COMMENTARY

Parent-Investigators

A Dilemma

David B. Resnik, JD, PhD

RECENT NEWS STORY REPORTED THAT SOME PSY

-chologistsandneuroscientistshaveusedtheirchil-drenasstudyparticipants.

Inonecase,aneurol-ogyprofessorhad3ofhischildrenundergoabrainscan with magnetic resonance imaging. In a language de-velopmentstudy,apsychologistvideotaped70%ofhisson’swaking hours for the first 3 years of his life.

Other inves-tigators have also conducted research on their own chil-dren, including Edward Jenner,

who tested his smallpoxvaccineonhis11-month-oldson;JonasSalk,whotestedhispolio vaccine on his own children; and psychologist JeanPiaget, who constructed his theories of child developmentbased on studies of his own children.Shouldinvestigatorsconductresearchontheirownchil-dren? Federal research regulations

do not address this is-sue and regulatory agencies, such as the Office of HumanResearchProtections,

havenopoliciesdealingwithit.Otherresearch guidelines, such as the Nuremberg Code

and theHelsinkiDeclaration,

alsodonotaddresstheissueofparent-investigators.Whenlawsandprofessionalcodesdonotpro-videguidanceconcerninghumanresearch,investigatorsandcommittees that oversee research, such as institutional re-view boards (IRBs), must rely on their ethical judgment todecide.

Parental Conflict of Interest

A strong ethical argument can be made that parents shouldnot conduct research on their own children because itcreates a conflict of interest that threatens the parent-childrelationship and the integrity of science. Parents areresponsible for deciding whether their children should par-ticipate in research studies and they have moral obligationsto support them and protect them from harm.

When par-ents enroll their own children in a study, their interest inpromoting their own research could undermine the obliga-tion to protect their children from harm. For example,parent-investigators might bend inclusion criteria for astudy to ensure that their child qualifies, fail to inform theIRB about adverse events or risks, or ignore an older child’sobjections to participating in a study.Clinical investigators whose children participate intheir research might compromise the integrity of thestudy to benefit their children. For example, if theirchildren are enrolled in a double-blind randomizedclinical trial, parent-investigators might attempt toascertain whether the children are receiving the experi-mental treatment, interfere with the randomization, orchange the dose of a medication the children are receiv-ing. Clinical investigators who act in this manner mightbe honoring their obligations as the children’s physician,but they would be violating legal and ethical rules for theconduct of research. Parent-investigators, in this case,would also succumb to the therapeutic misconceptionthat research studies are specifically designed to benefitresearch participants.

The temptation to manipulatea research protocol to benefit one’s child may also existin social/behavioral research, but it would occur lessfrequently than in clinical research because social/ behavioral research usually provides fewer direct benefitsfor participants.

The conflict of interest that occurs when parents enrolltheir children in their own research is similar in someways to the conflict that arises when physicians providemedical treatment for their own children, because in bothsituations, professional judgment could be affected bypersonal relationships.

One important differencebetween physicians who treat their own children andinvestigators who study their own children is that thegoal of medical treatment is to benefit patients, whereasthe goal of scientific research is to generate knowledgethat can benefit others.

Conducting research on one’sown child is therefore potentially more hazardous thantreating one’s own child because scientific research is notpatient centered, and some types of pediatric researchimpose risks on human participants without the prospectof direct benefit.

Clinical vs Social Behavioral Research

Somemightarguethattheethicalargumentagainstpartici-pation of children in their parents’ research applies only toclinical studies, not to social behavioral research becausethe risks in social behavioral research are generally muchlower in severity than the risks in clinical research. For ex-

Author Affiliation:

National Institute of Environmental Health Sciences, NationalInstitutes of Health, Research Triangle Park, North Carolina.

Corresponding Author:

David B. Resnick, JD, PhD, NIEHS/NIH, Box 12233 MailDrop CR 03, Research Triangle Park, NC 27709 (resnikd@niehs.nih.gov).

(Reprinted) JAMA,

May 27, 2009—Vol 301, No. 20

2159

ample,nogreatharmislikelytoresultwhenparentsenrolltheir children in studies in which parent-investigators ob-serve their children’s social interactions with other chil-drenorasktheirchildrenquestionsabouttheirknowledgeof television commercials for a survey. While some clinicalstudies involve only low-risk procedures or methods, suchas collection of biological specimens for biorepositories orphysical examinations, many clinical studies involve riskyprocedures or methods, such as administration of chemo-therapy drugs or experimental surgery.Although the risks of social/behavioral research are gen-erally lower in severity than the risks of clinical research,social/behavioral research is not risk free. Surveys some-times include questions that are embarrassing or emotion-ally disturbing. Social/behavioral investigators also dealwith sensitive, private information.

Parents may engage inlong-term studies of an intrusive nature, such as theproject in which the parent videotaped much of his son’sfirst 3 years of life. Moreover, participation of children intheir parents’ social/behavioral study can strain the parent-child relationship. For example, children may feel somepressure to participate in a study conducted by theirmother or father. The children may feel that they cannotrefuse to participate or stop participating at any time.Although this pressure also exists when parents who arenot the investigator want their child to participate in astudy, the pressure can be more intense when parents arethe investigators because of the parents’ interest in meetingenrollment goals.Even though the ethical arguments against childrenparticipating in their parents’ social/behavioral researchare weaker than the arguments against participation inclinical research, there are still some ethical problemswith children participating in their parents’ social/ behavioral research, and these ethical concerns are suffi-cient to warrant special attention by oversight commit-tees. Investigators who are planning to enroll their ownchildren in social/behavioral research should be able toexplain to the IRB why their children should be enrolledand how they plan to protect their own children and theintegrity of the science. While there should be a strong presumption against par-entsconductingclinicalresearchontheirownchildren,therearesomeexceptionstothisrule.First,ifparent-investigatorsare not able to determine whether their children are partici-pating in the study, the children’s participation would notthreaten the parent-child relationship or the integrity of thescience. For example, suppose a couple is divorced and themother is deciding whether her child should participate in aresearchstudythathappenstobeconductedbythechild’sfa-ther.Ifthemotherdoesnotinformthefatheraboutthechild’sparticipation,thefather’sconsentisnotrequired,andthefa-ther does not interact with research participants or have ac-cess to data or samples that personally identify research par-ticipants, it could be acceptable for the child to participate.Second, if a study could provide a child with significantmedicalbenefits,enrollingthechildcouldbejustifiable.Forexample, suppose that a mother, who is a pediatric oncol-ogy investigator, has a child who is dying of a rare form of childhoodleukemia.Suppose,also,thatthemotheriscon-ducting the only clinical trial involving a medication witha significant chance of curing the leukemia and that par-ticipationinthetrialisthebestchanceofcuringthechild’scancerorsignificantlyprolongingthechild’slife.Ifthechilddoesnotparticipateintheclinicaltrial,heorshehasahighprobability of dying within 3 months, but if the child par-ticipatesinthestudy,heorshehasachanceofliving1yearormore.Onecouldarguethatenrollmentinthisstudywouldbethebestoptionforthechild—parentalinvolvementnot-withstanding.Although enrollment of the child in the parent’s studymight be justified in this case, it is important to be mindfulof the potential for therapeutic misconception. The par-entsshouldunderstandthatalthoughthestudymightben-efit the child, the main purpose of the study is to developgeneralizable knowledge, not to help the child. The par-entsshouldunderstandthatthechildmaynotbenefitfromparticipation and may experience serious harms. Althoughone would think that the investigator would not fall preytothetherapeuticmisconception,evidenceshowsthatthisis often not the case.

Also, additional oversight steps maybe needed to protect the integrity of the research and thewelfare of the participant. For example, the IRB and datamonitoring board should be informed that the researcheris enrolling his or her own child and the noninvestigatingparent should also sign the consent, assuming he or she isavailable.IRBs should formulate policies addressing the partici-pation of children in their parents’ research. The policiesshould state when investigators may (and may not) enrolltheir own children in research, when additional monitor-ing is needed, and address additional protections for chil-dren, such as requiring both parents to sign the consentdocument. IRBs should inform investigators about thesepolicies through seminars, memos, and other educationaland training methods.It may be difficult for IRBs to enforce the policies theydevelop because IRBs usually do not know the identity of human participants. The documents the IRB reviews, suchasprotocols,informedconsentforms,questionnaires,amend-ments,adverseeventreports,andcontinuingreviewforms,donotmentionspecificresearchparticipants.AlthoughtheIRBhastheauthoritytoexaminesignedconsentformsandotherresearchrecords,mostIRBsdonotunlesstheyareau-diting a study.

Since IRBs will usually not know whetherinvestigatorshavetheirownchildrenparticipating,IRBswillneed to trust investigators to abide by the policies they set.Parents enrolling their own children in their own researchmust respond appropriately to the dilemma of protectingtheir children while simultaneously protecting the integ-

COMMENTARY

2160

JAMA,

May 27, 2009—Vol 301, No. 20

(Reprinted)