Medical Statistics from A to Z: A Guide for Clinicians and Medical Students PDF
Preview Medical Statistics from A to Z: A Guide for Clinicians and Medical Students
MEDICAL STATISTICS from A to Z From‘Abortionrate’to‘Zygositydetermination’,thisaccessibleintroductiontotheterminology ofmedicalstatisticsdescribesmorethan1500terms,allclearlyexplained,illustratedanddefined innon-technicallanguage,withoutanymathematicalformulae!Withthemajorityofterms revisedandupdatedandtheadditionofmorethan100brandnewdefinitions,thisnewedition willenablemedicalstudentstoquicklygraspthemeaningofanyofthestatisticaltermsthey encounterwhenreadingthemedicalliterature.Furthermore,annotatedcommentsareused judiciouslytowarntheunwaryofsomeofthecommonpitfallsthataccompanysomecherished biomedicalstatisticaltechniques.Whereverpossible,thedefinitionsaresupplementedwitha referencetofurtherreading,wherethereadermaygainadeeperinsight,sowhilstthedefinitions areeasilydisgestible,theyalsoprovideasteppingstonetoamoresophisticatedcomprehension. Statisticalterminologycanbequitebewilderingforclinicians:thisguidewillbealifesaver. Brian EverittisEditor-in-ChiefofStatisticalMethodsinMedicalResearchandProfessor EmeritusatKing’sCollege,London. i ii M E D I C A L S T A T I S T I C S from A to Z A Guide for Clinicians and Medical Students Second Edition B.S. Everitt InstituteofPsychiatry,King’sCollege,UniversityofLondon iii cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press TheEdinburghBuilding,Cambridgecb22ru,UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521867634 ©B.Everitt2006 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2006 isbn-13 978-0-511-25763-6eBook (NetLibrary) isbn-10 0-511-25763-5 eBook (NetLibrary) isbn-13 978-0-521-86763-4hardback isbn-10 0-521-86763-0 hardback isbn-13 978-0-521-68718-8paperback isbn-10 0-521-68718-7 paperback Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guaranteethatanycontentonsuchwebsitesis,orwillremain,accurateorappropriate. Everyefforthasbeenmadeinpreparingthispublicationtoprovideaccurateand up-to-date information which is in accord with accepted standards and practice at the time of publication. Although case histories are drawn from actual cases, every effort has been made to disguise the identities of the individuals involved. Nevertheless, the authors,editorsandpublisherscanmakenowarrantiesthattheinformationcontained herein is totally free from error, not least because clinical standards are constantly changing through research and regulation. The authors, editors and publishers therefore disclaim all liability for direct or consequential damages resulting from the use of material contained in this publication. Readers are strongly advised to pay careful attentiontoinformationprovidedbythemanufacturerofanydrugsorequipmentthat they plan to use. Preface to the second edition InthesecondeditionofMedicalStatisticsfromAtoZIhaveaddedmanynew definitionsandtakentheopportunitytocorrectandclarifyanumberofentries. Morereferencesarealsoprovidedthatpointreaderstomoredetailedaccountsof topics. Preface to the first edition Clinicians,researchworkersinthehealthsciences,andevenmedicalstudentsoften encountertermsfrommedicalstatisticsandrelatedareasintheirwork,particularly whenreadingmedicaljournalsandotherrelevantliterature.Theaimofthis guideistoprovidesuchpeoplewithnontechnicaldefinitionsofmanysuch terms.Consequently,nomathematicalnomenclatureorformulaeareusedinthe definitions.Thosereadersinterestedinsuchmaterialwillbeabletofinditinoneof themanystandardstatisticaltextsnowavailableandinTheCambridgeDictionary ofStatistics.Inaddition,readersseekingmoreinformationaboutaparticular topicwillhopefullyfindthereferencesgivenwiththemajorityofentriesofsome help;wheneverpossible,theseinvolvemedicalratherthanstatisticaljournals,and introductorystatisticaltextsratherthanthosethataremoreadvanced.(References arenotgivenfortermssuchasmean,varianceandcriticalregionforwhich furtherdetailsareeasilyavailableinmostintroductorymedicalstatisticstexts.) Severalformsofcross-referencingareused.Termsincourier newappear asaseparateheadwordelsewhereinthedictionary,althoughthisprocedureis usedinarelativelylimitedwaywithheadwordsdefiningfrequentlyoccurringterms suchasrandomvariable,probabilityandsamplenotreferredtointhisway.Some entriessimplyreferreaderstoanotherentry.Thismayindicatethatthetermsare synonymousorthatthetermisdiscussedmoreconvenientlyunderanotherentry.In thelattercase,thetermisprintedinitalicsinthemainentry.Entriesareinalphabetical orderusingtheletter-by-letterratherthantheword-by-wordconvention. OfthemanysourcesofmaterialIhaveconsultedinthepreparationofthisbook, Iwouldliketomentiontwothathavebeenofparticularhelp,namelythe EncyclopediaofBiostatisticsandtheDictionaryofEpidemiology. v REFERENCES Armitage,P.andColton,T.,1989,EncyclopediaofBiostatistics,J.Wiley&Sons,Chichester. Everitt,B.S.,2006,TheCambridgeDictionaryofStatistics,3rdedn,CambridgeUniversityPress, Cambridge. Last,J.M.,2001,DictionaryofEpidemiology,4thedn,OxfordUniversityPress,NewYork. vi A Abortion rate: Theannualnumberofabortionsper1000womenofreproductiveage (usuallydefinedasage15–44years).Forexample,intheUSAin1970theratewas five,in1980itwas25andin1990itwas24.[FamilyPlanningPerspectives,1998,30, 244–7.] Abortion ratio: Theestimatednumberofabortionsper1000livebirthsinagivenyear. Forexample,intheUSAin1970theratiowas52,in1980itwas359andin1990it was344.[FamilyPlanningPerspectives,1998,30,244–7.] Abscissa: Thehorizontal(orx-axis)onagraph,oraparticularpointonthataxis. Absolute cause-specific risk: Synonymforabsoluterisk. Absolute deviation: Synonymforaveragedeviation. Absoluterisk:Oftenusedasasynonymforincidence,althoughalsousedoccasionally forattributable risk, excess riskorriskdifference.Defined moreproperlyastheprobabilitythatadisease-freeindividualwilldevelopagiven diseaseoveraspecifiedtimeintervalgivencurrentageandindividualriskfactors, andinthepresenceofcompeting risks.Absoluteriskisaprobabilityand consequentlyliesbetween0and1.Seealsorelativerisk.[Kleinbaum,D.G., Kupper,L.L.andMorgenstern,H.,1982,EpidemiologicResearch:Principlesand QuantitativeMethods,LifetimeLearningPublications,Belmont.] Absolute risk reduction: Theproportionofuntreatedpeoplewhoexperiencean adverseeventminustheproportionoftreatedpeoplewhoexperiencetheevent.For exampleinaclinical trial ofmammographyitwasfoundthatoutof 129750womenwhowereinvitedtobeginhavingmammogramsinthelate1970s andearly1980s,511diedofbreastcanceroverthenext15years,adeathrateof0.4 percent.Inthecontrolgroupof117260womenwhowerenotinvitedtohave regularmammograms,therewere584breastcancerdeathsoverthesameperiod,a deathrateof0.5percent.Sotheestimatedabsoluteriskreductionis0.1percent. Seealsorelativeriskandnumberneededtotreat.[Sackett,D.L.,Richardson, W.S.,Rosenberg,W.andHaynes,R.B.,1997,EvidenceBasedMedicine:Howto PracticeandTeachEBM,ChurchillLivingstone,NewYork.] Absorbing barrier: Seerandomwalk. Accelerated failure time model: Ageneralmodelfordataconsistingofsurvival times,inwhichexplanatoryvariablesmeasuredonanindividualareassumedto 1 actmultiplicativelyonthetimescale,andsoaffecttherateatwhichanindividual proceedsalongthetimeaxis.Consequentlythemodelcanbeinterpretedinterms ofthespeedofprogressionofadisease.Thismodelwhichsimplyregressesthe logarithmofthesurvivaltimeonthecovariates,althoughusedfarlessoftenthat Cox's proportional hazards model,mightbeausefulalternativein manysituationsbecauseofthisintuitivephysicalinterpretation.[Collett,D.,2003, ModellingSurvivalDatainMedicalResearch,2ndedn,ChapmanandHall/CRC, BocaRaton,FL.] Acceptable quality level: Seequalitycontrolprocedures. Acceptable risk: Theriskforwhichthebenefitsofaparticularmedicalprocedureare consideredtooutweighthepotentialhazards.Forexample,islettransplantation wouldhelptocontrolthemanysecondaryeffectsoftype1diabetes,butwhatisthe appropriatelevelofrisktoimplementthistechnologyresponsiblyconsideringthe possibledangersfromretroviruses?[Nature,1998,391,326.] Acceptance region: Atermassociatedwithstatisticalsignificancetests,whichgivesthe setofvaluesofatest statisticforwhichthenullhypothesisistobe accepted.Suppose,forexample,thataz-testisbeingusedtotestthenull hypothesisthatthemeanbloodpressureofmenandwomenisequalagainstthe alternativehypothesisthatthetwomeansarenotequal.Ifthechosensignificance ofthetestis0.05,thentheacceptanceregionconsistsofvaluesoftheteststatisticz between−1.96and1.96.[Altman,D.G.,1991,PracticalStatisticsforMedical Research,ChapmanandHall/CRC,BocaRaton,FL.] Accident proneness: Apersonalpsychologicalfactorthataffectsanindividual’s probabilityofsufferinganaccident.Theconcepthasbeenstudiedstatistically underanumberofdifferentassumptionsforaccidents: (cid:2) purechance,leadingtothePoisson distribution; (cid:2) truecontagion,i.e.thehypothesisthatallindividualsinitiallyhavethesame probabilityofhavinganaccident,butthatthisprobabilitychangeseachtimean accidenthappens; (cid:2) apparentcontagion,i.e.thehypothesisthatindividualshaveconstantbut unequalprobabilitiesofhavinganaccident. Thestudyofaccidentpronenesshasbeenvaluableinthedevelopmentofparticular statisticalmethodologies,althoughinthelasttwodecadestheconcepthas,in general,beenoutoffavour.Attentionnowappearstohavemovedmoretowards riskevaluationandanalysis.[Shaw,L.andSichel,H.S.,1971,AccidentProneness, PergamonPress,Oxford.] Accrual rate: Therateatwhicheligiblepatientsareenteredintoaclinical trial, measuredaspeopleperunittime.Oftendisappointinglylow,forreasonsthatmay bebothphysicianandpatientrelated.[JournalofClinicalOncology,2001,19, 3554–61.] Accuracy: Thedegreeofconformitytosomerecognizedstandardvalue.Seealsobias. Accuracy versus precision: Anaccurateestimateisclosetothequantitybeing 2 estimated.Apreciseintervalestimateisanarrowone,butitmaynotbeaccurate evenwhenquotedtoalargenumberofdecimalplaces. ACES: Abbreviationforactivecontrolequivalencestudies. ACF: Abbreviationforautocorrelationfunction. ACORN: Acronymfor‘aclassificationofresidentialneighbourhoods’.Asystemfor classifyinghouseholdsaccordingtodemographic,employmentandhousing characteristicsoftheirimmediateneighbourhood.Derivedbyapplyingcluster analysisto40variables,includingage,class,tenure,dwellingtypeandcar ownership,usedtodescribeeachneighbourhood.[Dorling,D.andSimpson,S., 1999,StatisticsinSociety,Arnold,London.] Acquiescence bias: Thebiasproducedbyrespondentsinasurveywhohavethe tendencytogivepositiveanswers,suchas‘true’,‘like’,‘often’or‘yes’toaquestion. Atitsmostextreme,thepersonrespondsinthiswayirrespectiveofthecontentof thequestion.Thusapersonmayrespond‘true’totwostatementssuchas‘Ialways takemymedicineontime’and‘Ioftenforgettotakemypills’.Seealso end-aversionbias.[JournalofIntellectualDisabilityResearch,1995,39,331–40.] Active control equivalence studies (ACES): Studiesthataimtodemonstratethatan experimentaltreatmentisequivalentinefficacytoastandardtreatment.The justificationforundertakingsuchstudiesisthatevenifthenewtreatmentisno moreeffectivethantheexistingtreatmentinalleviatingaparticularcondition,it maystillbeofuseforpatientswhoareresistantto,orwhosimplycannottolerate, thestandardtreatment.Soclinical trialsaresometimesundertakenwhen theobjectissimplytoshowthatthenewtreatmentisatleastasgoodastheexisting treatment.[Senn,S.,1997,StatisticalIssuesinDrugDevelopment,J.Wiley&Sons, Chichester.] Active control trials: Clinical trialsinwhichthenewtreatmentiscompared withsomeotheractiveagentratherthanaplacebo.Forexample,aclinicaltrial investigatingtreatmentsforasthmamightcomparethelong-actingbeta-agonists salmeterolandformoterolwiththeshorter-actingbeta-agonistsalbutomol. [Senn,S.,1997,StatisticalIssuesinDrugDevelopment,J.Wiley&Sons, Chichester.] Active life expectancy (ALE): Definedforagivenageastheexpectedremainingyears freeofdisability.Inlife expectancytheendpointisdeath.Inactivelife expectancytheendpointisthelossofindependenceortheneedtorelyonothers forassistancewithdailyactivities.ALEisausefulindexofpublichealthandquality oflifeinapopulation.Interestinrecentyearshascenteredonwhethercurrent trendstowardslongerlifeexpectancyhavebeenaccompaniedbycomparable increasesinactivelifeexpectancy.Seealsodisability-freelifeexpectancy.[New EnglandJournalofMedicine,1983,309,1218–24.] Activities of daily living scale (ADLS): Ascaledesignedtomeasurephysical ability/disabilitythatisusedininvestigationsofavarietyofchronicdisabling conditions,suchasarthritis.Thescaleisbasedonscoringresponsestoquestions 3
The list of books you might like
