Co-enrollment of critically ill patients into multiple studies: patterns, predictors and consequences.

0.31 MB·English

by  Cook, Deborah

Cooketal.CriticalCare2013,17:R1
http://ccforum.com/content/17/1/R1
RESEARCH Open Access
Co-enrollment of critically ill patients into
multiple studies: patterns, predictors and
consequences
Deborah Cook1,2*, Ellen McDonald2, Orla Smith3, Nicole Zytaruk2, Diane Heels-Ansdell2, Irene Watpool4,
Tracy McArdle4, Andrea Matte3, France Clarke2, Shirley Vallance5, Simon Finfer6, Pauline Galt7, Tim Crozier7,
Rob Fowler3, Yaseen Arabi8, Clive Woolfe9, Neil Orford10, Richard Hall11, Neill KJ Adhikari3, Marie-Clauide Ferland12,
John Marshall3 and Maureen Meade1,2 and for The PROTECT Research Coordinators and for
PROTECT Investigators, for
Canadian Critical Care Trials Group and the Australian and New Zealand Intensive Care Society Clinical Trials
Group
Abstract
Introduction: Research on co-enrollment practices and their impact are limited in the ICU setting. The objectives
of this study were: 1) to describe patterns and predictors of co-enrollment of patients in a thromboprophylaxis trial,
and 2) to examine the consequences of co-enrollment on clinical and trial outcomes.
Methods: In an observational analysis of an international thromboprophylaxis trial in 67 ICUs, we examined the co-
enrollment of critically ill medical-surgical patients into more than one study, and examined the clinical and trial
outcomes among co-enrolled and non-co-enrolled patients.
Results: Among 3,746 patients enrolled in PROTECT (Prophylaxis for ThromboEmbolism in Critical Care Trial), 713
(19.0%) were co-enrolled in at least one other study (53.6% in a randomized trial, 37.0% in an observational study
and 9.4% in both). Six factors independently associated with co-enrollment (all P < 0.001) were illness severity
(odds ratio (OR) 1.35, 95% confidence interval (CI) 1.19 to 1.53 for each 10-point Acute Physiology and Chronic
Health Evaluation (APACHE) II score increase), substitute decision-makers providing consent, rather than patients
(OR 3.31, 2.03 to 5.41), experience of persons inviting consent (OR 2.67, 1.74 to 4.11 for persons with > 10 years’
experience compared to persons with none), center size (all ORs > 10 for ICUs with > 15 beds), affiliation with trials
groups (OR 5.59, 3.49 to 8.95), and main trial rather than pilot phase (all ORs > 8 for recruitment year beyond the
pilot). Co-enrollment did not influence clinical or trial outcomes or risk of adverse events.
Conclusions: Co-enrollment was strongly associated with features of the patients, research personnel, setting and
study. Co-enrollment had no impact on trial results, and appeared safe, acceptable and feasible. Transparent
reporting, scholarly discourse, ethical analysis and further research are needed on the complex topic of co-
enrollment during critical illness.
Introduction study. Restricting enrollment to only one study when
Clinical trials are essential to improve care and reduce patients are eligible for more than one is a potentially
morbidity and mortality inthe intensive care unit(ICU). modifiable barrier to recruitment [1]. Testing two inter-
Some critically ill patients are eligible for more than one ventions concurrently can be achieved with a factorial
designasusedsuccessfullybytheAcuteRespiratory Dis-
*Correspondence:[email protected] tress Syndrome Network. In other circumstances, when
1DepartmentofMedicine,McMasterUniversityHealthSciencesCenter, trials are initiated by different investigators at different
Hamilton,ONL8N3Z5,Canada
times, with different inclusion and exclusion criteria,
Fulllistofauthorinformationisavailableattheendofthearticle
©2013Cooketal.;licenseeBioMedCentralLtd.ThisisanopenaccessarticledistributedunderthetermsoftheCreativeCommons
AttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,andreproductionin
anymedium,providedtheoriginalworkisproperlycited.
Cooketal.CriticalCare2013,17:R1 Page2of11
http://ccforum.com/content/17/1/R1
co-enrollment can facilitate either sequential or simulta- Patients considered eligible were ≥ 18 years old, weighed
neousrecruitment(Figure1). > 45 kilograms, and were expected to remain in ICU
Co-enrollment in multiple trials, often driven by > 72 hours. Exclusion criteria were admission diagnosis
patientdemand,occursinpersonswithhumanimmuno- of trauma, neurosurgery or orthopedic surgery, need for
deficiency virus (HIV) [2], and was documented among therapeutic anticoagulation, receipt of > 72 hours of
23% of persons with HIV in six ongoing studies [3]. In heparin, contraindication to heparin, blood or pork pro-
thispopulation,co-enrollmentisactively encouragedby ducts, pregnancy, life support limitation, and prior
some research programs [3] but not others [2]. In pre- enrollment in this or a related trial. The primary out-
hospital resuscitation trials,co-enrollment occurs either come was proximal leg deep vein thrombosis (DVT).
in series or in parallel [4]. Half of the members of two Other outcomes were pulmonary embolism, venous
criticalcareresearchconsortiareportedco-enrollmentof thromboembolism, bleeding, heparin-induced thrombo-
apatientinmorethanonestudyinthelastyear[5].Ina cytopenia, duration of mechanical ventilation, ICU and
parental survey, 74% endorsed enrollment of their pre- hospital stay, and ICU and hospital mortality. PRO-
mature babies in 2 or more studies, 50% would consent TECT was conducted over four years from May 2006 to
to 3 or more studies, and 10% were willing to join more June 2010 in 67 ICUs in Canada, the United States, the
than10studies[6]. United Kingdom, Australia, Brazil and Saudi Arabia, as
Some Institutional Review Boards restrict the practice published previously [9].
of co-enrollment, while concerned about patient safety, Ethical approval was obtained from each participating
decisionalburdenorscientificintegrity.Giventhedearth Institutional Research Board (listed at the end of the
of evidence onthese issues, trialists have called for con- manuscript under PROTECT Collaborators). In-person
siderationof co-enrollment on a case-by-case basis, and informed consent was required prior to randomization.
reportingonitsimpact[7].Theprimaryobjectiveofthis Deferred consent was not permitted. For substitute deci-
study was to document the patterns and predictors of sion-makers not in hospital, initial telephone consent,
patientco-enrollmentinaninternationalheparinthrom- followed by in-person consent when possible, was
boprophylaxis trial. The secondary objective was to approved in 16 of the 67 (23.9%) centers.
examine the consequences of co-enrollment on clinical Beginning and throughout the trial, the PROTECT
andtrialoutcomes. Steering Committee reviewed each multicenter protocol
to decide whether co-enrollment was admissible, using
Materials and methods Canadian Critical Care Trials Group guidelines [10].
PROTECT (Prophylaxis for ThromboEmbolism in Criti- These guidelines outline important scientific (for exam-
cal Care Trial) (clinicaltrials.gov NCT00182143) was a ple, interacting interventions), psychosocial (for example,
randomized, blinded clinical trial comparing unfractio- family stress) and logistic (for example, research coordi-
nated heparin to dalteparin for thromboprophylaxis [8]. nator workload) factors to consider. The general
approach to co-enrollment was that all reasonable
efforts should be made to minimize the exclusion of
patients co-enrolled in another trial if they would likely
(cid:90)
(cid:38)(cid:258)(cid:272)(cid:410)(cid:381)(cid:396)(cid:349)(cid:258)(cid:367) represent those patients to whom trial results would
(cid:90)
(cid:24)(cid:286)(cid:400)(cid:349)(cid:336)(cid:374)(cid:3) possibly be applied in practice, as long as biologic inter-
(cid:90) (cid:894)(cid:400)(cid:349)(cid:374)(cid:336)(cid:367)(cid:286)(cid:3)(cid:90)(cid:18)(cid:100)(cid:895) action of the interventions being tested in the two trials
seemed highly implausible. Dialogue between the princi-
(cid:94)(cid:286)(cid:395)(cid:437)(cid:286)(cid:374)(cid:410)(cid:349)(cid:258)(cid:367)(cid:3) pal investigator and steering committees of each multi-
(cid:90) (cid:90) (cid:18)(cid:381)(cid:286)(cid:374)(cid:396)(cid:381)(cid:367)(cid:373)(cid:286)(cid:374)(cid:410) center study determined whether co-enrollment would
(cid:47)(cid:374)(cid:3)(cid:1006)(cid:3)(cid:90)(cid:18)(cid:100)(cid:400) impact the scientific integrity of either study. When
relevant, this was reasoned at the Canadian Critical Care
Trials Group or the Australian and New Zealand Inten-
(cid:90) (cid:94)(cid:349)(cid:373)(cid:437)(cid:367)(cid:410)(cid:258)(cid:374)(cid:286)(cid:381)(cid:437)(cid:400)
sive Care Society Clinical Trials Group meetings for
(cid:18)(cid:381)(cid:286)(cid:374)(cid:396)(cid:381)(cid:367)(cid:373)(cid:286)(cid:374)(cid:410)
refutation or ratification. If co-enrollment was endorsed,
(cid:90) (cid:47)(cid:374)(cid:3)(cid:1006)(cid:3)(cid:90)(cid:18)(cid:100)(cid:400)
each participating center handled the relevant study
governed by formal or informal co-enrollment policies
Time of their ICU or hospital Institutional Review Board.
Figure1Factorialandco-enrollmentdesigns.Inthisfigure,we Local policies could deny co-enrollment approved cen-
presentaschematicforafactorialdesignrandomizedtrial, trally. Local, single-center study co-enrollment could
sequentialco-enrollmentintworandomizedtrialsandsimultaneous also be approved after agreement with the PROTECT
co-enrollmentintworandomizedtrials.
Steering Committee and the relevant consortium.
Cooketal.CriticalCare2013,17:R1 Page3of11
http://ccforum.com/content/17/1/R1
Decisions were revisited if emerging evidence required variablesintothemodel,weselectedoneoffourmeasures
reconsideration (Figure 2). All other studies into which ofresearchcoordinatorexperience,oneofthreemeasures
patients were enrolled before, concurrent with, or subse- ofresearchinfrastructure,andresearchconsortiumaffilia-
quent to PROTECT were documented on case report tionrather thancountry.Thefollowingindependent fac-
forms. tors were analyzed: patient factors (age, sex, Acute
One examplewasco-enrollmentintotheAgeofBlood Physiology and Chronic Health Evaluation (APACHE) II
Evaluation Study (ABLE, ISRCTN44878718). ABLE is a score,medicalversussurgicalstatus);individualconsent-
randomizedtrialevaluatingmortalityfollowingtransfusion ing(substitutedecision-makerorpatient);researchcoordi-
ofredbloodcellsstoreduptooneweekversusstoredup nator factors (years of experience obtaining consent for
to 42 days [11]. Both PROTECT and ABLE investigators studiesintheICUwhenPROTECTbegan);centerfactors
initiallyendorsedco-enrollment.Monthslater,anobserva- (numberof ICU beds; numberof fulltime researchstaff;
tional trauma study suggested thatamonga subgroup of nationalresearchconsortiumaffiliation(CanadianCritical
patients transfused with more than five units, when Care Trials Group or the Australian or New Zealand
patientsreceivedbloodstoredlessthan,versusmorethan, Intensive Care Society Clinical Trials Group); and year
28 days, DVT rates (16.7% versus 34.5%, P = 0.006), and (pilottrialoryear1,2,3and4ofthefulltrial).Resultsare
mortality rates(13.9% versus26.7% P=0.02)were lower summarized using odds ratios (OR) with 95% CI. A P-
[12]. If prolonged blood storage is thrombophilic in valueof<0.01wasconsideredstatisticallysignificant.
trauma,thiscouldsimilarlyincreaseDVTriskincritically Wecalculatedtheproportionofpatientsineacharmof
ill medical-surgical patients.Inreconsidering PROTECT the PROTECT trial who were co-enrolled. To evaluate
andABLEco-enrollment,wesoughtadditionalevidence. whether co-enrollment affected patient safety, we re-
Usinganexistingprospectiveobservationalstudydata- analyzedtheproportionofpatientsineacharmwhohad
base of 261 medical-surgical ICU patients screened for serious adverse events. To evaluate whether trial results
DVT [13], we evaluated age of transfused blood as an would be any different without co-enrolled patients, we
additionalDVTrisk factor.Wealsoexaminedredblood re-analyzedoverallresultsexcludingthesepatients.
celltransfusionasapossibleriskfactorinthispopulation
because in 349 trauma patients, transfusions increased Results
DVTrisk[14].Wefoundthat126(48.3%)patientshadat In67participatingICUs,3,746patientswereenrolledin
least one transfusion, and patients had a median of four PROTECT. Consent was declined for 810 patients.
(interquartile range; IQR 2, 8) units. Multivariable ana- Patients who were not enrolled in PROTECT due to
lysesdocumentedthatneither redbloodcelltransfusion enrollment in another study represented 65 of 2,288
nor storage age predicted DVT in medical-surgical patients (2.8%) who were eligible but not randomized.
patients.Trendswere counterto findingsintrauma (for Those 65 patients were enrolled in 71 other studies, 41
example,redbloodcellsstoredfor≤7dayshadahigher (63.1%)ofwhichwereindustry-funded.
associated DVT risk compared to >7 days (hazard ratio Among the 3,746 patients in PROTECT, 713 (19.0%)
5.3; 95% CI 1.3 TO 22.1)) [15]. Based on inconclusive were co-enrolled in at least one other study (53.6% in a
research evidence, the PROTECT and ABLE Steering randomized trial, 37.0% in an observational study and
Committees affirmed co-enrollment into these trials. 9.4% in both types of studies). Co-enrollment rates
GiventhePROTECTsamplesize,we anticipatedsimilar across participating centers ranged from 0 to 53.9% and
transfusionratesandsimilarageofredbloodcellstrans- across participating countries from 1.1 to 26.0%. No co-
fused in the two arms. The ABLE trial now includes enrollment occurred in 30 of 67 (44.8%) centers.
venousthromboembolismasatertiaryoutcome. Factorsassociatedwithco-enrollmentinunivariateana-
lysisarepresentedinTable1.Patientswithhigherillness
Statistical analysis severityandmedicalconditionsweremorelikelytobeco-
We reported proportions with 95% confidence intervals enrolled than patients who were less ill and surgical.
(CI), and mean and standard deviation (SD) or median Substitute decision-makers were more likely to agree to
and IQR. We compared groups using Chi square, t-test co-enrollment than patients. Researchcoordinators with
and Fisher’sExact test.We examinedunivariateassocia- more ICU experience, and those with more experience
tions between co-enrollment rates (the dependent vari- obtaining consent in the ICU, were more likely to co-
able)andotherfactors(independentvariables)relatedto enroll patients than those with less experience. Centers
characteristicsofthepatient,researchcoordinator,center with more ICU beds and centers affiliated with national
and trial.AP-valueof<0.01was consideredstatistically research consortia were more likely to co-enroll than
significant. others. A higher proportionof patientswere co-enrolled
Weconductedmultivariablelogisticregressionanalyses. inCanada,theUnitedStatesandAustraliathaninBrazil,
To avoidincorporationofhighly correlated independent Saudi Arabia and the United Kingdom. Co-enrollment
Cooketal.CriticalCare2013,17:R1 Page4of11
http://ccforum.com/content/17/1/R1
START: Discuss potential study for
co-enrollment with respective PIs
and Steering Committees
Consider
New evidence
biological
bearing on previous
interaction Yes, likely No co-enrollment
co-enrollment decision
(e.g., ABLE)
Discuss with

research consortia
if relevent
CENTRAL Notify all relevant
Not likely; consider
participating centers ,
co-enrollment
LOCAL Consider local REB guideline
Local site considers (e.g., approved, annual report) No co-enrollment
co-enrollment of
centrally approved Yes, continue
study or local study
Consider local ICU guideline
(e.g., 2 study maximum) No co-enrollment
Questions and co-enrollment
Public reporting Yes, continue
reporting to
of co-enrollment
Methods Centers
Consider logistic issues
No co-enrollment
(e.g., no additional central access)
Yes, continue
Consider psycho-social issues
(e.g., family dynamics) No co-enrollment
Yes, continue
Obtain
Consider Coenroll
consent
co-enrollment
Figure2Co-enrollmentschema.Inthisfigure,weoutlinestepstakentoconsiderco-enrollmentofonepatientintooneormoreadditional
studies.ABLETrial,AgeofBloodEvaluationTrial;ANZICS,AustralianandNewZealandIntensiveCareSocietyClinicalTrialsGroup;CCCTG,
CanadianCriticalCareTrialsGroup;Fonda,fondaparinux;REB,ResearchEthicsBoard;UFH,unfractionatedheparin.
was less common in the pilot phase of the trial than the were co-enrolled into another academic investigator-
maintrial. initiatedstudy(97.5%).Of713patients,592(83.0%)were
In Table 2, we present the six factors independently co-enrolled in one other study, 93 (13.0%) were co-
associatedwithco-enrollmentinthemultivariableanaly- enrolled in two studies, and 28 (3.9%) were co-enrolled
sis. In order of decreasing strength of association, these inthreeormorestudies.
were: phase of the trial (all ORs > 8 for recruitment Of865co-enrollmentsinvolving713patients,themost
beyondthepilotphase);centeraffiliationwitharesearch commonotherinternationaltrialstestedpharmaconutri-
consortium(OR5.59,3.49to8.95);centersize(allORs> tion,intensiveglucosecontrol,sedationinterruptionand
10for ICUs with > 15 beds); substitute decision-makers high frequency oscillation (Table 1). Observational stu-
providingconsentratherthan patients(OR3.31,2.03 to dies were both quantitative (for example, registries,
5.41); experience of researchcoordinator (OR 2.67, 1.74 audits,qualityimprovementstudies,diagnostics,transla-
to4.11for>10yearsofexperiencecomparedtopersons tionalbiologyorlong-termfollow-upstudies),andquali-
whose first trial was PROTECT); and patient illness tative(forexample,interviews,focusgroups).
severity (odds ratio (OR), 95%CI 1.35 (1.19 to 1.53 for The proportion of patients co-enrolled in any study
each10-pointincreaseinAPACHEIIscore). wassimilarbetweenthedalteparingroupandtheunfrac-
Table 3 summarizes characteristics of the studies into tionatedheparin group(352(18.8%) versus361(19.3%),
whichPROTECTpatientswereco-enrolled.Themajority P = 0.74). There were no differences between groups in
Cooketal.CriticalCare2013,17:R1 Page5of11
http://ccforum.com/content/17/1/R1
Table 1Characteristics offactors associated andnot associated withco-enrollment
Total Not Co-enrolled P-value
patients co-enrolled patients
(N=3,746) (N=3,033) (N=713)
Patientcharacteristics
Age,mean(SD) 61.4(16.5) 61.4(16.6) 61.4(15.9) 0.996
Female,N(%)* 1,614(43.3) 1,319(43.8) 295(41.4) 0.257
APACHEIIscore,mean(SD)** 21.5(7.8) 21.1(7.8) 23.3(7.6) <0.001
Medicaladmissiontype,N(%) 2,831(75.6) 2,262(74.6) 569(79.8) 0.004
Personconsenting,N(%)***
Patient 354(9.5) 335(11.1) 19(2.7) <0.001
Substitutedecision-maker 3,380(90.5) 2,686(88.9) 694(97.3)
Researchcoordinatorcharacteristics
Yearsofnon-researchICUexperience,N(%)
0years 572(15.3) 491(16.2) 81(11.4) 0.003
>0to10years 1,254(33.5) 1,017(33.5) 237(33.2)
>10years 1,920(51.3) 1,525(50.3) 395(55.4)
YearsofprocuringconsentforclinicalstudiesinICU,N(%)
0years 302(8.1) 268(8.8) 34(4.8) <0.001
>0to10years 2,652(70.8) 2,276(75.0) 376(52.7)
>10years 792(21.1) 489(16.1) 303(42.5)
Centercharacteristics
NumberofICUbedsscreened,N(%)
<15beds 420(11.2) 414(13.6) 6(0.8) <0.001
15to20beds 1622(43.3) 1,218(40.2) 404(56.7)
>20beds 1,704(45.5) 1,401(46.2) 303(42.5)
FulltimeICUresearchstaff,N(%)
<1FTE 286(7.6) 280(9.2) 6(0.8) <0.001
1FTE 740(19.8) 622(20.5) 118(16.5)
>1FTE 2,720(72.6) 2,131(70.3) 589(82.6)
Formaltrialsgroupaffiliation,N(%)
Yes 3,224(86.1) 2,533(83.5) 691(96.9) <0.001
No 522(13.9) 500(16.5) 22(3.1)
Country,N(%)
Canada 2,456(65.6) 1,818(59.9) 638(89.5) <0.001
Australia 768(20.5) 715(23.6) 53(7.4)
Brazil 275(7.3) 272(9.0) 3(0.4)
SaudiArabia 138(3.7) 135(4.5) 3(0.4)
UnitedStates 91(2.4) 77(2.5) 14(2.0)
UnitedKingdom 18(0.5) 16(0.5) 2(0.3)
Characteristicofenrollmentphase,N(%)
Pilottrial 128(3.4) 126(4.2) 2(0.3) <0.001
Year1 556(14.8) 504(16.6) 52(7.3)
Year2 826(22.1) 626(20.6) 200(28.1)
Year3 1,009(26.9) 761(25.1) 248(34.8)
Year4 1,227(32.8) 1,016(33.5) 211(29.6)
Inthistable,wecomparepatients,personconsenting,approachestoconsentandresearchenvironmentassociatedwithco-enrollmentversusnoco-enrollment.
FormaltrialsgroupaffiliationreferstoacenterassociatedwitheithertheCanadianCriticalCareTrialsGrouportheAustralianandNewZealandIntensiveCare
SocietyClinicalTrialsGroup.P-valuesrefertoresultsofunivariateanalyses.APACHEIIscore,AcutePhysiologyandChronicHealthEvaluationIIscore;ICU,
intensivecareunit;SD,standarddeviation.
patients enrolled in any randomized trial (209 (11.2%) P = 0.20). Twenty PROTECT patients were co-enrolled
versus 239 (12.8%), P = 0.14), or the proportion in each in ABLE (9 of 1,873 (0.5%) in the dalteparin group and
group enrolled in any of the five most common co- 11 of 1,873 (0.6%) in the unfractionated heparin group),
enrollment studies (197 (10.5%) versus 223 (11.9%), P=0.82.
Cooketal.CriticalCare2013,17:R1 Page6of11
http://ccforum.com/content/17/1/R1
Table 2Factors independently associated withco-enrollment inmultivariate analysis
Oddsratio P-value
(95%CI)
Patientdemographics
Age(10-yearincrease) 0.96(0.91,1.02) 0.155
Female 0.92(0.77,1.11) 0.403
APACHEIIscore(10-pointincrease) 1.35(1.19,1.53) <0.001
Medicalversussurgical 1.26(1.01,1.57) 0.041
Individualconsenting
Substitutedecision-makerversuspatient 3.31(2.03,5.41) <0.001
YearsofprocuringconsentforclinicalstudiesinICU
>0to10yearsversus0years 0.83(0.55,1.25) <0.001
>10yearsversus0years 2.67(1.74,4.11)
Centersize(bedsscreenedforPROTECTpatients)
15to20bedsversus<15beds 20.06(7.56,53.25) <0.001
>20bedsversus<15beds 13.76(5.15,36.80)
FulltimeICUresearchstaff
1FTEversus<1FTE 1.13(0.41,3.11) 0.966
>1versus<1FTE 1.10(0.40,3.03)
Formaltrialsgroupaffiliation
Yesversusno 5.59(3.49,8.95) <0.001
YearofPROTECT
Year1versusPilot 8.22(1.95,34.61) <0.001
Year2versusPilot 32.89(7.95,135.98)
Year3versusPilot 38.15(9.24,157.51)
Year4versusPilot 24.53(5.94,101.25)
Inthistable,wepresenttheindependentfactorsassociatedwithco-enrollmentofonepatientintotwoormorestudiesidentifiedbymultivariateregression
analysis,presentedusingoddsratios(OR)and95%confidenceintervals(95%CI).P-valuesrefertoresultsofmultivariateanalyses.ANZICS,AustralianandNew
ZealandIntensiveCareSocietyClinicalTrialsGroup;APACHEIIscore,AcutePhysiologyandChronicHealthEvaluationIIscore;CCCTG,CanadianCriticalCare
TrialsGroup;FTE,fulltimeequivalent;ICU,intensivecareunit.
Among patients co-enrolled in other randomized additional studies in 83% and 13% of patients, respec-
trials, rates of serious adverse events were similar tively. These findings are consistent with membership
between the dalteparin (2 of 209, 1.0%) and unfractio- surveys of research consortia indicating that two was
nated heparin (0 of 239, 0.0%) groups, P = 0.14, as per the median number of randomized trials into which one
the main trial findings (7 of 1,873, 0.4%) versus 6 of patient was enrolled [5].
1,873, 0.3%), respectively, P = 0.74. Protocol violations Multivariate analysis showed that consent encounters
were also similar (data not shown). In Table 4, we show with substitute decision-makers were more likely to
that the overall PROTECT results excluding patients co- involveco-enrollmentthanthosewithpatients.Thissug-
enrolled in other randomized trials, which were no dif- geststhatmore seriouslyillpatientsare frequently eligi-
ferent than the results of all patients randomized [9]. ble for several studies, yet too sick to make decisions
That is, pulmonary embolism rates were lower in themselves,congruentwiththefindingthatpatientswho
patients receiving dalteparin compared to those receiv- wereco-enrolledweremore seriously illthan those who
ing unfractionated heparin; rates of DVT, venous throm- were not. Substitute decision-makers may seek several
bosis and major bleeding were similar. No patients were researchopportunities whilehelpingtoadvancescience,
withdrawn or lost to follow-up whether co-enrolled or so-calledconditionalaltruism[16].
not. Researchcoordinatorswithgreaterconsentexperience
weremorelikelytoco-enrollthanothers,suggestingthat
Discussion professional maturity may foster sound judgment about
In this international heparin thromboprophylaxis trial, approachingpersonsforco-enrollment,whethertraining
one-fifth of patients were co-enrolled in at least one enhances comfort and success with co-enrollment is
other study. Half of the co-enrollments were in rando- unclear. Co-enrollment occurred more often in larger
mized trials, although a variety of study designs were ICUs, and in centers affiliated with a national consor-
involved. Co-enrollment was limited to one or two tium, perhaps reflecting group norms. More frequent
Cooketal.CriticalCare2013,17:R1 Page7of11
http://ccforum.com/content/17/1/R1
Table 3Co-enrollment study characteristics. more likely to exclude individualsdue toage, comorbid-
N(%of713patients) itiesandconcomitantmedications,raisingconcernsabout
theirgeneralizability[17].However,ifindustrytrialsprohi-
Typeofotherstudy
bit co-enrollment in academic studies, selection bias in
Randomizedtrial 380(53.3)
academic trials may result, as well as slower completion,
Observationalstudy 265(37.2)
thereby delaying answers to publicly motivated research
Both 68(9.5)
questions. Certainly, co-enrollmentintrialsof investiga-
Genesisofotherstudy
tionaldrugsordevicesisimprudentduetodifficultymon-
Investigator-initiated 695(97.5)
itoringsafetyandinterpretingharm.Sincepatientsinthe
Industry-initiated 15(2.1)
investigator-informed, industry-funded trial comparing
Both 3(0.4)
drotrecoginalfatoplaceboinpatientswithpersistentsep-
CCCTGorANZICSaffiliatedstudy
ticshock(PROWESS-SHOCK,NCT00604214)[18]would
Yes 451(63.3)
typically receive heparin thromboprophylaxis in the
No 181(25.4)
absence of contraindications, PROTECT co-enrollment
Both 81(11.4)
waspermitted.Weidentifiedthreepatientswhowereeli-
Numberofstudiesintowhichpatientswereco-enrolled
gible for PROTECT but not recruited because of enroll-
1 591(82.9)
mentinPROWESS-SHOCK,andnoPROTECTpatients
2 94(13.2)
whowereco-enrolledinPROWESS-SHOCK.
3 25(3.5)
Onemajorfocusregardingpermissibleco-enrollmentin
4* 3(0.4)
twoacademic trialsis the biologic plausibility of the two
Inthistable,amongthe713patientswhowereco-enrolledinanotherstudy,
interventions having a potentiating or attenuating effect
wepresenttheco-enrollmentstudytype(randomizedtrial,observational
study),studygenesis(investigator-initiatedversusindustry-initiated),affiliation oneachother.Havingidenticalprimaryoutcomesintwo
witharesearchconsortia,andthenumberofstudiesintowhichPROTECT academictrialswouldnotbeasolecriterionforprohibit-
patientswereco-enrolled.Thebottomhalfofthetableoutlines,ofthe865
co-enrollments,whichstudieswereinvolved.ANZICS,AustralianandNew ing co-enrollment, especially when treatment effects are
ZealandIntensiveCareSocietyClinicalTrialsGroup;CCCTG,CanadianCritical expectedtobemodest,whichiscommonincriticalcare.
CareTrialsGroup.
Forexample,iftwotrialshad thesameprimaryoutcome
ofmortality(forexample,atrialcomparingstarchresusci-
duringthefulltrialthan thepilotphase, facilitywithco- tationvsnormalsalineinsepticshockandintensiveinsu-
enrollmentmayhaveincreasedovertime. lintherapyvsliberalglucosemanagementinheterogenous
Participationinanotherstudywasthereasonwhy65eli- ICUpatients),thetwointerventionswouldlikely becon-
gible patients were not enrolled in PROTECT; 63% of sideredunrelated,andco-enrollmentwouldbepermitted,
these studies were industry-initiated. Only 2% of PRO- because starch and antioxidants would not be known to
TECTpatientswereco-enrolledinindustrystudies.Gen- mediatetheireffectonmortalitythroughrelatedmechan-
erally,industry-fundedtrials,comparedtoothertrials,are isms. Several additional scientific issues need careful
Table 4PROTECT resultsexcluding patients co-enrolled in another randomized trial
N(%) Dalteparin UFH Hazardratio Dalteparin UFH Hazardratio
(N=1,873) (N=1,873) (95%CI) (N=1,664) (N=1,633) (95%CI)
Allpatients Allpatients Notco-enrolled Notco-enrolled
inRCT inRCT
Primaryoutcome:proximallegdeepvein 94(5.1) 108(5.9) 0.91(0.68, 83(5.0) 93(5.7) 0.87(0.63,
thrombosis 1.23) 1.21)
Anypulmonaryembolism 22(1.2) 42(2.3) 0.48(0.27, 19(1.1) 35(2.1) 0.48(0.26,
0.84) 0.89)
Anyvenousthromboembolism 150(8.2) 184(10.0) 0.87(0.69, 133(8.0) 161(9.9) 0.83(0.64,
1.10) 1.07)
Majorbleeding 100(5.5) 105(5.7) 0.98(0.73, 88(5.3) 93(5.7) 0.94(0.69,
1.31) 1.28)
Heparin-inducedthrombocytopenia 5(0.3) 12(0.7) 0.47(0.16, 5(0.3) 10(0.6) 0.56(0.18,
1.37) 1.67)
Hospitalmortality 395(21.7) 444(24.3) 0.91(0.79, 372(22.4) 391(23.9) 0.95(0.82,
1.05) 1.10)
Inthistablewereportthemainresultsoftheoriginaltrialincludingallpatients,andresultsincludingonlythosepatients,(209inthedalteparingroupand239
intheunfractionatedheparingroup),whowerenotco-enrolledinanotherrandomizedtrial(1,664inthedalteparinarmand1,633intheunfractionatedheparin
arm).CI,confidenceinterval;RCT,randomizedclinicaltrial;UFH,unfractionatedheparin.
Cooketal.CriticalCare2013,17:R1 Page8of11
http://ccforum.com/content/17/1/R1
considerationregardingco-enrollment,suchasprojected Limitationsofthisstudyincludeourinabilitytoexplore
impactonstatistical power,outcome ascertainmentbias, the decisional burden on substitute decision-makers,
increasedriskofadverseeventsandtheultimateinterpre- patients and research coordinators. However, in another
tation of study results. We recommend widespread four-month single center study, we found consent rates
consultationaboutthemeritsanddemeritsofvariousco- similarforanysingleenrollment(84%)andco-enrollment
enrollmentpairsbeforeandduringconductofatrial. (79%)opportunity[22].We couldnotdocument ratesor
Furthermore,specificapproachestodatacollectionand reasons for no co-enrollment, or which person declined
analysiscanbeestablishedaprioriifconcernsexistabout (forexample,patient,substitutedecision-maker,physician,
co-enrollment. Shared definitions and case report forms surgeon,anesthesiologist).Examiningthechoiceofwhich
foruseacrossstudiescouldhelp[3],asweusedforPRO- studytopursueifapatientwaseligibleformorethanone
TECTand ABLE. New researchinfluencing previous co- wasbeyondthescopeofthisproject.However,investiga-
enrollment decisions should prompt revisiting previous torsreportthat whenapproachingpersonsforco-enroll-
decisionsasstudiesunfold.Documentingconsecutiveand ment in a randomized trial, they consider trial rigor and
concurrent co-enrollment eligibility and consent rates relevance,potentialforbenefitorharm,consortiumaffilia-
throughoutatrial,andtransparentreportingofco-enroll- tionandremuneration[5].
mentuponcompletion,includingeffectmodificationand Strengths of this study include comprehensive docu-
risk ofharm,will help to disseminate co-enrollment pat- mentation of co-enrollment throughout a multicenter
terns, and provide data to examine its actual rather than trial.Investigatorsusedaprospective,transparentframe-
perceivedimpact.Ifconcernsexistandtrialistsarewilling workforco-enrollmentdecisions,independentlyexamin-
toexchangerandomizationcodes,unadjustedandadjusted ing each pair of studies, guided by independent
analysescanbeconductedtoevaluatetheimpactnotjust Institutional Review Boards, and research consortia.
ofsubstantialco-enrollmentofpatientsinonetrialonthe Using multivariate analysis, significant predictors of co-
resultsoftheothertrial,andviceversa,butalsotheimpact enrollment were identified, adjusting for confounding.
ofeachspecificallocationarm. We examined the impacton patient and trial outcomes.
Although an understanding of all available treatment Representation from diverse ICUs and countries
options is important for informed consent for medical enhances the generalizability of these findings, which
therapy,thereisnosimilarperceptionthatpatientsshould mayapplytootheracademictrialstestingcurrentlyavail-
beinformedofallavailablestudiesforwhichtheyareeligi- ableinterventions.
ble.Indeed,mosthumansubjectsresearchdiscoursedeals
with protection from harm rather than opportunity for Conclusions
participation [1], which is particularly germane to co- Co-enrollmentwascommoninthisthromboprophylaxis
enrollment. More open discussion will help to elucidate trial,andwasstronglyassociatedwithspecificfeaturesof
key ethical issues, since the vulnerability of critically ill the patients, research personnel, setting and study. Co-
patients[19]raisesconcernaboutadverseeffectsfromco- enrollmentwas an effective, feasible method to enhance
enrollment. We found that PROTECT patients co- recruitment,providedthatthepatientsorsubstitutedeci-
enrolledinrandomizedtrialswerenotmorelikelytohave sion-maker, clinicians, principal investigators, steering
seriousadverseeventsorprotocolviolationscomparedto committees, research consortium and local Institutional
patients who were not co-enrolled. Post hoc analyses of Review Boards agreed. Co-enrollment did not influence
PROTECT omitting patients co-enrolled in randomized overall trial results, patient safety or adverse events.
trialsyieldedthesameoverallresultsasthemainanalysis. Furtherscientificdebate,ethicalanalysisandresearchare
Finally, no PROTECT patients were lost to follow-up, needed on the complex topic of co-enrollment for criti-
thus,noco-enrolledpatientswerewithdrawn. callyillpatients.
In 2007, a tri-national survey showed that only 11% of
respondents indicated that their local Institutional Key messages
Review Board had a co-enrollment policy, whereas 35% (cid:129) In this international heparin thromboprophylaxis
reported a local ICU guideline [5]. Co-enrollment guide- trial of 3,746 patients, one-fifth of patients were co-
lines exist for adult resuscitation studies [4], pediatric enrolled in at least one other study. Half of the co-
[20] and adult [10] critical care. Public health mandates enrollments were in randomized trials, although a
to answer research questions quickly during pandemics variety of study designs were involved and almost all
have encouraged co-enrollment of patients in treatment co-enrollments were in academic investigator-
and observational studies [21]. Professional position initiated studies.
statements about whether, when, why and how to co- (cid:129) In decreasing strength of association, six factors
enroll will raise awareness and facilitate stakeholder were independently associated with co-enrollment:
dialogue. later phase of the trial compared to the pilot phase,
Cooketal.CriticalCare2013,17:R1 Page9of11
http://ccforum.com/content/17/1/R1
center affiliation with a research consortium, larger (cid:129)DrAndreasFreitag,ChristineWynne,MarkDuffett,MichelleKho,Nicole
centersize,substitutedecision-makersprovidingcon- Zytaruk;PharmacyTechnician:KarenCurrie;HamiltonHealthSciences,
McMasterHospital,Hamilton
sentratherthanpatients,greaterresearchcoordinator (cid:129)DrJohnGranton,AndreaMatte,PaulinaFarias,LeslieChu,NancyBrockest,
experienceandhigherpatientillnessseverity. StephanieGo,MargaretMcGrath-Chong,MadisonDennis,MarcLipkus,Emily
(cid:129)Co-enrollmentdidnotinfluenceoveralltrialresults, Stern,RyanAlbert;Pharmacy:RonSeto,MuhammadZuberi,JieMingand
LauraArus-Pampin,MuhammadWalid,RobertSolek,KimDefreitas;
patientsafetyoradverseevents. UniversityHealthNetwork,TorontoGeneralHospital,Toronto
(cid:129)Beforeandduringatrial,wesuggestwidespreadcon- (cid:129)DrsStephanLangevin,FrançoisLauzierandAlexisFTurgeon,MartineBlais,
sultationamonginvestigators,clinicians,trialsteering MaximeBeauparlant,JulieAsselin,CarolineRoy,ChantalGagné,Marie
Thibodeau;Pharmacists:AnikRiouxandTuong-ViTran;Hôpitaldel’Enfant-
committees,researchconsortiaandlocalInstitutional
Jésus,QuebecCity
Review Boards about the scientific, psychosocial and (cid:129)DrGermainPoirier,IsabelleNeas,SandrineSpearson;Pharmacist:BettyTon;
logisticeffectsofvariousco-enrollmentpairs. CharlesHôpitalCharles-LeMoyne,Longueuil
(cid:129)DrsLauralynMcIntyreandPaulHébert,IreneWatpool,TracyMcArdle,
(cid:129) Transparent reporting, scholarly discourse, ethical
ClaudeGaudert,PauleMarchand,CarsonDavidson;Pharmacists:Anne-Marie
analysisandfurtherresearchareneededonthecom- DugalandSusanFetzer;OttawaHospital,GeneralCampus,Ottawa
plextopicofco-enrollmentduringcriticalillness. (cid:129)DrJoePagliarello,Mary-JoLewis,ErinMurphy,JuliaFoxall;Pharmacist:
SherryWeir;OttawaHospitalCivicCampus,Ottawa
(cid:129)DrYoannaSkrobik,JohanneHarvey,StefaniaChitu;Pharmacists:Marceline
QuachandLindaPinet;HôpitalMaisonneuve-Rosemont,Montréal
Abbreviations (cid:129)DrMartinAlbert,CaroleSirois,CaroleNadon,StéphanieDolle,Audrey-Anne
ABLE:AgeofBloodEvaluation;APACHE:AcutePhysiologyandChronic Gosselin,PatriceDeroy;Pharmacists:AnneJulieFrenetteandDavid
HealthEvaluation;CI:confidenceinterval;DVT:deepveinthrombosis;HIV: Williamson;HôpitalduSacré-CoeurdeMontréal,Montréal
humanimmunodeficiencyvirus;ICU:intensivecareunit;IQR:interquartile (cid:129)DrSangeetaMehta,CherylEthier,SamTirgari,LindsaySteinberg,Rod
range;OR:oddsratio;PROTECT:ProphylaxisforThromboEmbolisminCritical McDonald,VidhyaSivanantham,KristoferBandayrel,FriederikeQuittnat
CareTrial;SD:standarddeviation Pelletier,MarnieKramer-Kile,MaedeanBrown,ScottKim;PharmacistHolly
Leung;MountSinaiHospital,Toronto
Authors’contributions (cid:129)DrRobertFowler,NicoleMarinoff,KarenCode,BorisBojilov,Derek
DC,EM,FCandNZconceivedofthestudy.DC,EM,FC,NZ,IW,TM,AM,FC Parsotam;Pharmacist:JohnIazzetta;SunnybrookHospital,Toronto
andMCFdesignedthestudy.NZandSVcoordinatedthestudy.DC (cid:129)DrJohnMarshall,OrlaSmith,BethFry,KerriPorretta,YoonLee,Jeanna
obtainedfunding.EM,OS,NZ,IW,TM,AM,FC,SVandMCFcollecteddata. Morrissey,VictoriaWen;PharmacyTechnicians:LauraParsonsandAnn
DC,RF,NA,JMandMMconsultedonmethods.DHAperformedthe Kosinski;StMichael’sHospital,Toronto
analyses.YA,NA,RFandDCdesignedthefigure.DHA,DC,SF,TC,RF,YA, (cid:129)DrJohnMuscedere,SusanFleury,NicoleGodfrey,SharleneHammond,
CW,NO,RH,NA,JMandMMinterpretedthedata.DC,EM,OS,DHAand ElizabethMann,MonicaMyers,AmberRobinson;Pharmacist:ChrisGrey;
MMwrotethedraft.Allauthorsreadandapprovedthefinalmanuscriptfor KingstonGeneralHospital,Kingston
publication. (cid:129)DrsSeanKeenanandStevenReynolds,MiroslavSvetik,MaryVanOsch;
Pharmacist:Anne-MarieLiberman;RoyalColumbianHospital,Westminster
Competinginterests (cid:129)DrsDeanChittockandVinayDhingra,MaureenGardner,SusanLogie,
Theauthorsdeclarethattheyhavenocompetinginterests. DeniseFoster,RogerAutio,DaraDavies,PiaGanz,LaurieSmith;Pharmacy:
JaneDay,KaldipMattuandJudyYip;VancouverGeneralHospital,Vancouver
Acknowledgements (cid:129)DrPeterDodek,BettyJeanAshley,SheilaghMans;Pharmacist:MaraPavan;
ThisstudywasfundedbyaNorthAmericanMeetingGrantoftheCanadian St.Paul’sHospital,Vancouver
InstitutesforHealthResearchwhichhadnoroleinthedesign,conduct, (cid:129)DrChipDoig,LindaKnox,CrystalWilson,KevinChampagne;Pharmacist:
analysis,interpretationorwrite-upofthisreport.DCook,NZytarukandD AngelaKayallPeters;CalgaryUniversityFoothillsHospital,Calgary
Heels-Ansdellhadfullaccesstoallthedatainthestudyandtake (cid:129)DrNiallFerguson,AndreaMatte,JamesStevenson,JoelElman,Madison
responsibilityfortheintegrityofthedataandtheaccuracyofthedata Dennis;Pharmacist:JennTung,RobertSolek,KimDeFreitas,NgaPham;
analysis. UniversityHealthNetwork,TorontoWesternHospital,Toronto
Weappreciatetheinterestoffamilymembersandpatientsinresearch, (cid:129)DrJimKutsogiannis,PatricaThompson,NorineWhalen;Pharmacist:Liz
whichenabledthemaintrialandthisstudy.Wethankresearchcoordinators Helboe;RoyalAlexandraHospital,Edmonton
andphysiciansfromallparticipatingPROTECTcentersfortheircollaboration. (cid:129)DrFrançoisLellouche,Marie-ClaudeFerland,PatrickDussault,Caroline
WearegratefultotheCanadianCriticalCareTrialsGroupandtheAustralian Jacob,Marie-ÈveMorneau,NancyLaberge;Pharmacist:NathalieChateauvert;
andNewZealandIntensiveCareSocietyClinicalTrialsGroupandall InstitutUniverstairedeCardiologieetdePneumologiedeQuébec,Québec
participatingcentersfortheirengagement. (cid:129)DrTimKarachi,AndreaTkaczyk;PharmacyTechnician:DianeLourenco;
DCookisaResearchChairoftheCanadianInstitutesforHealthResearch.R HamiltonHealthSciences,JuravinskiHospital,Hamilton
FowlerisaClinicianScientistoftheHeartandStrokeFoundationofCanada. (cid:129)DrMichaelJacka,MarleenIrwin,CarmenChan,LeecaSonnema,Kelly
MMeadeisanRCTMentoroftheCanadianInstitutesforHealthResearch. Marsh,JenniferMaurer,TamaraKreidl,CandiceVarden,CareyKinjerski;
WethankLHand,KPorretta,CWynne,DFosterandPFariasforhelpful Pharmacist:NoelleCarey;UniversityofAlberta,Edmonton
commentsonearlierdrafts. (cid:129)DrChipDoig,LindaKnox,CrystalWilson,KevinChampagne;Pharmacist:
PROTECTCollaborators AngelaKayallPeters;CalgaryUniversityPeterLougheedHospital,Calgary
ClinicalCollaborators (cid:129)DrKosarKhwaja,LauraBanici,CaroleSirois,LenaHavell;Pharmacists:
CanadianInvestigators GilbertMatteandKathleenNormandin;MontrealGeneralHospital,Montréal
(cid:129)DrDeborahCook(Lead),EllenMcDonald,AndreaTkaczyk,FranceClarke; (cid:129)DrGordonWood,FionaAuld,LeslieAtkins;Pharmacist:JohnFoster-Coull;
Pharmacist:ChristineWallace;StJoseph’sHealthcare,Hamilton VancouverIslandHealthAuthority,Vancouver
(cid:129)DrsRickHallandGraemeRocker,LisaJulien,DebbieWright,CarolineRoy, (cid:129)DrsOlivierLesurandFrançoisLamontagne,SandraProulx;Pharmacist:
JudyTheriault,SusanPleasance;PharmacyTechnicians:DebiSnowand SylvieCloutier,BrigitteBolduc,Marie-PierreRousseau,JulieLeblond;Centre
ShannonHerbert;CapitalHealthQueenElizabethIIHealthScienceCenter, HospitalierUniversitairedeSherbrookeandCentredeRechercheClinique
Halifax Étienne-LeBel,Sherbrooke
(cid:129)DrMaureenMeade,LoriHand;PharmacyTechnician:MayaBiljan;Hamilton (cid:129)DrKosarKhwaja,LauraBanici,CaroleSirois,LenaHavell;Pharmacists:
HealthSciences,HamiltonGeneralHospital,Hamilton GilbertMatteandKathleenNormandin;RoyalVictoriaHospital,Montreal
Cooketal.CriticalCare2013,17:R1 Page10of11
http://ccforum.com/content/17/1/R1
(cid:129)DrsGeraldHollingerandVasantiShende,VanessaBelcastro;Pharmacist: (cid:129)DrsNiltonBrandao,CassianoTeixeiraandCíntiaRoehrig,JulianaZeni;
JaneMartin;GuelphGeneralHospital,Guelph Pharmacist:DanielPanizzi;MoinhosdeVentoHospital,PortoAlegre
(cid:129)DrBillPlaxton,AndersFoss;PharmacyTechnicians:HeatherMcDougall, (cid:129)DrsSuzanaAlvesdaSilvaandRubensCostaFilho,RenatoCorreaAlves
SharonMorrisandGoranPetrovic;GrandRiverHospital,Kitchener MoreiraandPlínioN.GomesandRodrigoBiondi;Pharmacist:MarciaCaneca,
(cid:129)DrBojanPaunovic,KymWiebe,NicoleMarten;Pharmacist:Denise GrazieleSilva,PróCardíacoHospital,PROCEP,RiodeJaneiro
Sawatzky;StBonifaceHospital,Winnipeg (cid:129)DrsOtavioBerwanger(Co-Lead)andEdsonRomano,AnnaMariaBuehler;
(cid:129)DrJonathanEisenstat,TammyDoerle;Pharmacist:LindaSkinner;Lakeridge Pharmacist:MarceloMuradandPauloBuononato;HospitalCoracaoResearch
Health,Oshawa InstituteHCor,SãoPaulo
(cid:129)DrsStevenReynoldsandSeanKeenan,SheilaghMans;Pharmacist:Ray (cid:129)DrsHelioPennaGuimarãesandRenatoDLopes,AdrianoTruffa,Rosana
Jang;SurreyMemorialHospital,Surrey Nakamura,LillianMazzaBarbosa;Pharmacist:RosanaSuemiNakamura;
(cid:129)DrMichaelSharpe,MonaMadady;Pharmacist:ChandikaMankanjee; HospitalSãoPaulo,SãoPaulo
LondonHealthSciencesCenter,London SaudiArabianInvestigators
UnitedStatesInvestigators (cid:129)DrIsmaelQushmaq(Lead),JeanBrennick,SawsanBassi;Pharmacist:Amnah
(cid:129)DrJamesKlinger,BarbaraSmithson;Pharmacist:AndreaMonckeberg; Mukhtar,MajdahAttasandAmerSoliman;KingFaisalSpecialistHospitaland
RhodeIslandHospital,Providence ResearchCenter,Jeddah
(cid:129)DrNicholasEVlahakis(Lead),LaurieMeade;Pharmacist:DebbieBauer; (cid:129)DrMohammedAlsultanandYaseenArabi,RietteBrits;Pharmacist:Antoine
MayoClinic,Rochester Cherfan;KingSaudBinAbdulazizUniversityforHealthSciences,Riyadh
(cid:129)DrMichaelCox,JackieO’Brien,CatherineKrause;Pharmacist:Sandra (cid:129)DrJamalAlhashemi,SanaaShalabi;Pharmacist:RandaAinosah;King
Ahearn;StJohn’sMercyMedicalCenter,StLouis AbdulazizUniversityHospital,Jeddah
(cid:129)DrsJosephNatesandSajidHaque,DeidreMooring,RoseErfe,Paula (cid:129)DrsYasserMandourahandNadeemShaikh;Pharmacist:ShathaShosho;
Nickerson;Pharmacist:KimMcConnell;UniversityofTexasMDAnderson RiyadhMilitaryHospital,Riyadh
CancerCenter,Houston (cid:129)DrsManalAl-HazmiandM.AliAl-Azem,TrevorWyngaard;Pharmacist:
UnitedKingdomInvestigators YahyaMoustafa;KingFahadMedicalCityHospital,Riyadh
(cid:129)DrsMarliesOstermann(Lead)andDavidTreacher,TonySherry,JohnSmith, PROTECTMethodsCenter
BarnabySanderson,JosephineNg,JohnBrooks,LingLim,KatieLei; NicoleZytaruk,LoisSaunders,SuzanneDuchesne,JenniferSouthon,France
Pharmacists:PaulTunstellandDrCathyMcKenzie,FrancescoCicirello;King’s Clarke,BronwynBarlow-Cash,KatherineKrolicki,KristinaLutz,ShelleyKraus,
CollegeLondon,Guy’sandStThomas’Hospital,London AshleyBernotas,ChristaYeo,NealaHoad,TammyFrench,DianeHeels-
AustralianInvestigators Ansdell,LisaBuckingham,AravinDuraikannan,NormaBrown,ChrisCotoi,
(cid:129)DrsJamieCooper(Lead)andAndrewDavies,ShirleyVallance,Cindy LuqiWang,SandeepBhandari,LaurelGrainger,DeborahMaddock,Pat
Weatherburn,JasminBoard,VictoriaBennett;Pharmacists:AnneMakand Smith,ShawnBillington,MaureenMeade,GordonGuyatt,StephenWalter,
SookWernChua;AlfredHospital,Melbourne LindaOronatoandDeborahCook,McMasterUniversityandStJoseph’s
(cid:129)DrsSimonFinferandNareshRamakrishnan(deceased),SimonBird,Julie Healthcare,Hamilton,Ontario,Canada
Potter,AnneO’Connor,SusanAnkers;Pharmacist:MaggieGibson;Royal PROTECTSteeringCommittee
NorthShoreHospital,Sydney DeborahCook(PrincipalInvestigator),MarkCrowther,MaureenMeade,
(cid:129)DrJackCade,DeborahBarge,TaniaCaf,BelindaHowe;Pharmacist:Emma GordonGuyatt,WilliamGeerts,DJamieCooper,IsmaelQushmaq,Marcelo
Michael;RoyalMelbourneHospital,Melbourne Rocha,OtavioBerwanger,TheodoreEWarkentin,NicoleZytaruk(Project
(cid:129)DrRinaldoBellomo,GlennEastwood,LeahPeck,DonnaGoldsmith,Kim Manager,Global),ShirleyVallance(ProjectManager,Australia),DianeHeels-
O’Sullivan;LeadPharmacists:DrMichaelChing,JeanSchmidt,MeiHoand AnsdellandStephenWalter(Biostatisticians)
BaileyLim;AustinHospital,Melbourne DataMonitoringCommittee
(cid:129)DrsDavidErnest,SamRadford,AnnWhitfieldandAnthonyCross,Suzanne DrsRobinRoberts(Chair),ChristianBrun-BuissonandVictorMontori
Eliott,JaspreetSidhu,BelindaHowe,IngaMercer,AngelaHamilton(deceased);
Pharmacist:PaulaLee;BoxHillHospital,Melbourne
(cid:129)DrJohnBotha,JodiVuat,SharonAllsop,NinaFowler;PharmacistChuiYap; Authordetails
FrankstonHospital,Frankston 1DepartmentofMedicine,McMasterUniversityHealthSciencesCenter,
(cid:129)DrsTimCrozier,JonathanBarrettandChrisWright,PaulineGalt,Carly Hamilton,ONL8N3Z5,Canada.2DepartmentofClinicalEpidemiologyand
Culhane,RebeccaIoannidis,SueBurton,MarnieReily,CyveenWeeraratna; Biostatistics,McMasterUniversityHealthSciencesCenter,Hamilton,ONL8N
Pharmacist:HelenKopp;MonashMedicalCentre,Melbourne 3Z5,Canada.3InterdepartmentalDivisionofCriticalCare,Universityof
(cid:129)DrIanSeppelt,LeonieWeisbrodt,RobynBond;Pharmacists:StellaSuen Toronto,Toronto,ON,Canada.4DepartmentofCriticalCare,Ottawa
andJasonTrinh;NepeanHospital,Sydney University,Ottawa,ON,Canada.5IntensiveCareUnit,TheAlfred,Melbourne,
(cid:129)DrDavidEvans,JustineRivett,StephanieO’Connor,AlexPoole;Pharmacist: VIC,Australia.6TheGeorgeInstituteforGlobalHealth,UniversityofSydney,
PeterSlobodian;RoyalAdelaideHospital,Adelaide Sydney,NSW,Australia.7IntensiveCareUnit,MonashMedicalCentre,
(cid:129)DrCliveWoolfe,DorrilynRajbhandari,CaitlinRees;Pharmacist:JustineHay; Melbourne,VIC,Australia.8IntensiveCareDepartment,KingSaudBin
RoyalPrinceAlfredHospital,Camperdown AbdulazizUniversityforHealthSciences,Riyadh,SaudiArabia.9Intensive
(cid:129)DrsJohnEdingtonandJasonFletcher,JulieSmith,CatherineBoschert; CareUnit,RoyalPrinceAlfredHospital,Sydney,NSW,Australia.10Intensive
Pharmacist:RichardSummers;BendigoHealthCare,Bendigo CareUnit,BarwonHealth,Geelong,VIC,Australia.11Departmentsof
(cid:129)DrGrahamReece,TreenaSara,KiranNand;Pharmacist:RabsimaIbrahim; Anesthesiology,Medicine,SurgeryandPharmacology,DalhousieUniversity,
BlacktownHospital,Blacktown Halifax,NS,Canada.12IntensiveCareUnit,HopitalLaval,Quebec,QC,Canada.
(cid:129)DrsAndrewBerstenandAlexGallus,ElishaMatheson,MargieO’Callaghan;
Pharmacist:KellyWoolley;FlindersMedicalCenter,Adelaide Received:8August2012 Revised:26October2012
(cid:129)DrNeilOrford,TaniaElderkin,MelissaFraser,AllisonBone,TaniaSalerno, Accepted:21December2012 Published:8January2013
AnneKinmonth;Pharmacist:PaulMuir;BarwonHealth,GeelongHospital,
Geelong References
(cid:129)DrSubhashArora,BridgetO’Bree,KatherineShepherd;Pharmacists:Kerry 1. RandolphAG:Theuniquechallengesofenrollingpatientsintomultiple
GrayandTuVinh;DandenongHospital,Dandenong clinicaltrials.CritCareMed2009,37:S107-S111.
(cid:129)DrsAlanDavey-QuinnandMartinSterba,BronwynRuthJohnson,ReneeXu, 2. AbdoolKarimQ,KharsanyAB,NaidooK,YendeN,GengiahT,OmarZ,
FranciscoHill;Pharmacist:JulieThompson;WollongongHospital,Wollongong ArulappanN,MlisanaKP,LuthuliLR,AbdoolKarimSS:Co-enrollmentin
(cid:129)DrRajaramRamadoss,JosetteWood;Pharmacist:EricTahWaiYap;Lyell multipleHIVpreventiontrials-experiencesfromtheCAPRISA004
McEwinHospital,Adelaide Tenofovirgeltrial.ContempClinTrials2011,32:333-338.
BrazilianInvestigators 3. LarntzK,NeatonJD,WentworthDN,YurikT:Dataanalysisissuesfor
(cid:129)DrMarceloGarciadaRocha(Co-Lead),AndréaKramer,MarthaHädrich; protocolswithoverlappingenrolment.StatMed1996,15:2445-2453.
Pharmacist:PatríciaSoares,FernandoFrosi;SantaCasaHospital,PortoAllegre