Clinical trials handbook design and conduct

In the context of trials, the design is that portion of the plan that specifiesthe treatments to be studied, how persons will be assigned to treatment, the numbers to be assigned sample s

Clinical Trials Handbook

Clinical Trials Handbook DESIGN AND CONDUCT

Curtis L Meinert

Bloomberg School of Public Health

The Johns Hopkins University

Baltimore, MD

A John Wiley & Sons, Inc., Publication

Copyright  2013 by John Wiley & Sons, Inc All rights reserved

Published by John Wiley & Sons, Inc., Hoboken, New Jersey

Published simultaneously in Canada

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section

107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at

http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose No warranty may be created or extended by sales representatives or written sales materials The advice and strategies contained herein may not be suitable for your situation You should consult with a professional where appropriate Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in a variety of electronic formats Some content that appears in print may not be available in electronic formats For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

1 Drugs–Testing–Handbooks, manuals, etc 2 Clinical trials–

Handbooks, manuals, etc I Title.

RM301.27.M45 2013

615.1072
4–dc23

2012015097 Printed in the United States of America

10 9 8 7 6 5 4 3 2 1

Acknowledgments xi

Preface xiii

On planning xv

Explanatory notes, focus, and conventions xvii

Abbreviations and designations xxi

5 Design and flow schematics 13

6 Design and operating principles 15

7 Counting and analysis rules 17

8 Multi-study umbrella name 19

39 Masking, censoring, and shielding specifications 113

40 Drug masking procedure 115

41 Drug packaging and labeling 117

42 Drug supply 121

43 Masking safeguards 123

44 Unmasking treatment assignment 125

45 Results blackouts 127

VI Bias and Variance Control 129

46 Bias control procedures 131

47 Stratification 135

48 Variance control procedures 139

49 Separations 141

VII Treatment Assignment/Randomization .143

50 Assignment methods: Fixed vs adaptive 145

51 Treatment assignment: Random vs nonrandom 147

52 Randomization: Complete vs restricted 151

59 Consent: Disclaimers and notifications 181

60 Consent: Principles and purpose 183

61 Consent: Process 185

62 Consent: Types 189

63 Consent: Questions and answers 191

IX Enrollment and Followup 195

82 Exclusions from enrollment 245

83 Eligibility and exclusions by reason 249

X Sample Size .251

84 Sample size: Design 253

85 Sample size: Specifications 257

viii C O N T E N T S

86 Sample size: Calculation 259

87 Fixed vs sequential sample size designs 261

88 Fixed vs adaptive designs 263

89 Designed subgroup comparisons 265

XI Data Collection and Processing 267

100.Form design: Principles and procedures 295

101.Time window specifications 299

102.Data entry design 301

103.Data sharing: Internal 307

104.Data sharing: External 311

XII Study Centers 315

112.Standing and working committees 341

113.Committee rules and procedures 343

114.Study officers 347

115.Study chair/vice-chair 349

C o n t e n t s ix

116.Executive committee 353

117.Executive committee members 357

118.Steering committee 359

119.Steering committee members 361

120.Steering committee: Questions, answers, and observations 363

121.Steering committee representation models 367

XV Treatment Effects Monitoring .371

122.Treatment effects monitoring 373

123.Treatment effects monitoring: Purpose 375

124.Treatment effects monitoring: Approach 377

125.Treatment effects monitoring: Masking 379

126.Stopping rules and guidelines 381

127.Treatment effects monitoring: Questions and answers 383

128.Treatment effects monitoring committee 387

129.Treatment effects monitoring committee: Questions and answers 391

XVI Quality Control and Assurance 393

130.Quality control and assurance procedures 395

131.Performance monitoring 399

132.Training procedures 401

133.Assurances and certifications 403

134.Site visiting procedures 405

135.Audit procedures 409

XVII Data Analysis 413

136.Analysis datasets 415

137.Analysis questions regarding study results publications 417

138.Frequentist vs Bayesian analysis 419

x C O N T E N T S

XIX Policies 441

146.Policies 443

147.Publicity policy 445

148.Policy on access to study documents 447

149.Policy on access to study data and results 449

150.Policy on advertising for patients 453

151.Policy on incentive payments 455

152.Policy on payment of patient-related travel expenses 457

153.Ancillary study policy 459

154.Policy on patient-care-related payments 461

155.Policy on conflicts of interest 463

164.Study history log 491

165.Landmark events and dates 493

166.Registration 495

Appendices 497

appendix1.Design summaries for selected finished trials 499

appendix2.Sample design slide sets 517

References 551

Index 557

I wish to thank the people of the Johns Hopkins Center for Clinical Trials for help

in producing this document—especially Betty Collison, Jill Meinert, Annette Wagoner,Karen Collins, Mark Van Natta, and Susan Tonascia

xi

The idea underlying randomized trials is deceptively simple: Create comparable groups ofpersons by randomization, treat them according to a prescribed protocol, follow them forthe outcome of interest, analyze the results, write up the results, and pass on to the nexttrial

How hard can it be? Decide on the treatment, define who is eligible, calculate asample size, recruit and randomize, treat and follow, and wait for fame and fortune, or ifnot fortune, then at least fame!

If only it were so

The difference between success and failure in trials is hundreds of details Details

of design and conduct to the trialist is what blueprints are to builders But trials donot come with blueprints They have to be developed, almost always in a ‘‘rush.’’ Thishandbook is produced in the hope of making that process easier and less error prone

statistician–methodologist in coordinating centers formulticenter trials Its roots are

in materials posted to trialsmeinertsway.com It took the form of a handbook in early

2000 but then lay fallow for several years while I was working on the 2nd edition of my

textbook Clinical Trials: Design, Conduct, and Analysis60and An Insider’s Guide to Clinical

Trials.53It came back to life a couple of years back as work on those books ebbed.This handbook consists of 21 sections, each containing 2 to 20 parts, depending

on the section, three appendices, reference list and index

Curtis L Meinert

Towson, Maryland

8 March 2012

xiii

On Planning

The traditional classroom paradigm for instructing students in trials is one involving aseries of steps First, students formulate questions amenable to testing Next, they designtrials suitable for addressing the questions After that they produce protocols (study plans)and necessary forms for recording data from the trials Then, somehow, they do the trials,analyze the results, write up the results, and hand the write-ups in for a grade

Design in the paradigm of experimentation is the portion of a plan that serves to

specify the variable to be manipulated (experimental variable), the levels or states of thevariable to be studied (treatments), and the way in which the treatments will be applied(protocol) In the context of trials, the design is that portion of the plan that specifiesthe treatments to be studied, how persons will be assigned to treatment, the numbers

to be assigned (sample size), the proportions assigned to the different treatment groups(assignment ratio), and the details of treatment

Clearly, planning is an activity that starts before any work can be done andproceeds over the life of the activity In the context of trials, it starts with the first glimmer

of an idea for a trial and continues to completion of the trial ‘‘Planning’’ is over the ‘‘life’’

of the trial

Much of the time devoted to ‘‘planning’’ will be spent in deciding how the trial is

to be organized, directed, and monitored and in arriving at closure on contentious issuesrelating to matters of policy and practice The likelihood is that more time will be spent

on matters relating to organization and ‘‘policy’’ than on the actual ‘‘science’’ of the trial.Groups can be expected to become more interested and assertive regarding matters

of policy and procedures as they ‘‘mature.’’ Topics not on the ‘‘radar screen’’ early inplanning will show up later Sooner or later, one can expect groups to want policiesconcerning rights to data, authorship, credits, presentations, and ancillary studies.Planners need to have the mindset of marathon runners, always with still anotherhill to climb They have to maintain the flexibility and patience needed for repairingplans that ‘‘come apart’’ or for revising plans that have been ‘‘set.’’ They need to recognizethe politics of decision making, and they need to be practiced in the art of compromise.They need a sense of timing to be able to know when to ‘‘hold them’’ and when to ‘‘foldthem.’’ They need to be able to sense the tolerance of the group at any point in time foraddressing aspects of organization or operation They need to know that even if the detail

is important, no good will be achieved if it is seen as a ‘‘detail’’ by the group

xv

Explanatory Notes, Focus, and Conventions

A clinical trial is an experiment done on human beings to determine the feasibility, safety,

or efficacy of a treatment.‘‘Clinical trial’’ as a publication type in PubMed is a Pre-planned

clinical study of the safety, efficacy, or optimum dosage schedule of one or more diagnostic, therapeutic, or prophylactic drugs, devices, or techniques in humans selected according to predetermined criteria of eligibility and observed for predefined evidence of favorable and unfavorable effects.

In 2010 there were 36,650 full-length publications indexed in PubMed to ‘‘clinicaltrial’’; 35,000 in English language journals About half of the 35,000 (19,500) were alsoindexed to the publication type ‘‘randomized controlled trial’’ (defined in PubMed as a

trial that involves at least one test treatment and one control treatment, concurrent enrollment and follow-up of the test- and control-treated groups, and in which the treatments to be administered are selected by a random process, such as the use of a random-numbers table).

About 20% (4000) of the randomized controlled trials were also indexed to the

publication type ‘‘multicenter study,’’ defined in PubMed as work consisting of a controlled

study executed by several collaborating institutions.

The language in this handbook is that of clinical trials Hence, the designation for a person enrolled in trials is patient Treatment is the experimental variable The treatments may be test treatments or control treatments The group of persons assigned to receive a study treatment is referred to as a treatment group.

The methods and procedures of trials are the same across a broad spectrum of trialslargely without regard to size, length of treatment, or choice of treatments The emphasisherein is on comparative trials motivated by underlying states of clinical equipoise.33Design and methods unique to developmental phase I and II trials are not covered.Broadly, randomized trials are of two design types: parallel treatment designs(designs in which treatment groups are comprised of different persons) or crossovertreatment designs (designs that provide for the administration of two or more studytreatments to the same person, one after another in a specified or random order, oftenwith a washout period between administrations) Much of what is covered herein applies

to either type of design, but with emphasis on parallel treatment designs

Definitions in this handbook are adapted from Clinical Trials Dictionary:

Terminology and Usage Recommendations58 and a second edition by the same name

published by Wiley (2012) Etymologies are from Webster’s New Collegiate Dictionary.95,96Terms defined in the handbook are listed under the heading ‘‘Definitions’’ in theindex

xvii

SOCA Studies of Ocular Complications of AIDS (umbrella structure for trials and

observational followup studies)80

FGCRT Foscarnet-Ganciclovir CMV Retinitis Trial (treatment trial)85

CRRT Cytomegalovirus Retinitis Retreatment Trial (treatment trial)84

HPCRT HPMPC Peripheral Cytomegalovirus Retinitis Trial (treatment trial)83MACRT Monoclonal Antibody CMV Retinitis Trial (treatment trial)82

GCCRT Ganciclovir Cidofovir CMV Retinitis Trial (treatment trial)81

UGDP University Group Diabetes Program (treatment trial)93

All the trials in the list above are multicenter and randomized, and they haveparallel treatment designs All have person as the randomization unit except the GLTwith eye as the randomization unit The list is a mix of investigator- and sponsor-initiatedtrials All are NIH-funded, some in conjunction with drug companies

Trials are done all over the world, but counts of where they are done are hard tocome by The closest that one can get to place of conduct with PubMed indexing is fromcountry of publication (There are no codes to indicate where studies are done.) Abouthalf (16,300) of the 35,000 2010 English language publications were published in U.S.journals Another third (9,700) were published in UK journals

Location is important because trials are subject to different rules and regulations,depending on place of conduct and funding source Some regulations are distinctly UnitedStates-centric for NIH-funded studies—for example, as with the mandate for validgender analysis88 and data sharing (http://grants.nih.gov/grants/guide/notice-files/NOT-OD-03-032.html)

The structure for review and approval of trials differs by country In the UnitedStates, this responsibility is vested in IRBs of institutions housing investigators doingtrials This means, for multicenter trials, that there can be as many IRBs as centers in thetrial Presently, central review, e.g., as required with NCI-funded trials is in addition toindividual reviews The structure in states of the European Union is centralized with theClinical Trials Directive (effective 4 April 2001) Part of its purpose was to harmonizeadministrative procedures in conduct of clinical trials including a centralized review oftrials to streamline approval processes A similar structure is needed in the United States

to streamline the IRB review process here

Another area of difference is in licensure of drugs, biologics, and medical devices.The United States has a centralized structure vested in the Food and Drug Administration

E x p l a n a t o r y N o t e s , F o c u s , a n d C o n v e n t i o n s xix

The European Union has moved toward harmonizing differences in member states incontrol and approval of medicinal products with the establishment of the EuropeanMedicines Agency (formerly the European Agency for the Evaluation of MedicinalProducts)

The coverage of regulations, principally as detailed in Section VIII and XX, isU.S.-centric Covering the waterfront of regulations is beyond the scope of this effort.The number of locales involved in trials can be seen from counts of registrations inclinicaltrials.gov for intervention studies with at least one study site in the countries listed(counts as of 20 December 2011):

1 Trials are under the stewardship of study investigators.

2 Trials have treatment effects monitoring committees/data and safety monitoring

committees that report directly to study leaders or simultaneously to sponsors andstudy leaders

3 Study investigators have unfettered rights to publication.

4 Centers in trials are free-standing and independent of study sponsors.

Abbreviations and Designations

A

AACTG Adult AIDS Clinical Trials Group

AD Alzheimer’s disease

ADAPT Alzheimer’s Disease Anti-inflammatory Prevention Trial

ADE adverse drug experience

ADR adverse drug reaction

CAMP Childhood Asthma Management Program

CAST Cardiac Arrhythmia Suppression Trial

CBET Chemoprevention for Barrett’s Esophagus Trial

CC coordinating center

CCC clinical coordinating center

CDP Coronary Drug Project

CL central laboratory

CMV cytomegalovirus

CO chair’s office

CONSORT Consolidated Standards of Reporting Trials

CRF case report form

CRRT CMV Retinitis Retreatment Trial

FDA Food and Drug Administration

FGCRT Foscarnet–Ganciclovir CMV Retinitis Trialfem feminine

FEV forced expiratory volume

FTE full-time equivalent

HCFA Health Care Financing Agency

HDFP Hypertension Detection and Followup ProgramHIPAA Health Insurance Portability and Accountability ActHIV human immunodeficiency virus

HPCRT HPMPC Peripheral CMV Retinitis Trial

HPT Hypertension Prevention Trial

I

ICMJE International Committee of Medical Journal Editors

ID infectious dose

Id, ID identification

IDE Investigational Device Exemption

IND Investigational New Drug

INDA Investigational New Drug Application

IOP intraocular pressure

IP internet protocol

IRB institutional review board

ITT intention-to-treat

IU international unit

MPS Macular Photocoagulation Studies

MRFIT Multiple Risk Factor Intervention Trial

N

n noun

NCI National Cancer Institute

NCR no carbon required

NDA New Drug Application

NETT National Emphysema Treatment Trial

NEI National Eye Institute

NEJM New England Journal of Medicine

neut neuter

NHLBI National Heart, Lung, and Blood Institute

NIH National Institutes of Health

NLM National Library of Medicine

NSAID non-steroidal anti-inflammatory drug

O

O observed

OE Old English

OF Old French

OHRP Office of Human Research Protections

OHG Old High German

OIt Old Italian

OMB Office of Management and Budget

ONF Old North French

ORI Office of Research Integrity

P

PC personal computer

pdf probability distribution function

xxiv A B B R E V I A T I O N S A N D D E S I G N A T I O N S

PEFR peak expiratory flow rate

PI principal investigator

pl plural

PO project office; project officer

POAG primary open angle glaucoma

RFA request for application

RFP request for proposal

T

TCC treatment coordinating center

TEMC treatment effects monitoring committeet.i.d three times daily

trt treatment

U

UGDP University Group Diabetes Program

UK United Kingdom

URL uniform resource locator

U.S United States (of America)

USPHS United States Public Health Service

General

Patient for persons studied

Treatment (any treatment group including placebo treatment)

Outcome or outcome measure (as opposed to event, endpoint)

Variable (as opposed to parameter)

Center director (as opposed to principal investigator)

Study treatment (any of the assigned treatment regimens)

Test treatment (any of the assigned treatments, except control treatments)

Control treatment (placebo treatment)

Terms avoided

Treatment failure (presumptive)

Informed consent (wishful thinking in the absence of information to indicate thatconsent is truly informed)

Endpoint (operational implications usually inconsistent with requirements forcontinued followup)

Placebo (as an adjective, e.g., as in placebo patients, or as a synonym for notreatment)

in study protocols and publications is confusing

Troublesome terms in trials include dropout, drop-in, endpoint, missed visit,and treatment failure Concepts not well understood and subject to misuse include thenotation of stratification versus subgroup analysis, subgroup analysis vs data dredging,randomization, internal validity versus external validity, bias (e.g., selection bias vs.treatment-related bias), and ‘‘statistical significance.’’

Some terms are best avoided in trials because of their multidisciplinary natureand especially in multicenter trials For example, investigator as a synonym for clinicianinvestigator (investigatorship is not limited to the clinical side of activities in trials)

4 I G E N E R A L

and principal investigator (PI) in multicenter trials where there are as many ‘‘principal’’investigators as there are centers

Q U E S T I O N S

• What practices do you intend to follow to standardize terminology?

• Do you have a list of terms to be avoided? If not, should you prepare such a list?

• Do you have operational definitions for enrolled (randomized), dropout, and missed visits?

• Are you planning to produce a glossary of terms and definitions for inclusion in the trial handbook? If yes, who will produce it and who will maintain it?

Dropout: Person missing three consecutive FU visits; person unable or unwilling to

continue under followup

Scheduled visit: Any visit required in the data collection schedule of the trial

Missed visit: A scheduled visit not made

Interim visit: Any followup visit after randomization over and above scheduled visits Protocol deviation: Any departure from the treatment, examination, or data

collection protocol

Adverse drug reaction: A drug reaction that results in hospitalization, or

prolongation of hospitalization, or that has negative health implications for

a study patient

R E L AT E D E N T R Y

Terminology (page 3)

N A R R AT I V E

The expectation is that terminology will be problematic Terms subject to misuse include:

• Are you keeping a list of terms to be defined as planning proceeds (recommended)?

• Have you looked at glossaries of definitions prepared and maintained by other groups doing trials (recommended)?

to decide whether to live with the variation (and accommodate for it in the way datacollection forms are constructed) or to eliminate it by imposing unit standards and/or

by centralizing measurements wherever possible Indeed, one of the reasons for centrallaboratories in multicenter trials is for standardization of measurement Standardization

is usually out of the question for determinations done at local hospital laboratories.Even standardization of the unit of measurement for simple measures such asheight and weight can be complicated If measures are to be made in centimeters andkilograms, clinics may have to be supplied with equipment calibrated in metric units

An appealing expedient is to allow measurement in the unit of choice of a clinic,but also to require personnel to convert to the desired unit when completing study forms.This approach is to be discouraged because data are not improved by ‘‘conversions.’’ It isbetter, in such cases, to design forms to allow for unit variation in reporting and to makethe necessary conversion by computer when data are harvested for inclusion in the studydatabase

The unit represented in presentations and publications should be as reported ondata forms If conversions are made for standardization or to satisfy journal editors, thepublication should make note of the conversion

Trial as a noun is from Anglo-French, trier, meaning to try The term can mean58: 1.

An experiment designed and carried out to provide information on the merits of onetreatment or procedure relative to another treatment or procedure; controlled trial 2 Anytentative or experimental procedure or treatment carried out, performed, or administered

to obtain data to arrive at some judgment or conclusion concerning the procedure ortreatment 3 One of a number of replications of an experiment, process, or procedure(e.g., Bernoulli trial) 4 The action or process of putting something to a test or proof 5.Something tried

Typically, the modifier clinical is used in reference to trials done in clinical settings,

those involving people with a disease or an adverse health condition, and those involvingtreatment of people for cure, amelioration, or prevention of disease or for betterment of

an adverse health condition

Trials come in many forms Hence, the term trial or clinical trial, in the absence of

other descriptors or modifiers, is not informative Some of the dimensions of classificationlisted below overlap For example, randomized trials are, by definition, controlled Hence,

controlled —e.g., as in randomized controlled trial —is redundant.

Also, any trial involving two or more study groups is comparative, but onlythose involving simultaneous enrollment of persons to the different treatment groups arecomparative, concurrent Trials involving use of historical controls for comparison arecomparative, nonconcurrent

Deterministic (e.g., as in play the winner schemes)

Best medical judgment

Treatment Preference

N A R R AT I V E

Consider using graphic representations for depicting design and flow processes In addition

to schematics of the design, as pictured above, schematics can be used to depict stepsand decision points in screening patients for enrollment, the ordering of proceduresperformed in relation to a given examination, the steps in preparing and shipping blood

to a central laboratory, and the steps in receiving and processing data for inclusion in thestudy database

Avoid unnecessary clutter and words for crispness and simplicity

6 D e s i g n a n d O p e r a t i n g P r i n c i p l e s 15

6 Design and operating principles

S L I D E

Studies of Ocular Complications of AIDS (SOCA) trials80

Common study protocol; all clinics expected to participate

Randomization as line of demarcation between the baseline and followup periodsPersons counted as enrolled when randomized

Analysis by treatment assignment

Contiguous time windows for followup visits

Followup regardless of course of treatment

R E L AT E D E N T R Y

Counting and analysis rules (page 17)

Typical design and operating principles

• Common study protocol (a specification in multicenter trials in which all clinicsare expected to follow the same protocol; important in that the protocol has to

be written with a common understanding; requirement precludes substudies as ameans of accommodation, i.e., studies in which only certain clinics participate)

• Persons counted as enrolled when randomized

• Person counted as randomized when treatment assignment is revealed to clinicpersonnel

• Time of randomization used to mark end of the baseline period of data collectionand the start of the followup period of data collection

• Contiguous time windows for followup visits (operationally means a visit may not

be done before the designated window opens and is counted as missed if not donewithin the designated time window)

• Followup regardless of course of treatment (operational impact is to require clinics

to continue to follow persons even if they are no longer receiving the assignedtreatment)

• Followup regardless of intercurrent events (operational implication is to requireclinics to continue to follow persons once enrolled regardless of intercurrent events)

• Analysis by treatment assignment

3 All events occurring from the moment of randomization forward in time counted

4 Events counted to the assigned treatment group regardless of treatment being

administered at the time of the event and regardless of degree of treatmentcompliance at the time of the event

Analysis rules

1 Primary analysis by treatment assignment

2 Analyses for outcome subset (e.g., deaths due to cardiovascular causes) preceded by

analyses of the entire outcome set (e.g., deaths, regardless of cause)

3 Analyses for composite outcome preceded by analyses for the component parts of

the composite outcome

4 Subgroup analyses limited to variables observed at or prior to randomization

The rules outlined above cannot be satisfied if:

1 Followup and data collection terminates when a person experiences a nonfatal

‘‘endpoint’’ (see page 39 for usage note)

2 Persons are dropped or withdrawn from the trial when they are no longer able or

willing to take or receive the assigned treatment

3 There are not continuing efforts to collect minimal information on dropouts

Questions relevant to formulating rules

• When is a person considered enrolled?

• What is the event or act defining enrollment?

• Is the baseline period of observation closed when the person is randomized?(Reminder: A ‘‘yes’’ means that any observation made after randomization, evenmoments after randomization, is not used as baseline data.)

• Is followup dependent on treatment or compliance? (Reminder: A ‘‘yes’’ implies alikely violation of one or more of the rules listed above.)

18 I G E N E R A L

• Will clinics be allowed to drop persons from followup when they have an ‘‘endpoint’’

or when treatment ‘‘fails’’? (Reminder: A ‘‘yes’’ implies a likely violation of one ormore of the rules listed above.)

• Will an effort be made over the course of trial to remain in touch with dropoutsand to provide minimal information on them including vital status?

multi-study n - A study having two or more studies performed under the same

organizational structure; multi-trial when studies are trials Usage note: Not to be

confused with a study having a series of substudies

umbrella study name n - A study name encompassing a number of studies, e.g., Studies

of Ocular Complications of AIDS.80

N A R R AT I V E

Umbrella names are akin to family names, and study names are akin to given names, e.g.,Studies of Ocular Complications of AIDS as the family name and Foscarnet–GanciclovirCMV Retinitis Trial a given name to yield ‘‘Studies of Ocular Complications of AIDS:Foscarnet–Ganciclovir CMV Retinitis Trial.’’

A fair number of multi-trials emerge from the structures created at the outsetfor doing a single trial, e.g., as with the Macular Photocoagulation Study (MPS) TheMPS came into being as a result of investigator initiative to evaluate laser-inducedphotocoagulation of neovascularization associated with age-related macular degeneration.Ultimately, the Macular Photocoagulation Research Group carried out several trials underthe MPS structure.48 – 50One surmises, with the perspective of hindsight, that investigatorswould have preferred to have had an umbrella name for the collection of trials performed

As it was, they had to use the name used for the first trial also as the umbrella name;sometimes confusing

Often the name of a group comes to serve the function of an umbrella name, e.g.,

as with the Eastern Cooperative Oncology Group (ECOG) Originally the name referred

to a group of collaborating oncologists from the eastern region of the United States, butnow it refers to a network of researchers from public and private institutions from acrossthe country and beyond

Originally the Adult AIDS Clinical Trials Group (AACTG) was the surname for

a group doing AIDS trials in adult populations, but now it serves as the name for a muchbroader set of activities

Names, once established, are difficult to change—a fact to be kept in mind whenestablishing a name

Reminders and recommendations

• Choose in favor of brevity, crispness, and succinctness

• Avoid restrictive terms likely to render a name obsolete later on

• Choose a neutral, nonpromotional, name

• Consider in conjunction with likely names of particular studies; avoid redundancies

or contradictory terms when used in conjunction with study names

20 I G E N E R A L

• Keep in mind that the name chosen will be used in publication titles

• Keep other likely uses in mind, as in funding applications, presentations, and otherwritings

• Keep contractions of name and acrostics or acronyms in mind when choosing

• Avoid choosing to create a desired or ‘‘cute’’ acronym

• Avoid names producing undesirable shortened name from letters of the name (ameaningless sequence of letters in one language can have a specific meaning inanother)

A study name should be informative, succinct, accurate, and telegraphic The preferred

base term in the name is trial, because it is the most accurate descriptor of the study Other base terms like study (e.g., Coronary Artery Surgery Study20), project (e.g., Coronary

Drug Project18), or program (e.g., University Group Diabetes Program93) are used, but they

are less informative than trial.

The name should include appropriate modifiers, such as randomized, controlled, and masked or blind to characterize the nature of the trial The name may also include

terms to characterize the phase of the trial (phase I, II, III, or IV) and, perhaps, termsintended to convey information about the treatment structure (e.g., parallel, crossover,

or factorial) The name should contain terms intended to indicate the type of treatmentsbeing tested (e.g., drugs, vaccines, diets, etc.) and the condition or disease under study(e.g., hypertension, diabetes, prostate cancer) Also the name may contain demographicterms to indicate the population enrolled (e.g., women, elderly, African-Americans).The name should remain accurate in the presence of changes to the study designduring conduct Everyday life is rich in names that have been sapped of their originalmeaning; Big Ten, Motel 6, Dime Savings, and Dollar Car Rental to name a few Theneed for accuracy in names of trials argues for staying clear of descriptors related toselection criteria because they may change over the course of enrollment

One has to be careful in the use of descriptors of the treatment being tested The

term drug in Coronary Drug Project was well chosen because all the treatments tested were

drugs (if one is willing to accept placebo—the control treatment—as a ‘‘drug’’) Oneshould steer clear of such descriptors if there is likelihood of adding treatments during thetrial that are not members of the class reflected in the name

Names indicative of site (e.g., Oslo Diet and Exercise Study) are useful only so long

as the trial remains confined to the site The term National or International to indicate

spread is questionable Both terms are subject to being rendered inaccurate with expansion

of ‘‘National’’ to ‘‘International’’ or contraction from ‘‘International’’ to ‘‘National’’ withthe loss of sites during the trial

Redundant terms, such as National and Cooperative in National Cooperative

Gallstone Study44 or University and Group in University Group Diabetes Program,93 are

best avoided Repeating a term, e.g., as with Study in African American Study of Kidney

Disease and Hypertension Pilot Study,101should be avoided if the reference is to a singlestudy

Most study names are shortened for everyday use Hence, the National CooperativeGallstone Study is also the NCGS, the Coronary Drug Project is also the CDP, and theUniversity Group Diabetes Program is also the UGDP

Sometimes groups choose study names to generate pronounceable words fromletters of the names (acronyms) However, that practice is questionable, to the extent that

22 I G E N E R A L

it leads to contrived study names The practice should be to select an appropriate nameand then the shortened name, not the reverse

Reminders and recommendations

• Choose in favor of brevity; the fewer the characters, the better

• Choose being mindful of the use of the name in publications and study documents

• Choose in favor of a neutral, nonpromotional name

• Consider in conjunction with use in mastheads of study publications

• Keep likely uses in mind, as in funding applications, publications, presentations,and other study documents

• Keep likely contractions of name and acronyms in mind when choosing

• Avoid choosing to create a desired or ‘‘cute’’ acronym

• Avoid unnecessary words

Characteristics of a good name

• Succinct

• Neutral one not favoring a particular treatment over another

• Robust; does not become obsolete or inaccurate with changes to the trial, e.g., as in

use of National when the study is expanded to include sites in other countries

• Indicates nature of treatment and population being studied

• Includes the term trial and other currency terms like randomized

• Does not contain unprintable graphic characters

• Does not contain abbreviations

Numbering

• Generally unnecessary

• Avoid if trial is a descendent of one done by another group

• Questionable if follow-on trial involves a distinctly different or expanded studypopulation in which treatment regimens or treatment groups are different

• Numbering (e.g., the XYZ Trial II) acceptable when the trial is largely a repeat of

the previous namesake or where the treatments are the same but the population

is more restrictive (e.g., as in PARIS II43); otherwise use other means to showconnection, e.g., as the Coronary Drug Project Aspirin Study17

• Things to remember about numbering:

❑ Numbering helps remind readers of other related publications

❑ The connection may not be advantageous if the precursor study was a ‘‘bust.’’

❑ The connection can serve to erode trust in the precursor study if the one beingnamed turns out to be a ‘‘bust.’’

❑ Numbering usually indicates previous success; Rocky II came about because ofthe success of Rocky; in regard to ships there was a Queen Elizabeth 2 (QE2),but no Titanic 2

9 S t u d y N a m e 23

Shortened names

• Useful

• Avoid creating study names to produce pronounceable acronyms

• Focus on name first, then on producing a shortened name; avoid the reverse ofstarting with an acronym and then fashioning a name to match the acronym

• Stress test before adopting (i.e., by use in different types of settings and by screeningfor different meanings in other settings or languages)

Design Specifications

1 0 O b j e c t i v e 27

10 Objective

S L I D E

Objective: Coronary Drug Project (CDP)18

To evaluate the efficacy of different lipid-influencing drugs in prolonging life ofmen with a history of MI

R E L AT E D E N T R Y

Specific aims (page 29)

N A R R AT I V E

The objective, in the case of a trial, is the reason for undertaking it Typically, the reason

is to assess the efficacy, or the safety and efficacy, of a named treatment or set or class oftreatments

The objective is limited to issues of safety or to determination of dosage in phase Iand II trials It relates to issues of feasibility in pilot or feasibility trials, as well as ‘‘proof’’

or ‘‘demonstration’’ in the case of demonstration trials

The objective should be written with an appreciation of the limitations thatinvestigators are likely to face regarding their ability to recruit, treat, and follow persons

in the trial

Objectives change over the course of time The evolution is toward greaterspecificity as planning proceeds to implementation Usually, objectives are stated in moregeneral (and sometimes grandiose) fashion in funding initiatives than as written when thetrial is implemented

The objective should reflect the realities and limitations of the trial Therefore, itmust be revised and refined to reflect the facts and limitations of the trial, as they areimposed or become known In this sense, ‘‘planning’’ continues over the course of thetrial Changes to the protocol when the trial is underway may require changes in thestatement of objective

Normally, there will be only one objective However, there are occasions when it isnecessary or expedient to list several In such cases, they should be arranged in descendingorder of importance relative to the primary objective

The Coronary Drug Project18 was a grant-funded, investigator-initiated trial Itsprimary objective was as stated in the slide above Secondary objectives were:

1 To characterize the natural history and clinical course of coronary heart disease

2 To develop methods and procedures suited to the design and conduct of long-term,

large, multicenter clinical trials

Q U E S T I O N S

• Is the objective realistic, given the scope of the trial and resources available?

• Is the objective succinctly stated?

28 I I D E S I G N S P E C I F I C A T I O N S

• Is it ‘‘stand alone,’’ i.e., understandable and meaningful by itself, without other supporting statements or documents?

• Does it name the treatment or class of treatments to be tested?

• Does the statement indicate the outcome or measure used to assess the treatments?

• Does the statement indicate specify the population to be treated?

• If there are multiple objectives:

❑ Is the primary objective (first in the list) the fundamental reason for undertaking the trial?

❑ Are they arranged in descending order of importance relative to the primary objective?

❑ Are the objectives compatible?

❑ Are they relevant and realistic?

1 1 S p e c i fi c A i m s 29

11 Specific aims

S L I D E

Specific aims: Studies of Ocular Complications of

AIDS — 1988 grant application

• Develop and maintain core structure for support of multicenter studies

• Design and conduct multicenter randomized trials of treatments for AIDS-inducedocular complications

• Design and conduct multicenter epidemiological studies of persons with induced ocular complications

Q U E S T I O N S

• Do the aims relate to the stated objectives?

• Are they realistic?

• Are they consistent with available resources and scope of activities planned?

• Are they succinctly stated?

• Are they several in number? If yes, can the number be reduced by combining, rewording, or recasting?

• Are the methods and procedures for achieving the specific aims spelled out?