Disclosure of the Right of Research Participants toReceive Research ResultsAn Analysis of Consent Forms in the Children’s Oncology Group

Conrad V. Fernandez, M.D.1,2

Eric Kodish, M.D.3

Shaureen Taweel, R.N.4

Susan Shurin, M.D.2

Charles Weijer, M.D., Ph.D.5

1 Department of Pediatrics, IWK Health Centre,Dalhousie University, Halifax, Nova Scotia, Canada.

2 Department of Pediatrics, Dalhousie University,Halifax, Nova Scotia, Canada.

3 Rainbow Center for Pediatric Ethics, RainbowBabies and Children’s Hospital, Cleveland, Ohio.

4 Department of Nursing, IWK Health Centre, Hali-fax, Nova Scotia, Canada.

5 Department of Bioethics, Dalhousie University,Halifax, Nova Scotia, Canada.

Supported by a Children’s Oncology Group Chair-man’s Grant of services-in-kind and a grant fromthe ARG McNaught Fund, IWK Health Center Re-search Services, IWK Health Center, Halifax, NovaScotia.

The authors thank Marcelo Alvarez of the Chil-dren’s Oncology Group Operations staff for hiscontributions in contacting principal investigatorsand clinical research associates.

Address for reprints: Conrad V Fernandez, M.D.,IWK Health Centre, Box 3070, 5850 UniversityAvenue, Halifax, Nova Scotia B3J 3G9, Canada;Fax: (902) 470-7216; E-mail: conrad.fernandez@iwk.nshealth.ca

Received November 19, 2002; revision receivedJanuary 19, 2003; accepted February 17, 2003.

BACKGROUND. The offer of return of research results to study participants has

many potential benefits. The current study examined the offer of return of research

results by analyzing consent forms from 2 acute lymphoblastic leukemia studies of

the 235 institutional members of the Children’s Oncology Group.

METHODS. Institutional review board (IRB)-approved consent forms from 2 stan-

dard-risk acute lymphoblastic leukemia studies (Children’s Cancer Group [CCG]

1991 and Pediatric Oncology Group [POG] 9407) were analyzed independently by

2 reviewers.

RESULTS. The authors received replies from 202 of the 235 institutions that were

contacted (85%). One hundred eighty-one institutions had CCG 1991 (n � 96) or

POG 9905 (n � 85) protocols that were approved by an IRB. Most institutions

provided contact information for the principal investigator (n � 175; 97%) and a

member of the institution’s research services office (n � 154; 85%). Only 5 (2.8%)

institutions provided an indication of a participant’s right to receive a summary of

research results; most of these institutions provided details on how (n � 5) or when

(n � 5) this was to occur. All of these institutions (n � 162; 89.5%) provided a

specific statement offering new information that might affect a participant’s deci-

sion to continue to participate in a study. Only 2 institutional consent forms

offered participants the option to receive research results, and only 10 (5.5%)

consent forms contained an unambiguous, specific statement offering to provide

new information after the study was closed.

CONCLUSIONS. Few institutional review board–approved consent forms explicitly

indicate the right of research recipients to receive a summary of the results of the

research in which they have participated. Cancer 2003;97:2904 –9.

© 2003 American Cancer Society.

DOI 10.1002/cncr.11391

KEYWORDS: research results, disclosure, participants, research ethics, clinicaltrials, informed consent

Human clinical research is founded on the premise that carefulclinical observation, clinical trials, and/or biologic research will

enhance well-being, survival, and long-term quality of life, or con-tribute to an enhanced understanding of the biology of disease. Acentral tenet of human clinical research is respect for the researchparticipant.1–5 Respect for individuals requires that the choices ofthose who are capable of making decisions for themselves be ac-corded high regard, and this is reflected by careful solicitation ofinformed consent and promises of confidentiality.

The Children’s Oncology Group (COG) is one of the largest multi-institutional pediatric clinical trials groups in the world. It enrolls

2904

© 2003 American Cancer Society

more than 4000 children per year on treatment, trans-lational, and biologic studies. The COG adheres to thedetailed requirements of the consent process man-dated in the Code of Federal Regulations, Protection ofHuman Subjects (Common Rule) and by the Depart-ment of Health and Human Services.2,4 No guidance isgiven in the Common Rule (or, to our knowledge, inany other set of international guidelines) regardinghow the principle of respect for persons can be upheldafter a study has been completed.

Our group and others have argued that fulfillingthe ideal of respect for participants obligates the re-searcher to offer to provide research results to allparticipants upon completion of a study.6 – 8 Offeringresults has several potential benefits, including 1) con-firmation of the central nature of the participant’s rolein clinical research; 2) reduction of the potential for aresearch participant to perceive that he or she hasbeen exploited; 3) distribution of information thatcould enhance quality of life or lead to interventionsthat may decrease the risk of future harm;9 4) distri-bution of more accurate information than might beavailable in the lay press;10,11 and 5) dissemination ofknowledge that could improve the public’s perceptionof clinical research.12–14

Disclosure of research results, however, some-times is associated with risks to research participantsor their families. Potential drawbacks include 1) dis-tress for those who did not benefit or were harmed bythe research; 2) rekindling of old memories and emo-tions, especially in the setting of serious illness; 3)anguish for family and friends if the research partici-pant has died; 4) discrimination in obtaining employ-ment or insurance for individuals identified as havingdeveloped (or being at high risk of developing) com-plications; and 5) survivor guilt for participants as-signed to a superior treatment arm of a randomizedclinical trial. These risks emphasize that the return ofresearch results is complex and must be done in asensitive manner and with the consent of the partici-pant. Patient advocate groups and research ethicistsgenerally believe that study results should be offeredto participants or their survivors, who then can choosewhether or not to receive them.

To optimize the benefit for those who choose toreceive research results, a summary of results shouldbe offered to participants within the context of well-designed programs to facilitate the return of results,and it should be recognized that careful preparationand follow-up are needed; to our knowledge, no em-piric data regarding the prevalence of these programshas been reported. In addition, the needs and atti-tudes of participants and researchers with respect tothe return of research results has received little atten-

tion. We therefore used a systematic approach to ex-amine current activity and the requirements for estab-lishing formal programs. In the current study, usingthe consent form as a surrogate marker, we deter-mined the frequency with which research results wereoffered and surveyed the methods that were used tooffer them.

We report our examination of the offer to returnresearch results to study participants, as stated in cur-rent consent forms approved by an institutional re-view board (IRB) or a research ethics board, for twoacute lymphoblastic leukemia protocols that are openthrough the COG.

MATERIALS AND METHODSThe current study was approved by the IWK HealthCentre Research Ethics Board. Consent to use theconsent forms of a given institution was assumedupon reply of the institutional principal investigator orclinical research associate.

The Children’s Oncology Group is a clinical trialsgroup formed by the amalgamation of the Children’sCancer Group (CCG), the Pediatric Oncology Group(POG), the National Wilms Tumor Study Group, andthe Intergroup Rhabdomyosarcoma Study Group.Membership in the COG includes 236 institutions inthe United States, Canada, Australia, and Switzerland.Each institution has an identified principal investiga-tor. The COG disseminates research protocols, includ-ing a sample consent form template for each study, tomember institutions. Each member institution sub-mits the study and consent form to its local IRB, butboth the institution and the IRB have the prerogativeto change the consent document as they see fit. Manyinstitutions use the template verbatim in constructingthe local institutional consent form.

Each of the COG principal investigators was sentan informational e-mail with a request to return acopy of a consent form for the currently open COGacute lymphoblastic leukemia protocol (CCG 1991 orPOG 9905) at the investigator’s institution. The con-sent form had to have been approved by the local IRBor research ethics board (REB).

If no response was received by 3 weeks after thisinitial mailing, an e-mail was sent to the lead clinicalresearch associate (CRA) at the institution. Two re-minders were sent to the CRA and contained a requestthat the consent form be sent by e-mail, fax, or regularmail. Consent forms were collected for acute lympho-blastic leukemia protocols CCG 1991 and POG 9905.

For the CCG 1991 protocol, the CCG instituted a2-step consent process, in which consent for the studywas divided into consent for collection of data on astandardized induction (the first 28 days) and consent

Disclosure of Research Results/Fernandez et al. 2905

for collection of data and randomization for the re-mainder of the study. A number of institutions werenot participating in either CCG or POG leukemia trialsand had local protocols. These protocols were notcollected for analysis, but the participation of the in-stitutions in an alternate protocol was noted.

A questionnaire was designed to abstract informa-tion from consent forms. The questionnaire examinedthe length of consent forms, contact information forresearchers and research services, indication of rightsto receive research results, indication of a commit-ment to return research results upon discovery of newinformation, and indication of how research resultswould be returned to research participants. Data ab-straction and coding were completed independentlyby two of the investigators (C. F. and S. T.). In theevent of a disagreement in coding, the consent formsin question would be reviewed. If this review did notyield a resolution, the forms in question would bereviewed by a third investigator.

One of our goals was to determine whether a timeframe for receiving a summary of results was availableto research participants. The CCG 1991 study wasopened in June 2000, and completion of accrual isestimated to occur in June 2004. The statistical sectionhas an estimated follow-up of 3 years. Therefore, asummary of results can be expected to be available in2007 or 2008. Similarly, POG 9905 was opened in April2000. The statistical section has an estimated accrualperiod of up to 3.1 years and a follow-up of up to 4years, and so a summary of results for this study canbe expected to be available in 2007 or 2008.

Completed questionnaires were entered into EpiInfo software (Version 6.02b; Centers for Disease Con-trol and Prevention, Atlanta, GA), and data were ana-lyzed using descriptive techniques. Variables werepreselected for examination with chi-square analysis.These variables included CCG versus POG protocoland provision of REB or IRB contact versus no provi-sion of ethics board contact.

RESULTSThere were 236 independent institutions in the COG atthe time of the study. One institution recently mergedwith a sister institution and therefore was excludedfrom the analysis. As a result, 235 independent insti-tutions were surveyed.

Two hundred two replies were received (86%).Seventeen institutions participated in local protocolsand did not have a CCG or POG protocol open. There-fore, 181 (77%) COG institutional consent forms wereavailable for analysis.

The mean length of the POG 9905 consent formswas 17 pages (range, 9 –31 pages). Complete CCG 1991

consent forms were, on average, 33 pages long (range,8 –50 pages). Some CCG institutions submitted onlyPart I or Part II of the CCG 1991 consent form (n � 16).Information regarding the return of research resultswas determined to be the same for institutions thatsubmitted one part of the form compared with insti-tutions that submitted both parts. Therefore, we as-sumed that Parts I and II of the CCG consent formsprovided the same information about return of results.

The independent reviewers agreed in their origi-nal analysis of 171 of 181 consent forms. The 10 dis-crepancies were resolved by review of the consentforms in question, with subsequent complete agree-ment in all 10 cases. No consent forms required a thirdreviewer to reconcile the interpretations.

Contact information for the principal investigatorwas provided on 175 (96.7%) consent forms. Contactinformation for other members of the institution whowere involved in research (primarily other physiciansor patient advocacy groups) was provided on 123(68%) consent forms. Contact information for re-search services that were independent of the re-searcher (i.e., IRB or REB contacts) was provided onmost consent forms (n � 154; 85%).

We examined the consent forms for a specificstatement indicating that the research participant orthe participant’s guardian had a right to receive asummary of the results; only 5 (2.8%) forms containedsuch a statement. Five (2.8%) consent forms specifiedhow the research participant was to receive a sum-mary of results, and 2 forms offered participants andguardians the opportunity to indicate whether theywished to receive a summary of results.

In offering a summary of results to research par-ticipants, we believed that it was important to providean estimate of when those results might be available.Of the consent forms that indicated a right to receiveresearch results (n � 5), most (n � 4) provided aspecific statement of either the time frame for com-pletion of the study (i.e., accrual plus follow-up), andall 5 provided a specific statement regarding when theparticipant could expect to receive the results.

The majority of consent forms from both CCG andPOG indicated that if new information with the poten-tial to affect a decision to continue participation be-came available, this information would be shared withthe participant or guardian during the study (n � 162;90%).

Many CCG consent forms contained the followingstatement: “You have the right to know about newinformation that may affect your child’s health, wel-fare, or willingness to stay in this study. We will pro-vide you with this information.” This statement wasnot accompanied by any information about how

2906 CANCER June 1, 2003 / Volume 97 / Number 11

guardians could retain contact with the researcher,when the study would be completed, or in what formthe information would be provided. We believed thatthe statement would be interpreted by consent-giversas applying only to the time during which the childwas receiving treatment. Only 10 consent forms (5.5%)specifically indicated that new information would becommunicated to parents after the study was closed.

The statement “We will tell you/your child aboutnew information that may affect your/your child’shealth, welfare or willingness to stay in this study”appears in the POG template. The template contains aspecific statement indicating that the study will lastapproximately 21⁄2 years (the duration of active treat-ment) for the child; this period therefore does notinclude longer-term follow-up for the individual childor for the entire cohort. The statement “You have theright to know about new information that may affectyour child’s health, welfare, or willingness to stay inthis study” appears in the template for CCG consentforms. There is a specific statement in the templateindicating that the study lasts 2–3 years, again suggest-ing a lack of commitment to providing data outsidethe context of a child’s actual participation in thetreatment component of the study.

Chi-square analysis was not performed becausethe numbers per cell were not large enough.

DISCUSSIONRespect for individuals is enshrined in the Declarationof Helsinki and in the United States Common Rule(Code of Federal Regulations; Title 45, Part 464), andadherence to such codes is mandated by the NationalInstitutes of Health in the United States and by similarfunding and regulatory bodies in other countries.4 Theprinciple of respect is manifested in the consent pro-cess; most human research requires IRB or REB ap-proval and the consent of the participant or the par-ticipant’s guardian (or assent, in the case of anindividual who is unable to consent fully).

It is recognized that providing substantive infor-mation that can affect the decision to participate in aparticular trial is mandatory, both before enrollmentand during participation as new information becomesavailable. In the current study, most consent formsindicated that research participants had the right toreceive this information, and the consent form prom-ised to make the information available.

The offer to provide information after researchparticipation and study closure appears to be muchless universal. In the current study, few consent formsindicated that the research participant had any rightto a summary of results or to new information (such asadvice for prevention of adverse results), and few con-

sent forms specified how or when this informationcould be accessed. In a separate study examining theattitudes and needs of principal investigators in theCOG (the same setting as the current study), 3.3% ofrespondents indicated that their institution had a for-mal program for returning research results to partici-pants;15 this finding agreed with the low frequencyobserved in the current study.

A review of the consent form templates providedby the POG and CCG demonstrated that the studyperiod specifically referred to in the consent formswas the active treatment period for the individualchild, not the overall duration of the clinical trial andits follow-up. This finding, along with the scarcity offormal programs for the return of research results,supports our interpretation that the general statementfound in many consent forms (“You have the right toknow about new information that may affect yourchild’s health, welfare, or willingness to stay in thisstudy”) refers only to information from the periodduring which the child is receiving active treatmentand not to information that can be shared after thestudy is closed.

It seems appropriate that a summary of researchresults should not be shared with participants untildata are mature and have at least been peer-reviewedand accepted for publication. Because many studiesfirst are submitted as abstracts, and because the timeto actual publication may be long, a possible compro-mise would be to share information with participantsat the time of abstract publication; doing so alsowould help avoid the perception that research partic-ipants are the last to be informed of the results.16 –19

The timely sharing of research results is particularlyimportant for studies in which adverse outcomes arereported with implications for individual partici-pants.20 –22 The conflict between the importance ofdata maturity and the importance of informing studyparticipants of results as quickly as possible is not easyto resolve and will require ongoing discussion be-tween researchers and participants.8 Very few consentforms in the current study indicated when study ac-crual was to be completed or when data were expectedto be mature enough to be shared.

The question of how best to retain contact withresearch participants or guardians over an extendedperiod of time is not simple. This is particularly truefor the guardians or extended families of participantswho have died while on study or after the completionof study, as is the case in many oncology and Phase Itrials. Offering the option to receive research results atthe time of enrollment would provide an excellentopportunity to emphasize the availability of this infor-mation to participants. It also would afford partici-

Disclosure of Research Results/Fernandez et al. 2907

pants or guardians the opportunity to clearly decidewhether they wished to be contacted in the future withresearch results. This could lead to a discussion of howto foster contact and when to expect results to becomeavailable. Internet-based information systems providea possible method of maintaining contact after studyclosure.23,24

It is important to recognize that follow-up is po-tentially stressful and can have a negative impact onquality of life.25 Despite the potential negative conse-quences of disclosing research results, very few con-sent forms in the current study allowed participants toindicate whether or not they wished to receive a sum-mary of results. The option not to receive resultsshould be discussed in detail with all participants, andparticipants’ wishes should be respected.

There is evidence that survivors of childhood can-cer have an incomplete understanding of their cancerand their future risks. Byrne et al. interviewed 1928adults who survived childhood cancer.26 Fourteenpercent of survivors of malignancies at sites other thanthe central nervous system stated that they had nothad cancer. In addition, a study of women who hadmediastinal Hodgkin disease and radiotherapy at ayoung age demonstrated that up to 40% were notaware of their increased risk of breast cancer or theirgreater need for screening.27 Because the majority ofchildren are offered participation in clinical trials, thereturn of research results is an ideal opportunity tohighlight the need for appropriate medical follow-upbased on evidence acquired in research trials. Theoffer to provide a summary of results clearly is not acomprehensive solution to these problems, but thereceipt of such information may help survivors avoidsome of the pitfalls associated with an incompleteunderstanding of their cancer and their future risks.

A limitation of the current study is that the con-sent form is only one component of the interactionbetween researcher and participant. It is possible thatinstitutions offer to provide research summary resultsat other times during contact with participants, but webelieve that this is unlikely, as a study of the attitudesof principal investigators found that few institutionsacknowledged a formal program for returning re-search results.15 In addition, no formal program forgenerating summaries exists within the COG. We arein the process of surveying formal institutional prac-tices within the COG regarding the offer to provideresearch results to study participants.

It can be argued that both toxicity and outcomedata will be of interest to participants— did they par-ticipate in a study that was useful in advancing a curefor childhood cancer?, and did they experience imme-diate side effects, or can they expect to experience late

side effects of the therapy?— especially in the contextof whether or not there was a benefit to the partici-pant. It will be important to determine the needs ofresearch participants in this regard and tailor pro-grams to suit those needs.

There are many tangible benefits to offering dis-closure of research results to participants, both forindividual participants and for the research as awhole.28 The failure to address this issue in the con-sent forms that were reviewed in the current studymarks a lost opportunity to support research as awhole and to demonstrate ongoing respect for theparticipant. Researchers should consider revising theirconsent forms to reflect the need to offer summaryresults, including estimates of how and when theseresults may be available, to research participants.Prompting institutions in cooperative groups to do sowould be relatively simple, as most consent formsanalyzed in the current study followed an originaltemplate that was provided by the study committee.Aside from templates, the generation of groupwideformal guidelines, formal institutional programs, andlay summaries will be important in making the returnof research results routine practice. A detailed discus-sion of the theoretic aspects of developing formal pro-grams is provided elsewhere.8,29

Future directions of research in this area shouldexamine both researchers’ views regarding perceivedbarriers to the return of research results and partici-pants’ views regarding their needs. The findings ofthese studies should be used to generate guidelines forthe implementation of programs for returning re-search results to study participants. In turn, the im-plementation of these programs should be studied tomeasure their impact and to determine the optimalmeans of returning results to study participants.

REFERENCES1. U.S. Department of Health, Education, and Welfare. Protec-

tion of human subjects: Belmont Report— ethical principlesand guidelines for the protection of human subjects of re-search. Fed Regist. 1979;44:23192–23197.

2. Office of Human Research Protection. Code of Federal Regu-lations. Title 45, Part 46. Protection of human subjects. Avail-able at: URL http://ohrp.osophs.dhhs.gov/humansubjects/guidance/45cfr46.htm 46.116 [accessed 7 April 2003].

3. Medical Research Council of Canada. TriCouncil policystatement. Ethical conduct for research involving humans.Ottawa: Medical Research Council of Canada, 2000.

4. National Institutes of Health, Office of Human SubjectsResearch. OHSR information sheets. Bethesda: National In-stitutes of Health, Office of Human Subjects Research, 2002.

5. Council for International Organizations of Medical Sciences.International ethical guidelines for biomedical research in-volving human subjects. Geneva: CIOMS, 2002.

6. Fernandez CV. Informing study participants of research re-sults. Children’s Oncology News. 2001;2:10.

2908 CANCER June 1, 2003 / Volume 97 / Number 11

7. Partridge AH, Winer EP. Informing clinical trial participantsabout study results. JAMA. 2002;288:363–365.

8. Fernandez CV, Kodish ED, Weijer C. Informing study par-ticipants of research results: an ethical imperative. IRB. Inpress.

9. Bhatia S, Meadows AT, Robison LL. Second cancers afterpediatric Hodgkin’s disease. J Clin Oncol. 1998;16:2570 –2572.

10. Hoffman-Goetz L, MacDonald M. Cancer coverage in mass-circulating Canadian women’s magazines. Can J PublicHealth. 1999;90:55–59.

11. Reynolds T. Media coverage of cancer risks may mislead.J Natl Cancer Inst. 1993;85:1366 –1368.

12. Unson CG, Dunbar N, Curry L, Kenyon L, Prestwood K. Theeffects of knowledge, attitudes, and significant others ondecisions to enroll in a clinical trial on osteoporosis: impli-cations for recruitment of older African-American women.J Natl Med Assoc. 2001;93:392– 401; discussion, 402– 404.

13. Winett LB, Wallack L. Advancing public health goals throughthe mass media. J Health Commun. 1996;1:173–196.

14. Ratzan SC. Ethical research and practice of how we com-municate health and risk. J Health Commun. 1998;3:91–92.

15. Fernandez CV, Kodish ED, Shurin S, Weijer C. Disclosure ofthe right of research participants to receive research results:attitudes and needs of principal investigators in the Chil-dren’s Oncology Group (abstract). Blood. 2002;101:870a.

16. Daluiski A, Kuhns CA, Jackson KR, Lieberman JR. Publica-tion rate of abstracts presented at the annual meeting of theOrthopaedic Research Society. J Orthop Res. 1998;16:645–649.

17. Juzych MS, Shin DH, Coffey JB, Parrow KA, Tsai CS, BriggsKS. Pattern of publication of ophthalmic abstracts in peer-reviewed journals. Ophthalmology. 1991;98:553–556.

18. McCormick MC, Holmes JH. Publication of research pre-sented at the pediatric meetings. Change in selection. Am JDis Child. 1985;139:122–126.

19. Pitkin RM, Branagan MA, Burmeister LF. Accuracy of data inabstracts of published research articles. JAMA. 1999;281:1110 –1111.

20. Acquavella JF, Collins JJ. Perspective on the content ofworker notifications. Am J Ind Med. 1993;23:77– 83.

21. Boal WL, Friedland J, Schulte PA. Workers’ response to risknotification. Am J Ind Med. 1995;27:471– 483.

22. Schulte PA, Ringen K, Altekruse EB, et al. Notification of acohort of workers at risk of bladder cancer. J Occup Med.1985;27:19 –28.

23. Brennan PF. Health informatics and community health:support for patients as collaborators in care. Methods InfMed. 1999;38:274 –278.

24. Kaplan B, Brennan PF. Consumer informatics supportingpatients as co-producers of quality. J Am Med Inform Assoc[serial online]. 2001;8:309 –316. Available from URL: http://www.jamia.org/cgi/content/full/8/4/309 [accessed 7 April2003].

25. Dow KH, Ferrell BR, Anello C. Balancing demands of cancersurveillance among survivors of thyroid cancer. CancerPract. 1997;5:289 –295.

26. Byrne J, Lewis S, Halamek L, Connelly RR, MulvihillJJ. Childhood cancer survivors’ knowledge of their diagnosisand treatment. Ann Intern Med. 1989;110:400 – 403.

27. Diller L, Medeiros Nancarrow C, Shaffer K, et al. Breastcancer screening in women previously treated for Hodgkin’sdisease: a prospective cohort study. J Clin Oncol. 2002;20:2085–2091.

28. Dow KH, Ferrell BR, Leigh S, Melancon CH. The cancersurvivor as co-investigator. The benefits of collaborativeresearch with advocacy groups. Cancer Pract. 1997;5:255–257.

29. Di Blasi Z, Kaptchuk TJ, Weinman J, Kleijnen J. Informingparticipants of allocation to placebo at trial closure: postalsurvey. BMJ. 2002;325:1329.

Disclosure of Research Results/Fernandez et al. 2909