Ethical Principles & Guidelines for Research Involving Human Subjects

EthicalPrinciples&GuidelinesforResearchInvolvingHumanSubjectsScientificresearchhasproducedsubstantialsocialbenefits.Ithasalsoposedsometroublingethicalquestions.Publicattentionwasdrawntothesequestionsbyreportedabusesofhumansubjectsinbiomedicalexperiments,especiallyduringtheSecondWorldWar.DuringtheNurembergWarCrimeTrials,theNurembergcodewasdraftedasasetofstandardsforjudgingphysiciansandscientistswhohadconductedbiomedicalexperimentsonconcentrationcampprisoners.Thiscodebecametheprototypeofmanylatercodes(1) intendedtoassurethatresearchinvolvinghumansubjectswouldbecarriedoutinanethicalmanner.Thecodesconsistofrules,somegeneral,othersspecific,thatguidetheinvestigatorsorthereviewersofresearchintheirwork.Suchrulesoftenareinadequatetocovercomplexsituations;attimestheycomeintoconflict,andtheyarefrequentlydifficulttointerpretorapply.Broaderethicalprincipleswillprovideabasisonwhichspecificrulesmaybeformulated,criticizedandinterpreted.Threeprinciples,orgeneralprescriptivejudgments,thatarerelevanttoresearchinvolvinghumansubjectsareidentifiedinthisstatement.Otherprinciplesmayalsoberelevant.Thesethreearecomprehensive,however,andarestatedatalevelofgeneralizationthatshouldassistscientists,subjects,reviewersandinterestedcitizenstounderstandtheethicalissuesinherentinresearchinvolvinghumansubjects.Theseprinciplescannotalwaysbeappliedsoastoresolvebeyonddisputeparticularethicalproblems.Theobjectiveistoprovideananalyticalframeworkthatwillguidetheresolutionofethicalproblemsarisingfromresearchinvolvinghumansubjects.Thisstatementconsistsofadistinctionbetweenresearchandpractice,adiscussionofthethreebasicethicalprinciples,andremarksabouttheapplicationoftheseprinciples.[RETURNTOTABLEOFCONTENTS]PartA:BoundariesBetweenPractice&ResearchA.BoundariesBetweenPracticeandResearchItisimportanttodistinguishbetweenbiomedicalandbehavioralresearch,ontheonehand,andthepracticeofacceptedtherapyontheother,inordertoknowwhatactivitiesoughttoundergoreviewfortheprotectionofhumansubjectsofresearch.Thedistinctionbetweenresearchandpracticeisblurredpartlybecausebothoftenoccurtogether(asinresearchdesignedtoevaluateatherapy)andpartlybecausenotabledeparturesfromstandardpracticeareoftencalled"experimental"whentheterms"experimental"and"research"arenotcarefullydefined.Forthemostpart,theterm"practice"referstointerventionsthataredesignedsolelytoenhancethewell-beingofanindividualpatientorclientandthathaveareasonableexpectationofsuccess.Thepurposeofmedicalorbehavioralpracticeistoprovidediagnosis,preventivetreatmentortherapytoparticularindividuals.(2) Bycontrast,theterm"research'designatesanactivitydesignedtotestanhypothesis,permitconclusionstobedrawn,andtherebytodeveloporcontributetogeneralizableknowledge(expressed,forexample,intheories,principles,andstatementsofrelationships).Researchisusuallydescribedinaformalprotocolthatsetsforthanobjectiveandasetofproceduresdesignedtoreachthatobjective.Whenacliniciandepartsinasignificantwayfromstandardoracceptedpractice,theinnovationdoesnot,inandofitself,constituteresearch.Thefactthataprocedureis"experimental,"inthesenseofnew,untestedordifferent,doesnotautomaticallyplaceitinthecategoryofresearch.Radicallynewproceduresofthisdescriptionshould,however,bemadetheobjectofformalresearchatanearlystageinordertodeterminewhethertheyaresafeandeffective.Thus,itistheresponsibilityofmedicalpracticecommittees,forexample,toinsistthatamajorinnovationbeincorporatedintoaformalresearchproject.(3)Researchandpracticemaybecarriedontogetherwhenresearchisdesignedtoevaluatethesafetyandefficacyofatherapy.Thisneednotcauseanyconfusionregardingwhetherornottheactivityrequiresreview;thegeneralruleisthatifthereisanyelementofresearchinanactivity,thatactivityshouldundergoreviewfortheprotectionofhumansubjects.PartB:BasicEthicalPrinciplesB.BasicEthicalPrinciplesTheexpression"basicethicalprinciples"referstothosegeneraljudgmentsthatserveasabasicjustificationforthemanyparticularethicalprescriptionsandevaluationsofhumanactions.Threebasicprinciples,amongthosegenerallyacceptedinourculturaltradition,areparticularlyrelevanttotheethicsofresearchinvolvinghumansubjects:theprinciplesofrespectofpersons,beneficenceandjustice.1.RespectforPersons. --Respectforpersonsincorporatesatleasttwoethicalconvictions:first,thatindividualsshouldbetreatedasautonomousagents,andsecond,thatpersonswithdiminishedautonomyareentitledtoprotection.Theprincipleofrespectforpersonsthusdividesintotwoseparatemoralrequirements:therequirementtoacknowledgeautonomyandtherequirementtoprotectthosewithdiminishedautonomy.Anautonomouspersonisanindividualcapableofdeliberationaboutpersonalgoalsandofactingunderthedirectionofsuchdeliberation.Torespectautonomyistogiveweighttoautonomouspersons'consideredopinionsandchoiceswhilerefrainingfromobstructingtheiractionsunlesstheyareclearlydetrimentaltoothers.Toshowlackofrespectforanautonomousagentistorepudiatethatperson'sconsideredjudgments,todenyanindividualthefreedomtoactonthoseconsideredjudgments,ortowithholdinformationnecessarytomakeaconsideredjudgment,whentherearenocompellingreasonstodoso.However,noteveryhumanbeingiscapableofself-determination.Thecapacityforself-determinationmaturesduringanindividual'slife,andsomeindividualslosethiscapacitywhollyorinpartbecauseofillness,mentaldisability,orcircumstancesthatseverelyrestrictliberty.Respectfortheimmatureandtheincapacitatedmayrequireprotectingthemastheymatureorwhiletheyareincapacitated.Somepersonsareinneedofextensiveprotection,eventothepointofexcludingthemfromactivitieswhichmayharmthem;otherpersonsrequirelittleprotectionbeyondmakingsuretheyundertakeactivitiesfreelyandwithawarenessofpossibleadverseconsequence.Theextentofprotectionaffordedshoulddependupontheriskofharmandthelikelihoodofbenefit.Thejudgmentthatanyindividuallacksautonomyshouldbeperiodicallyreevaluatedandwillvaryindifferentsituations.Inmostcasesofresearchinvolvinghumansubjects,respectforpersonsdemandsthatsubjectsenterintotheresearchvoluntarilyandwithadequateinformation.Insomesituations,however,applicationoftheprincipleisnotobvious.Theinvolvementofprisonersassubjectsofresearchprovidesaninstructiveexample.Ontheonehand,itwouldseemthattheprincipleofrespectforpersonsrequiresthatprisonersnotbedeprivedoftheopportunitytovolunteerforresearch.Ontheotherhand,underprisonconditionstheymaybesubtlycoercedorundulyinfluencedtoengageinresearchactivitiesforwhichtheywouldnototherwisevolunteer.Respectforpersonswouldthendictatethatprisonersbeprotected.Whethertoallowprisonersto"volunteer"orto"protect"thempresentsadilemma.Respectingpersons,inmosthardcases,isoftenamatterofbalancingcompetingclaimsurgedbytheprincipleofrespectitself.2.Beneficence. --Personsaretreatedinanethicalmannernotonlybyrespectingtheirdecisionsandprotectingthemfromharm,butalsobymakingeffortstosecuretheirwell-being.Suchtreatmentfallsundertheprincipleofbeneficence.Theterm"beneficence"isoftenunderstoodtocoveractsofkindnessorcharitythatgobeyondstrictobligation.Inthisdocument,beneficenceisunderstoodinastrongersense,asanobligation.Twogeneralruleshavebeenformulatedascomplementaryexpressionsofbeneficentactionsinthissense: (1) donotharmand (2) maximizepossiblebenefitsandminimizepossibleharms.TheHippocraticmaxim"donoharm"haslongbeenafundamentalprincipleofmedicalethics.ClaudeBernardextendedittotherealmofresearch,sayingthatoneshouldnotinjureonepersonregardlessofthebenefitsthatmightcometoothers.However,evenavoidingharmrequireslearningwhatisharmful;and,intheprocessofobtainingthisinformation,personsmaybeexposedtoriskofharm.Further,theHippocraticOathrequiresphysicianstobenefittheirpatients"accordingtotheirbestjudgment."Learningwhatwillinfactbenefitmayrequireexposingpersonstorisk.Theproblemposedbytheseimperativesistodecidewhenitisjustifiabletoseekcertainbenefitsdespitetherisksinvolved,andwhenthebenefitsshouldbeforegonebecauseoftherisks.Theobligationsofbeneficenceaffectbothindividualinvestigatorsandsocietyatlarge,becausetheyextendbothtoparticularresearchprojectsandtotheentireenterpriseofresearch.Inthecaseofparticularprojects,investigatorsandmembersoftheirinstitutionsareobligedtogiveforethoughttothemaximizationofbenefitsandthereductionofriskthatmightoccurfromtheresearchinvestigation.Inthecaseofscientificresearchingeneral,membersofthelargersocietyareobligedtorecognizethelongertermbenefitsandrisksthatmayresultfromtheimprovementofknowledgeandfromthedevelopmentofnovelmedical,psychotherapeutic,andsocialprocedures.Theprincipleofbeneficenceoftenoccupiesawell-definedjustifyingroleinmanyareasofresearchinvolvinghumansubjects.Anexampleisfoundinresearchinvolvingchildren.Effectivewaysoftreatingchildhooddiseasesandfosteringhealthydevelopmentarebenefitsthatservetojustifyresearchinvolvingchildren--evenwhenindividualresearchsubjectsarenotdirectbeneficiaries.Researchalsomakesitpossibletoavoidtheharmthatmayresultfromtheapplicationofpreviouslyacceptedroutinepracticesthatoncloserinvestigationturnouttobedangerous.Buttheroleoftheprincipleofbeneficenceisnotalwayssounambiguous.Adifficultethicalproblemremains,forexample,aboutresearchthatpresentsmorethanminimalriskwithoutimmediateprospectofdirectbenefittothechildreninvolved.Somehavearguedthatsuchresearchisinadmissible,whileothershavepointedoutthatthislimitwouldruleoutmuchresearchpromisinggreatbenefittochildreninthefuture.Hereagain,aswithallhardcases,thedifferentclaimscoveredbytheprincipleofbeneficencemaycomeintoconflictandforcedifficultchoices.3.Justice. --Whooughttoreceivethebenefitsofresearchandbearitsburdens?Thisisaquestionofjustice,inthesenseof"fairnessindistribution"or"whatisdeserved."Aninjusticeoccurswhensomebenefittowhichapersonisentitledisdeniedwithoutgoodreasonorwhensomeburdenisimposedunduly.Anotherwayofconceivingtheprincipleofjusticeisthatequalsoughttobetreatedequally.However,thisstatementrequiresexplication.Whoisequalandwhoisunequal?Whatconsiderationsjustifydeparturefromequaldistribution?Almostallcommentatorsallowthatdistinctionsbasedonexperience,age,deprivation,competence,meritandpositiondosometimesconstitutecriteriajustifyingdifferentialtreatmentforcertainpurposes.Itisnecessary,then,toexplaininwhatrespectspeopleshouldbetreatedequally.Thereareseveralwidelyacceptedformulationsofjustwaystodistributeburdensandbenefits.Eachformulationmentionssomerelevantpropertyonthebasisofwhichburdensandbenefitsshouldbedistributed.Theseformulationsare (1) toeachpersonanequalshare, (2) toeachpersonaccordingtoindividualneed, (3) toeachpersonaccordingtoindividualeffort, (4) toeachpersonaccordingtosocietalcontribution,and (5) toeachpersonaccordingtomerit.Questionsofjusticehavelongbeenassociatedwithsocialpracticessuchaspunishment,taxationandpoliticalrepresentation.Untilrecentlythesequestionshavenotgenerallybeenassociatedwithscientificresearch.However,theyareforeshadowedevenintheearliestreflectionsontheethicsofresearchinvolvinghumansubjects.Forexample,duringthe19thandearly20thcenturiestheburdensofservingasresearchsubjectsfelllargelyuponpoorwardpatients,whilethebenefitsofimprovedmedicalcareflowedprimarilytoprivatepatients.Subsequently,theexploitationofunwillingprisonersasresearchsubjectsinNaziconcentrationcampswascondemnedasaparticularlyflagrantinjustice.Inthiscountry,inthe1940's,theTuskegeesyphilisstudyuseddisadvantaged,ruralblackmentostudytheuntreatedcourseofadiseasethatisbynomeansconfinedtothatpopulation.Thesesubjectsweredeprivedofdemonstrablyeffectivetreatmentinordernottointerrupttheproject,longaftersuchtreatmentbecamegenerallyavailable.Againstthishistoricalbackground,itcanbeseenhowconceptionsofjusticearerelevanttoresearchinvolvinghumansubjects.Forexample,theselectionofresearchsubjectsneedstobescrutinizedinordertodeterminewhethersomeclasses(e.g.,welfarepatients,particularracialandethnicminorities,orpersonsconfinedtoinstitutions)arebeingsystematicallyselectedsimplybecauseoftheireasyavailability,theircompromisedposition,ortheirmanipulability,ratherthanforreasonsdirectlyrelatedtotheproblembeingstudied.Finally,wheneverresearchsupportedbypublicfundsleadstothedevelopmentoftherapeuticdevicesandprocedures,justicedemandsboththatthesenotprovideadvantagesonlytothosewhocanaffordthemandthatsuchresearchshouldnotundulyinvolvepersonsfromgroupsunlikelytobeamongthebeneficiariesofsubsequentapplicationsoftheresearch.PartC:ApplicationsC.ApplicationsApplicationsofthegeneralprinciplestotheconductofresearchleadstoconsiderationofthefollowingrequirements:informedconsent,risk/benefitassessment,andtheselectionofsubjectsofresearch.1.InformedConsent. --Respectforpersonsrequiresthatsubjects,tothedegreethattheyarecapable,begiventheopportunitytochoosewhatshallorshallnothappentothem.Thisopportunityisprovidedwhenadequatestandardsforinformedconsentaresatisfied.Whiletheimportanceofinformedconsentisunquestioned,controversyprevailsoverthenatureandpossibilityofaninformedconsent.Nonetheless,thereiswidespreadagreementthattheconsentprocesscanbeanalyzedascontainingthreeelements:information,comprehensionandvoluntariness.Information. Mostcodesofresearchestablishspecificitemsfordisclosureintendedtoassurethatsubjectsaregivensufficientinformation.Theseitemsgenerallyinclude:theresearchprocedure,theirpurposes,risksandanticipatedbenefits,alternativeprocedures(wheretherapyisinvolved),andastatementofferingthesubjecttheopportunitytoaskquestionsandtowithdrawatanytimefromtheresearch.Additionalitemshavebeenproposed,includinghowsubjectsareselected,thepersonresponsiblefortheresearch,etc.However,asimplelistingofitemsdoesnotanswerthequestionofwhatthestandardshouldbeforjudginghowmuchandwhatsortofinformationshouldbeprovided.Onestandardfrequentlyinvokedinmedicalpractice,namelytheinformationcommonlyprovidedbypractitionersinthefieldorinthelocale,isinadequatesinceresearchtakesplacepreciselywhenacommonunderstandingdoesnotexist.Anotherstandard,currentlypopularinmalpracticelaw,requiresthepractitionertorevealtheinformationthatreasonablepersonswouldwishtoknowinordertomakeadecisionregardingtheircare.This,too,seemsinsufficientsincetheresearchsubject,beinginessenceavolunteer,maywishtoknowconsiderablymoreaboutrisksgratuitouslyundertakenthandopatientswhodeliverthemselvesintothehandofaclinicianforneededcare.Itmaybethatastandardof"thereasonablevolunteer"shouldbeproposed:theextentandnatureofinformationshouldbesuchthatpersons,knowingthattheprocedureisneithernecessaryfortheircarenorperhapsfullyunderstood,candecidewhethertheywishtoparticipateinthefurtheringofknowledge.Evenwhensomedirectbenefittothemisanticipated,thesubjectsshouldunderstandclearlytherangeofriskandthevoluntarynatureofparticipation.Aspecialproblemofconsentariseswhereinformingsubjectsofsomepertinentaspectoftheresearchislikelytoimpairthevalidityoftheresearch.Inmanycases,itissufficienttoindicatetosubjectsthattheyarebeinginvitedtoparticipateinresearchofwhichsomefeatureswillnotberevealeduntiltheresearchisconcluded.Inallcasesofresearchinvolvingincompletedisclosure,suchresearchisjustifiedonlyifitisclearthat (1) incompletedisclosureistrulynecessarytoaccomplishthegoalsoftheresearch, (2) therearenoundisclosedriskstosubjectsthataremorethanminimal,and (3)thereisanadequateplanfordebriefingsubjects,whenappropriate,andfordisseminationofresearchresultstothem.Informationaboutrisksshouldneverbewithheldforthepurposeofelicitingthecooperationofsubjects,andtruthfulanswersshouldalwaysbegiventodirectquestionsabouttheresearch.Careshouldbetakentodistinguishcasesinwhichdisclosurewoulddestroyorinvalidatetheresearchfromcasesinwhichdisclosurewouldsimplyinconveniencetheinvestigator.Comprehension. Themannerandcontextinwhichinformationisconveyedisasimportantastheinformationitself.Forexample,presentinginformationinadisorganizedandrapidfashion,allowingtoolittletimeforconsiderationorcurtailingopportunitiesforquestioning,allmayadverselyaffectasubject'sabilitytomakeaninformedchoice.Becausethesubject'sabilitytounderstandisafunctionofintelligence,rationality,maturityandlanguage,itisnecessarytoadaptthepresentationoftheinformationtothesubject'scapacities.Investigatorsareresponsibleforascertainingthatthesubjecthascomprehendedtheinformation.Whilethereisalwaysanobligationtoascertainthattheinformationaboutrisktosubjectsiscompleteandadequatelycomprehended,whentherisksaremoreserious,thatobligationincreases.Onoccasion,itmaybesuitabletogivesomeoralorwrittentestsofcomprehension.Specialprovisionmayneedtobemadewhencomprehensionisseverelylimited--forexample,byconditionsofimmaturityormentaldisability.Eachclassofsubjectsthatonemightconsiderasincompetent(e.g.,infantsandyoungchildren,mentallydisablepatients,theterminallyillandthecomatose)shouldbeconsideredonitsownterms.Evenforthesepersons,however,respectrequiresgivingthemtheopportunitytochoosetotheextenttheyareable,whetherornottoparticipateinresearch.Theobjectionsofthesesubjectstoinvolvementshouldbehonored,unlesstheresearchentailsprovidingthematherapyunavailableelsewhere.Respectforpersonsalsorequiresseekingthepermissionofotherpartiesinordertoprotectthesubjectsfromharm.Suchpersonsarethusrespectedbothbyacknowledgingtheirownwishesandbytheuseofthirdpartiestoprotectthemfromharm.Thethirdpartieschosenshouldbethosewhoaremostlikelytounderstandtheincompetentsubject'ssituationandtoactinthatperson'sbestinterest.Thepersonauthorizedtoactonbehalfofthesubjectshouldbegivenanopportunitytoobservetheresearchasitproceedsinordertobeabletowithdrawthesubjectfromtheresearch,ifsuchactionappearsinthesubject'sbestinterest.Voluntariness. Anagreementtoparticipateinresearchconstitutesavalidconsentonlyifvoluntarilygiven.Thiselementofinformedconsentrequiresconditionsfreeofcoercionandundueinfluence.Coercionoccurswhenanovertthreatofharmisintentionallypresentedbyonepersontoanotherinordertoobtaincompliance.Undueinfluence,bycontrast,occursthroughanofferofanexcessive,unwarranted,inappropriateorimproperrewardorotherovertureinordertoobtaincompliance.Also,inducementsthatwouldordinarilybeacceptablemaybecomeundueinfluencesifthesubjectisespeciallyvulnerable.Unjustifiablepressuresusuallyoccurwhenpersonsinpositionsofauthorityorcommandinginfluence--especiallywherepossiblesanctionsareinvolved--urgeacourseofactionforasubject.Acontinuumofsuchinfluencingfactorsexists,however,anditisimpossibletostatepreciselywherejustifiablepersuasionendsandundueinfluencebegins.Butundueinfluencewouldincludeactionssuchasmanipulatingaperson'schoicethroughthecontrollinginfluenceofacloserelativeandthreateningtowithdrawhealthservicestowhichanindividualwouldotherwisebeentitle.2.AssessmentofRisksandBenefits. --Theassessmentofrisksandbenefitsrequiresacarefularrayalofrelevantdata,including,insomecases,alternativewaysofobtainingthebenefitssoughtintheresearch.Thus,theassessmentpresentsbothanopportunityandaresponsibilitytogathersystematicandcomprehensiveinformationaboutproposedresearch.Fortheinvestigator,itisameanstoexaminewhethertheproposedresearchisproperlydesigned.Forareviewcommittee,itisamethodfordeterminingwhethertherisksthatwillbepresentedtosubjectsarejustified.Forprospectivesubjects,theassessmentwillassistthedeterminationwhetherornottoparticipate.TheNatureandScopeofRisksandBenefits. Therequirementthatresearchbejustifiedonthebasisofafavorablerisk/benefitassessmentbearsacloserelationtotheprincipleofbeneficence,justasthemoralrequirementthatinformedconsentbeobtainedisderivedprimarilyfromtheprincipleofrespectforpersons.Theterm"risk"referstoapossibilitythatharmmayoccur.However,whenexpressionssuchas"smallrisk"or"highrisk"areused,theyusuallyrefer(oftenambiguously)bothtothechance(probability)ofexperiencingaharmandtheseverity(magnitude)oftheenvisionedharm.Theterm"benefit"isusedintheresearchcontexttorefertosomethingofpositivevaluerelatedtohealthorwelfare.Unlike,"risk,""benefit"isnotatermthatexpressesprobabilities.Riskisproperlycontrastedtoprobabilityofbenefits,andbenefitsareproperlycontrastedwithharmsratherthanrisksofharm.Accordingly,so-calledrisk/benefitassessmentsareconcernedwiththeprobabilitiesandmagnitudesofpossibleharmandanticipatedbenefits.Manykindsofpossibleharmsandbenefitsneedtobetakenintoaccount.Thereare,forexample,risksofpsychologicalharm,physicalharm,legalharm,socialharmandeconomicharmandthecorrespondingbenefits.Whilethemostlikelytypesofharmstoresearchsubjectsarethoseofpsychologicalorphysicalpainorinjury,otherpossiblekindsshouldnotbeoverlooked.