in 21 patients.
(ii) In a trial study of skin reactions in 58 breast cancer patients, where both
internal mammary lymphatic chains were irradiated with electrons, using
upper and lower fields with total doses of 50–60 Gy, Wambersie (1984)
reported that electron and 60Co adjacent fields dose differences of 10% were
detected in 80% of the cases. For symmetric supraclavicular fields, even
dose differences of 5% could be clinically detected.
(iii) In a retrospective study of 314 patients with laryngeal cancer, Herring et al.
(1971) found a reduction of 20% in the probability of recurrence of the
primary tumour in three groups of T3 staged patients treated with doses
increased by steps of about 4.5%. The doses administered were 52.5 Gy to
the ‘low dose’ group, 55 Gy to the ‘middle dose’ group, and 57.5 Gy to the
‘high dose’ group. An unintended increase in dose, less than 10% above the
optimal dose, produced a readily observable increase in the incidence and
severity of complications. The increase in dose was due to the omission of a
conversion from roentgens to rads that, depending on the accelerator
energy, resulted in overdoses between 7 and 10%.
(iv) Differences in local tumour control and survival for groups of patients that
differed only by 5% in the average dose delivered have been reported by
Dische et al. (1993). In two review studies of the clinical basis for dosimetry
accuracy in radiotherapy Wambersie et al. (1994) and Mijnheer (1996) con-
cluded that differences in absorbed dose in the range of 5-10% could be
detected in a number of situations.

(14) Although the impact of differences in dose on the probability of tumour

control and normal tissue complications depends on the type of tissue irradiated, the
fractionation used, the volume irradiated, and other factors (see Chapter 3), from
the literature described above it appears that the lowest dose differences clinically
detected were in the order of  5–10%.
(15) On the other hand, it should be noted that the delivery of absorbed dose has
an associated uncertainty. The estimated combined standard uncertainty in good
radiotherapy practice is in the order of 5% (see Appendix A), i.e., it is of the same
order as the lowest dose difference clinically detectable. The larger the size of the
error, the larger the probability of the undesired outcome and, in the case of normal
tissue complications, the greater the severity. When the error size is large enough,
the unplanned detrimental outcome becomes virtually certain. This was the case in
some of the accidental exposures described in Chapter 2.
(16) Traditionally, only patient overdose has been considered of importance.
However, patient underdosage is as important as overdosage, and in certain instances
21
 ICRP Publication 86

it can be even more significant (see the case described in Section 2.2 and Chapter 3).
Unfortunately, cases of underdosage are usually more difficult to detect clinically
than overdosages, and they are usually only discovered long after the treatment has
been completed and there is a higher than expected incidence of tumour recurrence.

1.4. Other events of interest to the prevention of accidental exposures

(17) There is a group of events that can be considered accident ‘precursors’ (i.e.,
do not have clinical consequences but have the potential to develop into accidents
with such consequences). For example, events detected by ‘good luck’, or by a
quality assurance programme including redundant checks and systems, could result
in an accidental exposure if the same events occur in another institution.
(18) The compilation of information on such events, their evaluation to determine
generic implications and lessons that can be learned, and the dissemination of this
information is highly desirable to prevent accidental exposures. Some national sys-
tems for reporting, such as that in the USA, include provisions to report ‘events of
interest’ that are considered precursors, even if they do not involve a major reduc-
tion of the level of protection. Similar programs are being developed by professional
organisations.

22
 2. CASE HISTORIES OF MAJOR ACCIDENTAL EXPOSURES IN
RADIOTHERAPY

(19) This section provides a short description of some major accidental exposures,
as an illustration of how initiating events can develop into severe, large scale, acci-
dental exposures with fatal consequences (IAEA, 2000a).

60
2.1. Incorrect Co decay chart and lack of verification (USA, 1974-76)

(20) A cobalt (60Co) teletherapy unit had been calibrated correctly initially, but
subsequent dose calculations were based on an incorrect 60Co decay curve. No peri-
odic measurements of the machine output were made. The decay curve had been
drawn on a linear scale with an incorrect slope, resulting in a calculated decay more
rapid than that of the real source. The dose rate was therefore underestimated and
consequently longer treatment times were calculated from the erroneous decay curve.
(21) Initially, the incorrect machine output was attributed to a faulty measuring
system. The medical physicist produced ten calibration documents showing that the
machine output used was correct. When some patients showed symptoms of over-
exposure, the hospital asked an external consultant physicists to review the dosi-
metry. They found that the 60Co unit and measuring system were functioning
correctly, but the resident medical physicist had fabricated all but one of the reports
on machine output.
(22) Overdoses to patients had increased progressively over 22 months. The level
of overdose was 10% in the first 5 months and increased to as high as 50% by the
end of the period. During the last 16 months, 426 patients were treated.
(23) Approximately 20 years after the accidental exposure, Cohen et al. (1995)
evaluated the clinical consequences using the medical records of the 26 patients.
Three hundred patients died in less than one year, mainly due to their cancer. Of the
patients who survived the first year, 88 had immediate severe complications related to
the irradiated sites (including brain, spinal cord, skin, oropharyngeal mucosa, colon,
and rectum). The clinical follow-up notes pertaining to the period between one and
three years after the treatment showed that a high proportion of patients developed
significant, often life-threatening, complications. These included severe skin reactions
with ulceration, mucosal reactions with necrosis, stenosis of pharynx and esophagus,
ulceration and/or perforation of stomach and bowel, bone necrosis, and myelopathy.
(24) A contributory factor was insufficient staff resources to assure adequate
dosimetry and quality control. The actual beam output was not checked over a 22-
month period, and the medical physicist was dedicated to supporting a new accel-
erator; there was no independent check of the decay curves; the physicist falsified
records afterwards in an attempt to justify his calculations.

2.2. Lack of acceptance procedures for a treatment planning system (UK, 1982-1990)

(25) Until 1982, manual calculations were the only method available for treatment
planning in a specific hospital. Treatments were usually performed with a fixed SSD
23
 ICRP Publication 86

(source-skin-distance) technique. For treatments requiring an SSD other than the

usual 100 cm, a procedure had been established based on correction factors for
varying distances that radiotherapy technologists applied manually.
(26) A computerised treatment planning system was acquired in the autumn of
1982. This allowed isocentric treatments on a more frequent basis. Since isocentric
techniques are carried out with distances shorter than the usual 100 cm SSD, for the
first isocentric treatment the technologists (with the acquiescence of the medical
physicist) used the distance correction factors previously established. Every sub-
sequent isocentric treatment had this manual calculation for the distance correction
applied.
(27) It was not realised that the computer planning software already incorporated
corrections for varying source-skin distance. The correction was therefore dupli-
cated when the radiotherapy technologists continued to manually apply distance
correction for distances other than 100 cm. Since no quality control measurements
were made, the error was not detected. Thereafter, the practice of manually applying
the distance correction factors every time that an isocentric treatment was used
remained. At the end of 1991, almost a decade later, the error was discovered.
(28) The error caused delivered radiation doses to be lower than the prescriptions
in all isocentric treatments during that period of time, which involved 1,045 patients.
A study made by Ash and Bates (1994) on these patients concluded that 492 patients
developed a local recurrence, possibly as a result of the underdosage. For 189
patients it was not possible to be certain about the effects of the underdosage. For
320 patients, it was considered unlikely that they had suffered any effects from the
underdosage, and in the remaining 44 patients the outcome was unknown.
(29) An independent inquiry into the case found the following factors contributing
to this accident: the responsibility and authority of the medical physicist were not
clearly defined; the number of staff members was insufficient for the number of
patients treated and, in particular, the shortage of medical physicists led to over-
loaded working conditions; a systematic quality assurance programme that included
commissioning following the introduction of new equipment was not in place; and the
correction procedure followed for nearly a decade was never embodied in writing.

2.3. Six accidental exposures involving software problems in several accelerators of

the same type (Canada and USA, 1985-87)

(30) Between June 1985 and January 1987, six known accidental exposures
involved massive overdose by the same type of accelerator (Leveson, 1993). Several
flaws were identified in the software used to enter selected parameters for the treat-
ment such as type of radiation and energy. Two of these flaws caused the deaths of
three patients.
(31) The first accident leading to a fatality occurred in March 1986. After selection
of the treatment parameters for x-ray mode, the operator changed the selection
from x-ray to electron mode. During this time the equipment was executing the
initial request, which required the automatic rotation of a turntable assembly with
the bending magnet for beam scanning (electron mode) and the target and beam
24
 ICRP Publication 86

flattener (x-ray mode). The equipment did not execute the second request correctly
and the turntable had not reached the correct position. The equipment then operated
without the x-ray target and the beam flattener, thus leading to massive overdoses.
(32) The outcome was radiation-induced myelitis of the cervical cord causing
paralysis of the patient’s left arm and both legs, left vocal cord paralysis (which left
him unable to speak), neurogenic bowel and bladder, and paralysis of the left dia-
phragm. He died from complications of the overdose five months after the accidental
exposure.
(33) The maintenance engineer was unable to reproduce the malfunction and the
problem remained unresolved. The same problem reappeared in April 1986 in the
same hospital and a second patient was overexposed. He developed disorientation
that progressed to coma and neurological damage from high-dose radiation injury
to the right temporal lobe of the brain and the brain stem. The patient died three
weeks after the accident.
(34) The medical physicist in the hospital was finally able to reconstruct the fault.
This forced the manufacturer to respond (Rawlinson, 1987). The exact doses in these
cases are uncertain but they were estimated to be in the order of 150 Gy- 250 Gy in
one single treatment fraction (delivered within a few seconds).
(35) A different software problem caused the death of a patient in 1997 in another
hospital. An operator pressed the ‘set’ button at the precise moment when a variable
in the software rolled over to zero. This allowed the machine to start irradiation
without x-ray target and without scanning of the electrons over the radiation field.
The result was a highly concentrated electron beam.
(36) There were a number of problems with the design, testing, and process for
following up of equipment faults, as well as in the dissemination of warnings to
users of the same type of equipment. It is not possible to summarise these problems
in a few lines. Interested readers are referred to the detailed reports of Rawlinson
(1987) and Leveson (1993). Two of the major problems were:

. The software package for controlling some of the functions in the accelerator
was reused from an older type of accelerator of a significantly different design.
In the new accelerator type, safety of these functions relied entirely on the old
software. In a review study (Leveson, 1993), it was concluded that using soft-
ware modules designed for other equipment does not guarantee safety in the
new system to which they are transferred and sometime leads to awkward and
dangerous designs. Safety is a quality of the system in which the software is
used; it is not a quality of the software itself.
. There was no efficient mechanism to follow up reports of suspected accidents.
The first accident occurred in June 1985 and resulted in a mastectomy for
radiation complications. The patient completely lost the use of her shoulder
and arm. In spite of this, the manufacturer and operators of the machine
refused to believe that it could have been caused by the accelerator (Leveson,
1993). The accident was not reported to the US Food and Drug Administration
(FDA) until 1986. This considerably delayed the investigations by the FDA.
The reporting regulations at that time applied only to equipment manufacturers
25
 ICRP Publication 86

and importers, not users. The regulations were amended in 1990. Other users
of the same type of machine remained unaware of the problem and another
accident occurred seven weeks later in Canada.

60
2.4. Computer file not updated for Co source change (USA, 1987-88)

(37) A computer programme file used for the treatment of brain cancer with
‘trimmer bars’ was not updated in March 1987 when the therapy department
replaced the 60Co source. Other programme files had been revised, but this afore-
mentioned file had not, because ‘trimmer bars’ were not used at that time to treat
patients with whole brain irradiation. However, in September 1987, brain treatments
were restarted and the ‘trimmer bars’ computer file with the data corresponding to
the prior source was used.
(38) As a consequence of the error, 33 patients undergoing radiation therapy to
the brain had received radiation doses that exceeded the prescribed dose by 75%. By
the time of the notification of the case to the State Agency, 20 patients had died
either during or after the conclusion of treatment (NRC, 1988; 1996).

2.5. Incorrect repair followed by lack of communication (Spain, 1990)

(39) Following instability of the radiation beam of a linear accelerator, the inter-
lock system terminated the treatment of a patient. The problem had arisen at the
beginning of a long holiday weekend. An engineer from the maintenance company
who had been adjusting an adjacent 60Co therapy unit was requested to ‘have a
look’ at the accelerator. Over the long weekend the maintenance engineer made
several unsuccessful attempts to identify the cause of the machine fault. Eventually,
he manipulated the energy control system to be able to restore the radiation beam.
(40) The following Monday, patient treatments resumed. There were established
procedures for notifying the hospital’s maintenance engineer and medical physicists,
but these were not followed. If they had, beam tests would have been performed, as
required after any repair that could affect the dosimetry of the accelerator.
(41) A control panel meter indicated 36 MeV permanently, regardless of the energy
selected; i.e., for an electron energy selection of 10 MeV the selector light button
indicated 10 MeV but the control panel meter indicated 36 MeV. The radiotherapy
technologists operating the accelerator pointed out the discrepancy between these two
indications to the maintenance engineer, but his explanation was that the meter of the
control panel was mechanically stuck on 36 MeV. He did not realise that his attempts
to restore the radiation beam had forced the beam energy control to be at its max-
imum. He failed to compare the indication of the control panel meter with any other
parameter related to the accelerator energy (e.g., on the modulator in the machine
room). He simply assumed that the control panel meter was stuck and that the actual
energy was the one indicated by the selector light button.
(42) The reality was the opposite: the control panel meter indicated the actual energy
and the energy selection button did not work at all. Treatments, therefore, resumed
without any beam measurements. The medical physicists were only notified of the
26
 ICRP Publication 86

discrepancy between the two indicators on the 10th day after resuming treatments.
They measured the beam output and energy and concluded that both were incorrect.
(43) The consequences of the mistake were relevant only to treatments with elec-
tron beams, for which the delivered dose and beam energy (and therefore the elec-
tron dose distribution) were substantially different from the selection made at the
operating console of the accelerator. The penetration of the electron beams was that
of 36 MeV electrons even when smaller penetration depths were intended by select-
ing low energies; therefore, large doses were delivered to tissues deeper than inten-
ded. The dose delivered was also much higher than intended due to the ‘focusing’
effect of electrons into a smaller cross section of the beam. The dose rate was
between 3 and 7 times higher than intended, depending on the selected energy.
(44) The subsequent court proceedings (El Periódico, 1993) established that during
the 10 days of malfunction, 27 patients had been treated with electrons; 15 died with
radiation as the primary cause, two died with radiation as a major contributor to
their death and others suffered major disability. The main radiation effects con-
tributing to death were respiratory insufficiency, bronchopneumonia, radiation
myelopathy, oesophageal stenosis and perforation, renal insufficiency, cervical hae-
morrhage, and arteriosclerosis. Major effects on patients alive at the time of the
court decision were paralysis and severe disability.

2.6. Malfunction of brachytherapy high dose rate equipment (USA, 1992)

(45) A patient was to be treated using a high dose rate (HDR) brachytherapy unit
equipped with a 16 GBq (4.3 Ci) 192Ir source. The prescribed dose was 18 Gy in
three fractions. Five catheters were placed in the tumour and the source was to be
stepped through the pre-programmed positions in each catheter. During the first
fraction the radiation oncologist experienced difficulties in positioning the source
into the fifth catheter and decided to retract the source. The source became detached
from the driving mechanism while still inside the patient.
(46) The staff disregarded an alarm from an external area radiation monitor
because the console of the brachytherapy unit indicated ‘safe’. All three technolo-
gists and one physician who were attending the patient were aware of the alarm
condition but none of them conducted a survey with the available portable radiation
survey instrument (NRC, 1992).
(47) The patient, with the source still in the catheter, was transported back to the
nursing home. The source remained inside the patient for almost four days, until the
catheter containing the source fell out. The patient received a dose of 16,000 Gy at 1
cm distance from the source, instead of the prescribed 18 Gy. The nursing home staff
disposed of the catheter in an area used to store non-radioactive medical waste and
was removed later by an incinerator company. The source was discovered when it
tripped a radiation monitor located at the incinerator.
(48) The patient died shortly after the source was dislodged. The overexposure
was the major contributing cause of death. The lost source also caused radiation
exposure to 94 other individuals, including persons at the cancer clinic, nursing
home, ambulance staff, and workers at the waste-disposal company. A similar accident
27
 ICRP Publication 86

in another hospital was subsequently avoided because the medical physicist was
aware of the first case and immediately recognised the problem. This emphasises the
importance of incident reporting and dissemination of the lessons learned.

60
2.7. Beam miscalibration following the exchange of a Co source (Costa Rica, 1996)

(49) A hospital had two 60Co teletherapy units. Inconsistencies in dosimetry at the
hospital had been detected before the accident by the IAEA/WHO TLD postal dose
quality audit service. An external expert was sent who confirmed the deficiencies in
dosimetry procedures and the lack of appropriate quality assurance. These findings
were reported to the hospital during the audit, but were not acted upon. Insufficient
education in medical physics, lack of an independent calibration, the lack of a
quality assurance programme, the absence of documented procedures, and the lack
of awareness at the hospital, allowed mistakes to remain unnoticed.
(50) After a source exchange in one of the 60Co machines the medical physicist
incorrectly calibrated the beam output; using a time of 30 s instead of 0.3 min (18 s)
to determine the absorbed dose rate. This resulted in a 66% overestimation of the
exposure time and therefore in an underestimation of the dose per unit time, which
led to longer treatment times.
(51) The radiotherapy technologists questioned why patient treatment times
remained about the same as with the old source when a new source was in place. The
medical physicist told them that the delivered doses were correct according to his
calculations. The radiation oncologist saw the patients only occasionally and
ignored the high incidence of side effects such as vomiting, diarrhoea, and rectal
bleeding. Over a period of about one month 114 patients received considerable
overdoses, resulting in severe health effects including fatalities. The accident was
finally detected by another radiation oncologist at a different hospital who noted
severe complications in his patients that had been treated on that machine.
(52) Two assessments of the consequences were performed by an international
medical team, one and two years after the accident (IAEA, 1998b). Of 51 patients
who died within the two years after the accident, 13 were radiation related and 4
possibly related. Of the 52 patients who where alive two years after the accident, 4
showed ‘severe or catastrophic’ effects due to radiation overexposure and 12 showed
marked effects, with a high risk for future effects. The radiation effects found in
patients that survived two years after the accident were:

. Nervous system: Brain-atrophy, necrosis, decreased cognitive function, head-

aches, mood alteration, seizures, decreased intellectual function. Spinal cord
paralysis, quadriplegia, paraplegia.
. Skin: fibrosis, atrophy, contraction, induration, edema, pigmentation, puritis,
hypersensitivity, pain.
. Lower gastrointestinal: chronic or bloody diarrhea, bowel stenosis, stricture,
fibrosis, obstruction, fistula perforation.
. Bladder: dysuria, hematuria, contracture, incontinence.
. Vascular and lymphatic: stenosis, premature atherosclerosis.
28
 ICRP Publication 86

Figure 1 (see frontispiece) shows some of these effects. In addition, there were
considerable psychosocial effects (IAEA, 1987).

29
 3. CLINICAL CONSEQUENCES OF ACCIDENTS IN RADIOTHERAPY

3.1. Side effects and complications—normal therapy versus accidental exposures

(53) In radiotherapy, a number of normal tissue side effects or complications are

expected and accepted as part of treatment since some normal tissues will receive
doses of the same order as the tumour dose. This cannot be avoided in most cases.
Fig. 3 illustrates this principle (adapted from Pérez and Brady, 1992). The figure also
shows that a departure from the prescribed optimum dose will result in either an
increase in the number of complications or in a decrease in tumour control.
(54) Side effects are usually considered to be minor and transient. Complications
are more severe and long lasting. Side effects can be accepted at a high frequency
while complications will only be acceptable at a lower frequency. Traditionally, tol-
erance doses in radiotherapy have been set at a ‘5-5’ level, i.e. 5% complications at 5
years (Rubin, 1968). This approach is presently changing, for a number of reasons:
. Firstly, the ‘acceptability’ of side effects and/or complications actually varies
according to their type. In cancer treatments, even a high percentage of rela-
tively minor side effects, such as xerostomia or localised subcutaneous fibrosis,
is acceptable. In a survey, a 5% rate of very severe complications (such as

Fig. 3. The solid curves represent the probability of tumour control (TCP) and of normal tissue complica-
tion (NTCP) versus delivered dose. The dashed curve represents the probability of uncomplicated tumour
control with its maximum at dose B. This dose represents the optimal balance between local tumour abla-
tion and an acceptable incidence of side effects or complications. Dose C could achieve 100% of cure rate
but, at this high level, because of complications most patients would not survive the treatment. At dose A
there would be no complications, but the probability of controlling the tumour would be very small.

31
 ICRP Publication 86

radiation myelitis) was considered as unacceptable by most radiation oncolo-

gists in curative radiotherapy (Sebag-Montefiore, 1992).
. Secondly, techniques of radiotherapy are regularly being improved; the recent
developments in conformal radiotherapy have allowed radiation oncologists to
increase tumour dose to improve cure rate, without concomitant increases in
toxicity.

(55) These side effects and/or complications of radiotherapy (which are acceptable
in cancer treatment) should be clearly distinguished from the detrimental con-
sequences of an accidental overdosage. The former are an intrinsic part of the pre-
scription, representing a benefit-risk assessment. The latter are unplanned events
which have no relation to a benefit-risk analysis.
(56) An accidental exposure in radiotherapy can also be the result of an unplanned
and significant underdosage of the target volume. In this case, side effects or com-
plications will be absent or perhaps minor in normal tissues, but the probability of
cancer cure may be substantially decreased.

3.2. Consequences of accidental exposures in radiotherapy

(57) The consequences of accidental exposures can be categorised into three types:
(i) impact on local tumour control rate, (ii) early (or acute) complications, and (iii)
late (or chronic) complications. These consequences will be briefly discussed in the
following subsections, concentrating on the physical and biological parameters that
affect local tumour control and toxicity.

3.2.1. Impact on tumour control rate

(58) The delivery of a tumour dose significantly lower than the prescribed dose, as
a result of an accident, can severely jeopardise the probability of curing the patient.
Although infrequently reported in the literature, this can lead to the death of the
patient (or a group of patients; see case history in Section 2.2) due to disease pro-
gression. Such errors are frequently the cause of lack of tumour control. These
errors may not be identified for a long time, and consequently may involve a large
number of patients.
(59) In the case of an accidental overdosage, the tumour control probability may
increase. However, there will be an associated overdose to normal tissues, as
described in Sections 3.2.2 and 3.2.3, leading to death or to a severely reduced
quality of life. The choice of the prescribed dose is based on an acceptable benefit/
risk ratio (see Section 3.1), and any unintended deviation is undesirable.

3.2.2. Early (or acute) complications

(60) Acute complications are early deterministic effects (effects due to cell killing).
They are dose related and have a threshold (i.e., below a certain dose they are not
seen). These effects are usually observed in tissues or organs with rapid cell turnover
32
 ICRP Publication 86

rates (e.g., skin, mucosa, and bone marrow). These complications are observed
within days or weeks after irradiation. They are often transient.
(61) The main determinants of early effects are: (1) the delivered dose; (2) the total
duration (protraction) of the radiation treatment (because of the ability of rapidly
proliferating normal tissues to compensate for cell loss during the few weeks of a
conventional, or accidental, irradiation); and (3) the size and location of the irra-
diated volume- or percentage of the irradiated organ. In contrast, there is a low cor-
relation between early complications and fraction size or dose rate, except when the
latter is very high.
(62) In overdosage accidents, the severity of the early side effects or complications
is increased. Ultimately, if the overdose is very high, all tissues within the irradiated
volume will be destroyed (e.g., radionecrosis will occur with skin doses over 25 Gy in
one fraction, particularly if the field is larger than 2 cm diameter).

3.2.3. Late (chronic) complications

(63) Complications of this type are also deterministic effects and have a threshold
(see Fig. 3). These late effects are mainly observed in tissues or organs with slowly
proliferating cells. They can also be seen in organs with rapidly proliferating cells, a
consequence of very severe acute reactions (‘consequential’ effects; Maciejewski,
1990). The physiopathology of these late complications is complex, a combination of
cell loss, alteration of supportive tissues (e.g., capillaries, glial cells), and the pro-
gressive and chronic development of radiation fibrosis. These late complications
usually occur more than six months after the end of irradiation, but can be observed
much later (several years). They are usually considered as irreversible, and are often
slowly progressive.
(64) The delivered dose is a major determinant of these late effects. The slope of the
dose/effect curve is even steeper than for early (acute) effects. Fractionation or frac-
tion size (for external beam) or dose rate (for brachytherapy) has been shown, both
experimentally and in practice, to have a major impact on late effects, even with
relatively modest changes in fraction size or dose rate.
(65) In the case of an accidental increase in fraction dose, this ‘fractionation sen-
sitivity’ of late effects amplifies the toxicity of an already increased total dose
(Thames et al., 1987; Cosset et al., 1994). In some accidents, increased doses per
fraction have clearly amplified the severity of the injuries (see Section 2.7). Treat-
ment with only one beam, when the prescription is for several beams every day,
results in relatively high fractional doses to some tissues close to the entrance beam.
This increases the late effects, similar to using larger doses per fraction. In accidental
overdoses, this further increases the severity of complications. Similarly, in low-
dose-rate brachytherapy, an accidental increase in dose rate enhances the severity of
late complications for the same total dose.
(66) The severity of the observed late complications is closely related to the
volume of tissue irradiated above the threshold in organs where subunits are arran-
ged in parallel, such as lung and liver (see Fig. 4). In contrast, in serially arranged
organs such as spinal cord, intestinal tract or large arteries, a lesion in a subunit
33
 ICRP Publication 86

Fig. 4. Schematic examples of tissue organisation structures in the parallel-serial model: (a) a serial string
of subunits (e.g., the spinal cord); (b) a parallel string of subunits (e.g., the lungs); (c) a serial-parallel
string of subunits (e.g., the heart) (d) a combination of parallel and serial structures (e.g., a nephron).
From ICRU (1999).

significantly affects the function of the whole organ (see Fig. 4). In such serially
arranged tissues irradiation of a very small volume above the threshold may lead to
highly incapacitating complications. Complete paralysis caused by a radiation-
induced transverse myelitis is an example of a major complication related to the
irradiation of a very small10 volume in a serially arranged organ. In contrast, total
duration (protraction) of irradiation has a minor role in late complications (except in
the case of ‘consequential’ late effects, after very severe acute reactions).

10
While most published work regards spinal cord as a serially arranged tissue, some authors consider
that this may not be strictly correct (Powers, 1998). In practice, these considerations should not lead to
underestimation of the consequences of irradiation of these organs.

34
 ICRP Publication 86

(67) In summary, in overdosage accidents, where the patient survives the early/
acute period, late complications will mainly be related to total dose, to the dose per
fraction, to the type of tissue and to the volume of the organ irradiated. The duration
of the radiation treatment (protraction) has only a limited impact. These late com-
plications can lead to very severe detriment for the patient and can be fatal.
(68) In addition to the late (deterministic) complications, secondary malignancies
are also late (stochastic) effects of irradiation. These effects are not due to cell-killing
but are related to mutations. An increase in dose increases the probability of second
cancers. However, the precise relationship between dose and carcinogenic risk in the
radiotherapy dose range is still debated. When these secondary radiation-induced
malignancies occur their severity is independent of the radiation dose.

3.3. Impact of individual radiosensitivity

(69) In the past, the claim that some individuals were hypersensitive to ionising
radiation was frequently used as an explanation for some unexpected side effects or
complications. However, it is probable that most such cases were related to accidental
overdosages. The development of quality assurance programmes has demonstrated
that the majority of unexpected side effects and complications were due to the over-
doses (accidents) and not to individual increased sensitivity to radiation. A very small
subgroup of cancer patients can be identified for whom individual hypersensitivity
may pose a problem during (or after) radiotherapy. However, this can only be
established after careful and thorough elimination of the possibility of an overdose.
The number of hypersensitive patients in a cancer population may be in the range of
0.5–3%. This percentage may be much higher than that estimated for the normal
population who do not have cancer. The reasons for this are: (1) patients with well-
known and very rare hypersensitivity syndromes, such as ataxia-telangiectasia, are
much more likely to develop a cancer than non-affected individuals; and (2) patients
with genetic profiles predisposing to cancer have been related to alterations of DNA
repair mechanisms, and consequently they may be hyper-radiosensitive (ICRP,
1998).
(70) Within this small group of cancer patients, two subpopulations can be dis-
tinguished. The first one corresponds to the known rare hypersensitive syndromes
(essentially ataxia-telangiectasia and Nijmegen breakage syndrome). Since the diag-
nosis of these syndromes precedes the emergence of a cancer in almost all cases, it
should be possible for the radiation oncologist to modify the therapy (Sharp, 1999).
In some cases, radiotherapy will be simply avoided, in others, fraction size and total
dose will be reduced (by a factor of about 3–4).
(71) The second group of hypersensitive patients, which represent the majority,
cannot be clinically distinguished a priori from other patients. Currently, an increase
in early and/or late reactions is the only way to identify such patients. Recent studies
suggest that in some of these patients, alterations of DNA repair mechanisms could
be responsible for their enhanced radiosensitivity, but this is not the only factor. To
date, it has been difficult to find assays that will consistently predict radiosensitivity
in these groups before radiotherapy commences.
35
 ICRP Publication 86

3.4. Clinical detection of accidental exposures in radiotherapy

(72) Careful clinical follow up may detect an overdose accident before the total
treatment course is completed. Overdoses often cause early/acute, abnormal
(enhanced) reactions. These should be promptly detected by an experienced radia-
tion oncologist during regular weekly consultation. As mentioned earlier, dose var-
iations as low as 7-8% have been clinically detected (Dutreix, 1984). When an
unexpectedly severe reaction is seen, it is essential to assess all patients treated on the
same machine, and possibly in the same unit/department.
(73) Some overdoses will cause late severe effects without abnormal acute (early)
effects. Consequently it will be only possible to detect such accidents during long-
term follow-up. Regular follow-up is therefore of paramount importance. In the
case of unusual reactions in a single patient, it will be useful to recall all other
patients treated at the same period of time with the same machine, and possibly in
the same unit/department.
(74) In underdose accidents, clinical detection is much more difficult. However, an
experienced radiation oncologist should be able to identify a decrease in the nor-
mally observed early side effects, especially if this affects several patients at the same
time. When the decrease in side effects or complications is insufficient to attract
attention, decreased tumour control in a number of patients may raise concern that
underdoses have been given. Evaluation of the mid and long term results of radia-
tion treatments should be undertaken and may allow the detection, even if late, of
some types of accidents.
(75) When there are very severe consequences of an accident, such as massive and
extended radionecrosis, the management is often complex and very difficult, and the
results of treatment are frequently disappointing. Massive radiation-induced fibrosis
is almost irreversible, and large areas of necrosis may require complex therapies,
including reconstructive surgery. This reinforces the requirement to detect an acci-
dent at the earliest possible opportunity, in order to minimise complications and
fatalities.

36
 ICRP Publication 86

4. CAUSES OF AND FACTORS CONTRIBUTING TO

ACCIDENTAL EXPOSURES IN RADIOTHERAPY

(77) Event reporting is essential to improving safety. Dissemination of informa-

tion on radiotherapy accidents and their causes will help to prevent future acci-
dents. This will only be effective if the information has been compiled and analysed.
Each accident is triggered by an initiating event. Such events normally do not pro-
gress to an accident if a well-designed quality assurance programme is in place. Well-
designed quality assurance programmes contain sufficient layers of safety (physical or
procedural or both) to prevent progression of initiating events into accidents.
(78) This section provides a short description of initiating events together with a
summary of the main contributing factors and omissions. These are used to provide
recommendations on accident prevention that are given in Chapter 5. The informa-
tion is from published papers on individual accidents, as well as from reports issued
regularly by the U.S. Nuclear Regulatory Commission (NRC) and a safety report of
the International Atomic Energy Agency (IAEA, 2000a).
(79) The information in Sections 4.1 and 4.2 concerning external beam therapy
and brachytherapy is classified into the groups described in Table 3, ordered to align
with the treatment process. The topic of radiation sensitive patients was discussed in
Section 3.3 and will not be revisited below. Public exposures and environmental
contamination are discussed briefly in Section 4.3. Section 4.4 summarises generic
lessons learned.

4.1. External beam

4.1.1. Equipment problems

(80) There have been software problems leading to equipment faults under certain
operating conditions, as described in Section 2.3. Accidents with the same type of
accelerator occurred in six different occasions over a period of two years and three
patients died. The contributing factors were:

(i) a software package was transferred from an from an older accelerator type
and without effective assessment of all safety implications;
(ii) the equipment faults were difficult to reproduce; and
37
 ICRP Publication 86

Table 3. Classes and frequencies of accidental exposure in radiotherapy

Accidental exposures in external beam therapy No. of Percentage of cases

cases (rounded)

Equipment problems 3 6.5

Maintenance 3 6.5
Calibration of the beams 14 30
Treatment planning and dose calculation 13 28
Simulation 4 9
Treatment set-up and delivery 9 20 (**)
Total 46 (*) 100

Accidental exposures in brachytherapy No. of cases

Equipment and source problems 5 15
Source order and delivery, calibration, and acceptance 3 9
Source storage and preparation for the treatment 5 15
Treatment planning and dose calculation 6 18
Treatment delivery 11 34
Source removal and return 3 9
Total 33 (*) 100

*The number of accidents in the table are fewer in number than in the source publications, since the
source publications include events with unsealed sources and accidents involving the public.
**It is likely that errors in the treatment set-up are more frequent than tabulated, since many instances
probably remain unreported, especially if the consequences are moderate, i.e., affecting one or a few
fractions.

(iii) the difficulties in identifying the cause led to a long delay in disseminating
warnings and taking corrective actions

4.1.2. Maintenance

(81) There have been two major accidents related to maintenance problems. Both
accidents resulted in deaths. In the first of them, the initiating event was the mal-
adjustment of the electron energy in an accelerator (see the case history in Section
2.5). The second accident was due to intermittent failures, followed by frequent
interruption of treatments for repair work and several unsuccessful attempts to repair
the machine. This situation eventually led to interlocks being disabled. Contributing
factors were:

(i) Insufficient knowledge on the part of the maintenance engineer about the
consequences of manipulating beam parameters, such as the energy, and
deficiencies in training and expertise in diagnosing the cause of equipment
faults.
(ii) Transfer of the machine from the hospital staff to the maintenance engineer
and return without communication to the medical physicists about the
maintenance. Treatment resumed without dosimetric check of the beam.
38
 ICRP Publication 86

(iii) Operation with the energy selector or other key features disabled.
(iv) Conflicting display and signals were ignored.
(v) Lack of equipment for quick constancy checks.
(vi) Intermittent faults, difficult to reproduce, identify, and repair.

4.1.3. Beam calibration

(82) The most important factor in accidental exposure in radiotherapy is an error

in the determination of the machine ‘output’ (dose per time unit). Typically, these
occur when there is an error in the beam calibration, or incorrect calculation of
decay. Other errors involve misinterpretation of calibration certificates and errors in
correction for atmospheric pressure. In one case, data reported by the weather sta-
tion was corrected to sea level but the user assumed that the data was related to
pressure at the level of the weather station. In a different accident, a 60Co beam was
used without calibration (the exposure value in Roentgen from the radiation source
certificate was simply taken as cGy). In yet another accident, a plane-parallel
chamber was used incorrectly (upside down), because the label was misplaced and a
new physicist was unfamiliar with the chamber.
(83) Factors contributing to these accidents were:

(i) Insufficient training on beam calibration procedures; insufficient under-

standing of calibration certificates, dosimetry equipment, and conditions to
determine atmospheric correction factors.
(ii) Lack of redundant and independent absorbed dose determination.
(iii) Lack of clear procedures and protocols, and of overall supervision of com-
pliance with procedures.
(iv) Changes of personnel (physicist) with poor communication and transfer of
information between staff.

4.1.4 Treatment planning systems

(84) Two major subgroups of accidents can be distinguished in this area: (a) acci-
dents related to the commissioning of a radiotherapy treatment planning system
(TPS) or entry of incorrect basic common data, and (b) accidents due to mistakes
with individual patients. Two accidents of the (a) type were discussed in Chapter 2.
In the case in Section 2.2, the event affected 1,045 patients and was related to
operation of the TPS without proper commissioning. In the case in Section 2.4, there
was a failure after a source change to enter new data into one of the computer files
for certain type of treatment.
(85) Factors contributing to group (a) accidents were:

(i) Insufficient understanding of the treatment planning system, TPS (no train-
ing of the staff on the new TPS).
(ii) Lack of formal commissioning of the TPS (no comprehensive operational
tests before using the TPS for treating patients).
39
 ICRP Publication 86

(iii) Lack of an independent check of the planning (either by manual calculations

to selected points or by measurement in a phantom).

(86) Factors contributing to group (b) accidents were:

(i) Lack of understanding of the TPS.

(ii) Lack of independent check of the treatment planning.

4.1.5 Treatment simulation

(87) In one incident the wrong side of the patient was treated due to incorrect
labelling of the simulator film. As a result 2 Gy was given to the right side instead of
the (intended) left side.
(88) Contributing factors were:

(i) Treatment simulation in an unusual position.

(ii) No check of the anatomical site relative to the film.

4.1.6 Treatment set-up and delivery

(89) In one accidental exposure in this category, a patient responded when another
patient was called and a fraction of 2.5 Gy was given to the wrong patient. In other
accidental exposures patients have been treated at the wrong anatomical site as a
result of a variety of reasons, including using the wrong chart, using a tattoo port
marker from a previous treatment, and reliance on asking the patient to identify
the site. In the latter circumstance, the patient received a brachytherapy treatment
with a strontium plaque to the eye instead of the prescribed 10 Gy external beam
treatment; irradiation was initiated with the settings for rotational therapy from
a previous patient; a technologist continued treatment with more fractions than
prescribed.
(90) Contributory factors were:

(i) Procedures for identification of patients and the correct chart were not fol-
lowed, including lack of verification of the treatment site against anatomical
marks on the patient, and patient objections about being treated on the
wrong site were not thoroughly investigated prior to treatment.
(ii) The oncologist relied only on asking the patient for the treatment site.

4.2 Brachytherapy

4.2.1. Equipment problems

(91) In the most serious case, with fatal consequences, the source became detached
from the drive mechanism of a high dose rate (HDR) machine. In another equipment
40
 ICRP Publication 86

failure, the wrong site was treated because a kink in the catheter prevented the
source from reaching the correct position.
(92) Contributing factors in brachytherapy accidents due to equipment problems
were:

(i) Equipment was not tested for reasonably foreseeable conditions.

(ii) Conflicting signals were misinterpreted (equipment indicated ‘source
shielded’, but area monitor detected radiation) and the wrong indication
was accepted.
(iii) Previous radiation monitor malfunctions encouraged misinterpretation and
induced the staff not to trust the radiation monitoring device.
(iv) The patient, clothes, and room were not checked with a radiation monitor to
establish the location of the radiation source.

4.2.2. Source ordering, delivery, calibration, and acceptance

(93) In one accidental exposure, the hospital and manufacturer used different units
of activity (mCi and mg-Ra-equivalent respectively), and three cases of underdosage
occurred because of failure to check the actual source activity. In another accidental
exposure uncalibrated sources were used for many years, affecting many patients,
with dose deviations of 5% to 29%.

(94) Contributory factors were:

(i) Delivery of ‘wrong’ sources by the supplier, i.e., the sources did not corre-
spond to the accompanying source certificate.
(ii) Use of different units for source activity.
(iii) Lack of verification of the documents on source delivery against the docu-
ments of the purchase order.
(iv) Lack of verification of source strength before use.
(v) Interchangeable use of sources without consistency checks.
(vi) Usage of incorrect conversion factor on switching of sources from 226Ra to
137
Cs.

4.2.3. Treatment planning

(95) Errors in this area have included incorrect time calculations, resulting in dose
deviations ranging from 59% to +49%. Contributing factors were:

(i) Copies of an obsolete form were still available for clinical use.
(ii) Treatment plan did not include the reference point for the time calculation of
a brachytherapy treatment.
(iii) Miscommunication between radiation oncologist, physicist, and dosimetrist
(a treatment plan was modified but the unmodified plan was used).
(iv) Lack of independent time calculations.
41
 ICRP Publication 86

4.2.4. Source preparation

(96) Incorrect or defective source use has caused some accidental exposures. In one
case there was 50% of the prescribed dose, and doses were lower than intended
over three months in another case. In other different cases, a manufacturer had
delivered a source with essentially no activity; two 192Ir sources were lost (they were
separated from the ribbon and left unsecured), a leaking 125I source was re-used, and
sources that had been withdrawn from clinical use were used with an incompatible
applicator. Factors contributing to these accidents were:

(i) Personnel handling sources and applicators lacked proper training.

(ii) Sources withdrawn from clinical use were not removed and were re-used by
mistake.
(iii) Lack of verification of source activity.
(iv) Leakage of a source was not detected during preparation, and a similar
incident that had occurred in another hospital had not triggered verification.
(v) Failure to identify the correct source ribbon end.
(vi) Lack of survey of all radiation sources before implantation.

4.2.5. Treatment delivery

(98) Examples of this type include: a junior physician who failed to implant one of
the prescribed sources; the wrong patients were treated; a source ribbon was dis-
lodged from the catheter and a nurse taped it on the face of the patient; sources were
removed by a patient; and a source that did not match the applicator was loose and
fell out of the applicator.
(99) Contributing factors were:

(i) An untrained physician worked without supervision; ordinary nurses with no

special training were nursing brachytherapy patients; there were no written
procedures; and poor communication of instructions that were not understood.
(ii) A prescription was misunderstood.
(iii) Lack of procedures for accountability of the sources.
(iv) The wrong chart was left on the console of the remote afterloading machine
and used without verification.
(v) Errors were made with infrequently used treatment protocols.

4.2.6. Source removal

(100) In addition to the case described under 4.2.1 (HDR sources left inside the
patient), there were several cases in which sources were lost due to lack of radiation
monitoring after presumed source removal.
Contributing factors were:

(i) Sources were not accounted for after removal.

42
 ICRP Publication 86

(ii) The patient, clothes, and room and/or waste from the treatment room were
not monitored.
(iii) In one case, the sources were checked after removal against the total number
of sources implanted, but not against the total number of sources that were
sent to the room (more sources were sent than needed).

4.3. Accidents involving public exposure and environmental contamination

(102) The unsecured long-term storage of radiotherapy sources has led to cata-
strophic accidents with severe exposure and deaths among members of the public.
These occurred as a result of hospital management, source suppliers, and importers
relinquishing the accountability for sources.
(103) Contributing factors were:
(i) Non-compliance with regulations for transport and/or import.
(ii) Poor storage conditions before commissioning, or after decommissioning, of
significant radiation sources.
(iii) Provisional short-term storage arrangements turning into long-term storage.

4.4. Generic lessons learned

(104 ) In most of the accidents, a combination of contributing factors allowed an

initial mistake to escalate into an accidental exposure. In some cases this resulted in
very serious or fatal consequences.
(105) Often, the lack of concern of management was the underlying root cause.
When this situation is present it leads to many contributing factors including: lack of
appropriate staff resources; insufficiently qualified or untrained staff; lack of effective,
systematic quality assurance programmes/procedures, and lack of effective commu-
nication procedures. Often the need for reassessment of staff, resources, and training
were ignored when a new machine was purchased or a new technology was intro-
duced, or when the workload increased. Hospital management, source suppliers,
and importers can cause catastrophic accidents involving the public and severely
affecting the environment by relinquishing accountability for sources. These cases
usually violate transport, import, storage, and decommissioning regulations.

43
 5. RECOMMENDATIONS FOR THE PREVENTION OF ACCIDENTAL
EXPOSURE IN RADIOTHERAPY

(106) It has been emphasised in multiple publications (cf. Hanks, 1984; WHO,
1988; AAPM, 1994; Thwaites et al., 1995; ESTRO, 1995; Alletti and Bey, 1996;
IAEA, 1997; 1998a) that systematic quality assurance programmes in radiotherapy
(QART) can prevent systematic errors and decrease the frequency and size of ran-
dom errors. Prevention of most accidental exposures can be assured with minimal
effort and expense in a radiotherapy department when two conditions are fulfilled:
(i) a comprehensive and coherent quality assurance programme is in place, and (ii) for
external beam radiotherapy, some in-vivo dose measurements are routinely performed.
(107) The first quality assurance programmes in radiotherapy dealt only with the
verification of mechanical and electrical parameters of equipment and of dosimetric
data. QART programmes have been extended progressively to include the verification
of treatment planning procedures, patient set-up, and treatment delivery. Recently, the
concept of quality assurance in radiotherapy has been extended to the verification of
the entire radiotherapy process, from treatment prescription, patient data acquisition,
and target volume delineation, to patient follow-up and treatment records. Some
modern QART programmes also include recommendations for the structural orga-
nisation of a radiotherapy department and the qualifications and training of its staff.
(108) It is advisable that countries enforce regulations requiring radiotherapy depart-
ments to implement a comprehensive quality assurance programme. Some countries
have adopted regulations on quality assurance in radiotherapy, making it compulsory
to verify beam calibration by an external audit using mailed dosimeters (France) or
have implemented mandatory periodic quality audit site visits, where independent
calibrations and other type of measurements are made (Finland). However, it is
important to emphasise that the development of QART programmes and the verifica-
tion of their correct application is the responsibility of managerial and professional
staff, mainly radiation oncologists and medical physicists. Radiotherapy technol-
ogists and dosimetrists have an important role in the application of the programme.
(109) It is not the purpose of this Chapter to provide detailed guidance on the
contents of a quality assurance programme for radiotherapy, which are to be found
in numerous publications11 (c.f. WHO, 1988; AAPM, 1994; ESTRO, 1995; Alletti
and Bey, 1996; IAEA, 1998a). Existing QART programmes provide comprehensive
lists of parameters to be checked periodically, specifying the tolerance of the results
measured and recommending the frequency of various checks, etc. They also
emphasise the importance of double checks and independent verifications that pro-
vide defence in depth.
(110) The prevention of accidental exposures does not require an increase in the
frequency or the type of checks, beyond the normal recommendations given in modern

11
Most of the QART programmes are in accordance with guidelines provided by the ISO 9000 family
of standards for quality assurance (cf. ISO), 2000) and with IEC standards for radiotherapy equipment
(IEC, 1988; 1989; 1993; 1997; 1998; and in press), which have been extended to include recommendations
on the proper use of equipment by the radiotherapy staff.

45
 ICRP Publication 86

QART programmes, but rather the systematic application of the recommendations

of a given programme. QART programmes are generally designed to detect minor
errors and, as a consequence, they can also prevent the occurrence of major errors. The
majority of the accidental exposures in radiotherapy have occurred in departments
without a quality assurance programme or, when QART programmes existed, they were
not fully implemented (omitting, for example, some of the checks to be performed).
(111) The aim of this Chapter is to provide recommendations of particular impor-
tance for the prevention of accidental exposures. The text follows the normal sequence
of the radiotherapy process. Recommendations on staff functions, on equipment-rela-
ted requirements and patient workload requirements for staff resources, and on the
requirements for staff qualifications, can be found in numerous publications issued by
national and international organisations (cf. WHO, 1988; IAEA, 1998a; ISCRO, 1991
for USA; ESTRO, 1996 for Europe). These can be adapted, if necessary, to the
structure available in many countries. For the purpose of emphasising certain
aspects of the staff responsibilities that will help to prevent accidental exposures, a
summary of these responsibilities in relation to the education and training of the
radiotherapy staff is included in this Chapter.

5.1. Structural organisation

(112) Radiation therapy is a multidisciplinary speciality, which uses complex equip-

ment for the delivery of treatments to patients. A comprehensive quality assurance
programme has clinical, physical, and administrative components whose implementa-
tion requires teamwork from all the professionals involved in the radiotherapy process.
(113) Key staff functions include radiation oncologist, medical physicist, radio-
therapy technologist, and dosimetrist. Other categories of staff members that may be
associated with a radiotherapy department, or may participate in the different steps
of radiotherapy procedures, are mould room technician, nurse, maintenance engi-
neer, etc. Special consideration should be given to staff workload, which may
become excessive in the case of installation of new equipment, set-up of a new
technology, or occasionally an increase in patient load. This will lead to an increased
risk of accidental exposures, as a result of the difficulty of fully applying the quality
assurance programme in these situations.
(114) The structure of a radiotherapy department should be clearly defined, espe-
cially with regard to the role and responsibilities of each staff member. In addition,
the relationships between staff members and groups of professionals, either hier-
archic or functional, should be clarified so that each individual understands her or
his own position in the structure, as well as that of others. All decisions affecting the
structure of the radiotherapy department should be recorded and the information
should be disseminated effectively and quickly throughout the department.

5.2. Education and training

(115) A large number of accidental exposures have occurred because of the lack
of qualified and well-trained staff. The most important component of the entire
46
 ICRP Publication 86

radiotherapy process is qualified personnel. It is vital that all the staff dealing with
radiation sources and patients have the necessary educational background and
specialised training. Investment in equipment without concomitant investment in
training is dangerous. Training should not only include practical details of indivi-
dual procedures, but also the design of treatment approaches, which are compre-
hensive, reproducible, of high quality, and safe. Training should be consistent with
the responsibilities assigned to each group of professionals, and should include the
review and analysis of typical accidents together with a description of methods of
prevention. There should also be continuing professional development programmes.
Suggested guidelines for the education of staff members involved in radiotherapy are
given below (cf. IAEA, 1998a) together with some aspects of their responsibilities
related to accident prevention.

5.2.1. Radiation oncologists

(116) The physician practising radiation therapy should first be trained and
experienced in oncological practice with post-graduate training in radiation oncol-
ogy. The radiation oncologist will set the overall treatment policy for the radiation
therapy programme and should participate in the design of the facility and the pro-
curement of equipment. For individual patients, the radiation oncologist is respon-
sible for the patient’s care, including the details of the treatment and the patient’s
follow-up evaluation.
(117) The radiation oncologist who also practices brachytherapy should first be
trained as a radiation oncologist and experienced in oncological practice. He/she
should also have specific training in brachytherapy at an institution with an estab-
lished practice, so that the indications for patient selection, applicator insertion, and
dose prescription can be learned under the supervision of experienced mentors. The
length of this training will depend on many factors, but will usually be measured in
months. Such a training period should be undertaken whenever a substantially new
form of brachytherapy is introduced into an existing practice, for example when
adding high dose rate brachytherapy.

5.2.2. Medical physicists

(118) The medical physicist should have at least an advanced university degree in a
physical science or engineering, at least one year of academic and clinical training in
radiation oncology physics, and additional training of at least one month in bra-
chytherapy physics at an established centre, preferably the same centre as that vis-
ited by the radiation oncologist, if these treatments are to be undertaken. In this
manner, a consistent and comprehensive practice can be developed. It is also very
helpful for a junior physicist to have at least a part-time training in a department
with senior physicists. Special training courses are currently available for medical
physicists and are provided by international bodies in different parts of the world.
(119) The medical physicist’s responsibilities cover four major areas: dosimetry,
equipment acceptance and commissioning, quality control, and radiation safety. In
47
 ICRP Publication 86

dosimetry, the medical physicist helps minimise the probability of patient injury and
poor treatment outcome by assisting in devising, for each patient, an appropriate
treatment regimen, and reviewing all patients’ treatment plans. The medical physi-
cist is responsible for the calibration of the output of the treatment machine, during
commissioning and on a routine basis, and for assuring that all machine data used
for patient treatment is accurate and adequate. For quality control, the medical
physicist will be involved with establishing and running a quality control pro-
gramme which includes patient safety. Radiation safety also requires the establish-
ment and maintenance of a radiation protection programme designed to ensure the
safety of staff and the public. These duties will be the responsibility of the medical
physicist, and/or the Radiation Protection Officer who may or may not be the same
person. The administrative structure will vary depending on the nation, the facility,
and the resources; what matters is that the necessary authority is available.
(120) It should be understood that for the practice of radiation therapy is it man-
datory that a hospital have access to a suitable medical physicist. It is not sufficient
that the physics staff be trained; they should also be available in sufficient numbers
to carry out all the required duties.

5.2.3. Radiation therapy technologists, dosimetrists, and nurses

(121) In addition to the radiation oncologist and the medical physicist, a radiation
therapy programme requires radiation therapy technologists, dosimetrists, and
radiation oncology nurses. Radiotherapy technologists have the responsibility for
the set-up and delivery of the treatment, are involved in the simulation of the treat-
ment, and have, therefore, an essential function in noticing any abnormal reaction
of the patient or the machine and to report them as indicated in Section 5.4. Thus,
radiation therapy technologists play an important role in preventing accidents.
(122) Radiation therapy technologists, dosimetrists, and nurses should have a
degree, granted by a university or medical school, in academic studies and clinical
training for a period of three or four years. Nurses in charge of patients during
brachytherapy treatment preparation and/or treatment delivery should receive
training that emphasises the importance of technical features (e.g., the correct posi-
tion of sources). In addition, they should have clear instructions on the necessity to
report immediately unexpected events to a medical physicist, a radiation oncologist,
or a radiation protection officer.
(123) Although the radiation oncologist and medical physicist may delegate spe-
cific duties to these personnel as appropriate, they will retain the responsibility for
providing adequate supervision and training. For example, computerised dose cal-
culations may be performed by a ‘treatment planner dosimetrist’, or preparation of
low dose rate sources for patient treatments and maintenance of the source inven-
tory may be delegated to a ‘source curator’. Such individuals can perform valuable
service as technicians, especially where more highly trained persons are scarce, but
they should not be given responsibilities beyond their professional competence.
(124) A clear delineation of responsibilities is particularly important in the case of
dosimetrists and technologists. In some institutions, these professionals substitute
48
 ICRP Publication 86

for medical physicists, and treatment planning and delivery procedures are made with-
out the supervision of a qualified medical physicist. Whether this is done for economic or
practical reasons, such practices might have detrimental consequences for the patient.
For example, the lack of education in the specialised areas of mathematics and physics
restricts the understanding of the algorithms used in modern computerised treatment
planning systems; this can easily jeopardise the interpretation of results produced by
limitations of a treatment planning system. Furthermore, although the competence of a
dosimetrist may be adequate for most problems found in typical routine work, their
training may not be sufficient to identify the causes of abnormal or unexpected situations,
and to decide which checks are required and their degree of urgency.

5.2.4. Maintenance engineers

(125) If there is a large amount of equipment, e.g. several external therapy units
and simulators, block cutting equipment, treatment planning computers, tissue
compensation devices, etc., it might be advisable to ensure the immediate availability
of trained engineering maintenance staff. If remote after-loading devices are used in
the brachytherapy programme, then provision should be made for servicing the
devices. This may be best accomplished with service agreements with the manu-
facturer. Alternatively, staff will need to be trained in repair and preventative
maintenance of equipment, as well as in the basics of radiation protection.

5.3. Acceptance testing and commissioning of equipment

(126) A number of important steps should be taken before, during, and immedi-
ately after radiotherapy equipment is delivered to a facility. Radiation sources need
to be safely received, registered, and stored, the radiation measurement equipment
tested and calibrated, the shielding of special rooms measured, the radiation sources
tested and calibrated, and teletherapy units and remote afterloaders commissioned.
A record-keeping system should be in place.
(127) The acceptance tests demonstrate whether the equipment meets or exceeds
the procurement specifications. Frequently, acceptance tests follow a protocol sup-
plied by the manufacturer, but the purchaser may develop his/her own protocol. In
either case, the acceptance test protocol should be part of the purchase order for the
equipment. Acceptance tests protocols specify what tests will be performed, what
equipment is used to perform these tests, and what the results of these tests should
be. They constitute a legal document in which the medical physicist confirms that
the equipment met the procurement specifications.
(128) Acceptance tests are crucial to the prevention of accidental exposures, because
safety interlocks are tested for the first time in the hospital, and because some of these
interlocks cannot be tested under normal operation. At this stage, once the installation
has been completed, for the first time it is possible to test the safety of the system as a
whole, including interaction between different parts of the equipment. Acceptance
tests have been facilitated by the introduction of international standards for medical
electrical equipment (such as the IEC or equivalent national standards), which have
49
 ICRP Publication 86

started to specify the tests to be performed at the hospital level and the methods of
testing for the user. For those interlocks that cannot be tested at hospital level, these
standards also specify the type of evidence that the manufacturer has to provide of
the testing in factory. It is therefore essential that the hospital staff responsible for
the acceptance, usually the medical physicist, be acquainted with these standards.
(129) At the completion of acceptance tests, commissioning measurements begin.
During commissioning measurements, the physicist will measure all the data required
to place the unit into clinical service. The physicist should ensure that all data needed
to perform any anticipated clinical procedure is acquired at this time. The data should
be acquired in the format required for entry into the treatment-planning computer.
All data should also be entered into a logbook for archival purposes. The pages of the
logbook should be dated and signed by the physicist. Formal written acceptance
should only be signed after these tests have been performed.
(130) Immediately after the conclusion of the commissioning measurements,
quality control tests should be established, following the quality assurance pro-
gramme adopted by the institution. Checks should only be performed by qualified
and experienced persons, such as the medical physicist, who can sometimes delegate
the work to persons she/he has trained. Regardless of who performs the tests, the
medical physicist remains the responsible party for assuring the correct performance
of the equipment. The medical physicist should also verify that the data in the
treatment planning computer, any computer used to calculate treatment times, and
in the logbook, are correct and consistent.

5.4. Follow-up of equipment faults

(131) The safe repair of equipment is essential in the prevention of accidental

exposures. As shown in the case histories, some unsafe repairs, or lack of follow-up
of unexplained malfunctions, ended in fatal accidents. The most difficult situations
are those malfunctions that cannot be reproduced by the maintenance engineers,
either due to intermittent faults or faults that only appear in some particular cir-
cumstances. In these cases, it will be difficult for the maintenance engineer to isolate
the cause and to remove it. A procedure that facilitates interaction between users
and manufacturer, a follow-up of unexplained malfunctions, and dissemination of
the information to other users of the same type of equipment and to maintenance
engineers is the recommended approach to maintenance.

5.5. Communication

(132) Many accidental exposures could have been avoided by better communica-
tion between the different staff members or between staff and maintenance engi-
neers. Communication should be assured in the organisation of the department and,
for instance, a maintenance engineer should not be allowed to work on a treatment
unit without clearance by those responsible for a unit. Most of the problems of
communication identified in some of the accidental exposures reported in this
document could have been solved by using one of the following procedures:
50
 ICRP Publication 86

(i) If anomalous behaviour is observed in a treatment unit, this should be

reported immediately to the medical physicist. If the behaviour is suspected
to be caused by a change in machine performance, the physicist should carry
out suitable checks for assessment promptly.
(ii) If an unexpected reaction is observed in a patient by a radiotherapy technol-
ogist, or indicated by the patient, the radiation oncologist should immedi-
ately request the medical physicist to perform a verification to detect any
possible error in the different steps of the treatment procedure.
(iii) If unexpected reactions are observed for multiple patients, the medical phy-
sicist should be informed immediately and requested to verify the dosimetry
of the treatment unit.

(133) In any of these circumstances, the medical physicist should recommend the
radiation oncologist not to continue patient treatments until equipment performance or
treatment parameters have been verified. A heavy workload on a treatment unit should
never be a reason to resume patient treatments without the necessary verifications.
(134) It is essential that every radiation treatment be carefully documented.
Treatment charts, radiographs, and all patient documents should be clearly identi-
fied and dated to avoid confusion between patients or between different stages of the
treatment for one patient. If an accident occurs, these documents will be essential for
the radiation oncologist in any decision relative to further patient care.

5.6. Patient identification and patient chart

(135) Some of the accidental exposures resulted in the delivery of a treatment or

part of it to the wrong patient, or in treating the wrong site. Errors with the patient
chart resulted in mistakes in the number of fractions or the dose or with the changes
in the course of the treatment, for example, new fields, or prolonging the treatment
beyond the plan. It is important that the quality assurance programme includes
provisions and procedures for an effective identification of the patient (photograph,
ID) and site, and that a chart check protocol is implemented. The chart review
should be at least once a week, before the third fraction following the start of a new
treatment field or modification and at completion of treatment (AAPM, 1994).

5.7. External beam radiotherapy

(136) In external beam therapy most accidental exposures involving a large num-
ber of patients have occurred in relation to erroneous calibration of the beam and,
to a lesser extent, with the calculation of the dose (or treatment time) to be delivered.
Other accidents involving one patient or a few patients were related to simulation
and treatment delivery.

5.7.1. Beam calibration

(137) The calibration of a radiotherapy beam using a well established dosimetry

protocol is mandatory, whenever a new machine is installed or the source of a 60Co
51
 ICRP Publication 86

teletherapy unit exchanged. Dosimetry protocols like IAEA TRS-277 (IAEA, 1987)
and AAPM TG-21 (AAPM, 1983) can be recommended for air kerma-based dosi-
meter calibrations whereas IAEA TRS-398 (IAEA, 2000b) and AAPM TG-51
(AAPM, 1999) can be used for absorbed-dose-based dosimeter calibrations. Atten-
tion must be paid to the correct understanding of the type of dosimeter calibration
(NX, NK, ND,w) prior to the selection of a dosimetry protocol. Periodical verification
measurements should be made in accordance with the recommendations of the
quality assurance programme adopted.
(138) A redundant independent measurement, to verify the calibration, should be
performed before the first patient is treated. It is recommended that national reg-
ulations require independent calibrations for all radiotherapy treatment units and
that regulations be enforced. The redundant calibration can be done by staff of
another radiotherapy department, by an external audit using a postal dosimetry
service (for example with TLD, see Appendix B) or by a site visit conducted by an
independent organisation.
(139) The results of the redundant calibration should not differ from the original
calibration by more than 2 to 3%, depending on the type of detector used. If the
difference is more than 5%, the reasons for the discrepancy should be carefully
investigated and patient treatment not initiated until the discrepancy is resolved.
With discrepancies around 3–5%, patient treatments can start, but still the reasons
for the deviation must be identified. It is essential to understand that the purpose of
independent verifications made with TLD mailed dosimeters is to discover possible
errors in the calibration, but these should never be used as a substitute for a proper
local calibration, using an instrument with a valid traceable calibration certificate
and using a widely accepted dosimetry protocol.
(140) Every radiotherapy centre should participate regularly in an external audit
programme to verify the calibration of treatment units, ideally with the periodicity
of one year, but not less frequently than every five years. It has been reported
(Dutreix et al., 1993) that the size and number of discrepancies in beam calibration
in centres that have participated regularly in external audits is much smaller than for
centres that have not participated in such programmes.
(141) Quality control checks, particularly the verification of the dose delivered in
reference conditions, should be made at regular intervals. In addition, the calibration
of a treatment unit should be performed after any repair or adjustment of the unit that
may affect the characteristics of a beam. Other parameters related to dose delivery in
clinical conditions should be checked regularly, following in detail the quality assur-
ance protocol adopted by the institution.

5.7.2. Treatment planning systems

(142) Severe accidents in radiotherapy have been associated with the incorrect
commissioning of computerised treatment planning systems. Careful verifications of
absolute and relative doses delivered should be performed, including in-phantom
measurements for a comprehensive number of typical configurations. These are
included in the majority of quality assurance protocols for treatment planning sys-
52
 ICRP Publication 86

tems (cf. ICRU, 1987; AAPM, 1998). The physical basis on which the algorithms in
the treatment planning system are based should be clearly understood; this recom-
mendation applies also to any additional correction factors introduced manually in
dose calculation procedures.

5.7.3. In-vivo dose measurements

(143) Many of the accidents described in this publication could have been avoided
if in-vivo measurements had been performed on a selected group of patients. In-vivo
measurements (Leunens et al., 1990; Garavaglia et al., 1993; Van Dam and Mar-
inello, 1994) are an effective way of verifying the quality of the entire radiotherapy
treatment procedure. The additional cost of in-vivo dosimetry does not require a
considerable increase in funding even in a small hospital (Kesteloot et al., 1993). It is
an especially valuable investment, but to be effective, it requires careful preparation
in terms of equipment, staff training and quality assurance.
(144) Diodes and thermoluminescent dosimeters can be used for in-vivo measure-
ments. It is important to realise that when the detector used for in-vivo dosimetry
has been calibrated in the same treatment unit where patients are treated, the results
from in-vivo measurements will be correlated with the calibration of the machine
and, therefore, will not be able to show a potential error in the calibration of the
machine. A correct calibration of the dosimeter is thus an essential necessity.
(145) Patient entrance dose measured on the central axis of the incident beam,
performed systematically on each patient during the first session of each field, pro-
vides information on a combination of parameters from treatment preparation to
treatment delivery. Combined with exit dose measurements, they provide information
on patient data and on the performance of the treatment planning system used. When
results are carefully analysed large systematic errors can be detected rapidly, even
with a limited number of measurements.

5.8. Brachytherapy

(146) The majority of the accidental exposures that occurred with brachytherapy
treatments can be linked to source parameters, to dose calculation procedures and
to insufficient training of personnel.

5.8.1. Source activity and identification

(147) The activity of each brachytherapy source should be checked individually,

before it is used on a patient. Some of the major accidents in brachytherapy have
been caused by errors in the manufacturer’s specification of the activity of one or
several sources. Radiation oncologists should never use a source until its activity has
been verified. The unit of activity used at the hospital may differ from the unit stated
by the manufacturer. The documentation should be checked carefully. It is essential
that the unit of activity used for source calibration be the same as the unit of activity
used in the treatment planning system. The international recommendations for the
53
 ICRP Publication 86

specification of source strength (cf. ICRU, 1985) should be implemented for all
brachytherapy sources. After verification of the source activity, the source or source
holder should be marked with unique identifiers, to prevent the possibility of con-
fusion between different sources.

5.8.2. Dose calculation and treatment planning

(148) The calculation of the decay of the source is not always incorporated
into the dose calculation software, and a calculation by hand to take into account
this effect may be necessary for each patient. As information on the nature of
the sources for which the software has been developed is not always clearly stated,
it is essential that this radionuclide be identified. The use of radionuclide specific
software with a different radionuclide may lead to severe errors on dose calcu-
lation and to accidental exposures. Dose specification should be done following
the recommendations issued by international bodies (e.g., ICRU, 1985; 1993;
1999).

5.8.3. Source positioning and removal from the patient

(149) It is essential to verify that a source has been positioned correctly in the
patient and remains so during the treatment time. In remote afterloading devices
where sources are inserted automatically several times, the reproducibility of the
position should be checked before each application. Patients should be informed of
the necessity to avoid movements that may cause a source displacement during the
treatment. In some hospitals, brachytherapy treatments are not performed on
patients who are unable to understand these requirements, unless there is a way to
control patient movements. When a brachytherapy treatment is completed, the
sources should be removed from the patient either automatically or manually,
depending on whether a remote afterloader device is used. The patient and the
patient’s bedroom should be checked carefully with a simple survey monitor to
assure that all sources have been removed and properly stored.
(150) Fig. 5 summarises the recommendations given in this Section.

5.9. The potential for accidental exposures in the future

(151) The recommendations included in this Chapter are based on a retrospective

analysis of accidental exposures in radiation therapy with past and current types of
equipment. There are, however, a number of factors that may cause a change in this
picture in the future:

. With the worldwide expansion of radiotherapy there may be more accidents

related to inadequate staff training, mainly in countries where education pro-
grammes are still not widely implemented.
. There is a common misperception that modern equipment is safer and will
require less quality assurance.
54
 ICRP Publication 86

Fig. 5. A checklist for accident prevention.

55
 ICRP Publication 86

. An increasing number of clinical accelerators are being installed, either as new

therapy units or as a replacement of 60Co teletherapy machines, many of them
in developing countries. Accidents may occur due to inadequate accelerator
maintenance, as many developing countries may not have the infrastructure
required for the maintenance of complex equipment. The increased number of
computer-controlled systems may also lead to more computer related acci-
dents, compared to mechanical failures.
. The increased use of accelerators will require the decommissioning of 60Co
teletherapy units and perhaps some old 137Cs units, which may result in an
increased accident rate due to improper disposal. After a decade or so the fre-
quency of these events will probably decrease.
. The new technologies of high dose rate (HDR) brachytherapy, ‘gamma knife’
therapy units, multi-leaf collimators, and intensity modulated conformal
radiotherapy (IMRT) may produce new types of accidental exposures12.

(152) The consequences of these factors and new types of events can only be
anticipated by fault tree analysis. However, such analysis may not be able to take
fully into account human factors (e.g. education and training). It is expected that an
improved education/training system and enhanced quality assurance culture, toge-
ther with carefully designed regulations and adequate enforcement, will be the seed
to prevent the occurrence of severe accidents in the field of radiation therapy.

12
Note that in some countries neurosurgeons with little or no training on radiation effects, physics, or
protection are the major operations of ‘gamma knife’ units. A similar situation occurs with endovascular
brachytherapy sources, which often are manipulated by cardiologists.

56
 APPENDIX A. UNCERTAINTY IN RADIOTHERAPY

(A1) The physical quantity ‘absorbed dose’, like any other quantity, coefficient or
parameter, has an uncertainty associated to its numerical value that characterises the
dispersion or spread of values determined around an estimated value of the quan-
tity. This is described by a parameter like the standard deviation, which reflects the
lack of exact knowledge of the value of the quantity. Note that a result can have a
negligible error (be very close to a reference value), but yet it may have a large
uncertainty (ISO, 1995). The overall uncertainty in a process is obtained from the
combination of the uncertainties in all the steps involved in the process, yielding the
so-called combined standard uncertainty, uc (ISO, 1995).12

A.1. The influence of dose uncertainty in radiotherapy treatments

(A2) Uncertainties in absorbed dose arise during all steps of a treatment, from
beam calibration to dose delivery. The dispersion or spread of dose values received
by patients (Fig. A.1) is assumed to be a Gaussian distribution around the pre-
scribed dose. The width of the distribution is given by its standard deviation or
standard uncertainty. When a mistake or a systematic error causes a difference

Fig. A.1. Gaussian function representing the dispersion of the dose received around the prescribed dose
value (100%).

12
An expanded uncertainty can be obtained using a coverage factor k, U=k uc(k=1, 2, 3). With k=2,
i.e. two standard deviations, the expanded uncertainty corresponds approximately to the conventional 95
% confidence limit.

57
 ICRP Publication 86

Fig. A.2. Prescribed and delivered doses differing by 5%, each having an associated uncertainty described
by a Gaussian distribution of width 5%. The width of the distributions poses a limit to the ‘resolution’
with which the two distributions can be observed as clearly separated.

Fig. A.3. Gaussian distributions illustrating dose differences of 10% and 25%.

58
 ICRP Publication 86

between the dose received and the dose prescribed, the distribution of doses to the
affected patients is shifted away from the prescribed value (Fig. A.2).
(A3) As shown in this Appendix, the combined standard uncertainty in good
radiotherapy practice is of the order of 5%, i.e., approximately of the same order as
the lowest dose differences that can be detected clinically. The smaller the differences
between the dose received by patients and the dose prescribed, the more difficult it is
to detect errors in clinical practice, and they can only be detected when a relatively
large number of patients are affected and carefully observed. Dose uncertainties thus
pose a limitation to the ‘resolution’ in detecting an error in dose. When the differ-
ences between the dose received by patients and the dose prescribed are large, pro-
blems in the clinical outcome become more noticeable, as the distributions are more
separated (Fig. A.3).

A.2. UNCERTAINTY IN EXTERNAL BEAM RADIOTHERAPY

(A4) Following the work by Loevinger and Loftus (1977), estimates of the com-
bined standard uncertainty of the dose distribution in the target volume, along the
different steps of a treatment with external beams, have been given by Dutreix
(1984), Svensson (1984), and Brahme et al. (1988). These have been updated, mainly
with regard to the calibration process, by the evaluations of Andreo (1990, 1992)
and Thwaites (1994).
(A5) Uncertainty estimates are summarised here where the various steps of a
radiotherapy treatment are classified into four major groups according to:

(i) absolute dose determination at the reference point in a water phantom (beam
calibration), performed with a calibrated ionisation chamber according to a
code of practice or dosimetry protocol;
(ii) procedures involving relative beam dosimetry (field size output dependence,
influence of beam modifiers like wedges, secondary collimation, and blocks,
etc.), performed with any type of detector suitable for measurements in water
or in a plastic phantom;
(iii) the calculation of the dose delivered to the patient (monitor units or irradia-
tion time to deliver the prescribed dose) and its distribution, usually per-
formed with a computerised Treatment Planning System; and
(iv) the process of treatment delivery along a complete treatment, which accounts
for daily variations in patient and machine set-up, patient movements, and
machine instability during several weeks of treatment.

(A6) The uncertainties estimated for the various groups are given in Table A.1 for
radiotherapy treatments in which high-energy photon beams produced by clinical
accelerators are used, along with the various contributions to the uncertainty of each
group. Under the heading ‘Overall process’ in Table A.1, the column lists all possi-
ble contributions to the combined standard uncertainty in each step. The resulting
combined standard uncertainty is about 6 %.
59
 ICRP Publication 86

Table 1. Estimated relative standard uncertainties (uc) in the various steps of a radiotherapy treatment
with high-energy photon beams produced by clinical accelerators. Estimates correspond to clinical envir-
onments where high performance in dose determination and treatment delivery can be achieved. Data
marked as ‘-’ in the rightmost column are considered to be practically identical within an institution or
when several institutions adopt a common protocol for clinical comparisons; they do not contribute to
observed differences in clinical results.

Step in the radiotherapy treatment Overall Clinical

process comparisions

1.0 Absolute dose determination at the reference point in a water

phantom (beam calibration)
1.1 dosimetry recommendations 2.0 –
(protocols, codes of practice, standards of practice, etc.)
1.2 experimental procedure (including monitor errors) 1.0 1.0
combined standard uncertainty in step 1 2.2 1.0
2.0 Relative dosimetry
(water or plastic phantom measurements)
2.1 beam size output factors (including monitor errors) 1.0 1.0
2.2 beam modifiers (wedges, blocks, etc.) 1.0 1.0
2.1 measurement of isodose distributions 1.5 1.5
combined standard uncertainty in step 2 2.1 2.1
3.0 Dose calculation to the patient
3.1 patient input data (contour, CT-data, etc.) 1.0 1.0
3.2 Treatment Planning System (beam data, algorithms, 3.0 3.0
inhomogeneity corrections, etc.)
3.3 water to tissue transfer (including physical data) 2.0 –
combined standard uncertainty in step 3 3.7 3.2
4.0 Dose delivery (along complete treatment)
4.1 reproducibility in beam set-up, beam alignment, etc 1.0 1.0
4.2 reproducibility in patient set-up 2.0 2.0
4.3 contour and organ motion 1.5 1.5
4.4 machine instability 2.0 2.0
combined standard uncertainty in step 4 3.4 3.4
combined standard uncertainty in all steps 5.9 5.1

(A7) It is important to realise that not all the items appearing under the column
‘Overall process’ are included in published analyses of clinical results and estimates
of accuracy requirements in treatment delivery. Several components correspond to
steps whose uncertainties are dominated by those of fundamental physical quantities
and constants; this is mainly the case with correction factors and coefficients given in
dosimetry recommendations (protocols and codes of practice) for beam calibra-
tion13, as well as for data to convert dose from water to various types of tissue.
(A8) Uncertainties in such factors will, in general, be practically the same in most
clinical intercomparisons and they are assumed not to influence the results of relative

13
The uncertainty assigned to the calibration of therapy machines is the main factor responsible for the
differences in uncertainty estimates in radiotherapy treatment delivery published by different investigators
(Svensson, 1984; IAEA, 1987; Mijnheer et al., 1987; Brahme et al., 1988; Andreo, 1990, 1992; Thwaites,
1994; etc).

60
 ICRP Publication 86

clinical studies. Under the column ‘Clinical comparisons’ these components have
been omitted, yielding a combined standard uncertainty of about 5%. It is possible to
arrive at situations where still some other contributions may cancel out in a compar-
ison, for example for treatment outcomes within a single institution, sometimes even
with the same treatment unit. In those very special cases it is possible to arrive at a
combined relative standard uncertainty of about 3.5%. On the other extreme, when
beams other than high-energy photons are used (for example electrons), measurements
with instruments other than reliable ionisation chambers are made, or patients are
treated with sophisticated procedures, uncertainties (and the likelihood for errors in
treatment delivery) are expected to increase, so that an estimate of 5% may be difficult
to achieve. These cases are unfortunately not infrequent in routine clinical practice.
(A9) The conclusion from this analysis is that an estimated standard uncertainty of
5-6% seems to be a figure achievable at institutions with clinical practice of high
standard for carefully delivered radiotherapy treatments, but it is difficult to decrease
the combined uncertainty beyond this level. Studies assuming a narrower uncertainty,
based on the assumption that contributions cancel out, can only be justified under
very special circumstances.

61
 ICRP Publication 86

APPENDIX B. QUALITY AUDITS OF THE CALIBRATION OF RADIO-

THERAPY BEAMS

(B1) It is of interest to analyse the actual accuracy in dose determination in

radiotherapy beams worldwide. The information presented in this Appendix is
based on dose quality audits of radiotherapy centres (sometimes referred to as
‘intercomparisons’) performed with mailed thermoluminescence dosimeters (TLD)
under the IAEA/WHO TLD postal service (Izewska and Andreo, 2000)14. Since
1969, this postal service has been used to check more than 3000 radiotherapy beams,
and in many instances significant errors have been detected in their calibrations.
These calibration errors have often resulted in patient mistreatment and some could
even be included in the category of accidents. Fortunately, the number of substantial
calibration errors has decreased during recent years.
(B2) Results for the period of 1990–1997 (about one third of the total number of
beams checked during 30 years) are summarised in Fig. B.1, which reflects world-
wide data, mainly from developing countries. There are no substantial differences
between the various geographical regions. It can be seen that approximately 75% of
the results correspond to beam calibrations with deviations within  5% (previously
this figure was 65%), leading to treatment outcomes most likely within clinical tol-
erance. Fourteen percent of the results have deviations between 5–10%. The
remaining eleven percent of the results with deviations larger than 10% will have
severe detrimental consequences, and of these, four percent of the beam calibrations
can be considered associated with the likelihood of a radiotherapy accident.
(B3) Many severe discrepancies are solved through follow-up verifications and site
visits by national or IAEA experts, although in some instances the lack of further
reply by the institution has made it impossible to solve the problem detected. Of
special concern is the large amount of dosimeters not returned for evaluation
(approximately 25–30% of the dosimeters distributed, although the trend is
decreasing; previously this figure was 40%). In many cases this corresponds to
institutions not willing to show their level of performance. It is not known how
many radiotherapy machines exist in the world that have not been involved in such
quality audits.
14
The IAEA operates a cost-free service to verify the calibration of radiotherapy units in hospitals and
oncology centres. The TLD postal service is implemented through a collaboration between the IAEA and
the WHO (PAHO in Latin America) and is known as the IAEA/WHO TLD postal service. In all instan-
ces the service provides an independent and impartial quality audit of the dosimetry procedures used at
the hospitals. The TLD procedures receive the support of the Bureau International des Piods et Mesures
(BIPM), various Primary Standard Dosimetry Laboratories, and some advanced radiotherapy centres and
institutions in Europe and USA. The service warrants the confidentiality of the results, and only the per-
sons responsible for the radiotherapy departments or for the calibrations have access to the outcome of
the verification.

63
 ICRP Publication 86

Fig. B.1. Results of the IAEA/WHO TLD postal service for the calibration of therapy beams during the
period 1990–1997 (approximately 1100 beams verified). Measured deviations relative to the IAEA refer-
ence dose are grouped according to the levels in the legend. Data shown reflect worldwide figures, mainly
from developing countries; there are not substantial differences between the various geographical regions.
Approximately 25–30% of the dosimeters distributed by the TLD postal service were not returned for
evaluation.

(B4) There are several other national and international organisations providing
this type of inexpensive quality audit service to various countries and regions (cf.
Hanson et al., 1997; Ferreira et al, 2000). This is probably the simplest method
aimed at preventing radiotherapy accidents. These TLD-mailed procedures have
been recently adopted by a number of international and national organisations (cf.
IAEA, 1997). Participation in such quality audits is recommended in numerous
protocols for quality assurance in radiotherapy as a complement to or replacement
of the more efficient, but more expensive, site visits performed by an external audit
group (cf. Hanson et al., 1997; Järvinen, 1997).

64
 APPENDIX C. CASE HISTORIES OF ACCIDENTS RELATED TO
DECOMMISSIONING OF RADIOTHERAPY EQUIPMENT AND SOURCES

60
C.1. Illegal import, storage, and disposal of a teletherapy Co unit (Mexico, 1984)

(C1) A second-hand teletherapy unit with a 1,000 Ci (37 TBq) 60Co source was
purchased and imported, but did not comply with all the existing import regulatory
requirements. It was stored in a warehouse for six years; then its scrap value
attracted the attention of a maintenance technician, who dismantled the head of the
unit and removed the cylinder containing the 60Co (CNSNS, 1985).
(C2) The technician removed the cylinder and other metal parts from the unit,
loaded them into a pickup lorry, drove to a scrap yard and sold the parts as scrap.
Before arriving at the scrap yard, he deliberately ruptured the cylinder containing
the source. The pickup lorry thus contained a large quantity of radioactive material
from the source. The source consisted of about 6,000 tiny pellets (1 mm diameter) of
60
Co. When the cylinder was ruptured, several pellets were dispersed and remained in
the lorry when the heavy parts were unloaded at the scrap yard. Owing to mechanical
defects, the pickup lorry contaminated with 60Co pellets remained parked on the
street for 40 days. The lorry was then moved to another street where it stood for a
further ten days.
(C3) When the ruptured cylinder was moved by cranes in the scrap yard, together
with the other metal pieces, the 60Co pellets were spread over the scrap yard,
attracted by the magnetic field of the crane, and mixed with the other metal mate-
rials. Consequently, pellets and pellet fragments also were transferred to other
vehicles used for transporting scrap to various foundries. The main purchaser of the
scrap was a firm that manufactured construction reinforcing rods and connecting
rods for motor vehicles.
(C4) It was later found that scrap contaminated with 60Co had been used by steel
production plants to manufacture reinforcing rods and metal table bases. A lorry
transporting contaminated rods to be used for construction entered a nuclear
laboratory where radiation detectors were used to monitor removal of radioactive
material. The detectors not only indicated the presence of radioactivity, but also
activated a camera that photographed the contaminated vehicle. Two days later the
authorities ascertained the origin of the contaminated rods.
(C5) An extensive investigation showed that 30,000 table bases and 6,600 tons of
reinforcing rods had been made from the contaminated material. Aerial radiation
surveys of an area of 470 square kilometres were conducted, resulting in the recovery
of 27 cobalt pellets. Visits were made to 17,636 buildings to determine whether
contaminated material was used in their construction. Unacceptable radiation levels
were measured in 814 buildings that were then partly or completely demolished. The
accident exposed approximately 4,000 people to radiation, about 80 of whom
received doses greater than 250 mSv. Apparently, five people received doses of 3 to 7
Sv over a 2-month period.
(C6) Although the triggering event was that a person dismantled an insecurely
stored head of a 60Co teletherapy unit and broke the source capsule, other factors
65
 ICRP Publication 86

contributed, such as non-compliance with regulations, as the radiotherapy unit was

illegally imported and transported and remained in storage for six years under
unsafe and unsecured conditions.

137
C.2. Abandonment of a teletherapy Cs unit (Brazil, 1988)

(C7) A private radiotherapy institute moved to new premises, leaving a caesium

(137Cs) teletherapy unit without notifying the licensing authority. Because of partial
demolition of the building, the teletherapy unit became totally unsecured. Two people
entered the building and removed the source assembly from the radiation head. They
tried to dismantle the source assembly at home and in the attempt, the source capsule
was ruptured. Because the radioactive source was in the form of caesium chloride pow-
der, which is highly soluble and easily dispensed, there was considerable contamination
of the environment, resulting in external irradiation and internal contamination of sev-
eral persons (IAEA, 1988).
(C8) After the source capsule was ruptured, remnants of the source assembly were
sold for scrap to a scrap-yard owner. He noticed that the source material glowed blue
in the dark. Several persons were fascinated by this, and over a period of days, friends
and relatives came and saw the phenomenon. Fragments of the source the size of rice
grains were distributed to several families. Five days later, a number of persons showed
gastrointestinal symptoms that were not recognised initially as being due to exposure to
radiation. However, one of the persons irradiated connected the illnesses with the
source capsule and took the remnants to the public health department. This action
began a chain of events that led to investigation of the accident.
(C9) Many individuals received external and internal radiation exposure exceeding
acceptable levels. In total, some 112,000 persons were monitored; 249 persons were
contaminated either internally or externally. Some suffered very high internal and
external contamination owing to the way they had handled the caesium chloride
powder, such as daubing their skin, eating with contaminated hands, and handling
various objects. Four persons died within four weeks of admission to hospital, hav-
ing received total body doses of at least 5 Gy.
(C10) Seven main sites of contamination were identified, including the scrap-yards
concerned, some of them with dose rates of up to 2 Sv.h 1 at one metre. Aerial surveys
were conducted over 67 square kilometres. Of 159 houses monitored, 42 required
decontamination. The final volume of waste stored was 3500 cubic metres or 275 lorry
loads. The radioactivity accounted for in the decontamination was estimated at 1,200
Ci (44.4 TBq), compared with the known activity of the source before the accident
of 1375 Ci (51 TBq).
(C11) This case is similar to that discussed in Section C.1. The following factors
contributed to the accident: lack of compliance with regulations in that there was no
safe, formal, and complete decommissioning of the facility, and the 137Cs teletherapy
unit, no longer in use, was abandoned unsecured.

66
 REFERENCES

AAPM (1983) A protocol for the determination of absorbed dose from high-energy photon and electron
beams. Report of the American Association of Physicists in Medicine Radiation Therapy Committee
Task Group No. 21. Med. Phys. 10, 741–771.
AAPM (1993) Medical accelerator safety considerations. Report of the American Association of Physi-
cists in Medicine Radiation Therapy Committee Task Group No. 35. Med. Phys. 20, 1261–1275.
AAPM (1994) Comprehensive QA for radiation oncology. Report of the American Association of Phy-
sicists in Medicine Radiation Therapy Committee Task Group 40. Med. Phys. 21, 1–38.
AAPM (1998) Quality assurance for clinical radiotherapy treatment planning. Report of the American
Association of Physicists in Medicine Radiation Therapy Committee Task Group 53. Med. Phys. 25,
1773–1829.
AAPM (1999) Protocol for clinical reference dosimetry of high-energy photon and electron beams. Report
of the American Association of Physicists in Medicine Radiation Therapy Committee Task Group No.
51. Med. Phys. 26, 1847–1870.
Alletti, P., Bey, P. (1996) Recommendations for a Quality Assurance programme in external radiotherapy.
ESTRO Booklet 2, Garant, Leuven, Belgium.
Andreo, P. (1990) Uncertainties in dosimetric data and beam calibration. Int. J. Radiation Oncology Biol.
Phys. 19, 1233–1247.
Andreo, P. (1992) Absorbed dose beam quality factors for the dosimetry of high-energy photon beams.
Phys. Med. Biol. 37, 2189–2211.
Andreo, P., Brosed, A., Gultresa, J., et al. (1997) Tolerance, effective tolerance and action levels in a
quality assurance program for equipment used in radiotherapy (in Spanish). In: Proc. XI Annual
Meeting of the Spanish Society of Medical Physics, Valencia, Spain. SEFM, Madrid, pp. 39–65.
Ash, D., Bates, T. (1994) Report on the clinical effects of inadvertent radiation underdosage in 1045
patients. Clinical Oncology 6, 214–225.
Brahme, A., Chavaudra, J., Landberg, T., et al. (1988). Accuracy Requirements and Quality Assurance of
External Beam Therapy with Photons and Electrons Acta Oncol. Suppl. 1.
CNSNS (1985) Accidente por contaminación con 60Co México 1984. CNSNS-IT-001. Comisión Nacional
de Seguridad Nuclear y Salvaguardias, México.
Cohen, L., Schultheiss, T.E., Kennaugh, R.C. (1995). A radiation overdose incident: initial data. Int. J.
Radiat. Oncol. Biol. Phys. 33, 217–224.
Cosset, J.M, Dubray, B., Girinsky, T., et al. (1994) Fractionation sensitivity of mammalian tissues. Adv.
in Radiation Biol. 18, 91–121.
Dische, S., Saunders, M. I., Williams, C., et al. (1993) Precision in reporting the dose in a course of
radiotherapy. Radiother. Oncol. 29, 287–293.
Dutreix, A. (1984) When and how can we improve precision in radiotherapy? Radioth. Oncol. 2, 275–292.
Dutreix,A., van der Schueren, E., Derreumaux, S., et al. (1993) Preliminary results of a quality assurance
network for radiotherapy centres in Europe. Radiother. Oncol. 29, 97–101.
El Periódico (1993) Account of the court proceedings and verdict for the case of the accelerator accident
in Zaragoza (Spain), reproduced in the newspaper of April 7 (in Spanish). El Periódico, Madrid.
ESTRO (1995) European Society for Therapeutic Radiology and Oncology: Quality assurance in radio-
therapy. Radiother. Oncol. 35, 61–73.
ESTRO (1996) Quality assurance in radiotherapy: the importance of medical physics staffing levels.
Recommendations from an ESTRO/EFOMP joint task group. Radiother. Oncol. 41, 89–94.
Ferreira, I. H., Dutreix, A., Bridier, A., et al. (2000) The ESTRO-QUALity assurance network (EQUAL).
Radiother. Oncol. 55, 273–282.
Garavaglia, G., Johansson, K., Leunens, G., et al. (1993) The role of in-vivo dosimetry. Radiother.Oncol.
29, 281–282.
Goitein, M. (1983) Nonstandard deviations. Med. Phys. 10, 709–711.
Hanks, G. (1984) Quality Assurance in Radiation Therapy: Clinical and Physical Aspects. Pergamon
Press, New York, NY.
Hanson W.F., Aguirre J.F., Stovall, M. (1997) Radiotherapy physics quality audit networks in the USA.
In: Quality Assurance in Radiotherapy: Proceedings of an IAEA-ISRO working meeting held in Vienna
May 8-9, 1995. IAEA Tec-Doc 989, IAEA, Vienna, Austria, pp. 195–203.

67
 ICRP Publication 86

Herring, D.F., Compton, D.M.J. (1971) The degree of precision required in the radiation dose delivered
in cancer therapy. In: Brit. J. Radiol. Special Report No. 4. British Institute of Radiology, London, UK,
pp. 56–78.
Horiot, J.C. (1995). Radiotherapy: still young and almost a hundred years. The Magazine of the World
Health Organization. 48th year, No. 3. May/June, 1995.
IAEA (1988) The radiological accident in Goiania. STI/PUB/815. International Atomic Energy Agency,
Vienna, Austria.
IAEA (1987) Absorbed Dose Determination in Photon and Electron Beams: An International Code of
Practice, Technical Report Series, TRS-277. International Atomic Energy Agency, Vienna, Austria.
IAEA (1996) International Basic Safety Standards for Protection against Ionizing Radiation and for
the Safety of Radiation Sources, IAEA Safety Series, No. 115. International Atomic Energy Agency,
Vienna, Austria.
IAEA (1997) Quality Assurance in Radiotherapy, Proceedings of the Working Meeting on National Pro-
grammes: Design, Harmonization and Structures, jointly organized by the IAEA and the International
Society for Radiation Oncology (ISRO), Vienna 1995. IAEA Tec-Doc 989. International Atomic
Energy Agency, Vienna, Austria.
IAEA (1998a) Design and Implementation of a Radiotherapy Programme: Clinical, Medical Physics,
Radiation Protection and Safety Aspects. IAEA Tec-Doc 1040. International Atomic Energy Agency,
Vienna, Austria.
IAEA (1998b) Accidental overexposure of radiotherapy patients in San José, Costa Rica. IAEA Special
publication series. International Atomic Energy Agency, Vienna, Austria.
IAEA (2000a) Lessons learned from accidental exposures in radiotherapy. IAEA Safety Report Series 117.
International Atomic Energy Agency, Vienna, Austria.
IAEA (2000b) Absorbed dose determination in external beam radiotherapy: An international Code of
Practice for dosimetry based on standards of absorbed dose to water. IAEA Technical Report Series
398, International Atomic Energy Agency, Vienna, Austria.
ICRP (1991) 1990 Recommendations of the International Commission on Radiological Protection. ICRP
Publication 60, Annals of the ICRP 21 (1-3).
ICRP (1997) Protection from potential exposure: Application to selected radiation sources. ICRP Pub-
lication 76, Annals of the ICRP 27 (2).
ICRP (1998) Genetic susceptibility to cancer. ICRP Publication 79, Annals of the ICRP 28 (1/2).
ICRU (1976) Determination of absorbed dose in a patient irradiated by beams of X or gamma rays in
radiotherapy procedures. ICRU Report 24. International Commission on Radiation Units and Mea-
surements, Bethesda, MD.
ICRU (1985) Dose and Volume Specification for Reporting Intracavitary Therapy in Gynaecology, ICRU
Report 38. International Commission on Radiation Units and Measurements, Bethesda, MD.
ICRU (1987) Use of computers in external beam radiotherapy procedures with high-energy photons and
electrons. ICRU Report 42 International Commission on Radiation Units and Measurements,
Bethesda, MD.
ICRU (1993) Prescribing, Recording and Reporting Photon Beam Therapy, ICRU Report 50. International
Commission on Radiation Units and Measurements, Bethesda, MD.
ICRU (1999) Prescribing, Recording and Reporting Photon Beam Therapy (Supplement to ICRU Report 50),
ICRU Report 62. International Commission on Radiation Units and Measurements, Bethesda, MD.
IEC (1988) Medical Electrical Equipment- Part 1: General Requirements for Safety. IEC 601-1. Interna-
tional Electrotechnical Commission, Geneva, Switzerland.
IEC (1989) Medical Electrical Equipment- Part 2: Particular Requirements for the Safety of Remote-
Controlled Automatically-Driven Gamma-Ray Afterloading Equipment. IEC 60601-2-17. Interna-
tional Electrotechnical Commission, Geneva, Switzerland.
IEC (1993) Medical Electrical Equipment- Part 2: Particular Requirements for the Safety of Radiotherapy
Simulators. IEC 60601-2-29. International Electrotechnical Commission, Geneva, Switzerland.
IEC (1997a) Medical Electrical Equipment- Part 2: Particular Requirements for the Safety of Gamma
Beam Therapy Equipment. IEC 60601-2-11. International Electrotechnical Commission, Geneva,
Switzerland.

68
 ICRP Publication 86

IEC (1997b) Medical Electrical Equipment. General Requirements for Safety. 4. Collateral Standard:
Programmable Electrical Medical Systems. IEC 60601-1-4. International Electrotechnical Commission,
Geneva, Switzerland.
IEC (1998) Medical Electrical Equipment- Part 2: Particular Requirements for the Safety of Electron
Accelerators in the Range 1 to 50 MeV. IEC 60601-2-1. International Electrotechnical Commission,
Geneva, Switzerland.
IEC (in press) Medical Electrical Equipment- Part 2: Particular Requirements for the Safety of Treatment
Planning Systems. IEC 60601-2-xxx. International Electrotechnical Commission, Geneva, Switzerland.
ISCRO (1991) Radiation Oncology in Integrated Cancer Management (Blue Book). Inter-Society Council
for Radiation Oncology, Philadelphia, PA.
ISO (1993) International vocabulary of basic and general terms in metrology [Published by ISO in the
name of BIPM, IEC, IFCC, IUPAC, IUPAP and OIML]. International Organization for Standardi-
zation, Geneva, Switzerland.
ISO (1995) Guide to the expression of uncertainty in measurement, 2nd ed. [Published by ISO in the name
of BIPM, IEC, IFCC, IUPAC, IUPAP and OIML]. International Organization for Standardization,
Geneva, Switzerland.
ISO (2000) ISO Standards Compendium ISO 9000- Quality Management, 8th edition. International Orga-
nization for Standardization, Geneva, Switzerland.
Izewska I., Andreo, P. (2000) The IAEA/WHO TLD postal programme for radiotherapy hospitals.
Radiother. Oncol. 54, 65–72.
Järvinen, H. (1997) Implementation of the quality assurance programme at the national level. In: Quality
Assurance in Radiotherapy: Proceedings of an IAEA-ISRO working meeting held in Vienna May 8-9,
1995, IAEA Tec-Doc 989, IAEA, Vienna, Austria, pp. 205–213.
Kesteloot, K., Dutreix, A., Van der Schueren, E. (1993) A model for calculating the costs of in vivo
dosimetry and portal imaging in radiotherapy departments. Radiother. Oncol. 28, 108–117.
Leunens, G., Van Dam, J., Dutreix, A., et al. (1990) Quality Assurance in radiotherapy by in-vivo dosi-
metry 1. Entrance dose measurements, a reliable procedure. Radiother.Oncol. 17, 141–151.
Loevinger, R., Loftus T.P. (1977) Uncertainty in the delivery of absorbed dose. In: Casnati, E. (Ed.):
Ionizing Radiation Metrology, Editrice Compositori, Bologna, Italy, pp 459–473.
Leveson, N., Turner, C.S. (1993) An investigation of the Therac-25 accidents. IEEE Computer, 26 (7), 18–
41.
Maciejewski, B., Withers, H.R., Taylor, J.M.G., et al. (1990) Dose fractionation and regeneration in
radiotherapy for cancer of the oral cavity and oropharynx. Part 2. Normal tissue responses: acute and
late effects. Int. J. Rad. Oncol. Biol. Phys. 18, 101–111.
Mijnheer, B.J. (1996) The clinical basis for dosimetric accuracy in radiotherapy. In: Radiation incidents:
Uncertainties, errors and accidents in radiotherapy, a meeting organised by the British Institute of
Radiology. British Institute of Radiology, London, UK, pp. 16–20.
Mijnheer, B.J., Battermann, J.J., Wambersie, A. (1987) What degree of accuracy is required and can be
achieved in photon and neutron therapy? Radiother. Oncol. 8, 237–252.
Mijnheer, B.J., Battermann, J.J., Wambersie, A. (1989) Reply to Precision and accuracy in radiotherapy
(Letter to the Editor). Radiother. Oncol. 14, 163–167.
NRC (1988). Report to Congress on Abnormal Occurences. AS-88-8. Multiple Medical Teletherapy
Misadministrations at Sacred Heart Hospital in Cumberland, Maryland. NUREG-0090, Volume 11,
No. 4. US Nuclear Regulatory Commission, Washington, DC.
NRC Nuclear Regulatory Commission. (1992). Report to Congress on Abnormal Occurences. 92-18. Loss
of Iridium-192 Source and Medical Therapy Misadministration at Indiana Regional Cancer Center in
Indiana, Pennsylvania. NUREG-0090, Volume 15, No. 4. Washington DC.
NRC (1996). Report to Congress on Abnormal Occurences. AS-88-8. Multiple Medical Teletherapy
Misadministrations at Sacred Heart Hospital in Cumberland, Maryland. NUREG-0090, Volume 19.
US Nuclear Regulatory Commission, Washington, DC.
Ortiz, P. (1998) Lessons learned from accidents in radiotherapy. In: International Conference on the
Safety of Radiation Sources and the Security of Radioactive Materials, Dijon 1998, IAEA TEDCOC-
1045. IAEA, Vienna, Austria.

69
 ICRP Publication 86

Pérez, L.A., Brady, L.W. (1992). Principles and Practice of Radiation Oncology. J.B. Lippincott Com-
pany. Philadelphia, PA.
Powers, B.E., Thames, H.D., Gillette, S.M., et al. (1998) Volume effects in the irradiated canine spinal
cord: do they exist when the probability of injury is low?. Radioth. and Oncol. 46, 207–306.
Rawlinson, J.A. (1987) Report on the Therac-25. OCTRF/OCI Physicists Meeting. Kingston, Ont.
Canada, May 7, 1987.
Rubin, P., Casarett, G.W. (1968) Clinical Radiation Pathology. W.B. Saunders, Philadelphia, PA.
Sebag-Montefiore, D.J., Maher, E.J., Young, J., et. al. (1992) Variation in mantle technigue:implications
for establishing priorities for quality assurance in clinical trials. Radiotherapy and Oncology 23 (3),
144–149.
Sharp, C., Cox, R. (1999). Genetic susceptibility to radiation effects: possible implication for medical
ionising radiation exposures. Eur. J. Nucl. Med. 26, 425–428.
Svensson, H. (1984) Quality assurance in radiation therapy: physical aspects. Int. J. Radiat. Oncol. Biol.
Phys. 10, 59–65.
Thames, H.D., Hendry, J.H. (1987). Fractionation in Radiotherapy. Taylor and Francis, London., UK.
Thwaites, D.I. (1994) Uncertainties at the end point of the basic dosimetry chain. In: Proc. Symp. Mea-
surement Assurance in Dosimetry, Vienna, 1993. IAEA Proceedings Series STI/PUB/930. IAEA,
Vienna, Austria, pp. 239–255.
Thwaites, D., Scalliet, P., Leer, J.W., et al. (1995) Quality Assurance in Radiotherapy. Radiother.Oncol.
35, 61–73.
Van Dam, J., Marinello, G. (1994) Methods for in-vivo dosimetry in external radiotherapy. ESTRO
Booklet 1, Garant, Leuven, Belgium.
Wambersie, A., Van Dam, J., Hanks, G., et al. (1994). What accuracy is needed in dosimetry? In: Proc.
Sem. Radiation dose in radiotherapy from prescription to delivery, Leuven, 1991. IAEA-TECDOC-734.
IAEA, Vienna, Austria, pp. 11–35.
Wambersie, A., Zreik, H., Prignot, M., et al. (1984) Variation of RBE as a function of depth in a high
energy electron beam in the first millimetres of irradiated tissues determined by observation of skin
reactions on patients (a clinical trial). Strahlentherapie 148, 279–287.
WHO (1988) Quality Assurance in Radiotherapy. World Health Organization, Geneva, Switzerland.

70
 ',(doc red 00'I T 1e 'ut1ue1en rq '.,fue1erceg
cgquelcs dUCI eql q?norql pe8ueue oq u€c (serdoc erotu ro 9) seue,rtlep >11n9
'uepeA\s
rqotplcols gI IrI-as
dIISI
urluep1 1rq ',fte1e.rces cg4uarcs eql
uro4 e8reqcJo ee{ pe6enbsr eq,{eru 'qcuerg ro qs113ug ut 'satdoc IunpIAIpuI
'sJeqrJcsqns ol pelnqrrNlp
,(ylecrleruolne lou sI Vre '1VJI ary {o slDuuY eI{} Jo ged }ou sI }ePlooq sIqI
'pessncslp,(gerrq IIe eJ? ,,(4re1c pue 'f,cuelstsuoc ',$rltqe5 JoJ peeu
eql pue 'uorlcelord go ue1s,ft eql ut uotlucrldde pu" uolpleJd.relul 3lo surelqord
.uorlcelold go rue1s,{s eql
Jo aJnlcn4s eq1 'peureldxe fgetrq eJB seleurl}se {slr JoJ
srseq e1rpllluenb eq; puu 's1decuo3 ',(cr1od scuolssllutuoJ eqlJo srseq luctSolotq
eq; 'seqrluenf ilno plul eJe suollepueuuoceJ s.uolssluuoC eql Jo sentlcelqo
eql 'uorssruuoc eql Jo uotleredo Jo apoul pu€ eJnlcn4s eql pue uollslJuu?
pue ,(rolslg eql slueserd 1r 'uollcnporlul u? JoUV 'selcllod luaunc pue 'sldecuoc
'suorleredo;o epotu ',{rolstq sI Jo A\eIAeJ prur*uopqEt?Jls 3 se lueuncop slq} penssl
,(11uecer (aUCD uoqcelord lucl8olotpeg uo uolssllutuoC Isuolletuelul eql
df,JI &u1q1nsap pplooq aa'{ 7
soJnpefo.rd'sap11od'fu olst;1
: uollfalold lellBololpeu
uo uolsslrutuo) leuollPuJalul
 III
'sSuIleoru oJoru lou pue 'eerql lseel l" 'por.red
8urpue1le Jo slsoc rloql 01 Sutlnqtrluoc ,^ eu € Jo uorsscco qcee uo ',(p1 1 uro,rJ
'sosec . u€ur ur 'pu€ e3;eqc lnoqll.^a 'spoued ;ee.{-rnog JoJ unJ soo}}rurruoJ
elqelrel" eurl .sJeqlueur eq1 Surleru .{q slr pu€ uorssruruoS OqI 'qffie sJeqlue(u
il{JI ol poddns epltord osl€ suollnl 0z puB sI uee,rleq ur€luo3 soellrluruoJ
-rlsur .sreqluelN 'suou0)llqnd ddJI uo eql elq,{o. 'ueuuteq3 € pu€ sJoqluew
ser11e,(o; ruo4 enJc3€ spunJ Ieuolllpp€ zl Jo slsrsuoc uorssruluoJ ur€w eqJ
oruos 'uorloelord lecrSololp€J uI lsoJelul
u€ qll4A selpoq I€uoll€ulelul pu€ I€uoll€u 'l?rJ€leJces curluorcs eql pu"
ruo{ suoqnquluoc fuelun1o,t ,{q ,(pteur 'serge4 3uqro16 pue sdnorg 1se1
pesu€ug er€ ddJI Jo solll^llce eql )oq pD '(suorlepueurruoco1 il{JI Jo
uoqecrlddy) t eellruruoJ'(aunrpelq
'duJI ol elqelle^B ur uorlceloJd) g eellrruuro3 '(e;nsodxg
seruoru ruo4 .{resseceu s€ pepunJ uorlurp"d ruor3 sesoq) 7 eellrururo3
e.re s8urleetu eseql 'uolssruruoJ eql '(slcegg uorlerpeg) 1 eellrururo3'uors
q1r,lo. reqleSol pue .{llurol pelcnpuoc st -sruruoJ urBI tr eql Jo slsrsuos duJI
seollrtutuo) eq] 3o Surleetu I€nuu€ oql
'poued ;ee. -;no3 l{c€e uI ect,ll '.tee,( 'peleurSrro 1r e;eq,tr 'eurc
e eJuo sloeu osll[uruoJ qJuS ';ea,( e -rpeu ur ernsodxe leuorledncco Jo 1eql
0cr.^d.1 Jo ecuo sleeul uolsslluluoJ eql eprslno s?eJB ur luetuellolur 8ur,ro.r8
slr lcegeJ 01 096l ur eru€u luese;d
'1661 KIry 1 ue8eq poued .,rteu ? qcns s1r peldope 1nq 6.eellruruoJ runrp€d
'soellruruoC egl JoJ l{8nos sI I€^\oueJ pue .,(er-y leuorlBuJelul aql s€ gZ6I
Jo e1€J Jslrrurs y 'pe8ueqc eq lsntu ur peuuoJ s€.tr 1I '41{lI sr uorlcelo-rd
uolssrruruoJ eql Jo sJJqr.ueu 'ang uuql lect8olorper ur ,(poq .,(reurud eqa
'uoJlulpuJ 8ugs1uo.r 1sup8e uopcelord;o spadsu 1u uo
acuupp8 puz suopspualuluoJar Supptord ,{q rqncqrud ut
'uolpalord popololpur Jo arualJs aql
luouoq rgqnd eq1 roJ acuc^pu ol poqsllquNo
'flgruq3 parappag luapuedapul uu sg
ouol1ae1or4 pe;8o1o;puA uo uolssltutuoJ
'dUJI luuolluuralul oqJ
taaraarDls aorsstw nno
uorlceloJd Iecr8ololp€u
uo uorss[uruoJ IeuorlBurelul er{J
 ICRP Publication 86

The Commission uses Task Groups Thus, ICRP is an independent inter- MAIN COMMISSION
and Working Parties to deal with specific national network of specialists in var- Chairman:
areas. Task Groups are formally ious fields of radiological protection. Professor R H Clarke
appointed by the Commission to perform At any one time, about 100 eminent
a defined task, usually the preparation of scientists are actively involved in the Vice-Chairman:
a draft report. A Task Group usually work of ICRP. The four-tier structure Dr L-E Holm
contains a majority of specialists from described provides a rigorous Quality
outside the Commission's structure. It Management system of peer review for
is funded as necessarv from monies the production of ICRP Publications. Members:
available to ICRP. Professor R Alexakhin
In preparing its recommendations, DrJDBoiceJr.
Working Parties are set up by the Commission considers the funda- Dr R Cox
Committees to develop ideas, some- mental principles and quantitative Ms G J Dicus
times leading to the establishment of a bases on which appropriate radiation Dr A J Gonzalez
Task Group. The membership of a protection measures can be established, Professor F A Mettler
Working Party is usually limited to while leaving to the various national Dr Y Sasaki
Committee members. Working Parties protection bodies the responsibility of Professor C Streffer
receive no funding of their own, i.e. formulating the specific advice, codes Dr A Sugier
they operate primarily by correspon- of practice, or regulations that are best Professor Z QPan
dence and by meetings in direct con- suited to the needs of their individual Mr B C Winkler
junction with meetings of the Committee countries. The aim of the recommen-
concerned. dations of ICRP is to Emeritus Members: (as
an appropriate standard of Mr H J Dunster
These activities are co-ordinated with a -provide
protection Jbr mankind from sources of Professor B Lindell
minimum of bureaucracy by a Scientific ionising radiation, without unduly limiting Dr W K Sinclair
Secretary, ensuring that ICRP recom- beneficial practices that give rise to Dr L S Taylor
mendations are promulgated. exposure to radiqtion.
Scientific SecretarY
Dr J Valentin
Composition of the International Commission on Radiological
Protection and Committees, 1997-2001
 t
ii
]*
t
I
I
noqz I rc
Jol.{€I w c JosseJoJd
reqlels A\ f r(I
usqperd s v rcl
e{zler€d H rcl
rel^Ilgw H.ICI
IezueIN c H rcl
urelzsdrT'I f JosssJoJd
^or€lq>lrT
v I rcl rosseJord
€qeul f rc
,fu.{ V.{ sslntr
u€rure>lcg d ) rcl
e4el(7 D rC rosseJord
elll^nog v rc
re>lceog g rc
I
^ouole8
w rc lucJ8oloJpuy
sreqluol tr
urluel"A f r(I
frele;ce5 cglluelcs
JeJeJls ) JosseJoJd .uouDlpi
. :U?IUJI"qJ ;o1,{u1 S I JCI o^l ast't aa!3 toqt sattl.
rlelculs ) 1v\ rCI 8-utttur11 [ppun
rnoqrtlu ,uot
Z AAIIIIAIhIO) IIopulT fl rosseJord Jo saJtnos tuo.tt put4uoul
relsun(IIHrItr Jo p"tDpaDN 4ot.tdo.tddo
noqz )-d rc (tOOZ fpt;o se) :srequetr l snlrrorug 01 sl du
qclrlln d rc -uourluoceJ Oql
JO rrrr€ o
eI{cJ€luJII W JC rePlurl[ J g rw lPnpl^rput Jtaql Jo spseu
ile,4Aels d rcl u€d O Z rosseJord lseq oJ" 1uq1 suorleln8et r
eroqs a u rosseJord ;e€n5 y;q sepoc .ocr pe cgrcads eql
ueue [e teue tz>lu€S ) Jo sseJo Jd JeJaJlS J JosseJoJd
Jo {tilqlsuodser eq1 serpo
uou g rc DIBSES A JC
Jpuorleu snorJel aqt o1
uolserd f d r(I rellleltrvdrossaJord 'peqsrlqelse aq uPc
sarnsuo
uolserd cI rcl zeIEZvo{) I v r(I uorlerpeJ eleudordde qcr
EAIN O JC sncrc f D strN errrlelrluenb pu" seldrcu
p"ol{rlnw ) rc xoJ d rcl -PpunJ eql SJeprsuoJ
'rf ocrog (I f r(I Uorss
epll'I f rossoJord 'suorlBpueuuocor s1r
Eur.r
pu€T J rc urq>lBxolv l{ rosseJoJd
rerelle) ntr v rcl rosseJord :sJeqruetr tr 'suorlucrlqnd
duJJ Jo uor:
,{rpuag H I r(I JoJ ,rerlal .roed
3o uels,(s l
Joullolg t{ JosseJoJd Alllun| snoro8rr u sepr^oJ
,re,{41>1y A V rCI urloH g-'I r(I ernlsnJls ren-rnoJ oqJ
:slequetr l :u€ruJregJ-ecrA el{l ut pelJo^ut .{Je,u1cedu:e
lueurluo 00J lnoqe ,eur1 e
xoJ d rcl OITBIJ H d rossaJord 'uorlcalord
lurrEolorper go
:u€ruJIeqJ :u"IrrJI€qJ --r?A ur slsrlerceds
Jo lro.!\le
I AAJIII^IWOJ NOISSINWO] NIVIAI -Jelur luopuodepur uu sl
d1
98 uolleclTqnd dl{31
 ICRP Publication 86

COMMITTEE 3 COMMITTEE 4
Chairman: Chairman:
Professor F Mettler Mr B C Winkler

Members: Members:
Professor J M Cosset Dr E d'Amato
Dr C Cousins Dr D Cancio
Dr M J Guiberteau Dr M E Clark
Dr I Gusev Dr D Cool
Dr L K Harding Dr J Cooper
Dr M Hiraoka Dr T Kosako
Professor Dr J Liniecki Dr J F Lecomte
Professor S Mattsson Mr J Lochard
Dr P Ortiz-Lopez Dr G C Mason
Dr L V Pinillos-Ashton Dr A C McEwan
Professor M M Rehani Dr M Measures
Professor H Ringertz Mr M Savkin
Dr M Rosenstein Dr J E Till
Dr C Sharp Dr K Ulbak
Professor W Yin Dr W Weiss
Dr Y Xia
Dr CZuur

VI
 i':?::.2:.:
..',

{&i&
l)ilr:lliiiilt$!l

Is$$...n*n!ry916p|1,,9n4t€t ;WW,:,| ::::i

nilD$$x&(|!9,,r1],::':'''r']:rr'11r1r1'1rr111,111]1,,''i,"'-rriritr:ltititliiiilt,rtiii

Publicatisn 88, Doses

ICRP Publicaqgu:q8, ,@'E*bryo ona
Doses to t@"Embryo a iei, jrom Intakes oJ' Radionuclides by
(2001 ). 1
ii :
1

leke:g.*b1 a69l a*lb.:::,.jlnat6 ::PWI

YQK?,:gxb1@i&:*:,8a*lbt"Anet6 ::p&.i.ra.lb,gii,&F&liM4l:ll*:f,.w'u*:.,ii{|&iidio
"
orlc,libn,
Part 2. Anatomy, Physiology, and Elem.ental Composition (2002).
ICRP Pub[cation-, Dosimetric Model.for the Gastro-lntestinal Tracr (20021.

Guide,f,ar"the Pt'ai:tiid!Af$]i{ation bf the,,I.eRP.,,Eurnan Respirataly'Tr{tc/:Mt}del {2001).:

ilililll til illlilil llllllll ll llll il lll lilllllll lll

01 46-6453(200009)30:3; 1 -l

Printed and bound in Great Britain by Polestar Wheatons Ltd, Exeter