Equipoise and Randomized Controlled Testing Patrick Beach
Missouri State University Philosophy Department 901 S. National Ave.
Springfield, MO 65897
Tel.: +1 208-870-5240
Submitted: February 15, 2016
Word Count: 991
Equipoise and Randomized Controlled Testing
[Author Blinded] February 15,2016
CharlesFriedholdsthatphysicianshavea“dutyofpersonalcare”andthattheymust meeta“demandforundividedloyaltytotheinterestsofthepatient”(Fried,1974,p.148). Physicians have an absolute obligation to do what is best for their patients. They cannot dividetheirloyaltybetweentheirpatients,futurepatients,andsoon.
Thus,randomizedcontrolledtesting(RCT)presentsaproblemforphysicianswhoalso actasresearchers.Itseemsthatresearchers(physicians)cannotpermissiblyassignsubjects (their patients) to either test or control arms of an RCT unless they know that they will therebyreceivethebestavailabletreatment.
SupposethatDoctorisconductingaPhaseIIItrialofanewcancertreatment(Therapy A),whichisbeingtestedagainsttheprovenstandard(TherapyB).Theargumentagainst themoralpermissibilityofthisstudygoeslikethis:
1.Doctor ought to give Patient the best treatment shecan.
2.DoctordoesnotknowwhetherTherapyAisatleastasgoodasTherapyB.
3.Thus,DoctorisnotwarrantedingivingeitherTherapyAorTherapyBtoPatient.
4.Thus, Doctor ought to give neither Therapy A nor Therapy B toPatient.
This conclusion can be avoided. What’s forbidden is for Doctor to know thatTherapy B is worse than Therapy A and then to administer Therapy B anyway. If Doctor does not knowthatonetherapyisworse,thenit’spermissibletogiveeither.
The state of not knowing which treatment arm has more therapeutic value is called experimentalequipoise.WalterChiongsaysthat“Equipoiserequiresgenuineuncertainty about the relative therapeutic merits of the interventions studied” (Chiong, 2006, p. 37). DavidChamberssaysthat“Equipoiseisthepositionthatsubjectsshouldnotberandomized totreatmentconditionswhereitisreasonabletobelievebetteralternativesexist”(Chambers, 2011, p.133). Thisinsightweakensthefirstpremiseofthepreviousargument.Itoughttobeamendedto:
1. Doctor ought not to knowingly give Patient less than the best treatment she can.
At the start of the RCT, Doctor does not know which arm offers a better treatment.Thus, she is morally permitted to let chance decide which treatment Patient gets. As theresults comein,however,shemaycometohaveaninformedopinionthatTherapyB,forexample, islesseffective.Inthatcase,sheoughtnottoallowPatientthebadluckofbeingrandomly selected for TherapyB.
AsDonMarquispointsout(Marquis,1983),ifDoctordiscontinuesthestudyonpreliminaryresults,TherapyAhasnochanceofgettingtomarket,sincethemedicalcommunity willneveracceptTherapyAmerelyonpreliminaryresults.Butifresearcherscontinuetouse ofthelesseffectivearm(inordertocompletetheRCT)thentheyareknowinglysubjecting patientstothebadluckofgettingthesuboptimalarm.
Benjamin Freedman acknowledges that a study has to stop as soon as Doctor knows whichtreatmentisbest,butuntilthentheresearchcancontinue(Freedman,1987).Solong as Doctor’s opinions amount to something less than conviction, the equipoise ofmedical experts in the field is enough, according to Freedman. Thus, Freedman distinguishes between researcher equipoise and clinicalequipoise.
On the other hand, Doctor’s obligation to Patient (who may be exposed to harm in thetrial)overridesherobligation(asresearcher)tothirdparties,includingFuturePatient.
Thus,oncesheisoutofresearcherequipoise,Doctoroughttostoprandomizedtreatmentof herpatients,nomatterwhatthestateofclinicalequipoise.Researcherandclinicalequipoise areindividuallynecessaryrequirementstoconductrandomizedtesting.So,wearebackwith Marquis’sproblem:preliminaryresultsoughttohalttesting.Onthisview,inanyconflict ofobligationbetweenPatientandFuturePatient,DoctormustsidewithPatient.
WinstonChiongrejectsthisabsolutefidelitytoPatient,andarguesforalimitedobligation toPatientthatleavesroomforsomeobligationtoFuturePatient.Chiongrecognizessome contextsinwhichitismorallypermissibletoprovidecarethatislessthanthebest,butwhich stillmeetsastandardofbeinggoodenough(Chiong,2006,p.44).Inteachinghospitals,for example,proceduresaredonebystudentsunderthesupervisionofinstructorswhocould perform the procedures better. Patients sometimes get worse care than they might have, butthereisnomoralwrongdone,solongasthecareisgoodenough.Voluntaryinformed consentrequiresthatpatientsbeinformedthattheymayreceivecarethatismerelygood enoughandrespectforautonomyrequiresthatpatientsbeabletogivesomeweighttothe goodoffuturepatientsattheexpenseoftheirgood.
Like Chiong, I endorse a limited range of suboptimal treatment. I develop a notion of limitedequipoiseinwhichsubjectsshouldnotberandomizedtotreatmentarmswhereitis reasonabletobelieveonearmisnoteffectiveenough.SolongasTherapyAandTherapyB arebelievedtobesomewhatcloseintherapeuticbenefit,RCTcancontinue,evenifDoctor knows that one treatment arm is having better results. Rather than Chiong’s Kantian underpinnings,IarguethatPatientcanbeasatisficeraboutherowntreatment.
My view faces the following challenges (which I believe I can meet):
•Itobligatesresearcherstoreal-timemonitoringofwhethercareis“goodenough”.
•It obligates researchers to clarify to subjects/patients the operative notion of being goodenoughsuchthatpatientscangivevoluntaryinformedconsent.
•Inaddition,thismayextendthenotionofsatisficing.HerbertSimon’snotionwasthat thereisanadvantagetoacceptingwhat’s“goodenough”onlywhenyoudonotknow what is better. In research applications of limited equipoise, that conditionmay not bemet.
•Finally,non-optimizingmoralpermissionshavefacedanumberofchallengesthatought tobeaddressedinthecontextofconsideringexperimentalequipoise.
References
Chambers, D. W. (2011). Confusions in the equipoise concept and the alternative of fully informed overlapping rational decisions. Medicine, Health Care and Philosophy, 14 (2), 133–142.
Chiong, W. (2006). The real problem with equipoise. American Journal of Bioethics, 6(4), 37–47.
Freedman,B.(1987).Equipoiseandtheethicsofclinicalresearch.NewEnglandJournalof Medicine,317(3),141–145.
Fried, C. (1974). Medical experimentation: personal integrity and social policy. New York: American Elsevier Publishing Co., Inc.
Marquis,D.(1983).Leavingtherapytochance.HastingsCenterReport,13(4),40–47.doi:10.
2307/3561721