Dr Harbajan Chadha-Boreham (CH) Tel/Fax: +41 22 700 6380 · 54 Philippines 1 55 South Korea 3 1 ... past Presidents (Nancy Geller and Karsten ... many people for their contributions

Number 32 December 2001 Editor: David W. Warne

Executive Committee 2002

OfficersPresident: Mr Simon Day (UK)Vice-President: Prof. Maria Grazia Valsecchi (I)Secretary: Prof. Emmanuel Lesaffre (B)Treasurer: Prof. John Whitehead (UK)MembersNews Editor: Dr David W. Warne (CH)Webmaster: Miss Silvia Codony (DK)2001-2002: Dr Elia Biganzoli (I)

Dr Harbajan Chadha-Boreham (CH)Prof. Stephen Evans (UK)Dr Siem Heisterkamp (NL)Prof. Carol Redmond (USA)Dr Julia Singer (H)Prof. Elisabeth Svensson (S)

2002-2003: Prof. Norbert Victor (D)

SubCommittee Chairs 2002

Communications Dr David W. Warne (CH)Education Prof. Carol Redmond (USA)National Groups Prof. Michael Schemper (A)Regulatory Affairs A/Prof. Jørgen Seldrup (SGP)Student Awards Dr Marie Reilly (S)

Conference Chairs 2002

LOC Dr Harbajan Chadha-Boreham (CH)

SPC Prof. Stephen Senn (UK)

Message from the President, Simon Day

11 September 2001: my hopesThe events in New York and Washington were

devastating. Thousands of people lost their lives and, fora great many of their relatives and friends, that loss willcontinue for months and years to come. Whethersubsequent events are justifiable depends on your pointof view; your point of view may depend on your proximity(or that of your loved ones) to ongoing events. Myopinions or judgements, as a small individual, are of littleconsequence.

We are all privileged to belong to aninternational society; through the ISCB, many of us havefriends, collaborators and colleagues throughout theworld. My hope is that we all continue to get along andwork together. A simple hope but not always easy toachieve.

Editorial

P; �@ob vlr bkglvfkd qeb `lkcbobk`b>� @; �H tfii tebk H�sb dfsbk jv m^mbo"� Mlq e^sfkd ql dfsb ^ qîh fp ^ luxury and I thoroughly enjoyed the whole Stockholmweek. For the first time, computer presentations werethe norm, only a few years after hand-written OHP slideswere common. The mini-symposium was especially well-^qqbkaba tfqe ^_lrq 150 lc qeb jbbqfkd�p 420 môqf`fm^kqp pq^vfkd lk. �Vbii Clkb� ql qeb lod^kfpbop.

Thanks to the other contributors to this News:the numerous book reviewers, our book review reviewer,Caroline Jackson, and Carol Redmond, Marie Reilly,John Whitehead, Emmanuel Lesaffre, Elia Biganzoli,Stephen Evans, Michael Schemper, Jørgen Seldrup,Nancy Geller, and the ubiquitous Simon Day.

Correspondence Address:

Dr David W. Warne (ISCB News)Home: Chemin Frank-Thomas 40, CH-1208 Geneva,

Switzerland. Tel/Fax: +41 22 700 6380

Work: (Rbolkl, Afljbqof`p Cbm�q);Tel: +41 22 739 3374, Fax: +41 22 739 3330

E-mail & WWW Addresses:

Permanent: [email protected] [email protected]: http://WWW.ISCB-HOMEPAGE.ORGEditor: [email protected]

IndexISCB Membership ----------------------------------------------------------------------------------- 2Oobpfabkq�p Qbmloq ql the AGM ------------------------------------------------------------------ 3Sob^probo�p Qbmloq 2001--------------------------------------------------------------------------- 4AGM Minutes - ISCB22 Stockholm------------------------------------------------------------- 6SC Report 2001: Communications ------------------------------------------------------------- 8SC Report 2001: Education ---------------------------------------------------------------------- 8SC Report 2001: National Groups -------------------------------------------------------------- 9SC Report 2001: Operating Procedures ------------------------------------------------------ 9SC Report 2001: Student Conference Awards---------------------------------------------- 9SC Report 2001: Statistics in Regulatory Affairs ----------------------------------------- 10Books for Review ---------------------------------------------------------------------------------- 11Book Review by Marc Saez (Spain)---------------------------------------------------------- 12Book Review by Dario Gregori (Italy) -------------------------------------------------------- 13Book Review by Victor Moreno (Spain) ----------------------------------------------------- 13Book Review by Bob Gibberd (Australia) --------------------------------------------------- 14Book Review by Duolao Wang (UK) --------------------------------------------------------- 15Book Review by Tiberiu Postelnicu (Romania) ------------------------------------------- 16Stockholm Review 1 � Nancy Geller --------------------------------------------------------- 17Stockholm Review 2 � David W. Warne ---------------------------------------------------- 18

Book Review by Rainer Muche (Germany) ------------------------------------------------ 21Book Review by Matthias Wahl (Germany) ------------------------------------------------ 22Workshop on Regulatory Issues in Medical Decision Support ----------------------- 23ISCB Golf at ISCB Conferences -------------------------------------------------------------- 24Details of the Student Conference Awards 2002 ----------------------------------------- 25Details of the Conference Awards for Scientists 2002 ---------------------------------- 27How to Contact the ISCB Executive Committee (2002) -------------------------------- 29How to contact the ISCB Subcommittees -------------------------------------------------- 30NEW! Reduced Subscription Rate for Statistics in Medicine and Wiley Books -- 31Wiley Announcement ----------------------------------------------------------------------------- 32Blackwell Announcement------------------------------------------------------------------------ 32ISCB General Information: ---------------------------------------------------------------------- 33Advertising Rates---------------------------------------------------------------------------------- 33Aims --------------------------------------------------------------------------------------------------- 33Changes of Address or E-mail ----------------------------------------------------------------- 33Information on Submitting Articles ------------------------------------------------------------ 33ISCB Membership Information----------------------------------------------------------------- 34Membership Subscription ----------------------------------------------------------------------- 35Calendar --------------------------------------------------------------------------------------------- 36

ISCB News # 32 Page 2 December 2001

ISCB Membership

The Stockholm conference was attended by members from 34 countries, many new for ISCB. The end-of-year membership ishigher and more international than ever before.

end end Dec Dec Dec Dec Dec Dec Dec Sep Nov May Aug Nov89 92 93 94 95 96 97 98 99 00 00 01 01 01

Trento StockholmTotal 261 596 715 698 725 702 685 729 818 360 797 518 426 837

# Countries 23 32 32 31 33 34 37 37 41 31 40 36 34 45# Country1 UK 50 90 176 120 144 121 128 169 135 78 151 95 93 1532 USA 18 45 40 39 41 40 79 66 76 34 77 52 36 893 Sweden 23 51 53 54 58 64 51 45 38 19 44 31 76 884 Germany 30 67 75 84 71 78 72 70 186 31 90 61 25 875 Netherlands 14 30 38 33 36 29 31 39 35 17 33 20 29 386 France 30 52 62 50 73 67 52 52 49 21 53 21 14 377 Poland 11 11 24 24 30 21 19 26 6 34 35 9 378 Belgium 13 22 27 30 30 32 35 29 25 12 33 24 15 369 Denmark 4 58 38 31 30 32 26 35 38 16 39 22 21 36

10 Hungary 1 21 17 18 19 25 27 29 29 3 33 34 5 3411 Italy 16 33 37 32 32 33 26 33 26 64 63 20 16 2912 Switzerland 14 25 22 80 33 29 24 25 23 5 18 16 11 2313 Finland 2 7 7 9 9 9 7 5 10 2 9 6 16 1814 Norway 13 18 25 22 12 18 10 10 11 4 10 6 9 1615 Austria 4 9 11 13 11 16 13 11 15 15 18 11 10 1516 Israel 1 3 4 4 4 4 3 3 4 7 10 5 10 1317 Spain 10 12 18 12 46 23 14 16 12 5 11 7 5 1118 Japan 2 6 7 5 7 4 10 13 20 1 12 10 1 1119 Canada 6 12 14 14 11 13 15 14 9 4 9 10 4 1020 Singapore 3 6 4 2 5 3 4 821 Australia 6 9 11 6 9 8 11 9 10 1 12 9 1 822 Portugal 1 3 5 2 2 2 2 5 5 3 1 1 423 Ireland 1 2 3 4 3 4 4 2 3 1 2 2 3 324 New Zealand 1 1 2 1 2 2 1 2 1 1 325 South Africa 1 4 1 3 2 2 2 2 2 2 1 326 Lithuania 2 227 China Hong Kong 1 1 2 3 3 3 3 3 3 2 1 328 Slovenia 1 2 3 2 1 1 3 2 1 1 1 229 Estonia 1 230 Cuba 2 2 2 2 231 Thailand 1 1 1 1 2 1 1 2 2 232 Argentina 1 1 133 Brazil 2 1 134 Kuwait 1 1 135 Ukraine 1 1 136 Czech. Rep. 1 1 1 1 1 1 2 1 2 1 137 Greece 1 1 1 1 1 4 3 1 138 India 1 1 1 1 1 1 1 1 1 2 1 139 Malaysia 2 1 2 2 1 1 140 Mexico 1 1 1 1 1 1 141 Romania 2 4 1 1 1 142 Russia 1 3 3 3 2 1 2 143 Saudi Arabia 144 Sudan 145 Taiwan 1 1 1

46 Egypt 147 Colombia 1 1 1 1 148 Croatia 1 1 149 Indonesia 150 Iran 1 151 Kenya 1 152 Oman 153 Pakistan 1 1 154 Philippines 155 South Korea 3 156 Turkey 1 1 157 Zimbabwe 1


=RGSKFGOT\S >GQPRT TP TJG 069

From Simon Day

I thank Nancy Geller for guiding ISCB through1999 and 2000 and for passing on to me athriving and exciting Society. The transitionwent very smoothly thanks to a lot ofcommunication in the months leading up tothe hand-over and thanks to her continuedsupport and advice since then.

We have quite a new Executive Committee;there are six new Ordinary Members althoughall of the Officers have served on theExecutive for at least two years. I particularlywelcome the new members and hope they willwork at least as hard for the Society as theretiring members have done (and in somecases, still are).

Not everything is a look to the future: the firstfew months of this year saw the finishing ofthe Trento meeting. I thank Maria GraziaValsecchi for her work and also Bruno Cesanafor giving the accounts an independent reviewand reporting back to us that everything wasaccounted for thoroughly and properly.

The other big task for me, our two immediatepast Presidents (Nancy Geller and KarstenSchmidt) and Tony Johnson has been toconsider operational aspects of ourConstitution in great detail. The revisions thatwe are suggesting were outlined at the AnnualGeneral Meeting. Based on feedback from theExecutive Committee meeting and from theAGM, several formal proposals for changeswere sent to all Members of the Society for apostal ballot in the autumn*.

The other subcommittees have also beenbusy.

) Immediately foiiltfkd i^pq vbô�p conference, we ran a course onepidemiological methods in Prague andrecently we ran two simultaneouscourses (one on cluster randomisationand one on Bayesian methods inhealthcare) in Cuba.

) The Student Conference Awardsprogramme has been successfully runagain, as has the Scientists Awardsprogramme (under the banner of theNational Groups subcommittee). Iwould like to encourage moresubmissions for both of these valuableawards.

) The Regulatory Affairs subcommitteehave commented (and had theircomments incorporated) into severaldocuments.

) The Communications subcommitteehas clearly produced further excellenteditions of the News and is consideringfurther developments for that and forqeb Rl`fbqv�p tb_ pfqb.

) The Fraud subcommittee has not beenso active; there is plenty of enthusiasmbut some difficulty in making progress.We decided to end this subcommittee. Ithank them and congratulate them ontheir achievements.

Indeed, I express my thanks andcongratulations to all these groups. Their ownreports are presented elsewhere in the News.

I want to mention much of the (as yet) unseenwork of future conference organizers: theDijon team for 2002, the joint team (with theSociety for Clinical Trials) for the Londonmeeting in 2003 and the Leiden team for2004. We are considering a proposal for 2005from Krakow so there is a lot to look forwardto. Moreover, of course, my great thanks andcongratulations to everyone involved inlod^kfwfkd ^ka orkkfkd qefp vbô�p jbbqfkd fk Stockholm. It was first class. Well done!

Finally, whilst I take this opportunity to thankmany people for their contributions to ISCB, Iparticular thank the other Officers of theSociety with whom I have had a great deal ofcontact and sharing of thoughts and ideas. Itrust this will continue throughout 2002.

*with the December News


@RGCSURGR\S >GQPRT +))*

From John Whitehead

1. An nb_ _h^ i` 2000, nb_ Si]c_ns~m _kocns ][jcn[fstood at £104,305.50, compared with £118,850.89at the end of 1999. This is a reduction of£14,545.39 compared with an increase of £6,785during 1999.

2. The Trento conference produced a surplus of£3,742.87 for the Society.

3. During the first six months of 2001, the equitycapital has fallen to £95,724.70, compared with£117,715 at this time last year.

4. Membership was 521 at the end of June 2001, 18up on this time last year.

5. Tb_ Si]c_ns~m `ch[h]c[f l_jiln ìl 2000 cmpresented in the new format approved last year.This clarifies the costs to the Society of itemssuch as the ExCom and Officers, the StudentConference Awards, the Conference Awards forScientists and the Newsletter. As the figures werenot prepared in this way during 1999, on thisoccasion direct comparisons are not available forall items.

6. The Society's equity capital has fallen by £14,545during 2000 and a further £8,580.80 during thefirst half of 2001. This has eaten a little awayfrom our large reserve. The remaining reserve of£95,724.70 is slightly less than the total cost ofthe Trento conference (about £100,000).Although it would be hard actually to lose thatamount on one conference, it would be prudent to

ensure that erosion of our funds does not continueat the present rate.The way in which we have been collectingsubscriptions for Statistics in Medicine on behalf ofWiley has cost us £2,850 during 2000 due tocurrency fluctuations, and a further £2,976.55during the first half of 2001. Although suchfluctuations might move in the other direction infuture years, the Executive Committee havedecided that the arrangement cannot continue in itspresent form. A motion to this effect was put to theAGM of the Society at the Stockholm meeting, andwas accepted. Thus, the current form ofarrangement will end this year: discussions areproceeding with Wiley about whether a new formof arrangement can be introduced for 2002.

7. The AGM in Stockholm also approved a motion toraise the annual subscription fee from £15 to £20.

8. The finances of the Society remain healthy, andattempts to make good use of our funds have beensuccessful. However, it is clear from the reductionin the Society's equity capital that we must becautious about entering into new schemes forfunding activities for the next year.

Acknowledgement

I thank the staff of the ISCB Permanent Office in Denmark,and in particular Rita Schou, for their efforts inadministering the accounts of the Society and for preparingthe material for this report.

ISCB Membership, 1998-2001483 729 488 817 503 797 521

0

200

400

600

800

1000

JUNE DEC JUNE DEC JUNE DEC JUNE

1998 1999 2000 2001


@RGCSURGR\S >GQPRT +))* %EPOTKOUGF&/ 5KOCOEKCM >GQPRT HPR *... COF +)))

2000 1999

£ £

Equity from previous year 118,850.89 112,065.16

IncomeMembership fees 10,860.18 11,055.31Conference surplus 3,742.87 20,813.46Course surplus 1,189.00Advertising 1,500.00 1,476.40Earned interest 4,789.07 4,713.00

22,081.11 44,427.34Statistics in Medicine 36,762.09 39,162.73

Total income 58,843.20 77,220.90

ExpenditurePermanent Office:Consumables 472.04 906.85Postage & freight 7,626.68 3,586.28Telecommunication & internet 300.66 621.83Printing 5,788.05 3,442.79Temporary staff, Spadille 767.29 428.02Administration, Spadille 6,547.92 4,764.65

21,502.64 13,750.42Officers & ExCom:Travel expenses 1,586.98 2,609.56Conference fees 1,242.13 285.77Other expenses 1,032.14 57.97

3,861.25 2,953.31Student Conference Awards:Travel expenses 772.48Conference fees 644.75Other expenses 1,313.72

2,730.95Conference Awards, Scientists:Conference fees 967.12Other expenses 1,237.91

2,205.03Workshops / Courses:Honorarium 624.64Travel expenses 701.21

1,325.85Newsletter:Office expenses 80.82 53.37

Other items:Bank charges 1,029.88 909.51Audit 888.51 837.41Movement of equity funds 11.26Currency fluctuations* 159.93 11,828.06

2,070.32 12,793.78

Total expenditure(excl. Statistics in Medicine)

33,776.86 30,460.39

Statistics in Medicine 39,603.73 39,974.78

Total expenditure 73,388.59 70,974.78

NET INCOME: -14,545.39 6,785.73

AssetsBank accounts:Current account, Barclays 5,124.19 576.45High interest account, Barclays 5,224.81 2,919.10Unibank/Nordea, Denmark 32,569.43 47,157.06Bonds, Danish State 7% 2004 57,150.67 57,230.12

100,069.10 107,882.73

Others:Accounts receivable 5,974.99 150.00Surplus, ISCB-20 Heidelberg 11,279.82Seed money, ISCB-21 Trento 2,834.90Seed money, ISCB-22 Stockholm 1,764.31Seed money, ISCB-23 Dijon 5,101.81

12,841.11 24,264.72

Total Assets 112,910.21 122,147.45

LiabilitiesOwing to Spadille 5,541.71 2,122.22Audit 888.51 837.41Other 2,174.49 336.93

Total Liabilities 8,604.71 3,296.56

EQUITY 104,305.50 118,850.89

*) Rate of exchange DKK/£, 1998.12.31: 1058.43Rate of exchange DKK/£, 1999.12.31: 1194.16 DKK 104,063.87 x (1 - 1058.43 / 1194.16) = £ 11,828.06

*) Rate of exchange DKK/£, 1999.12.31: 1194.16Rate of exchange DKK/£, 2000.12.29: 1195.82 DKK 115,205.34 x (1 - 1194.16 / 1195.82) = £ 159.93


AGM Minutes - ISCB22 Stockholm

Tuesday 21 August

The meeting started at 12.15

The agenda of the meeting was:

1. President's Report

2. Treasurer's Report

3. Subcommittee Reports

4. Nominations for the ExCom

5. Future meetings

6. Other business

FaTbXST]cnb aT_âc %I9&

The president presented the agendafor the meeting and described brieflythe changes in the membership of theExecutive Committee as well of theboard of Officers. Namely, there aresix new Ordinary Members to theExCom, although all of the Officershave served on the ExecutiveCommittee for at least two years. Thepresident thanked NG for guiding theISCB through 1999 and 2000 and forpassing on a thriving and excitingSociety. This transition went verysmoothly thanks to a lot ofcommunication in the months leadingup to the hand-over and because ofher continued support and advicesince then. Further, he overviewedbriefly the activities of theSubcommittees. He also announcedthat the Subcommittee on Fraud hasbeen stopped by the ExCom after asuccessful period of 4 years.However, he stressed that the ExComis always open to new ideas for raisinga new Subcommittee on Fraud in afew years time should there be a needfor this and if there would be enoughnew activities to warrant its existence.Further, the president alsoencouraged everyone to stimulatestudents to take part in theConference Student Awardscompetition and the ConferenceAwards for Scientists scheme. Finally,he ended by giving an overview offuture ISCB conferences. The reportwas approved by the Membership.

JaTPbdaTanb aT_âc %@M&

JW showed the membership over thelast 4 years. In June 2001, there were521 ISCB members, 18 up on thistime last year. Hence, in this respectthe society is doing quite well. JWthen presented the financial figures ofqeb Rl`fbqv. Seb Rl`fbqv�p cfk^k`fî report for 2000 was presented in thenew format approved last year. Thisclarifies the costs to the Society ofitems such as the ExCom andOfficers, the Student ConferenceAwards, the Conference Awards forScientists and the Newsletter. As thefigures were not prepared in this wayduring 1999, on this occasion directcomparisons are not available for all

items. The financial balance showedthat the society has to be a bit carefulon the finances. At the end of 2000,qeb Rl`fbqv�p bnrfqv `^mfqî pqlla ^q £104,305.50, compared with£118,850.89 at the end of 1999. This isa reduction of £14,545.39 comparedwith an increase of £6,785 during1999. Despite the fact that the Trentoconference produced a surplus of£3,742.87 for the Society, during thefirst six months of 2001, the equitycapital has fallen to £95,724.70,compared with £117,715 at this timelast year.

JW expressed his concern about thefact that the Society's equity capitalhas fallen by £14,545 during 2000. Theremaining reserve of £95,724.70 isslightly less than the total cost of theTrento conference (about £100,000).He advised to be prudent to ensurethat the erosion of our funds will notcontinue at the present rate. In thesearch for possible causes for thistrend in financial losses, he mentionedthat the collection of subscriptions forStatistics in Medicine on behalf ofWiley cost the Society £2,850 during2000 due to currency fluctuations, anda further £2,976.55 during the first halfof 2001. Although such fluctuationsmight move in the other direction infuture years, he felt that it would bewise to avoid the risk. Furthermore,dealing with that subscription takes asubstantial amount of Rita Schou'stime at the Permanent Office. JWstressed that the Society's schemes tosupport students and scientists fromEastern European and Third Worldcountries are important and thereforeshould be continued. It must also benoted that there is a cost of theSociety's democracy, in the form ofmailing ballot papers. Furthermore, thetreasurer for the 2001 meeting hasindicated there may be a loss of about£20,000. In the light of the currentfinancial situation, JW proposed that(a) the Society ends the reducedsubscription arrangement for Statisticsin Medicine and (b) the annualsubscription fee is raised from £15 to£20. Both proposals were unanimouslyaccepted by the ISCB members.

Finally, JW thanks Rita Schou for herwork in preparing the detailed financialbalances. The report was approved bythe Membership.

Subcommittee Reports

The six Subcommittee Chairs reportedon the activity of their subcommittee,and their complete reports arepublished together with these minutesin the Newsletter, with terms ofreference and list of members. Asmentioned above, the Subcommitteeon Fraud has ended after being veryproductive, especially in the first 2

years of its existence. The reports ofthe 6 Subcommittees were:

Jørgen Seldrup on RegulatoryAffairs

The Subcommittee has prepared (a) a`ljjbkq lk �Qbsfpba Frfabifkbp clo Manufacturers and Sponsors ofTechnologies Making Submissions toqeb Hkpqfqrqb�, (_) ^ `ljjbkq lk BOLO/DVO/2330/99 ao^cq �Olfkqp ql consider on validity and interpretation ofmeta-analyses, and one pivotai pqrav� (c) a comment on CPMP/EWP/1776/99ao^cq �Olfkqp ql `lkpfabo lk jfppfkd a^q^� ^ka fp mobmôfkd ql `ljjbkq lk the Council for InternationalOrganisations of Medical Sciences(CIOMS) draft revision of the 1993International Ethical Guidelines forBiomedical Research Involving HumanSubjects. Further, the Subcommittee isaware of the existence of three otherDLD@ �Blk`bmq O^mbop� lk tefè qebv plan to comment when they becomeavailable. SD asked the members tolook out for other guidelines (e.g. USAguidelines) and to report back to theSubcommittee.

Carol Redmond on Education

CR thanks the previous chairmanMichael Campbell (UK) and alsoMichael Schemper (Austria) for theirefforts in the past. Two ISCB sponsorededucational programs have beencarried under their supervision. On 9and 10 September 2000, Dr. MartinGulliford and Professor NormanBreslow conducted a very successfulworkshop with 45 students, entitled�Oi^kkfkd ^ka Rq^qfpqf`î @kîvpfp lc Dmfabjflildf`î Rqrafbp,� pmlkploba _v the ISCB in conjunction with theNational Institute of Public Health, theDepartment of Biostatistics andInformatics and the Czech StatisticalSociety, in Prague. Further, on 10 and11 April 2001 Drs. Allan Donner andDavid Spiegelhalter taught two-day pre-conference courses in Veradero, Cuba,on Cluster Randomization andBayesian Methods in Health Research,respectively. The first course wasattended by over thirty participants fromCuba, Spain, Portugal and Chile. Therewere 25-30 participants, mostlyCubans, who attended the course onBayesian Methods.

CR also pointed out that there were nodefinite plans for new courses and thatproposals for new topics and locationsare very welcome. However, during thepreceding ExCom meeting several newideas were generated. For instance,one considered the possibility oforganizing a Website-based trainingprogramme. Further, a course forstatisticians on how to referee statisticaland medical papers was suggested.


AGM Minutes - ISCB22 Stockholm (continued)

Michael Schemper on NationalGroups

MS reported that a total of 12applications for the ConferenceAwards for Scientists for theStockholm ISCB meeting werereceived and according to a rankingby the National Groups Subcommittee10 award winners are: M.Gorkiewicz,Poland, J.Reiczigel, Hungary,H.Pisarev, Estonia, K.Fischer,Estonia, H.Stanuch, Poland, R.Chen,China, P.Vargha, Hungary,M.Shcherbatyy, Ukraine, Dawidowicz,Poland, M.Jurkowska, Poland. Thesame kind of support will be given toscientists wishing to attend the ISCBmeeting in Dijon in 2002. Therefore,SD suggested that enough publicityshould be made for this ISCBinitiative.

David Warne on Communications

First of all, DW mentioned that he hasbeen responsible for the Newsletterfor almost 10 years. He then thankedBjarne Nielsen for updating theRl`fbqv�p Vb_ pfqb obdriôiv. Gb îpl mentioned that the Subcommittee hasthought of new ways ofcommunicating with the ISCBmembers. For instance, theNewsletter could be put on theWebsite. In addition, the subscriptionto the ISCB society could be handledelectronically as well as the voting fornominations and or changes to theconstitution. The Subcommittee willcarefully examine which of thesesuggestions can be installed withoutmuch cost.

John Whitehead on StudentConference Awards

Last year it was decided that theSubcommittee would support up to 5students. A total of nine applicationswere made from six countries(Belgium: 3, Finland: 2, France: 1,Greece: 1, The Netherlands: 1, UK: 1).The winning awards for 2001 are (1)Fazil Baksh (University of Reading,UK); (2) Tom Loeys (University ofGhent, Belgium) (3) Didier Renard(Limburgs University, Diepenbeek,Belgium); (4) Samuli Ripatti(University of Helsinki, Finland) and(5) Sarah Zohar (Hôpital Saint-Louis,Paris, France). The total cost to ISCBof the 2001 scheme is likely to bearound £3,605, compared to £2,731 in2000.

JW mentioned that two winners in2001 also won in 1999, which wascompletely according to rules asstipulated. However, the purpose ofthis award is to support youngstudents for the first time. Thereforethe ExCom decided that the rulesshould be adapted so that a student

`^k ob`bfsb qeb Rqrabkq�p @tôa lkiv once in his lifetime.

Simon Day on Operations (on behalfof Karsten Schmidt)

The Operations Subcommittee hasreviewed all of the Constitution andfocused attention on several areas thatare in most need of clarification orchange. These are (a) 6.04: co-optingthe Webmaster onto the Committee, (b)7.03: appointment of Chairs ofSubcommittees, (c) 7.08: length ofserving on a Subcommittee, (d) 10.02:ensure Chair of programme committeeis a member of ISCB, (e) 10.07:audit/review of conference accounts, (f)13.02: reimbursement of expenses ofCommittee members, (g) 13.03:reimbursement of expenses of Officers,(h) 13.05: reimbursement of expensesfor other reasons, (i) 15.04:appointment of an Election Officer, (j)15.05: appointment of a NominationsCommittee, (k) 15.06: prohibit oneMember nominating more than threecandidates for the ExecutiveCommittee and clarify remit ofNominations Committee, (l) 16.01:content of Web site and (m) 16.02:appointment of Webmaster.

Full details will be sent to theMembership and a postal vote will beheld in accordance with Article 19.01 ofthe Constitution in the autumn of thisyear*.

Nominations

One ExCom position is vacant and wasannounced in the June issue of theNewsletter. SD invited everyone tonominate ISCB members for thisvacant position by signing a letternominating the candidate and lettingalso the candidate sign. Letters can besent to the Secretary by 31 August2001. There is already one candidatefor this position, namely Prof. Dr.Norbert Victor has been nominated byhis colleague Prof. Dr. U. Mansmann.Prof. Victor, who has been a member ofthe ExCom for 2 years can be re-elected.

Future ISCB meetings

ISCB 2002 meeting in Dijon byHarbajan Chadha-Boreham. HCBreported that most of the preparativework for the ISCB meeting in Dijon hasbeen done. The LOC has beeninstalled and consists of 5 members.The scientific committee consists of:Stephen Senn (Chair), MichaelAbrahamowicz (Canada), Peter Diggle(UK), Val Federov (USA), PhilipHougaard (Denmark), Witold Kupsc(Poland), Thierry Moreau (France),Michael Proschan (USA), Amy Racine(Switzerland) and Helmut Schaefer(Fboj^kv). Eroqebo, ^ �Efopq @kklrk`bjbkq� ib^cibq lk qeb Cfglk

meeting has been prepared and is inqeb môqf`fm^kq�p m^`h^db lc qeb mobpbkq ISCB conference. She invited allmembers to attend the ISCB meetingin Dijon and promised that it will beagain an exciting meeting both fromthe scientific as well as from the socialpoint of view.

ISCB 2003 meeting in London

SD reported that the ISCB meeting inLondon will be organised incollaboration with the Society forClinical Trials. The ISCB organiser isDiane Elbourne with Deborah Ashbyco-chairing the Scientific Committeewith Mary Foulkes.

ISCB 2004 meeting in Leiden

Hans Van Houwelingen announcedthat the dates for the Leiden meetinghave been fixed to 15-19 August 2004.The venue will be the Holiday Inn inLeiden.

ISCB 2005 meeting in Krakow

SD reported that the ISCB meeting in2005 has not been fixed yet, but thatcontacts have been made with peoplefrom Krakow. More informationconcerning this meeting will follow inthe coming year.

Other business

No questions were raised from thefloor. The meeting closed at 13:05.About 80 ISCB members attended themeeting:

Attendance at AGM

Andersen Marc; Archer Philip; ArmitagePeter; Babinski; Baras Mario; Biganzoli Elia;Brattström Gudrun; Chadha-BorehamHarbajan; Colton Theodore; Dalesio Otivia;Day Simon; Dawidowicz Antoni Leon;Decarli Adriano; Elton Robert; Engl Werner;Evans Stephen; Fidler Vaclav; FosenJohan; Francis Mildred; Fransson Eleonor;Fritsch Sandor; Gallus Giuseppe; GehanEdmund; Geller Nancy; Hagman Anna;Heisterkamp Simon; Hougaard Philip;Hutton Jane; Johansson Anna; KlersyCatherine; Kupsc Witold; Kuss Oliver;Lanke Jan; Lesaffre Emmanuel; LombardCarl ; Machin David; Mansmann Ulrich;Marshall Roger; Mowery Richard; MulderPaul; Murphy James; Olson Melvin; OlssonJonny; Palmer Chris; Palmgren Juni;Petersen Janne; Plna Kamila; PrescottRobin; Prescott Gordon; Quantin Catherine;Raab Gillian; Redmond Carol; RoystonPatrick; Sauerbrei Willi; Schemper Michael;Schouten Hubert; Seldrup Jørgen; SennStephen; Shcherbatyy Mykhaylo; SingerJulia; Stallard Nigel; Stevens John; StijnenTheo; Svensson Elisabeth; SvenssonHanna; Thalabard Jean-Christophe; VachWerner; Vaittinen Pauli; Valsecchi MariaGrazia; Van Houwelingen Hans; Van PuttenWim; Van Strik Roel; Vargha Peter; VastrupPernille; Warne David; Whitehead John;Wentzel-Larsen Tore; Zdravkovic Slobodan;Zhou Yinghui.

*with the December News


SC Report 2001: Communications

From David W. Warne

The SC met on Mon 20 Aug 01 during theStockholm meeting (present NG, ES,DWW, apologies from BN, CJ, SD).

Future uses of the Internet site werediscussed:

1. Publish the News electronicallyMight alienate some members if onlydone on web and not on paper. So keep2 methods for now.Paper version still needed fordistribution at Conference.Members could elect to only receiveNews electronically and not by post -would reduce costs for ISCB.How would e-News be delivered ?By email would be complicated usingthe egroups as we'd need 2 lists...messy. But no password would beneeded.If on Web then would we want to protectthe News document from possibility ofcopying?Can already do this anyway, but couldput on web as non-editable pdf file.To stop anyone getting the News, i.e.non-members, could have a singlepassword for members emailed to themeach year.No reduced fee if no paper News sentas would be too complicated toadminister.Question of size of pdf? Web would bebetter than individual emails.Web would be quicker for people withslow Post e.g. USA.

2. Pay Subs via WebLooked into last year and apparentlyexpensive - should be reviewed.Should be possible technically but someon ExCom 2000 doubted security.

3. Vote on ExCom, Constitution ChangesMight increase participation.Could be quite complicated forpasswords.

4. WebsiteShould be dynamic and exciting and asup-to-date as possible (SC to monitorwebsite and inform WM 4x/year).SCs should publish their reports onwebpages:- pre-conference for AGM discussionand- end-of-year as seen in News.SCs should use egroup to contactmembership e.g. when new guidelinesare open to comments.Should be used as a central storagesite for ISCB documents, accessible toOfficers, ExCom etc. Need secureaccess and passwords!

5. News QualityThis was examined with someexamples of papers of different qualitiesbeing compared. Extra costs notthought to be useful.Heavier paper might be more expensiveto post.Will continue to use colour for centrepages of December News. We won'tchange the News' design yet.The distribution costs currently aremuch more than those of printing (tocheck with Office).Ideas: history, index incl. book reviews,golf, walking.

Other ideasWould be good to acknowledge (new^ka obkbtfkd) jbj_bop� m^vjbkqp ^p they often don't know if they'remembers or not!Difficult to administer membership list(ISCB Office) and Egroup (DWW)separately - need to explain the systemto Office.Operations SC vote will be sent with theDecember News.Books sent to DWW and this takes a lotof time to sort out, choose reviewers,send out books, collect them (nagging)BUT changing my address in the

publishers' databases can take years...only now after 10 years is JorgenSeldrup not getting any! Maybe need toseriously decide to have another personresponsible.Annual Meeting's webpages should bestored by the ISCB webpages after themeeting so that they're there in futureyears as models for future organisers.ES gave DWW ISI list of events. He hasregistered next few ISCB conferenceson it. WM will use this list to updateISCB website with EFSPI documents.

Post-meeting thoughtsShould we continue policy of egroupsallowing membership for 2 years? i.e.we need to think whether people can be"egroup members" only without beingISCB members and paying GBP 20.We need to explain the egroups systembetter.We still need to explain that attending aconference grants membership for ayear (and a bit), and means the newmember can attend the AGM at thatconference. In fact, all people attendingthe conference, whether new members,Student or Scientist Award winners aremembers.

And finally,Thanks to Elisabeth for help over thelast few years - she's leaving to joinEducation SC.Thanks to Caroline for the bookreviews...Thanks to Bjarne for the webpage workwhich has now be passed on to SilviaCodony ([email protected])Stephen Evans has agreed to join theSC.Next time we need a secretary to writeup the minutes quickly!

SC Report 2001: Education

From Carol Redmond

The Education Subcommittee has continued to promote thedevelopment of courses that are consistent with its Terms ofReference.

On 10 and 11 April 2001, Allan Donner and David Spiegelhaltertaught two-day pre-conference courses in Veradero, Cuba, onCluster Randomization and Bayesian Methods in Health Research,respectively. The courses preceded the Third Biostatistics NationalConference (Cuba) and the First Medical Statistics Iberian-American Meeting. A total of over thirty participants from Cuba,Spain, Portugal and Chile attended the Cluster Randomizationcourse. With excellent assistance in translation from Dr. RosaJiménez, an interesting dialogue on some importantmethodological issues was established. There were 25-30participants, mostly Cubans, who attended the course on BayesianMethods, which was greatly enhanced by online softwaredemonstrations using the data projector available. Both instructorswere gratified that the subjects of their courses were directlyrelevant to current activities of the participants that led tointeresting discussions. We are pleased at the success of thesecourse offerings and thank Allan and David for their finecontributions.

Michael Schemper has produced a register of likely targetcountries and contacts for potential future course offerings.

Members of the Education Subcommittee present met during theannual ISCB meeting in Stockholm. We discussed follow-up withindividuals in target countries who had previously expressed aninterest in educational offerings. We will continue to pursue thesepossibilities, as well as new opportunities for educational activitiesthat may arise during the year. If any ISCB members havesuggestions for future courses or contacts in target countries, welook forward to receiving your contribution. Please send your ideasfor courses or other educational activities to a member of ourEducation Subcommittee for consideration. Terms of Reference forISCB sponsored courses are available for guidance online at:

http://www.iscb-homepage.org/subcommit.htm.

As the incoming Chair of the Education Subcommittee, I wish toextend special thanks to Michael Campbell for his outstandingleadership of the Subcommittee over the past several years. He hasbeen most helpful during the transition in the chair and remainsactive as a member of the Subcommittee. Nancy Geller, who hasbeen a member of the subcommittee for several years, has decidedto stand down. We thank her for her contributions and support for thepr_`ljjfqqbb�p bkab^slrop. Efkîiv, tb ôb mib^pba ql tbi`ljb several new members. They are: Maria Grazia Valsecchi, ElisabethSvensson, and Nicole Close.


SC Report 2001: National Groups

From Michael Schemper

A total of 12 applications for the ConferenceAwards for Scientists for the Stockholm ISCB22meeting were received and according to a rankingby the National Groups Subcommittee thefollowing 10 award winners were selected:M. Gorkiewicz, Poland,J. Reiczigel, Hungary,H. Pisarev, Estonia,K. Fischer, Estonia,H. Stanuch, Poland,R. Chen, China,P. Vargha, Hungary,

M. Shcherbatyy, Ukraine,Dawidowicz, Poland,M. Jurkowska, Poland.

A maximum of 12 awards will be granted for theISCB2002 meeting in Dijon.

The subcommittee continues to encourage theformation of new National Groups, carefullyconsidering benefits for the biostatisticians of theregion and for ISCB.

SC Report 2001: Operating Procedures

Separate papers are enclosed with this mailing.

We urge all members to consider the motions presented and to vote.

After the closing date for votes, changes approved will be written into the Constitution and theSubcommittee will then come to an end.

The deadline for votes is 28 February 2002.

SC Report 2001: Student Conference Awards

From John Whitehead

MembersJohn Whitehead (Chairman)Bjarne Nielsen (Secretary)Marc BuyseBruno CesanaSimon Day

Terms of reference

Student conference awards are available forregistered postgraduate students to attend theannual meeting and present a paper. TheSubcommittee shall receive submissions, judgethem, and administer the awards. The rules areannounced in a timely issue of the Newsletter.

Report of the Subcommittee, 2001

1. The students winning awards for 2001, and theirtalks, were

Fazil Baksh, University of Reading, UK

Design considerations in the sequential analysis ofmatched case-control data

Tom Loeys, University of Ghent, Belgium

A causal inference with survival data in a proportionalhazards framework

Didier Renard, Limburgs University, Diepenbeek,Belgium

Validation of a longitudinally measured surrogatemarker for a time-to-event endpoint

Samuli Ripatti, University of Helsinki, Finland

Joint modelling of genetic association and populationstratification using latent class models

Sarah Zohar, Hôpital Saint-Louis, Paris, France

Phase I and II dose-ranging clinical trials: Proposal ofa two-stage Bayesian design

2. The total cost to ISCB of the 2001 schemewas £3769, made up of accommodation (£1183),daily expenses (£296), registrations (£993),conference dinners (£195) and travel (£1102).

3. A total of nine applications were receivedfrom six countries (Belgium: 3, Finland: 2, France: 1,Greece: 1, The Netherlands: 1, UK: 1). Thesubcommittee felt that the standard of the entrieswas satisfactory, but would have hoped for moreparticipation.

5. Two winners in 2001 also won in 1999. TheExecutive Committee decided that the rules of thescheme should be amended to specify that previouswinners are not eligible to apply again.

6. After four years in the job, the Chairman ofthe Subcommittee has decided to step down. DrMarie Reilly of Karolinksa Institute, Stockholm,Sweden has kindly agreed to take on this role for2002.


SC Report 2001: Statistics in Regulatory Affairs

From Jørgen Seldrup

Members: Harbajan Chadha-Boreham, Geneva, SwitzerlandAnna Petroccione, Milano, ItalyHelmuth Schäfer, Marburg, GermanyKarsten Schmidt, Fredensborg, DenmarkJørgen Seldrup, Singapore (Chair)Stephen Senn, London, United Kingdom (Secretary)

The Sub-committee was reconstituted at the Annual General Meeting in Trento, Italy with two new membersjoining: Harbajan Chadha-Boreham, Switzerland and Anna Petroccione, Italy.

The current terms of reference are: "The subcommittee on Regulatory Affairs will review, comment upon and seekto influence the development of regulatory requirements, guidelines and other documents concerning the scientificaspects of data generation, collection, management, analysis, and reporting. In general, the Subcommittee will seekout and handle all regulatory issues in the name of the Society with the approval of the President or in his absence,the Vice-President"

During the time since the 2000 Annual Meeting in Trento, Italy, the Sub-Committee has prepared a comment on:1. {R_pcm_^ Goc^_fch_m ìl M[ho`[]nol_lm [h^ Sjihmilm i` T_]bhifiac_m M[echa So\gcmmcihm ni nb_

Ihmncnon_|, Fclmn Dl[`n, 12 O]ni\_l 2000.The Comment was submitted to NICE, National Institute for Clinical Excellence, London, UnitedKingdom, on 9 November 2000.The revised Guideline was issued in March 2001 and the ISCB gets a mention in Appendix B. It[jj_[lm nb[n nb_ Ihmncnon_ niie hinc]_ i` nb_ Si]c_ns~m ]igg_hnm [m cn q[m jimmc\f_ ni c^_hnc`simprovements relevant to several of our major points (and even a minor one).

2. CPMP/EWP/2330/99 ^l[`n {Pichnm ni ]ihmc^_l ih p[fc^cns [h^ chn_ljl_n[ncih i` g_n[-analyses, andih_ jcpin[f mno^s| Lihîh, 19 O]ni\_l 2000.The Comment was submitted to the EMEA Secretariat, London, United Kingdom, on 27 January2001.The final version adopted by CPMP in May 2001 appears to have taken virtually all the pointsmade by the Sub-Committee into consideration.

3. CPMP/EWP/1776/99 ^l[`n {Pichnm ni ]ihmc^_l ih gcmmcha ^[n[|, Lihîh. 25 J[ho[ls 2001.The Comment was submitted to the EMEA Secretariat, London, United Kingdom, on 27 April2001.

4. Council for International Organisations of Medical Sciences (CIOMS) draft revision of the 1993International Ethical Guidelines for Biomedical Research Involving Human Subjects.The Comment was submitted to CIOMS c/o WHO, Geneva, Switzerland, on 16 July 2001.

Copies of the above four documents are available from the Chairman.

The Sub-Committee met again on 20 August 2001 on the occasion of ISCB 22 in Stockholm.Plans to comment on some of the many disease specific guidelines have not yet been realised partly due to the[]ncpcnc_m chpifp_^ ch jl_j[lcha ìl [h^ ]igg_hncha ih nb_ [\ip_ }aoc^_fch_m~. Am mo]b, aoc^_fch_m ]ihncho_ nibecome available and as time allows we will pursue these plans further.

Tb_ So\]iggcnn__ cm [q[l_ i` nb_ _rcmn_h]_ i` mig_ nbl__ EMEA }Cih]_jn P[j_lm~ ih qbc]b q_ jf[h ni ]igg_hnas and when they become available:

) Adjustment for Multiplicity and Related Topics (CPMP/EWP/908/99); draft being worked on by theEfficacy Working Party; consultation date unknown.

) Adjustment for Baseline Covariates (CPMP/EWP/2863/99) - no draft yet.) Choice of Delta (CPMP/EWP/2158/99) - no draft yet.

On a personal note, the Chairman wishes to thank all Sub-Committee members for their hard work and timelyeffort without which NICE, EMEA and CIOMS would not have had the benefit of ISCB expertise.


Books for Review

In this issue:Author(s) Title Publisher (year) Reviewer

Darlene K. Stangl & Donald A. Berry (eds.) Meta-Analysis in Medicine and Health Policy Marcel Dekker (2000) Marc SaezEdward L Korn & Barry I Graubard Analysis of Health Surveys John Wiley (1999) Dario GregoriChi-Lun Cheng & John W Van Ness Statistical Regression with Measurement Error Arnold (1999) Victor MorenoSung H Park & G Geoffrey Vining Statistical Process Monitoring and Optimisation Marcel-Dekker (2000) Bob GibberdA H Leyland & H Goldstein Multilevel Modelling of Health Statistics John Wiley (2001) Duolao WangPhilip Hougaard Analysis of Multivariate Survival Data Springer (2000) Tiberiu PostelnicuMartin Bland An Introduction to Medical Statistics (3rd ed) Oxford (2000) Rainer MucheTerry M Therneau & Patricia M Grambsch Modeling Survival Data: Extending the Cox

ModelSpringer (2000) Matthias Wahl

Reviews awaited:Author(s) Title Publisher (year) Reviewer

Peter Armitage (ed) Encyclopedia of Biostatistics: Vol. 4: Med-Pre John Wiley (1998) Aurelio TobiasMichael R Hamrell (ed) The Clinical Audit in Pharmaceutical

DevelopmentMarcel-Dekker (2000) Marc Buyse

Shein-Chung Chow & Jen-Pei Liu (1) Design and Analysis of Bioavailability andBioequivalence Studies

Marcel-Dekker (2000) Laszlo Endrenyi

James E. De Muth Basic Statistics and Pharmaceutical StatisticalApplications

Marcel Dekker (1999) Ann Martin

CF Jeff Wu & Michael Hamada Experiments: Planning, Analysis, andParameter Design Optimisation

John Wiley (2000) Gilg Seeber

Shein-Chung Chow & Jen-Pei Liu (2) Design and Analysis of Bioavailability andBioequivalence Studies

Marcel-Dekker (2000) Graham Kimber

Books recently sent for review:Author(s) Title Publisher (year) Reviewer

Michael Healy Matrices for Statistics Oxford (2000) Istvan JanosiDonald C Monkhouse & CT Rhodes (Eds.) Drug Products for Clinical Trials Marcel Dekker (1998) Koos LubsenMartin Bland and Janet Peacock Statistical questions in Evidence Based

MedicineOxford (2000) John Carlin

Douglas G Altman, David Machin, Trevor NBryant, Martin J Gardner (eds)

Statistics with Confidence (2nd ed) BMJ (2000) Simon Day

Joseph L Gastwirth (ed) Statistical Science in the Courtroom Springer (2000) Carla RossiGeert Verbeke & Geert Molenberghs Linear Mixed Models for Longitudinal Data Springer (2000) Harry SouthworthMitchell H Gail & Jacques Benichou Encyclopedia of Epidemiologic Methods John Wiley (2000) Theo StijnenDavid W Hosmer & Stanley Lemeshow Applied Logistic Regression (2nd ed) John Wiley (2000) Jørgen SeldrupAndrew B Lawson Statistical Methods in Spatial Epidemiology John Wiley (2001) Maciej GorkiewiczSimon Day Dictionary for Clinical Trials John Wiley (1999) Rolf HolleAlex J Sutton, Keith R Abrams, David R Jones,Trevor A Sheldon & Fujian Song

Methods for Meta-Analysis in MedicalResearch

John Wiley (2000) Tim Friede

Reviewers wanted:Author(s) Title Publisher (year) Reviewer

JK Lindsey Nonlinear Models in Medical Statistics Oxford (2001)DM Titterington & DR Cox Biometrika: 100 Years Oxford (2001)Stephen W Duffy, Catherine Hill & JacquesEsteve

Quantitative Methods for the Evaluation ofCancer Screening

Arnold (2001)

Fotunato Pesarin Multivariate Permutation Tests John Wiley (2001)

Yes! The last 4 books are new and are available for review. Deadline for requests: 31 January 2002.

7^^Z _dQ[XbWTabn fTQ_PVTb4

Arnold: http://www.arnoldpublishers.comChapman & Hall: http://www.crcpress.com/www/chaphall.htm#msJohn Wiley: http://catalog.wiley.com/index.cgi?Marcel Dekker: http://www.dekker.com/catalog/catalog_top.htmOxford: http://www4.oup.co.uk/Springer: http://www.springer.de/statistic/books/newbooks.htmlBlackwell http://www.medirect.com/~cgilib/editorial.asp?subj=genm

Important note to potential reviewers:We regularly receive books from publishers for review in the Newsletter. We are most grateful for these�alk^qflkp�, qeb obsfbtp lc tefè tb obdôa ^p ^ pbosf`b ql vlr, lro jbj_bop. Qbdobqcriiv, pljb fkafsfarîp, despite repeated reminders, neither return a review, nor the book to ISCB... When requesting a book, pleaseobjbj_bo qe^q vlr�ob j^hfkd ^ `ljjfqjbkq ql qeb Rl`fbqv ql al ^ ifqqib tloh fk obqrok clo hbbmfkd qeb _llh. Please do a little work in return for keeping the book and your name will be published in the News!For the format and length, please see recent issues of ISCB News. You can send the review in a variety of formatsbut plain text e-mail, html, RTF or Word are preferred. The reviews may be edited for clarity (English grammar andspelling, punctuation etc.).


Book Review by Marc Saez (Spain)

Meta-Analysis in Medicine and Health Policy, By Stangl, DK and Berry, DA (eds), Marcel Dekker, New York, 2000.

This book can be considered a referencetextbook on meta-analysis. The book istofqqbk _v 32 `lkqof_rqlop clo �^mmifba statisticians, students of statistics andbiostatistics, and others who usestatistical methods in their professionalifcb�. Hk b^è è^mqbo, qeb ^rqelop mobpbkq a common problem, developmethodology to address the problem,and follow up with one or moreexamples. Each chapter starts from firstprinciples and proceeds to state-of-the-art techniques.Hk qeb cfopq è^mqbo, �Lbq^-analysis: past^ka mobpbkq èîibkdbp�, Stangl, DK andBerry, DA review the progression ofstatistical methodology for meta-analysis,highlight complexities encountered,summarise how the chapters of the bookaddress heterogeneity between studiesand suggest future directions formethodological development. The bookthen presents a further fourteenunordered (and this is its main drawback)chapters.Three chapters, 3, 8 and 14, and, to alesser extent chapter 10, provideillustrations and examples of meta-analyses.Brophy, J and Joseph, L fk �@ A^vbpf^k meta-analysis of randomised mega-trialsfor the choice for thrombolytic agents inacute myocardial ikcb`qflk� (è^mqbo 4), use Bayesian meta-analytic models,including bias adjustments for possibledifferences between the trials, toevaluate the evidence from three largerandomised trials which directly comparestreptokinase and tissue-plasminogenactivator in acute myocardial infarction.The authors conclude that the clinicalsuperiority of tissue-plasminogenactivator over streptokinase remainsuncertain.Hk �@ A^vbpf^k jbq^-analysis of therelationship between duration ofoestrogen exposure and occurrence ofbkaljbqofî `^k`bo� (è^mqbo 8), Larose,DT presents a Bayesian random-effectsmodel for combining exposure-responseinformation from several studiesexamining possible association betweenoestrogen exposure and endometrialcancer. Results indicate evidence that asignificant exposure-responserelationship exists.Stoto, MA fk �Qbpbôè pvkqebpfp clo public health policy: experience of thefkpqfqrqb lc jbaf`fkb� (è^mqbo 14), illustrates the meta-analysis approachthat the Institute of Medicine of theNational Academy of Sciences has takenin order to help clarify politically chargedpolicy issues. He focuses on statisticalaspects of the analysis, and to discussthe potential applicability of the approachin other settings.Also applied, Rahman, N andWakefield, J fk �Lbq^-analysis ofmlmri^qflk meôj^`lhfkbqf` a^q^� (chapter 10), describe a model which theauthors have recently developed to carryout population pharmacokinetic meta-analysis, and illustrate its applicationusing data from four phase I studies. Theauthors emphasise the assessment ofmodel adequacy, since, they argue, it isan area that is infrequently addressed in

hierarchical modelling, and the Bayesianapproach is ideally suited to theconsideration of informative diagnostics.The rest of the chapters, with theexception of chapter 15, are mainlymethodological although they areillustrated in all cases.Abrams, KR, Lambert, PC, Shaw, C andSansó, B fk �Lbq^-analysis ofheterogeneously reported study results: aA^vbpf^k ^mmol^è� (èapter 2), considerthe quantitative synthesis of publishedcomparative-studies when the outcomemeasures used in the individual studiesand the way in which they are reportedvaries between studies. While the formerdifficulty may be overcome, at least to alimited extent, by the use of standardisedeffect sizes, the latter is often moreproblematic. Two potential solutions tothis problem are sensitivity analyses anda fully Bayesian approach, in whichpertinent background information isincluded. Although the authors describeand illustrate both, they prefer the latter.Additionally they provide BUGS code forthe full Bayesian analysis.Hk �Lbq^-analysis versus large trials:obplisfkd qeb `lkqolsbopv� (è^mqbo 3), Berry, SM argues that large randomisedtrials are not superior to meta-analysis.He points out that they disagree onlybecause a key variance is ignored. Hepresents hierarchical meta-analysismodels which capture the heterogeneity inthe treatment effect across trials. Bymeans of two examples, he shows thatthe hierarchical approach resolves � or atleast sheds light on � the meta-analysiscontroversy.Hk �Blj_fkfkd pqrafbp tfqe `lkqfkrlrp and dichotomous responses: a latent-sôf^_ibp ^mmol^è� (è^mqbo 5), _v Dominici, F and Parmigiani, G analysethe problem in meta-analysis of combiningstudies in which similar medical outcomesare captured in some studies ascontinuous variables and in others asbinary variables. The authors propose astrategy that overcomes both of thedifficulties on using the standardapproach, i.e. to dichotomise thecontinuous responses and proceed as inqeb pfjmibo _fkôv `^pb. Seb ^rqelop� approach is based on assuming that thebinary responses are the result ofdichotomising some underlyingunobserved continuous variable.Bayesian reconstruction of theunobserved continuous variablepreserves the full information from thestudies reporting continuous variables anddoes not require the choice of arbitrarycut-off points.DuMouchel, W and Norman, SL in�Bljmrqbo-modelling and graphicalstrategies for meta-^kîvpfp� (è^mqbo 6), describe exploratory graphical methodsand present a unified modelling approachto meta-analysis, one that integratesfixed-, random- and mixed-effect models,as well as Bayesian hierarchical models,into a single framework. Their approachplaces emphasis on model selection,estimation, fitting, diagnostics, andinterpretation of results.Hk �Lbq^-analysis for 2x2 tables withjriqfmib qob^qjbkq dolrmp� (è^mqbo 7), _v

Gleser, LJ and Olkin, I the authors proposea regression procedure that yields anestimate of the overall effect in large studieswhen more than one competing treatmentmay be tested for its effectiveness versus acontrol and, further, different sites may usedifferent subsets of treatments. This leadsto a model with missing data and withcorrelated effect sizes.Hk �Llabiifkd ^ka fjmibjbkq^qflk fpprbp fk Bayesian meta-^kîvpfp� (è^mqbo 9), Pauler, DK and Wakefield, J outline ageneral hierarchical model for meta-analysis, introduce a new samplingalgorithm for estimation of its parameters,and provide diagnostics for examination ofdistributional assumptions. The authorsderive a simple rejection algorithm whichgenerates independent samples from theposterior distribution under arbitrary priors,and can be easily programmed in popularstatistical packages.Hk �Lbq^-analysis of individual-patientsurvival data using random-bccb`q jlabip� (chapter 11), Sargent, DJ, Zee, BC, Milan,C, Torri, V and Francini, G present a classof models that allow for random-effectanalyses of individual-patient survival data.The random-effect survival models that theauthors propose allow for between-trialheterogeneity in treatment effects and/orbaseline hazard functions, and in addition,allow for multilevel modelling of thetreatment effects. The models are fitted in aBayesian setting using Markov chain MonteCarlo methods.Smith, DD, Givens GH and Tweedie, RL in�@agrpqjbkq clo mr_if`^qflk _f^p ^ka nrîfqv bias in Bayesian meta-^kîvpfp� (è^mqbo 12), within a Bayesian hierarchical modelallowing stratification on quality, develop adata-augmentation technique to estimateand adjust for the numbers and outcomes ofstudies that may be missing in eachstratum. The authors point out that thispermits inferences to account for potentialpublication bias.Hk �Lbq^-analysis of clinical trials:lmmloqrkfqfbp ^ka ifjfq^qflkp� (è^mqbo 13), Simon, R reviews the important basics forconducting meta-analyses of clinical trials.He tries to describe frequent objectives ofmeta-analyses, why meta-analyses areimportant, and why they are frequentlyflawed and unreliable. He points outimportant methodological features thatshould be observed in meta-analyses andcomments on the importance of goodclinical trials to achieve the objectives ofmeta-analyses.Finally, Sutton, AJ, Lambert, PC, Abrams,KR, Jones, DR and Hellmich, M fk �Lbq^-analysis in practice: a critical review of^s^fi^_ib plcqtôb� (è^mqbo 15), qov ql inform the researcher of the softwarepackages and routines currently available,and describe the capabilities of each. Theyalso provide a critical overview of thechapter, highlighting some meta-analysismethods for which no software is presentlyknown to be available.Summing up, and citing the series editor,Chow, SC �qeb slirjb klq lkiv fkqolar`bp important statistical concepts, designs, andmethodologies of meta-analysis but alsoprovides applications in clinical researchqeolrde mo^`qf`î bu^jmibp�.


Book Review by Dario Gregori (Italy)

Analysis of health surveys by Edward L. Korn and Barry I. Graubard, John Wiley (1999)

This book is one of the few available sources to findorganized and systematic ideas and methods aboutlarge field surveys in health research. From the pointof view of the statistical methods, the subject iswidely explored and many textbooks already existdealing with survey methodology, but no connectionsusually exists with real and specific problems arisingin health research.

The book is organised in 9 chapters, plus 3appendices; the first one illustrating the surveys usedin the examples in the text and the other two givingdetails about the mathematics behind linearisation forimplicit functions of weighted sums and cubic splineinterpolation.

Chapter 2, 4 and 5 illustrate the basic surveymethodologies, covering the material about studydesign up to multi-stage sampling plans. Estimationis discussed deeply with respect to both weightedand un-weighted methods. Variance estimation is firstdiscussed in the basic methodology. Then, moreadvanced issues are presented like Jack-knife andbootstrap estimators, and super-population modelinference. In addition, the basic details about non-response and missing value treatment are presented.

The subsequent chapters are dealing with the design andthe analysis of some designs which are specific to thehealth field research. In particular, Chapter 6 describesthe cross-sectional analysis of a classical sampling plan,whereas Chapter 7 is discussing some issues inlongitudinal studies, both from the point of view of paneldata analysis, and from the point of view of transitionalmodel estimation. Actually, this is one of the fewsystematic descriptions of the transitional models, startingfrom the basic issues of design up to highlighting the mostrecent techniques of analysis of such data. Chapter 9includes a very useful description of the design of largescale, population based case-control studies. This is againa quite uncommon perspective for the clinician, that couldbenefit from reading this chapter in getting more criticismabout what he/she is currently doing in his/her research.

Finally, the weighting problem is again highlighted in thediscussion about the combination of multiple surveys,which is described in Chapter 8, and illustrated withseveral real-life examples.

The book is a very nice and valuable source of informationfor several people. It can be used as a reference manualfor people working in small agencies and registries, but itcan be also adopted as a textbook for post-graduatestudents in public health, epidemiology and medicalstatistics.

Book Review by Victor Moreno (Spain)

Statistical Regression with Measurement Error, by Cheng CL, Van Ness JW, Arnold, London (1999)

This book, the 6thth GL� RFC� QCPGCQ� / CLB?JJbs Library ofStatistics covers in depth the statistical theory ofregression models with measurement error. It is not anintroductory book. It has a high level and requires a fairamount of previous knowledge of probability andtheoretical statistics. The book is divided into eightchapters and one appendix. The introductory chapterpresents the general linear model with measurementerror & ( &

( are not directly observable but measured with additiveerrors: x = (

the assumptions about (

considered with varying degree of complexity. First, thefunctional model, where (i are unknown constants.Second, the structural models, where (i are independentidentically distributed random variables and

) . And third, the ultrastructural models, where (i arareindependent random variables, but not identicallydistributed, having possibly different means andcommon variance. The rest of the chapter deals with theproperties of these models, identifiability of theparameters and methods of estimation paying specialattention to the three possible structures of the models.

Chapter 2 is very theoretical, covering the properties ofthe estimates and predictions derived from thesemodels. Chapter 3 and 4 introduce other estimationmethods: generalised least squares, modified leastsquares and the use of instrumental variables. Inchapter, 5 models with more than one covariatemeasured with error are introduced. Chapter 6 coversthe difficulty of dealing with non-linear models withmeasurement error. Emphasis is done on the specialcase of one covariate showing a non-linear relationshipthat can be modelled with polynomials. The finalsubstantive chapter is dedicated to robust methods ofestimation to avoid bias when data have outliers or donot satisfy the assumptions used to derive theestimation procedures. The methods proposed aresimilar to those used in ordinary regression adapted tohandle the added problem of measurement error andparameter identifiability. The book finishes with achapter of miscellaneous topics and an appendixdealing with the subject of identification of parametersin linear measurement error models. Each chapter has alist of unsolved exercises.

In summary, I found this book very theoreticallyoriented with little interest for those looking a morepractical approach about how to deal with real datameasured with error.


Book Review by Bob Gibberd (Australia)

Statistical Process Monitoring and Optimization, by Sung H Park and G Geoffrey Vining. Eds. Marcel Dekker (1999)

The editors of this book have provided anopportunity for 25 researchers to documentsome of the advanced statistical methods forprocess monitoring and optimization used toimprove quality and productivity. The areascovered are control charts, multivariateprocess monitoring, process capability index,design of experiments and empirical modelsfor process optimization. Six authors alsocover the statistical approach to qualitymanagement, with the first being an overviewby Genichi Taguchi, one of the pioneerscontributing to the Japanese success story.

The advances in these areas have primarilybeen developed by or for engineers andstatisticians working in industry. Hence, thebook is primarily written for engineers andindustrial statisticians, and provides a usefuloverview of current issues.

The uses of these industrial statisticaltechniques have drifted to the serviceindustries (tourism, education and health) inrecent years. In particular, there has been anincreased interest in using these techniquesin improving the quality of health care inhospitals, since recent studies have shownthat about ten percent of patients suffer anadverse event due to the health careprovided to them. Generally, thebiostatisticians who are involved in theclinical quality improvement studies inhospitals are not overly familiar with thesetechniques, and are often rediscovering thesame methods developed earlier byengineers. This is particularly the case whenusing control charts to monitor hospitalprocesses.

Can this book also be used by clinicalbiostatisticians to improve their knowledge?Overall, the articles assume that the readeris quite familiar with the topics listed above.Elo bu^jmib, qeb è^mqbo lk �Rljb ob`bkq developments in control charts for monitoring^ molmloqflk� tluld be useful in hospitals formonitoring mortality or infection rates oroccurrences of rare events such aspulmonary embolism resulting from electivesurgery. The chapter assumes that thereader is familiar with the p and np controlcharts and the approach used for CUSUMS.What it does provide, successfully, is moreadvanced techniques to detect a shift in asmall proportion: methods that have beenrecently published in the Journal of QualityTechnology, probably not a standardreference journal for biostatisticians. Onreading this chapter, it is likely that thesenewer techniques will have applications forhealth.

The same comments can be made for thesubsequent chapter (e.g. process monitoringwith auto correlated data), which also haspotential applications to using control chartsin hospitals.

The uses of quality principles and tools aremore advanced in the pharmaceuticalindustry (monitoring the production of drugs)and hence this text would be extremelyrelevant, even though the 25 authors onlyprovide one example from this industry.

This is a book most biostatisticians would notneed on their bookshelves, but is number160 fk qeb pbofbp �Rq^qfpqf`p; Sbuq_llhp ^ka Llkldo^mep� ^ka ebk`b tlria _b lkb clo University libraries to purchase.


Book Review by Duolao Wang (UK)

Multilevel Modelling of Health Statistics, by A H Leyland & H Goldstein Eds., John Wiley (2001)

This very useful book gathers together 13 articles bysome well-known figures working on theoretical andpractical aspects of multilevel modelling, with anemphasis on application of the methodology to datasets relating to the health of individuals orpopulations. Most of are adapted from journals orbooks published previously, but a few were writtenspecially for this volume.

The book begins with an introductory article by Healyon the components of a multilevel model, explainingwhy such a model is useful and defining the variousconceptions, introducing both variance componentsand random coefficient models and providing anexample using data on children's sleep patterns.

In Chapter 2, Goldstein and Woodhouse deal withthe modelling of repeated measures or longitudinaldata. A basic two-level growth curve is fitted to theheights of a sample of boys. The chapter thenproceeds to consider different ways in whichautocorrelation between measurement occasions canbe modelled and concludes with a discussion ofvarious extensions to repeated measures designs.

The following two chapters are on generalised linearmultilevel models where the responses are notcontinuous. In Chapter 3, Rice focuses on multilevelbinomial response regression. After introducing aseries of examples in the modelling of healthstatistics, a linear multilevel model for binomial datais developed and its limitations are detailed. Ricethen moves to a discussion of extra-binomialvariation and model diagnostics and presents anexample concerning equity in the provision of healthcare services.

In Chapter 4, Langford and Day also considerdiscrete responses in their chapter on multilevelPoisson regression. They detail two examples: theanalysis of deaths from testicular cancer amongmales throughout Europe in the 1970s and theweekly incidence of food poisoning in England andWales in 1989-90.

McLeod shows in Chapter 5 how the standardmultivariate regression model in which the responsevariable is a vector can be extended to a multivariatemultilevel model using examples relating to the bloodpressure of low-birth weight children and therelationship between a patient' length of stay andprobability of readmission.

Procedures for carrying out diagnostics in multilevelmodels are covered by Lewis and Langford inChapter 6. They present methods for detecting outlierinfluential data points and leverage points that affectthe precision of the estimates and explain how thesequantities are calculated at each level and presentexamples of their use relating to mortality fromprostate cancer in Europe.

Chapter 7 by Rasbash and Browne illustrates hownon-hierarchical structures that can be manipulatedto be multilevel models. Two important non-hierarchical structures, cross-classifications andmultiple membership, are the focus of chapter andillustrated with examples.

Yang introduces in Chapter 8 multilevel models formultiple categorical responses with multinomial errordistributions for both nominal and ordinal data. Sheuses two examples from public health to illustratethese models, their underlying assumptions and theinterpretation of parameter estimates.

In Chapter 9, Marshall and Spiegelhalter describehow institutional performance within health servicescan be compared using multilevel models. They lookat so-called 'league tables' and their statisticallimitations, and demonstrate the methodology ondata relation to surgical performance in New York.

Kbvi^ka�p Be^mqbo 10 bumilobp elt jriqfibsbi jlabip can be used for the analysis of spatially organiseddata, and presents models for the case where thereis spatial autocorrelation. He shows how suchmodels can be expanded to include multivariateresponses as well as longitudinal data. Leylandillustrates the model with an example using Scottishlip cancer data.

Chapter 11 by Snijders deals with the problem ofsampling in a multilevel context. Snijders showsmathematically how the study design is influenced bythe level at which the explanatory variables areobserved, and discusses optimal sample sizes inmultilevel models.

In Chapter 12, Goldstein and Leyland discuss furtherissues such as meta-analysis, the modelling ofsurvival data, and contextual and compositionaleffects. The chapter finishes with brief sources ofreferences for readers interested in measurementerrors, structural equation modelling and missingdata in multilevel analysis.

Finally, in Chapter 13, de Leeuw and Kreft reviewsome of the important packages most commonlyused for fitting multilevel models. They compare indetail the availability, documentation, interfaces andalgorithms for MLwin, HLM, VARCL, MIXFOO, MLA,BMDP5-V and PROC MIXED.

Each chapter contains an introductory note that givesthe basic background information about the subjectmatter, followed by methodology and numericalexamples. Recommended further readings for eachsection together with references are provided.

In summary, this book is to be highly recommendedto biostatisticians, health care professionals andpublic health researchers in the application ofmultilevel model. It can also be used as a referencebook for postgraduate students studying medicalstatistics.


Book Review by Tiberiu Postelnicu (Romania)

Analysis of Multivariate Survival Data, by Philip Hougaard, Springer (2000)

This book is aimed at investigators who need toanalyse multivariate survival data and can beused as a textbook for a graduate course inmultivariate survival data. It is written from anapplied point of view and covers all the essentialaspects of applying multivariate survival models.More theoretical evaluations, such as asymptotictheory, are also described, but only to the extentuseful in applications and for understanding themodels. To read the book, it is useful, but notnecessary, to have an understanding ofunivariate survival data.

Survival data or more general time-to-event dataoccur in many areas, including medicine, biology,engineering, economics, and demography, butpreviously standard methods have required thatall time variables are univariate and independent.This book extends the field by allowing formultivariate times. Applications where such dataappear are survival of twins, survival of marriedcouples and families, time to failure of right andleft kidneys for diabetic patients, life history datawith time to outbreak of disease, complicationsand death, recurrent episodes of diseases, andcrossover studies with time responses.

As the field is rather new, the concepts and thepossible types of data are described in detail,and basic aspects of how dependence canappear in such data are discussed. Four differentapproaches to the analysis of such data arepresented. The multistate model where a lifehistory is described as the subject moving fromstate to state, is the most classical approach. TheMarkov model makes up an important specialcase, but it also described how easily moregeneral models are set up and analysed. Frailtymodels, which are random effects models forsurvival data, make a second approach; theyextend from the most simple shared frailtymodels, which are considered in detail, to modelswith more complicated dependence structuresover individuals or over time. Marginal modellinghas become a popular approach to evaluate theeffect of explanatory factors in the presence ofdependence, but without having specified astatistical model for the dependence. Finally, thecompletely nonparametric approach to bivariatecensored survival data is described.

The book starts with an Introduction to survivaldata, and to the various types of data to be

considered as multivariate survival data and alsolists a number of examples. Some time is spenton clarifying the concepts, which is particularlyimportant as the field is rather new.

Standard methods for Univariate Survival Dataare described in Chapter 2. The variousmechanisms that create dependence inmultivariate survival data are discussed inBe^mqbo 3 bkqfqiba �Cbmbkabk`b Rqor`qrobp�. Model-independent measures of dependence forbivariate data are described in Chapter 4,�Afsôf^qb Cbmbkabk`b Lb^probp�. Hk Be^mqbo 5, �Ool_^_fifqv @pmb`qp lc Lriqf-Rq^qb Llabip�, qeb most classical way of analysing life history data,are depòf_ba. Seb �Rq^qfpqf`î Hkcbobk`b clo Lriqf-Rq^qb Llabip� fp dfsbk fk Be^mqbo 6. �Reôba Eo^fiqv Llabip� ôb o^kalj bccb`qp jlabip< survival data models similar in concepts to thenormal distribution one-way variancecomponents models, and described in Chapter 7.Statistical inference is described in Chapter 8.Special aspects for the analysis of recurrentevents are described in Chapter 9. Extensions ofthe shared frailty model are described inChapters 10 and 11. Competing risks models,which are models for data on cause of death, areconsidered as a special case in Chapter 12.Regression models emphasizing the marginaldistributions, allowing for dependence asdescribed by some nuisance parameters, aredescribed in Chapter 13, together with the copulaapproach, modelling the dependence for fixedmarginals. The most general bivariate non-parametric methods are described in Chapter 14.Finally, Chapter 15 gives a Summary of thetheory and an overview of the most relevantmethod for each of the major applicationsdiscussed.

All chapters start with an introduction, which isless technical than the rest of the chapter andends with a summary of what has been found. Itis a new important contribution in the series�Rq^qfpqf`p clo Aflildv ^ka Gbîqe�. Seb cfbia ofapplication is the biological and medical field, butthe theory is also applicable to technicalreliability, demography, actuarial science andother fields, where multivariate times areobserved.


Stockholm Review 1 m Nancy Geller

The ExCom, following its annual meeting

Stephen Evans, Emmanuel Lesaffre, Siem Heisterkamp, Elia Biganzoli, Julia Singer, John Whitehead, Harbajan Chadha-Boreham, Simon Day, Maria Grazia Valsecchi, Carol Redmond, David W. Warne, Elisabeth Svensson

(L. to R.) Stephen Senn, David Warne, John Whiteheadbkglvfkd qebjpbisbp �Ml_iv� ^q qeb Qb`bmqflk ^q Bfqv Gîi

Astra-Zeneca Chorus performing at the Vasa Museum

Ed Gehan, Peter Armitage, Nancy Geller (all ISCB formerpresidents)

President Simon Day presenting champagne to Juni Palmgren

(L. to R.) Elisabeth Svensson, Nancy Geller, Harbajan Chadha-Boreham at dinner at the Vasa museum


Stockholm Review 2 m David W. Warne

Conference Site: Nobel Hall 1:

Vasa: Nobel Hall 2:


Stockholm Review 3 [ David W. Warne

City Hall 3:

City Hall 1: City Hall 2:

Gamla Stan from City Hall 1: Gamla Stan from City Hall 2:


Stockholm Review 4 [ David W. Warne

Skansen 1: Skansen 2:

Skansen 3:

Götha Canal 1: Götha Canal 2:


Book Review by Rainer Muche (Germany)

An Introduction to Medical Statistics, by Martin Bland, 3rd edition, Oxford (2000)

The book of Martin Bland is the third edition ofhis book about the introduction to medicalstatistics. Anticipating, for readers, who knowthe first and/or second edition, this book givessome new topics as the concept of numberneeded to treat, design and analysis of cluster-randomised trials, consent in clinical trials,ecological studies, conditional probability,repeated testing, random effects models,intraclass correlation, and conditional oddsratios. So, this edition can be a useful tool forreaders of the fist two editions.

The book covers an up-to-date description of themethods used in analysing medical data,especially for students, medical researchers,and others concerned with medical dataanalysis. It focuses on practical applications andnot on mathematical theory and proofs. Forunderstanding of the formulas no reallyfundamentals of mathematics are required. Theformulas are accompanied by well presentedexamples, so everyone can follow thecalculations. Beyond this, some of theunderlying mathematics are presented inappendices. Exercises are given at the end ofeach chapter with annotated solutions at the endof the book.

After a short introduction the books begins with^ è^mqbo `îiba �qeb abpfdk lc bumbofjbkqp�. H prefer to begin introductory texts in medicalstatistics to do so, because knowing thefundamentals and pitfalls of studies will preservedoing poorly research. In this chapter theemphasis is laid on allocation of patients, somebiases, and blinding. A new chapter stresses theconsent in clinical trials. Some comments inbehalf of realisation of clinical studies (studyprotocol, ICH-GCP, ethical problems) aremissing. Inserting them would rise the value forthe practice of medical research.

The coverage of the book reaches fromdesigning of clinical and observational studies toanalysis of variance. In the chapter aboutdescriptive analysis I like to see, that theintroduction of the median and box-plot comesfirst before introducing the mean and variance.But this focus on methods based on percentilesor ranks is not carried out through the book.Confidence intervals, significance tests, andcorrelation were given using normal probabilitymethods. After this in chapter 12 methods basedon ranks are introduced. Martin Bland discusses

the use of parametric or nonparametric methodsfk è^mqbo 12.7 (�mô^jbqof` lo klk-parametricjbqelap>�) ^ka fk è^mqbo 14 (�Bellpfkd qeb pq^qfpqf`î jbqela�). Rl, qeb obâbo `^k ab`fab by their own, what he prefer in analysing thedata.

Another chapter covers introduction to somemultifactorial methods (e.g. multiple-, logistic-and Cox-regression). The methods are welldescribed, but on 2-3 pages no one can presentall aspects. Some references would helpinterested readers to go further in this topics,e.g. Hosmer-Lemeshow for logistic regressionanalysis.

One important chapter is chapter 5, called�mobpbkqfkd a^q^�. Seb mobsentation of data is avery useful tool in analysing and interpretingmedical data. Bland gives in his application-oriented textbook many hints to do so.Dpmb`fîiv qeb è^mqbo 5.8 �Lfpibâfkd do^mep� fp worth reading.

Some relative new methods in medical statisticsare equivalence tests as special significancetests and the determination of sample sizes inthis field. The need for such methods isincreasing because in many clinical fields newtherapies cannot show superiority againststandard therapies but equal effects in theprimary outcome variable and superiority in thesecondary ones. The methods are now wellestablished but completely missing in thistextbook.

@klqebo ôb^ lc mo^`qf`î obibs^k`b te^q�p missing are hints to statistical software. MartinBland gives in the introduction a downloadaddress of his little MSDOS program CLINSTATbut I think that in an up to date textbook crossreferences to modern, windows based softwareis needed knowing the problem of beingantiquated soon.

Some chapters are marked by the author asbeyond the need of many students, which wouldnot usually be included in first courses. Myfeeling is ambivalent: on the one hand you getmore as the title of the book promises but on theother hand readers, who want a first introductionin the field can be confused by this.

Overall, I think that this book can berecommended as a good introductory statisticaltextbook for students, postgraduate studentsand medical researchers.


Book Review by Matthias Wahl (Germany)

Modeling Survival Data � Extending the Cox Model, by TM Therneau, PM Grambsch, Springer (2000)

The first sentence of the prefacesays it all: "This is a book forstatistical practitioners whoanalyse survival and event historydata and would like to extend theirstatistical toolkit beyond theKaplan-Meier estimator, log-ranktest and Cox regression model totake advantage of recentdevelopments in data analysismethods motivated by countingprocess and martingale theory."Additionally one can say it is alsofor those who do have thetheoretical knowledge already butwant to know how to implement itin S-Plus or SAS because nearlyevery example in this book isillustrated with the necessary codefor S-Plus, SAS or both of them. Ifthe code given during the regularten chapters of this book is notenough the interested reader willfind another five appendices filledwith source code.

The first three chapters give anintroduction to the Cox model, theNelson-Aalen and Kaplan-Meierestimator and offer a first view intocounting processes and how theywill be used later in the book. Thereader is presented with the basicsnecessary for the later chaptersbut for somebody who has noprevious knowledge of the Coxmodel this presentation of theory,about which other authors havewritten whole books, is probablytoo compact. These three chaptersare more for refreshing yourknowledge and linking it to thecounting processes. This "countingprocess representation" of the Coxmodel will be needed in the laterchapters. Chapter four deals withresiduals and which role thedifferent residuals have in theprocess of model validation. Thefollowing three chapters givedetailed examples for the use ofthe residuals introduced in thischapter but a few examples of howresiduals could be used and howthey should better not be used arealready given here. The fifthchapter explains how the rightfunctional form of a covariate canbe found. It is discussed howsmoothed plots of the martingaleresiduals can be used to detect

the appropriate functional form ofa covariate. In a more theoreticalpart, the necessary assumptionsare explained. As an alternative tosmoothed residual plots thepoisson approach and splinemethods, of which the smoothingsplines are more useful than theregression splines, are introduced.Again, detailed computer code isgiven to show how easy theseanalyses can be calculated. Thesixth chapter is devoted to the testof the proportional hazardassumption of the Cox model.Graphical and formal test methodsare introduced and explained.Apart from showing how to detectnon-proportionality, it is alsoshown what can be done if onehas a non-proportional model andwhich causes there are for non-proportionality. This chapter showsalso that there is still work to bedone on this topic because thereare limitations to both formal testsand graphical methods. At onepoint, the authors suggest thatthere are situations where theproportional hazard model workseven in the case of non-proportionality. The last applianceof the residuals is described inchapter seven where it is shownhow the so-called dfbeta residualscan be used to visualise theinfluence of individualobservations. Another use of theseresiduals is in the construction of arobust estimator for the covariancematrix of the estimated regressioncoefficients. That estimator isneeded in chapter eight, wheremethods for multiple events persubject are described. The chapterfocuses on three common variantsof marginal models, the Andersen-Gill, Wei-Lin-Weissfeld and thePrentice-Williams-Petersen (a.k.a.conditional approach) approach.All approaches have in commonthat they are based on three steps:Decide on a model, fit an ordinaryCox model, and replace thestandard variance estimate withone corrected for possiblecorrelations. There are a lot ofexamples given and they are veryhelpful in understanding thedifferences between the threemodels because it is shown how

the same data have to be modifiedfor each model and how theresults differ (or not). Following thevery illustrative way it is shownhow these models could also beused for multistate models andhow it is even possible to constructcombination models that havecombined properties of differentapproaches. This chapter showsthe tremendous flexibility that themarginal survival models offer.Nevertheless, it also shows thatthese models are still imperfectand that here, too, is muchresearch still to be done. Chapternine shows how models withrandom effects (such called frailtymodels) could be fitted with S-Plus. Again, a number ofexamples are given to show inhow many different fields ofresearch these models could beapplied and again it is stated thatresearch on this area is ongoing.The last chapter is devoted to thecalculation of expected survivaland how this tool can be linked tothe Cox model.

I have already used the countingprocess approach to the Coxmodel in 1999, although just forexploratory purposes. At that time,I had to download differenttechnical reports from TMTherneau's webpage at the Mayoclinic. Therefore, I am now veryhappy to have it all in one book.The idea to use a countingprocess approach to survivalanalysis and thus generalising it toanalysis of any event data isoutlined very clearly and illustratedby a lot of examples. Therefore,this book is recommended toeveryone who deals with any kindof event data and needs someadvise of how to model andanalyse them. It is also a valuableguide on how to do survival/eventanalysis with S-Plus (and to somelimited degree with SAS).Nevertheless, it should bementioned that the countingprocess approach is not the onlyone and that there are other bookswith focus on other approaches aswell.


Workshop on Regulatory Issues in Medical Decision Support

From Elia Biganzoli

19 October 2001, University College London

This one-day workshop was jointlyorganised by Prof. Paulo Lisboa, JohnMoores University, Liverpool, Dr RichardDybowski, Dybowski Associates, Pinner,and Dr Jeremy Wyatt, School of PublicPolicy, UCL, London, on behalf of the AIME(Association of Institutions concerned withMedical Engineering) and the BritishMedical Informatics Society. The workshopwas part of the IPEM (Institute of Physicsand Engineering in Medicine) Meetings. Sixinvited speakers with different backgroundexperiences developed a multidisciplinaryprogramme for twenty attending delegates.

The aims of the workshop are summarisedin the announcement, "Europe and theUSA have rapidly growing markets formedical equipment, whose operationincreasingly relies on complex software.The explosion of networking anddatabases, coupled with user demands forbetter quality, personalisation and remotecare, all point to a fast commercialdevelopment of healthcare informatics.However, there are many unresolvedissues surrounding the application of theMedical Devices Directives, especially inMedical Decision Support (MDS). Thisworkshop brings together the maincomplementary perspectives that impactdirectly on regulatory procedures,introduced by speakers who arerecognised authorities in the design,evaluation and practical use of statistical,rule-based and neural network systems.The purpose of the workshop is to identifyand discuss key issues, so as to providepointers to effective and practical ways toresolve them."

The workshop was opened by Prof. Lisboa,who addressed the issue of the qualityassurance of MDS software withalgorithmic components. The workshopprogramme consisted of two main parts. Inthe morning session the focus was onregulatory and practical aspects. The firstspeaker, Mr. P. Stonebrook, MedicalDevices Agency, London, gave a lectureentitled The Medical Devices Directives:Application of the directives to medicaldecision support systems (scope andconformity assessment). He presented theEuropean Directives (90/385/EEC,93/42/EEC, 98/79/ECC) [1] and theirimplementation in the UK with reference tothe software providing for proper medicaldevices, intended as diagnostic ortherapeutic tools. It should be noted thatsoftware for handling general patient-related data is a distinct entity that does notfall within the scope of the directives, sinceit has no direct medical purpose. Examplesof MDS software to which the requirementsof the directives are applicable are thealgorithm for implantable defibrillators,image enhancement for MRI applications,risk evaluation of trisomy 21, and signalprocessing for correlation of a diagnosticsignal with the presence of tissueabnormalities. The test for trisomy 21 isimportant because it may result in therecommendation of a furtheramniocentesis. This operation of this

particular test is of interest because a Y2Kerror caused an error in the calculation of thejlqebo�p ^db, tefè fk qrok afpormqba qeb operation of an algorithm to estimate the risklc qeb _^_v prccbofkd colj Cltk�p pvkaoljb. This incident has since been thoroughlyinvestigated and is reported in [2]. Amongthe essential requirements at the basis of thedirectives, the principle that the risksconnected to the application of the medicaldevice and the corresponding software areoutweighed by the benefits appears to befundamental. With reference to this principle,the aspects of major relevance tobiostatisticians are contained in Article 15(Clinical Investigation) and Annexes VIII andX (Clinical Evaluation) of 93/42/EEC [2],since they provide general indications for thedesign of clinical studies.

From the discussion it emerged that stand-alone MDS systems, i.e. those not involvinga specific medical device, have a separateregulatory framework which is yet to becovered by suitable directives. This woulddevelop a new regulatory area with criticalaspects related to the definition of clinicalinvestigation procedures for the assessmentof the risk/benefit ratio, possibly involvingrandomised controlled trials.

The second lecture, Key issues forcomputer-based decision support in clinicalpractice, was given by Dr Bipin Vadher,Consultant Haematologist, BromleyHospitals NHS Trust, Farnborough Hospital,Orpington. He presented the developmentand evaluation of a simple, fast and easilyimplemented MDS system for warfarindosing and monitoring as an aid to juniordoctors. The system has been validated inrandomised controlled trials showing itssafety and effectiveness [3]. However, DrVadher pointed out several study limits,including the adoption of surrogate outcomesfor some patients. Further development ofthe system is ongoing but in the meantimeother MDS systems have been proposed forthe same task [4], although they have not yetbeen adequately evaluated. In theconclusions a clinician's view of thedevelopment of regulation for MDS wasprovided regarding the description andevaluation of MDS, including the addedvalue of MDS with respect to currentpractice, its educational value, and itsintegration with other systems. Thediscussion dealt with the comparison ofretrospective and prospective studyapproaches for the assessment of MDS,noting a correspondence between the role ofretrospective studies for model optimisationin MDS, with phase I in drug development,and prospective studies for generalityestimation, with phase II. This approachwould follow a framework for thedevelopment and evaluation of RandomisedControlled Trials for complex interventionsoutlined in [5].

The last lecture of the morning session,Experience with longstanding decisionsupport and perspectives of newdevelopments in Europe, was given by DrSusan E. Clamp, Clinical InformationScience Unit, The University of Leeds. She

presented the experience of a MDS systemdeveloped in Leeds 30 years ago for thediagnosis of patients with acute abdominalpain (AAP). Treatment of AAP is still adifficult clinical problem, frequently involvingunnecessary hospitalisation and surgery. TheMDS system has been evaluated in severalclinical trials both in the UK and Europe. Thefeatures of the AAP MDS for inexperiencedclinicians were modelled along the lines ofthe evaluation studies [6]. The same clinicaloutcomes were measured in all studies andevery study showed improvements in theclinician's diagnostic accuracy and decision-making. However, this MDS system has notbeen widely applied in clinical practice forseveral reasons, including the generallyunfavourable attitude of clinicians towardsMDS tools, that provide a "black-box"presentation of the results. Lastly, Dr Clampaddressed the problem of the evaluation ofMDS and the lack of standard approaches tothis task, with respect to different outcometypes in terms of clinical issues and generalpatient care. The discussion dealt with theoptimal interaction between MDS andclinicians, and the particular difficulties thatarise in the design of clinical trials when thebenefit of MDS is mediated by humanjudgement. This so-called Hawthorne, orchecklist effect, needs to be taken intoaccount in the performance evaluation that isneeded to demonstrate either equivalence orimprovement of clinical practice by theadoption of MDS. A plea was made forpractical regulatory systems that are not soheavy as to thwart the development of usefulinnovations in MDS.

The afternoon session was dedicated to themethodological aspects underlying thedevelopment of MDS systems, providingperspectives both from biostatistical andinformation science viewpoints.

The first lecture was given by Prof. DougAltman, Institute of Health Sciences, Oxford,and was entitled Issues in the developmentof good statistical models for prognosis. Prof.Altman stressed the importance of the designof studies to develop prognostic classificationschemes, in particular with respect to sampleselection and size, the choice of relevantpredictor variables, the quality ofmeasurements, and missing data. Dataanalysis introduces further problems and thusthe issue of model selection was presented.Other important issues concerned thepractice of categorising continuous predictorsand the possible misuse of advanced, non-linear modelling techniques including artificialneural networks (ANNs). Bias should beanticipated and either avoided or correctedfor in analysis. In a further phase, modelvalidation is necessary to assess the qualityof average or individual predictions on newpatients [7], accounting for the differentconcepts of statistical and clinical validation.In the ensuing discussion it was pointed outthat proper handling of ANNs has led toimproved results in several applications, anissue that was taken-up by the last speaker,Dr Ian Nabney.


Workshop on Regulatory Issues in Medical Decision Support (continued)

The lecture of Professor John Fox, Head ofthe Advanced Computation Laboratory,Imperial Cancer Research Fund, London,entitled Clinical decision-making bymachine: maximising safety and limitingrisks, focused on the evidence thatcomputers may contribute significantly toimproving consistency and quality of care.Reviews of technologies and applicationsare available [6, 7]. Now that internet-basedMDS services are expanding, with theincreased use of and access to suchapplications and their growing complexitythere is a clear responsibility to demonstrateboth their efficacy and safety in practice.Prof. Fox presented several implementedMDS systems and discussed possibleproblems related to the need for quality self-regulation and/or external regulation. Withreference to the latter, validation, verificationof effectiveness by clinical trials andpossible certification procedures should beconsidered. The last part of the lecture dealtwith legal liability questions and approachesto addressing them. In particular, it waspointed out that it is not possible to avoidliability for death or personal injury causedby the use of a medical device, including theeffect of the software, whatever disclaimeraccompanies the device.

The final lecture, entitled Validation of neuralnetwork medical systems, was presented byDr Ian Nabney of the Cardionetics Instituteof Bioinformatics, Aston University [10]. Hediscussed the development and validation ofANN models as useful alternatives forproblems without algorithmic solutions. DrNabney stated that, although based oninductive principles, ANNs can suitablysubstitute algorithmic procedures, providedthat a proper statistical framework isadopted to control model complexity. Thekey issues to be considered in this respectinclude (i) design verification, (ii) uncertainty,(iii) risk assessment, and (iv) demonstrationof performance. These issues are strictlyrelated to those of predictive regressionmodelling, namely, data collection, pre-processing and checking the modelassumptions, parameter estimation, andassessment of generalisation performance(predictive ability).

With regard to model training, the speakerpaid particular attention to the minimisationof suitable error functions corresponding tothe noise distribution of the dependentvariables. In addition, he stressed thatmodels with a high curvature may have poorgeneralisation performance since they couldbe too complex and therefore overfitted; theuse of regularisation (penalised estimation)

is critical in limiting this risk. As for modeltesting, the need to detect automaticallywhether the test set is representative or notwas pointed out according to the conceptsof interpolation and extrapolation. Theadoption of Bayesian techniques for modeltraining and testing was introduced withspecific attention to the uncertainty of modelpredictions. Lastly, case studies werepresented in the framework of MDS for EEGand ECG analysis. Some aspects related toFDA approval were also introduced. Theconclusions stressed the need for a goodunderstanding of the basic properties ofANNs to allow their assessment; moreover,recent ANN theory in relation to advancedstatistical methods has been shown tosubstantially improve the robustness of ANNmodels.

At the end of the workshop there was ageneral discussion chaired by Dr JeremyWyatt, UCL, to lead on practical waysforward in the evaluation and certification ofsoftware for medical decision support. Hesummarised the issues that emerged fromthe different lectures; in particular, theprinciple that benefits should outweigh risksin MDS systems, which underpins theprocedures to be undertaken for theirevaluation.

ISCB Golf at ISCB Conferences

From Stephan Evans

Mats Lörstad, who was later President of ISCB in 1988, instituted what has become a traditional part of the ISCBconference in Cardiff in 1986. As far as we can tell, he, Jørgen Seldrup (also President in 1992-3), and StuartCummings played golf there at an unknown course.

There was no golf at the Gothenburg or Innsbruck Conferences, but in Maastricht in 1989, Mats, BernhardHuitfeldt, Stuart Pocock, Jørgen, Stuart C and Stephen Evans played at the Heerlen/Maastricht course namedHobru.

There was also golf at Nimes in 1990, with Mats and Stuart Pocock, among others. Mats played with a single clubwith adjustable loft (illegal by the formal rules of golf) but won. At Brussels (1991), Stuart Cummings organised anexcellent outing to a course with a chateau in the middle of it, and competitions were introduced with a long drivingcompetition - the lucky winner was Stephen E. Jørgen also played there, Bernhard, Mats and Stuart P were alsoparticipants with some others whose names have faded from the writer's memory. Jørgen will no doubt remember -it is believed that might have been the last time he played at ISCB - the experience was clearly traumatic.

The visit to Copenhagen, organised again by Mats, was memorable with a visit to the Royal course, founded in1898 and set in a deer park. The deer did not prove hazardous, but when they used their antlers to tear of lowerbranches of trees, the noise put some people off their putting.

Cambridge, the next year, had a visit to Brampton Park, and it is believed that Stuart P had another of his victoriesthere. Many golf balls were lost in the lakes and streams that criss-crossed the course. There was no golf in Basel,Barcelona, Budapest or Boston from 1994 to 1997. Is there anything statistically significant about having foursuccessive conferences in places that begin with a letter B? Does it have an effect on the likelihood of playing golf?

Since then, each ISCB conference has had a game of golf organised on the "social" afternoon. Dundee in 1998and Heidelberg in 1999 both advertised golf and Catherine Klersy was the first woman golfer to grace the ISCBmeetings. She has attended all the golf games since, winning regularly. She organised the golf for the Trentomeeting in a delightful place in the Dolomites called Petersberg. It is one of the Italian Tyrol places where Germanis spoken. Gareth Ambler, Tony Brady and Hanno Ulmer played in two or more of these meetings.

In 2001, Bernhard Huitfeldt organised a game at the home course (Bro Balsta, 40 km North of Stockholm) ofAnnika Sorenstam, arguably the world's leading woman golfer. Mats Lörstad, who has not been seen at an ISCBconference for a while, joined Stephen E so there were four, and again Catherine won by the complicated rules ofMats.

Golf might seem to be a natural game for statisticians, and we hope that this tradition will continue. Those of uswho have played on a number of occasions would certainly welcome others joining us. Errors in this article areSE's responsibility! He didn't play at all the venues!


Details of the Student Conference Awards 2002

Introduction

Student Conference Awards are available for registered postgraduate students to attend andpresent a paper at ISCB23 in Dijon, France, 9-13 September 2002. It is intended that five Awardswill be made. Selection will be made on the basis of a summary of the paper to be presented whichmust concern the application of statistical methodology to clinical or epidemiological research.Applications, prepared as described in Rule 4 below, should be sent to

Dr. Marie ReillyChair, ISCB Student Conference Awards Subcommittee

Dept. of Medical EpidemiologyKarolinska Institutet

PO Box 281Berzelius vag 8

SE-171 77 StockholmSweden

Tel: +46 8 728 61 56Fax: +46 8 31 49 57

e-mail: [email protected]

The closing date for applications will be Friday, 15 March 2002

Rules1. The ISCB Student Conference Award Scheme is administered by a Subcommittee established bythe Executive Committee of the ISCB.

2. Any student registered for a postgraduate degree is eligible to apply for an ISCB StudentConference Award, provided they have not been the winner of a previous ISCB Student ConferenceAward.

3. Applications should be received by the Chair of the Awards Subcommittee, prior to the closingdate: Friday, 15 March 2002.

4. Applications should consist of(a) a short abstract of the paper prepared according to the standard conference format.(b) a summary of the paper to be presented of between 3 and 5 sides of A4 paper in length.(c) [ mn[n_g_hn `lig nb_ mno^_hn~m moj_lpcmil, ih i``c]c[f hin_j[j_l i` nb_ mno^_hn~m chmncnoncih,confirming that the student is registered for a postgraduate degree, and that the paper to be presented isilcach[f, hin jl_pciomfs jo\fcmb_^, hin nb_ mo\d_]n i` [hs inb_l [q[l^ [h^ cm jl_îgch[hnfs nb_ mno^_hn~mown work.

5. Applications for Awards will be judged by the Awards Subcommittee, and applicants will benotified of the results as soon as possible, usually within six weeks of the closing date for applications.Judgement will concern the quality of the research described, and its relevance to the application ofstatistics to clinical and epidemiological research. The presentation of the application documents and theinterest of the work to ISCB members will be taken into consideration. The decisions made by theAwards Subcommittee will be final.


Details of the Student Conference Awards 2002 (continued)

6. In selecting papers for Awards, each member of the Awards Subcommittee will independentlygrade each paper submitted. In the case of ties in total grades, the Awards Subcommittee will make thefinal selection through discussion.

7. Members of the Awards Subcommittee must declare an interest in submissions from students withwhom they have had substantial contact. Such contact would certainly include students supervised by orin the same department as Subcommittee members and might also include professional contact with theproject as well as family links. The Subcommittee member with an interest in the submission would notgrade it, other grades being scaled accordingly, and would not participate in discussions of thatsubmission.

8. For 2002, it is intended to make five awards. However, the Awards Subcommittee reserves theright, subject to the approval of the President of the ISCB, to make more or fewer Awards, or even noneat all.

9. The Award Winners will have their papers scheduled in appropriate Contributed Paper Sessions.Their status as Award Winners will be indicated in the Conference Programme.

10. The ISCB will waive the registration fee of Award Winners, and will pre-pay their hotelaccommodation and Conference Dinner. The ISCB will not pay for any excursions for the student.Normally, it is expected that students will pay for travel, insurance and meals in the first instance. TheISCB will make a reimbursement soon after the conference, on the basis of appropriate receipts. A fixedper-diem to cover meals will be paid. Full travel costs will be reimbursed, provided that these are incurredeconomically. Thus full advantage of Apex fares, student discounts and so on should be taken. Adequateinsurance cover should be arranged, and will be reimbursed.

11. On being selected, a student would submit a travel budget to the Awards Subcommittee forapproval. If it proves more economic to make a longer visit in order to secure bargain fares, the cost ofextra accommodation and meals will be borne by the ISCB. Once a budget is approved, additional travelcosts will be reimbursed only if a convincing case of their necessity is made.

12. In the case of hardship, advance payments will be considered. If for good reason a student has tocancel the visit, then the ISCB will consider the reimbursement of costs incurred other than thoserecoverable from insurance.


Details of the Conference Awards for Scientists 2002

Introduction

Conference Awards for Scientists are available for biostatisticians from ISCB target countries (in particularcountries of Central and Eastern Europe as well as Third World countries) to attend and present a paper atISCB23 in Dijon, France, 9-13 September 2002. Up to twelve such awards will be granted. An award consistsof free accommodation and registration paid by ISCB. Scientists should submit the application form and aone-page summary of the paper to be presented. For full details of the scheme, which should be studied priorto preparing an application, and for the application form please write to:

Professor Michael SchemperChairman, ISCB Subcommittee on 'National Groups'

Department of Medical Computer Science, Section of Clinical BiometricsVienna University

Spitalgasse 23A-1090 Vienna, AUSTRIA

Tel: +43 1 40400 6689Fax: +43 1 40400 6687

e-mail: [email protected]

The closing date for applications will be Friday, 15 March 2002

Rules1. The ISCB Conference Awards for Scientists Scheme is administered by the Subcommittee on 'NationalGroups' established by the Executive Committee of ISCB. The purpose of the scheme is to support and encouragebiostatisticians working under financial constraints, and to enable them to participate in international dialogue.

2. Who can apply?S]c_hncmnm `lig {(ISCB n[la_n ]iohnlc_m| )ch j[lnc]of[l ]iohnlc_m i` C_hnl[f [h^ E[mn_lh Eolij_ [m q_ff [mThird World countries) whose work will benefit by attending an ISCB conference.

3. What is covered by the award?The registration fee is waived and inexpensive accommodation is organised and paid for by ISCB.Furthermore, a conference course may be attended free of charge.

4. What are the conditions for application?(a) Submission of a one-page summary of an intended oral presentation or poster which must concern the

application of statistical methodology to clinical or epidemiological research. Applicants who do nothave a specialist topic to pr_m_hn [l_ chpcn_^ ni jl_j[l_ [ gil_ a_h_l[f j[j_l qcnb [ ncnf_ mo]b [m {Tb_h__^ ìl \cimn[ncmnc][f ^_p_fijg_hn ch gs ]iohnls/l_acih/^_j[lng_hn|. Ohfs ih_ mo\gcmmcih j_lapplicant is permissible.

(b) Submission of a completed application form (item 6).The apjfc][ncih b[m ni \_ l_]_cp_^ \s nb_ Cb[clg[h, ISCB So\]iggcnn__ ih {N[ncih[f Gliojm| \s 15March 2002.

5. Who will decide on an application?Tb_ {N[ncih[f Gliojm| So\]iggcnn__ qcff ^_]c^_ [m miih [m jimmc\f_, omo[ffs qcnbch mcr q__em i` nb_closing date for applications, based on the following criteria: Quality of summary, financial need, value ofattendance at meeting to region (to National Groups), preference to those not supported by ISCB for theprevious meeting, possible set-up of future National Groups. In selecting award winners, each member ofthe Subcommittee will independently grade each application. In the case of ties in total grades, theSubcommittee will make the final selection through discussion. The Chairman of the Subcommittee willinform each applicant on the Subcommittee's decision.

6. Application form =>


Details of the Conference Awards for Scientists 2002 (continued)

Application

Conference Awards for Scientists

23rd Annual Conference of the International Society for Clinical Biostatisticsin Dijon, France, 9-13 September 2002

First Name

Surname

Title

Country of residence

Current affiliation(complete address)

Current position

Research interests

Language skills

I have submitted anabstract (give title)

Two most relevantpublications

Give details why ISCBshould support you toattend this conference

I would prefer to give an oral presentation / a poster.I agree to become a member of the ISCB for the year 2002 if I am granted the award. There will be nocharge for this.

Date: Signature:


How to Contact the ISCB Executive Committee (2002)

Who Address Tel: Fax: E-mail:

ISCB PermanentOffice:

ISCB, PO Box 25, DK-3480Fredensborg, Denmark

+45 48484100

+45 48484200

[email protected]

President:Mr Simon Day (UK)

Medicines Control Agency, MarketTowers, 1 Nine Elms Lane, LondonSW8 5NQ, UK

+44 207273 0112

+44 207273 0819

[email protected]

Vice-President:Prof. Maria GraziaValsecchi (I)

Dept. of Clinical Medicine andBiotechnology, University ofMilano-Bicocca, Via Cadore 48, I-20052 Monza, Italy

+39 039233 3074

+39 039233 2314

[email protected]

Secretary:Prof. EmmanuelLesaffre (B)

Catholic University Leuven,Biostatistical Centre, U.Z. St.Rafael, Kapucynenvoer 35, B-3000Leuven, Belgium

+32 16 336896

+32 16 336900

[email protected]

Treasurer:Prof. JohnWhitehead (UK)

MPS Research Unit, The Universityof Reading, PO Box 240, EarleyGate, Reading RG6 6FN, UK

+44 1189318027

+44 1189753169

[email protected]

News Editor:Dr David W. Warne(CH)

Chemin Frank-Thomas 40,CH-1208 Geneva, Switzerland

+41 22 7006380+41 22 7393374

+41 22 7393330+41 22 7006380

[email protected]

Webmaster:Miss Silvia Codony(DK)

Spadille Biostatistik ApS,Jernbanegade 34C, DK-3480Fredensborg, Denmark

+45 48484100

+45 48484200

[email protected]

Dr Elia Biganzoli Unit of Medical Statistics &Biometry, Istituto Nazionale per loStudio e la Cura dei Tumori, ViaVenezian 1, I-20133 Milano, Italy

+39 0223902456

+39 022362930

[email protected]

Dr HarbajanChadha-Boreham

Serono International, Chemin desMines 15bis, CH-1202 Geneva,Switzerland

+41 22 7393613

+41 22 7393330

[email protected]

Prof. Stephen Evans Medicines Control Agency, 1 NineElms Lane, London SW8 5NQ, UK

+44 207273 0279

+44 207273 0205

[email protected]

Dr Simon (Siem)Heisterkamp (NL)

RIVM, Dept. IMA, PB 86,P.O. BOX 1, NL-3720 BABilthoven, Netherlands

+31 30 2742587

+31 30 2744456

[email protected]

Prof. CarolRedmond

318A Parran Hall, 130 DeSotoStreet, Graduate School of PublicHealth, University of Pittsburgh,Pittsburgh, PA 15261, USA

+1 412 6241319

+1 412 6243047

[email protected]

Dr Julia Singer Chinoin Pharm. Sanofi-SynthelaboGroup, Budapest, To u. 1-5, 1045Hungary

+36 13692500/2189

- [email protected]

Prof. ElisabethSvensson

Department of Statistics (ESA),Örebro University, SE-701 82Örebro, Sweden

+46 19303339

- [email protected]

Prof. Norbert Victor(D)

Institut für Medizinische Biometrieund Informatik, Im NeuenheimerFeld 305, D-69120 Heidelberg,Germany

+49 622156 4140

+49 622156 4195

[email protected]


How to contact the ISCB Subcommittees

Please contact the chairs of these subcommittees for further information.

Title Terms of Reference Members E-mail addressesCommunications 1. To consider the future of the Newsletter, including ways to support the

Editor, procedures for transition of editorship.2. To maintain the ISCB homepage on the World Wide Web and facilitateplacement of annual meeting information on the homepage.3. To consider other communications with members, such as throughe-mail or the World Wide Web.

Chair: Dr DavidWarne (CH),Secretary: Mr BjarneNielsen (DK),Members:Prof. StephenEvans (UK),Dr NancyGeller (USA),Ms Caroline Jackson(UK),Miss Silvia Codony(DK),Mr Simon Day (UK)

[email protected]@[email protected]@[email protected]@[email protected]

Education To organise one or two day courses on contemporary methods in clinicalbiostatistics which will involve one or several members as lecturers whichwill be presented in locations represented by the Society. Guidelines andplans of previous courses are available.

Chair/Secretary:Prof. Carol Redmond(USA),Members:Prof. MichaelSchemper (A),Dr Albert Cobos (E),Prof. MikeCampbell (UK),Dr Shai Linn (ISR),Dr ElisabethSvensson (S),Dr Nicole Close (USA),Prof. Maria GraziaValsecchi (I).Mr Simon Day (UK)

[email protected]@[email protected]@[email protected]@[email protected]@[email protected]

National Groups 1. To help those who are interested in forming a National Group throughthe approval process.2. To review the arrangements with the current National Groups,specifically regarding financial matters.3. To set rules and standards for funding of ISCB members of NationalGroups and others from countries with exchange control restrictions orbarriers.

Chair/Secretary:Prof. MichaelSchemper (A),Members:Prof. JohnWhitehead (UK),A/Prof. JørgenSeldrup (SGP),Dr SiemHeisterkamp (NL),Prof. NorbertVictor (D),Dr Julia Singer (H),Dr Ewa Kawalec (PL),Dr Elia Biganzoli (I),Mr Simon Day (UK)

[email protected]@[email protected]@[email protected]@[email protected]@[email protected]

Statistics inRegulatory Affairs

The subcommittee on Regulatory Affairs will review, comment upon andseek to influence the development of regulatory requirements, guidelinesand other documents concerning the scientific aspects of datageneration, collection, management, analysis, and reporting. In general,the subcommittee will seek out and handle all regulatory issues in thename of the Society with the approval of the President or in his absence,the Vice-President.http://www.ucl.ac.uk/~ucaksjs/Guidance.html

Chair: A/Prof. JørgenSeldrup (SGP),Secretary:Prof. StephenSenn (UK),Members:Prof. HelmutSchäfer (D),Mr KarstenSchmidt (DK)Dr Harbajan Chadha-Boreham (CH),Dr Anna Petroccione(I),Mr Simon Day (UK)

[email protected]@[email protected]@[email protected]@[email protected]

StudentConferenceAwards

Student conference awards are available for registered postgraduatestudents to attend the annual meeting and present a paper. TheSubcommittee shall receive submissions, judge them, and administer theawards. The rules are announced in a timely issue of the Newsletter.

Chair:Dr Marie Reilly (S),Secretary: Mr BjarneNielsen (DK),Members:Dr Marc Buyse (B),Dr Bruno Cesana (I),Mr Simon Day (UK)

[email protected]@[email protected]@[email protected]


NEW! Reduced Subscription Rate for Statistics in Medicine and Wiley Books

From John Whitehead

A new arrangement between Wiley and the Societyhas been agreed, which will entitle ISCB membersto a generous discount on subscriptions to thejournal Statistics in Medicine and on Wiley books.The previous arrangement involved a reductionrelative to the dollar price, and was administeredby the ISCB Office. As explained in the Treasurer'sreport, currency fluctuations eroded the value ofthe discount and led to losses being borne by theSociety. At the AGM in Stockholm, it was agreedthat the old arrangement should be discontinued.

The new arrangement will be more generous andnot tied to the dollar rate: it will be administered byWiley. For members, the subscription rate for the2002 volume of Statistics in Medicine will be £235(US$355), compared with a personal subscriptionrate which will be c. £265 (US$398) for the year.The discount on books is 15% of the list price inpounds sterling (please see the new Wileycatalogue or visit the webpagewww.wileyeurope.com for current prices).

In order to take advantage of this offer, membersshould first renew their subscription to the Societyearly in 2002, and then use one of the Wileysubscription contact points detailed below. Youshould clearly indicate that you are a member ofnb_ ISCB [h^ [l_ ]f[cgcha nb_ g_g\_lm~ ^cm]iohn.

To subscribe to the journal, please contact WileyEurope: John Wiley & Sons Ltd., JournalsAdministration Department, 1 Oldlands Way,Bognor Regis, West Sussex PO19 1UD, UK;Phone: +44 (0) 1243 779777; Fax: +44 (0) 1243843232; E-mail: [email protected]

Members in North, Central, and South Americashould subscribe directly from John Wiley & SonsInc., Attn: Subscription Department, 605 ThirdAvenue New York, NY 10158; Phone: (212)850-6645 or (800) 825-7550; Fax: (212) 850-6021;E-mail: [email protected] Hours are Weekdays,8:30 a.m. to 5:30 p.m. Eastern Time.

To receive a 15% discount on Wiley books,please contact Rob Calver ([email protected],tel: +44 (0)20 8326 3867), or contact WileyCustomer Services Department: UK Dial free 0800243407; Overseas +44 1243 843294, giving thetitle and ISBN of the books required, and statingyou are a member of the ISCB and would like toclaim the 15% discount.

The histories of the ISCB and of the journalStatistics in Medicine have been intertwined fornearly 20 years now. The first conference of theSociety was held in 1979 while the first volume ofStatistics in Medicine appeared in 1982. In theirdifferent ways, both serve similar purposes. Otherconnections include the journal's publication eachyear of papers from ISCB conferences and thehosting of one of the journal's annual EditorialBoard meetings at the Society's conference. TheSociety is extremely grateful to Wiley for this newand generous contribution to our collaboration.


Wiley Announcement

I. Book News

- 3 for 2 Statistics Series Offer

- Introducing all the Latest Publications

II. Journal News

- Latest Developments with Statistics inMedicine

I. Book News

Special Offer in the Statistics in PracticeSeries.

The leading edge titles in the Statistics inPractice series provide detailed coverage ofstatistical concepts, methods and workedcase studies in specific fields ofinvestigation and study.

With sound motivation and many workedpractical examples, the books show indown-to-earth terms how to select and usean appropriate range of statisticaltechniques in a particular practical field.

The books meet the increasingly vital needfor statistical support required byprofessionals and research workers acrossa range of employment fields and researchenvironments. Subject areas coveredinclude medicine and pharmaceutics;industry, finance and commerce; publicservices; the earth and environmentalsciences.

This month we are offering you the chanceto buy three of the leading edge biostatisticstitles from this series at a special price. Buyany two of the titles from the list below, andreceive a third, absolutely free! For moreinformation on any of these titles, simplyclick on the links provided.

Analysing Survival Data from Clinical Trialsand Observation Studies � ETTOREMARUBINI and MARIA GRAZIAVALSECCHI

http://catalog.wiley.com/remtitle.cgi?isbn=0471939870&country=826

Design and Analysis of Sequential ClinicalTrials, Revised Second Edition � JOHNWHITEHEAD


Statistical Issues in Drug Development -STEPHEN SENN


Applied Mixed Models in Medicine � HELENBROWN and ROBIN PRESCOTT

http://catalog.wiley.com./remtitle.cgi?isbn=0471965545&country=826

To claim your discount, contact ZoeMitchell, [email protected]

Look out for the latest title on the Statisticsin Practice Series, due in February 2002:

Modelling in Medical Decision Making � ABayesian Approach, written by GIOVANNIPARMIGIANI, of the John Hopkins Schoolof Medicine, USA.

As more complex problems are being facedin medical decision making, simulationbased Bayesian methods are becomingincreasingly popular, as they are simple toimplement and can help to address themost pressing practical and ethicalconcerns arising in medical decisionmaking. This leading edge text:

Provides a unified framework for datacollection, inference and decision making

Discusses a range of methodologies

Features case studies, examples andexercises

Provides further examples, simulations andteaching material via a website

For more information, visit the Wileycatalogue at


and place your order today.

Gain an insight into all the latest Statisticstitles from Wiley with the NEW 2001/2002catalogue. Featuring both books andjournals, all the latest titles are presentedcovering everything from Bayesian

Analysis, Survey Research Methods andSampling, through to, Biostatistics,Computational and Graphical Statistics.

Be one step ahead with the NEW catalogueand order your FREE copy today by simplyemailing [email protected]

II. Journal News

Statistics in Medicine continues to be one ofthe most prestigious journals in its field, andwe are proud to announce a current impactfactor of 1.717.

Providing 24 issues a year of highly cited,peer reviewed science, the journalcontinues to provide leading edgeinformation in a rapidly evolving field,publishing papers on practical applicationsof statistics and other quantitative methodsto medicine and its applied sciences.

The frequent special issue program,together with the comprehensive bookreviews and rapid dissemination of thelatest information, make this journalessential reading for all statisticians,clinicians, epidemiologists, healthresearchers, mathematicians and computerscientists interested in medicine.

Bljfkd sbov pllk�

A Special Issue publishing papers from theInternational Meetings of the ISCB.

[No. 20-24, 31 December 2001]

See for yourself what Statistics in Medicinehas to offer you � visit the journalhomepage today and download a FREEonline sample copy:

www.interscience.wiley.com/statisticsinmedicine

Zoe Mitchell

Marketing Executive

[email protected]

Blackwell Announcement

Statistical Methods in Medical ResearchFourth EditionPublished September 2001. 832 pages, 98illustrations. ISBN 0 632 05257 0. Hardback£55.00By Peter Armitage, Geoffrey Berry, J N SMatthews

The explanation and implementation ofstatistical methods for the medical researcher orstatistician remains an integral part of modernmedical research. This book explains the use ofexperimental and analytical biostatisticssystems. Its accessible style allows it to be usedby the non-mathematician as a fundamentalcomponent of successful research.

Since the third edition, there have been manydevelopments in statistical techniques. Thefourth edition aims to provide the medicalstatistician with an accessible guide to thesetechniques and to reflect the extent of theirusage in medical research.

The new edition takes a much morecomprehensive approach to its subject. Therehas been a radical reorganization of the text toimprove the continuity and cohesion of the

presentation and to extend the scope by coveringmany new ideas now being introduced into theanalysis of medical research data. The authorshave tried to maintain a modest level ofmathematical exposition which characterized theearlier editions, essentially confining themathematics to the statement of algebraicformulae rather than pursuing mathematicalproofs.

Further information on this title can be found athttp://www.medirect.com/~cgilib/book.asp?subj=genm&book=5758. The book is priced at£55.00, with postage charges of £2.50 within theUK, £5.00 within Europe and £7.00 for the Restof the World.

Details of other Blackwell Publishing titleswhich may be of interest to you can be found inthe General Medicine section of ourmedirect.com website, athttp://www.medirect.com/~cgilib/editorial.asp?subj=genm z I really would urge you to take alook.

If you decide that you would like to place anorder, you can contact us using the details givenbelow. Alternatively, should you have any

queries regarding Blackwell Publishingproducts, please do not hesitate to contact us.We look forward to hearing from you soon.

Yours faithfully

Philip SaugmanMedicine Direct

Telephone: +44 (0)1865 206233Fax: +44 (0)1865 206026E-mail: [email protected]: www.medirect.com

Post:Medical Marketing DepartmentBlackwell PublishingOsney MeadOxfordOX2 0ELUK


ISCB General Information:

Advertising Rates

The prices are:Full A4 page: £ 200Half A4 page: £ 150Quarter A4 page: £ 100

Additionally, we will include loose flyers with thedistribution of the newsletter at an initial handling cost of£ 150. However, if the addition of the flyers increases thepostal charges, the advertiser will also be charged thedifference in distribution costs. For further information,please contact the Editor.

Publishing dates 2002(and deadlines)

(end April) June,(end October) December.

IMPORTANT NOTE: E-mail Lists and Personal InformationISCB has a strict policy not to give out any information concerning its members to any organisation which requestsfq. Hc ^ `ljm^kv tfpebp ql pbka j^qbofî ql qeb jbj_bop, qeb _olèrobp jrpq _b pbkq ql qeb Rl`fbqv�p Oboj^kbkq Office and News Editor for distribution with the News (see above). Alternatively, small announcements can be sentas an e-mail to members by the ISCB egroup (currently free of charge, but under review and subject to theDuBlj�p ab`fpflk).

Aims

The Society is organised and shall be operated for educational and scientific purposes with the following Aims:) to stimulate research on the biostatistical principles and methodology used in clinical research;) to increase the relevance of statistical theory to clinical medicine;) to promote high and harmonised standards of statistical practice;) to work with other societies and organisations in the advancement of biostatistics;) to promote better understanding of the use and interpretation of biostatistics by the general public, and by

national and international organisations and agencies within the public and commercial sectors with an interestin, and/or responsibilities for, public health; and

) to provide a common forum for clinicians and statisticians through meetings, seminars and publications

Changes of Address or E-mail

Please inform the Permanent Office that looks after the membership and mailing list databases.Also, if your e-mail address changes, please inform the Office and the News Editor so that your address ischanged in the ISCB database and egroup.

Information on Submitting Articles

Articles sent via e-mail or on diskette (Word, HTML or text) on almost any topic are most welcome. This is anfkclojî kbtpibqqbo clo vlr qeb obâbop, pl mib^pb glfk fk ^ka j^hb HRBA Mbtp ^ j^d^wfkb qe^q�p bsbk jlob interesting and fun to read.


ISCB Membership Information

The International Society for Clinical Biostatistics(ISCB) was founded in 1978 to stimulate research intothe principles and methodology used in the design andanalysis of clinical research and to increase therelevance of statistical theory to the real world of clinicalmedicine.

The ISCB organises an annual scientific meeting whichmembers and non-members are able to attend. Themain objective of the annual scientific meetings is tocreate an opportunity for the exchange of knowledge,experience and ideas among clinicians, statisticiansand members of other disciplines, such asepidemiologists, clinical chemists and clinicalpharmacologists, working or interested in, the field ofclinical biostatistics.

The scientific meetings cover a broad spectrum ofbiostatistical interests and regularly include sessions onthe design and analysis of clinical trials, epidemiologyand statistical methodology, as well as from time to timeconsidering more specialist issues such as, forexample, education of biometricians and biometricsusers, pharmacokinetics, medical data-bases andpharmaco-epidemiology. Each meeting includes a mini-symposium devoted to a particular medical or statisticalfield.

Previous meetings in recent years have been held inBoston (1997), Dundee (1998), Heidelberg (1999),Trento (2000) and Stockholm (2001). A selection oftalks at the meetings, for which papers are submittedfor review and which are eventually accepted, arepublished in Statistics in Medicine. The ISCB benefitsfrom a special journal concession from John Wiley &Sons Limited, the publishers of Statistics in Medicine,so that members are able to subscribe to the journal ata preferential rate.

The ISCB also organises courses to cover particularstatistical topics. These are run to precede or follow onfrom the annual scientific meeting and are given by theforemost researchers in the field. Recent courses haveincluded Analysis of Ordered Categorical Data, Cross-over Trials in Clinical Research, Analysis of RepeatedMeasures, and Survival Analysis, Extending the CoxModel, and Statistical Methods for GeneticEpidemiology.

The composition of the Executive Committee (ExCom)for 2002 is as follows:

Officers:President: Mr Simon Day (UK),Vice-President: Prof. Maria Grazia Valsecchi (I),Secretary: Prof. Emmanuel Lesaffre (B),Treasurer: Prof. John Whitehead (UK), andMembers:Newsletter Editor: Dr David W. Warne (CH),Webmaster: Miss Silvia Codony (DK),Dr Elia Biganzoli (I), Dr Harbajan Chadha-Boreham(CH), Prof. Stephen Evans (UK), Dr Siem Heisterkamp(NL), Prof. Carol Redmond (USA), Dr Julia Singer (H),Prof. Elisabeth Svensson (S) and Prof. Norbert Victor(D).

The annual general meeting of the ISCB is organised tocoincide with the scientific meeting. Membership of theSociety is drawn from more than 40 countries worldwideand the number of members is nearly 800.

The ISCB also has special Subcommittees dealingwith particular aspects of biostatistics.

The Society publishes a Newsletter 2 or 3times a year. The ISCB News editor is Dr David W.Warne, Chemin Frank-Thomas 40, CH-1208 Geneva,Switzerland. Items for inclusion in the Newsletter shouldbe sent to him via e-mail to:

[email protected]

Membership of the Society is open to all with aninterest in biostatistics. The current annual (to31 December 2002) Ordinary membership fee is £20.The Full-time Student Membership fee is £10.

Applications for membership should be sent to:ISCB Permanent Office, PO Box 25,DK-3480 Fredensborg, DENMARKTel: +45 48 484 100,Fax: +45 48 484 200,e-mail: [email protected]

[email protected]

WWW: http://WWW.ISCB-HOMEPAGE.ORG

Have you visited the ISCB Homepage recently ? http://www.iscb-homepage.org

Please check often for updated information between issues of the ISCB News!


Membership Subscription

INTERNATIONAL SOCIETY FOR CLINICAL BIOSTATISTICS2002 Membership Subscription

Surname ________________________________ Initials/Name _________ Occupation (please tick):

Title (Prof./Dr/etc.) _________________________ Post held ____________ & Statistician

Business address ______________________________________________ & Medical Doctor

_____________________________________________________________ & Both

Post code and country ___________________________________________ & Neither

Phone No. and Fax No. _____________________ e-mail: ______________ www:_____________________

SUBSCRIPTION: & £ 20.00 Ordinary membership of ISCB (to 31 December 2002).(please tick one only) & £ 10.00 Full-time Student Membership of ISCB (to 31 Dec. 2002).

PAYMENT IS MADE BY:

& Credit card authorisation: & VISA or & Master Card

Name on credit card: _____________ Expiry date ________ Signature __________

Card number to debit: - - -

& The following cheques, made payable to the International Society for Clinical Biostatistics are acceptable:

A certified cheque drawn on a bank in Denmark or London /Bestätigten Bankcheck auf dänisches oder London Geldinstitut /Chèque de banque certifié sur une banque danoise ou Londres.

A Sterling Eurocheque, or any cheque drawn in Sterling & payable on a named Danish or London bank.

Note: Non-Sterling cheques, bank cheques not drawn on a Danish bank or a London bank,and cheques not made payable to ISCB will be returned to sender.

Cheque / Money Order No.: (if known) _______________Date sent ___________

& Bank transfer

Please transfer direct to:Nordea A/SDK-1786 Copenhagen V, DenmarkSWIFT code: UNIB DK KKTelex Nr.: 27543 unib dkfor the credit of account2229 0392 562 109Bank code Account No.

Bitte überweisen Sie direkt an:Nordea A/SDK-1786 Kopenhagen V, DänemarkSWIFT-Kode: UNIB DK KKTelex Nr.: 27543 unib dkZugunsten des Kontos2229 0392 562 109Bankleitzahl Konto Nr.

Veuillez effectuer le virement directement à:Nordea A/SDK-1786 Copenhague V, DanemarkSWIFT: UNIB DK KKTélex: 27543 unib dkau crédit du compte indiqué2229 0392 562 109Code bancaire Numéro de compte

Please return this form to:ISCB Permanent Office

P.O. Box 25DK-3480 Fredensborg

DenmarkTel: +45 48 48 41 00, Fax: +45 48 48 42 00


Calendar

9-13 September 2002 Dijon, FRANCE

ISCB23Info: Harbajan Chadha-Boreham e-mail: [email protected]

20-24 July 2003 London, ENGLAND

ISCB24 joint with Society for Clinical TrialsInfo: Diana Elbourne e-mail: [email protected]

15-19 August 2004 Leiden, Netherlands

ISCB25Info: Hans Van Houwelingen e-mail: [email protected]

For the latest conf. info, see the ISCB webpage or the ISI webpage: www.cbs.nl/isi/calendar.htm2002:March 15-21 ENAR/IMS Eastern Regional to be

held in Washington, DC, USA. Program Chair:Jiayang Sun, [email protected],Case Western Reserve University LocalArrangements Chair: Colin Wu,[email protected], John Hopkins UniversityContributed Papers Chair: Nidhan Choudhuri,[email protected], Website CaseWestern Reserve University sun.cwru.edu/ims

March 19-22 German Open Conference onProbability and Statistics, to be held at theUniversity of Magdeburg, Germany. Address:Magdeburger Stochastik-Tage 2002, c/o Prof.Dr. Gerd Christoph, Otto-von-Guericke-Universität Magdeburg, Institut fürMathematische Stochastik, Universitätsplatz 2,D-39106 Magdeburg, Germany. Fax: 0049-(0)391-67 11172 Phone: 0049-(0)391-67 18652.Email: [email protected] Website:www.math.uni-magdeburg.de/stoch2002/

May 13-17 34th Journées de Statistique of theSociété Française de Statistique, jointlyorganized by the Institutes of Statistics of theUniversité libre de Bruxelles and of theUniversité catholique de Louvain, will be held inBrussels and Louvain-la-Neuve, Belgium.Information: JSBL 2002, Institute of Statistics,Université catholique de Louvain, 20 Voie duRoman Pays, B-1348 Louvain-la-Neuve,Belgium. Telephone : +32 10 47 43 54, Fax :+32 10 47 30 32, Websitewww.stat.ucl.ac.be/jsbl2002

May 26-29 Annual Meeting of the StatisticalSociety of Canada, Hamilton, Ontario, Canada.Local Arrangements Chair: Peter MacdonaldDepartment of Mathematics and Statistics,McMaster University, 1280 Main Street West,Hamilton, Ontario, L8S 4K1, Canada. email:[email protected] Phone: (905)525-9140 x 23423 Fax: (905) 522-0935.Program Chair: Bruce Smith Department ofMathematics and Statistics, DalhousieUniversity, Halifax, Nova Scotia, B3H 3J5,Canada email: [email protected] Phone:(902) 494-2257 Fax: (902) 494-5130.

May 27-29 Work Session on Statistical DataEditing of the Conference of EuropeanStatisticians of the UN, to be held in Helsinki,Finland. Contact: Mrs. Jana Meliskova,UNU/ECE Statistical Division, tel. +41 22 917 4150; fax +41 22 917 00 40, [email protected] Website:www.unece.org/stats/

June 2-6 Seventh Valencia International Meetingon Bayesian Statistics, to be held at the Playade las Americas,Tenerife, Canary Islands,Spain. Website www.uv.es/valencia7 orwww.stat.duke.edu/valencia7

June 5-9 Hawaii International Conference onStatistics and Related Fields, to be held at theSheraton Waikiki Hotel, Honolulu, Hawaii, USA.Contact: Hawaii International Conference onStatistics, 2440 Campus Road #517, Honolulu,HI, 96822, USA. Tel. (808) 223-1748, fax (808)947-2420, email [email protected]: www.hicstatistics.org

June 23-26 International Symposium onForecasting, ISF2002, to be held in TrinityCollege, Dublin, Ireland. Information: JohnHaslett, email: [email protected] Website:www.isf2002.org

June 23-29 The 8th International VilniusConference on Probability Theory andMathematical Statistics, to be held in Vilnius,Lithuania. Chair of Committee: ProfessorVytautas Statulevicius, Institute of Mathematicsand Informatics, Akademijos str. 4, 2600 Vilnius,Lithuania. Information: Dr. Aleksandras Plikusas,Institute of Mathematics and Informatics,Akademijos str. 4, 2600 Vilnius, Lithuania. Tel.:370-2-729207; Fax: 370-2-729209; Email:[email protected] Website:www.science.mii.lt/vilconf8/

July 2-5 MCQT'02 - First Madrid Conference onQueueing Theory, to be held at the Departmentof Statistics and OR, Faculty of Mathematics,University Complutense of Madrid, Spain.Information: Jesus R. Artalejo, email:[email protected]. Websitewww.mat.ucm.es/deptos/es/mcqt/conf.html

July 7-10 The 2002 Taipei International StatisticalSymposium and Bernoulli Society EAPRConference, to be held in the Activity Center,Academia Sinica, Taipei, Taiwan, China. Email:[email protected] Website:www.stat.sinica.edu.tw/2002symp

July 7-11 16th Australian Statistical Conference, tobe held at the National Convention Centre,Canberra, Australia. Email: [email protected]: www.statsoc.org.au/asc16.html

July 7-12 The Sixth International Conference onTeaching Statistics (ICOTS6), to be held inDurban, South Africa. Information: Maria-Gabriella Ottaviani - IPC [email protected], BrianPhillips - International [email protected], Dani Ben-Zvi - IPCScientific Secretary [email protected]. IPC Website:www.beeri.org.il/icots6 LOC Website:icots.itikzn.co.za/.

July 8-12 17th International Workshop onStatistical Modelling: Statistical Modelling inSociety, to be held in Chania-Crete, Greece.Email: [email protected] Website:www.unl.ac.uk/iwsm/

July 15-19 Current Advances and Trends inNonparametric Statistics, to be held on Crete,Greece. Organizers: Michael G. Akritas andDimitris N. Politis IMS Representative: MichaelG. Akritas, [email protected] Websitewww.stat.psu.edu/~npconf/

July 21-26 IBC 2002 - International BiometricConference 2002, to be held at the University ofFreiburg, Germany. Program Chair: RobertCurnow, e-mail [email protected] ChairLocal Organizing Committee: MartinSchumacher, e-mail: [email protected]: www.ibc2002.uni-freiburg.de

July 22-24 26th Annual Conference of theGesellschaft für Klassifikation (GfKl), to be heldat the University of Mannheim, Germany.Information: local organizer Prof. Dr. MartinSchader, website: www.wifo.uni-mannheim.de/gfkl2002/index.html

July 27- August 1 IMS Annual Meeting/FourthInternational Probability Symposium, to be heldin Banff, Canada. IMS Program Chair TomDiCiccio, Cornell, [email protected], SymposiumChair: Tom Kurtz, U. Wisconsin,[email protected], IMS Local Chair:Subhash Lele, U. Alberta, [email protected]

August 4-9 Fourth International Conference onStatistical Data Analysis based on the L1-Normand Related Methods - to be held at theUniversity of Neuchâtel, Switzerland.Information: Prof. Yadolah Dodge, ConferenceOrganizer Statistics Group, Case Postale 1825,CH-2002 Neuchatel. Phone +41 32 718 13 80Fax +41 32 718 13 81 Email:[email protected]

August 11-15 Joint Statistical Meetings, NewYork, New York. To be held at the New YorkHilton and Sheraton New York. Sponsored byASA, ENAR, WNAR, IMS, and SCC.Information: ASA, 1429 Duke St., Alexandria ,VA 22314-3415; (703) 684-1221; [email protected]

August 15-17 Symposium on Stochastics andApplications (SSA) to be held at the NationalUniversity of Singapore. Email:[email protected] Website:www.math.nus.edu.sg/ssa

August 19-23 24th European Meeting ofStatisticians, to be held in Prague, CzechRepublic. Information: Martin Janzura, Instituteof Information Theory and Automation, POB 18,182 08 Praha 8, Czech Republic. Tel: 420 26605 2572 Fax: 420 2 688 4903 Email:[email protected]

August 24-28 Compstat2002 to be held in Berlin,Germany. E-mail: [email protected],website www.compstat2002.de

August 25-28 International Conference onImproving Surveys (ICIS-2002), to be held at theUniversity of Copenhagen. ConferenceSecretariat: International Conference Services,P.O. box 41, Strandvejen 171, DK-2900Hellerup, Copenhagen, Denmark. Telephone:+45 3946 0500, Fax +45 3946 0515. E-mail:[email protected]

September 2-6 RSS 2002 - InternationalConference to be held at the University ofPlymouth, Plymouth, England. The 2002Conference of the Royal Statistitical Society (4-6September) will preceded by short courses (2-3September). Email: [email protected]/maths/research/stats/RSS2002.html

November 14-17 International Conference onQuestionnaire Development, Evaluation andTesting Methods, to be held in Charleston,South Carolina, USA. Information: JenniferRothgeb ([email protected]), orvisit the web site www.jpsm.umd.edu/qdet.

December 28-30 International Conference on"Ranking and Selection, Multiple Comparisons,Reliability, and Their Applications". TentativeVenue: Hotel Savera, Chennai, Tamilnadu,India. Organizers: N. Balakrishnan, McMasterUniversity, [email protected]; N.Kannan, University of Texas at San Antonio,[email protected]; H. N. Nagaraja, Ohio StateUniversity, [email protected].

Dr Harbajan Chadha-Boreham (CH) Tel/Fax: +41 22 700 6380 · 54 Philippines 1 55 South Korea 3 1 ... past Presidents (Nancy Geller and Karsten ... many people for their contributions

Documents