[Federal Register Volume 60, Number 163 (Wednesday, August 23, 1995)]
[Notices]
[Pages 43910-43934]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 95-20813]



      

[[Page 43909]]

_______________________________________________________________________

Part IV





Department of Health and Human Services





_______________________________________________________________________



Food and Drug Administration



_______________________________________________________________________



International Conference on Harmonisation; Draft Guideline on Structure 
and Content of Clinical Study Reports; Availability; Notice

  Federal Register / Vol. 60, No. 163 / Wednesday, August 23, 1995 / 
Notices   

[[Page 43910]]


DEPARTMENT OF HEALTH AND HUMAN SERVICES

Food and Drug Administration
[Docket No. 95D-0218]


International Conference on Harmonisation; Draft Guideline on 
Structure and Content of Clinical Study Reports; Availability

AGENCY: Food and Drug Administration, HHS.

ACTION: Notice.

-----------------------------------------------------------------------

SUMMARY: The Food and Drug Administration (FDA) is publishing a draft 
guideline entitled ``Structure and Content of Clinical Study Reports.'' 
This guideline was prepared under the auspices of the International 
Conference on Harmonisation of Technical Requirements for Registration 
of Pharmaceuticals for Human Use (ICH). The draft guideline is intended 
to facilitate the compilation of a single worldwide core clinical study 
report acceptable to all regulatory authorities.

DATES: Written comments by October 10, 1995.

ADDRESSES: Submit written comments on the draft guideline to the 
Dockets Management Branch (HFA-305), Food and Drug Administration, rm. 
1-23, 12420 Parklawn Dr., Rockville, MD 20857. Copies of the draft 
guideline are available from the CDER Executive Secretariat Staff (HFD-
8), Center for Drug Evaluation and Research, Food and Drug 
Administration, 7500 Standish Pl., Rockville, MD 20855.
FOR FURTHER INFORMATION CONTACT: 
    Regarding the guideline: Robert Temple, Center for Drug Evaluation 
and Research (HFD-100), Food and Drug Administration, 5600 Fishers 
Lane, Rockville, MD 20857, 301-594-6758.
    Regarding the ICH: Janet J. Showalter, Office of Health Affairs 
(HFY-20), Food and Drug Administration, 5600 Fishers Lane, Rockville, 
MD 20857, 301-827-0864.

SUPPLEMENTARY INFORMATION: In recent years, many important initiatives 
have been undertaken by regulatory authorities and industry 
associations to promote international harmonization of regulatory 
requirements. FDA has participated in many meetings designed to enhance 
harmonization and is committed to seeking scientifically based 
harmonized technical procedures for pharmaceutical development. One of 
the goals of harmonization is to identify and then reduce differences 
in technical requirements for drug development among regulatory 
agencies.
    ICH was organized to provide an opportunity for tripartite 
harmonization initiatives to be developed with input from both 
regulatory and industry representatives. FDA also seeks input from 
consumer representatives and others. ICH is concerned with 
harmonization of technical requirements for the registration of 
pharmaceutical products among three regions: The European Union, Japan, 
and the United States. The six ICH sponsors are the European 
Commission, the European Federation of Pharmaceutical Industries 
Associations, the Japanese Ministry of Health and Welfare, the Japanese 
Pharmaceutical Manufacturers Association, the Centers for Drug 
Evaluation and Research and Biologics Evaluation and Research, FDA, and 
the Pharmaceutical Research and Manufacturers of America. The ICH 
Secretariat, which coordinates the preparation of documentation, is 
provided by the International Federation of Pharmaceutical 
Manufacturers Associations (IFPMA).
    The ICH Steering Committee includes representatives from each of 
the ICH sponsors and the IFPMA, as well as observers from the World 
Health Organization, the Canadian Health Protection Branch, and the 
European Free Trade Area.
    At a meeting held on March 29, 1995, the ICH Steering Committee 
agreed that a draft guideline entitled ``Structure and Content of 
Clinical Study Reports'' should be made available for public comment. 
The draft guideline is the product of the Efficacy Expert Working Group 
of the ICH. Comments about this draft will be considered by FDA and the 
Expert Working Group. Ultimately, FDA intends to adopt the ICH Steering 
Committee's final guideline.
    The draft guideline is intended to facilitate the compilation of a 
single worldwide core clinical study report acceptable to all 
regulatory authorities. In general, once the guideline is in use, the 
only differences between submissions to various authorities should be 
in the amount of data submitted initially in tabular listings, in the 
number of case report forms submitted, and in specific supplemental 
analyses requested for particular cases by an authority, if any. The 
clinical study report described in this draft guideline is an 
integrated full report of an individual study of any therapeutic, 
prophylactic, or diagnostic agent conducted in patients. The clinical 
and statistical description, presentations, and analyses should be 
integrated into a single report, incorporating tables and figures into 
the main text of the report, or at the end of the text. Appendices 
should contain the protocol, sample case report forms, investigator 
information, trial material information, technical statistical 
documentation, related publications, patient data listings, and 
technical statistical details such as derivations, computations, 
analyses, and computer output, and so on. The draft guideline is 
intended to assist sponsors in the development of a report that is 
complete, free from ambiguity, well organized, and easy to review.
    In the past, guidelines have generally been issued under 
Sec. 10.90(b) (21 CFR 10.90(b)), which provides for the use of 
guidelines to state procedures or standards of general applicability 
that are not legal requirements but are acceptable to FDA. The agency 
is now in the process of revising Sec. 10.90(b). Therefore, this 
guideline is not being issued under the authority of Sec. 10.90(b), and 
it does not create or confer any rights, privileges, or benefits for or 
on any person, nor does it operate to bind FDA in any way.
    Interested persons may, on or before October 10, 1995, submit 
written comments to the Dockets Management Branch (address above) on 
the draft guideline. Two copies of any comments are to be submitted, 
except that individuals may submit one copy. Comments are to be 
identified with the docket number found in brackets in the heading of 
this document. The draft guideline and received comments may be seen in 
the office above between 9 a.m. and 4 p.m., Monday through Friday.
    The text of the draft guideline follows:

Structure and Content of Clinical Study Reports

Introduction to the Guideline

    The objective of this guideline is to allow the compilation of a 
single worldwide core clinical study report acceptable to all 
regulatory authorities. The regulatory authority specific additions 
will consist of modules to be considered as appendices, available 
upon request.
    The clinical study report described in this guideline is an 
``integrated'' full report of an individual study of any 
therapeutic, prophylactic, or diagnostic agent (referred to herein 
as drug or treatment) conducted in patients, in which the clinical 
and statistical description, presentations, and analyses are 
integrated into a single report, incorporating tables and figures 
into the main text of the report, or at the end of the text, and 
with appendices containing the protocol, sample case report forms, 
investigator information, trial material information, technical 
statistical documentation, related publications, patient data 
listings, and technical statistical details such as derivations, 
computations, analyses, and computer output. The integrated full 
report 

[[Page 43911]]
of a study should not be derived by simply joining a separate clinical 
and statistical report. Although this guideline is mainly aimed at 
efficacy and safety trials, the basic principles and structure 
described can be applied to other kinds of trials, such as clinical 
pharmacology studies.
    The guideline is intended to assist sponsors in the development 
of a report that is complete, free from ambiguity, well organized, 
and easy to review. The report should provide a clear explanation of 
how the critical design features of the study were chosen and enough 
information on the plan, methods, and conduct of the study so that 
there is no ambiguity in how the study was carried out. The report 
with its appendices should also provide enough individual patient 
data, including the demographic and baseline data, and details of 
analytical methods, to allow replication of the critical analyses 
when authorities wish to do so. It is also particularly important 
that all analyses, tables, and figures carry, in text or as part of 
the table, clear identification of the set of patients from which 
they were generated.
    Depending on the regulatory authority's review policy, 
abbreviated reports using summarized data or with some sections 
deleted may be acceptable for uncontrolled studies or other studies 
not designed to establish efficacy (but a controlled safety study 
should be reported in full), for seriously flawed or aborted 
studies, or for controlled studies which examine conditions clearly 
unrelated to those for which a claim is made. However, a full 
description of safety aspects should be included in these cases. If 
an abbreviated report is submitted, there should be enough detail of 
design and results to allow the regulatory authority to determine 
whether a full report is needed. If there is any question regarding 
whether the reports are needed, it may be useful to consult the 
regulatory authority.
    In presenting the detailed description of how the study was 
carried out, it may be possible simply to restate the description in 
the initial protocol. Often, however, it is possible to present the 
methodology of the study more concisely in a separate document. In 
each section describing the design and conduct of the study, it is 
particularly important to clarify features of the study that are not 
well-described in the protocol and identify ways in which the study 
as conducted differed from the protocol, and to discuss the 
statistical methods and analyses used to account for these 
deviations from the planned protocol.
    The full integrated report of the individual study should 
include the most detailed discussion of individual adverse events or 
laboratory abnormalities, but these should usually also be 
reexamined as part of an overall safety analysis of all available 
data in any application.
    The report should describe demographic characteristics of the 
study population and, where the study is large enough to permit 
this, present data for demographic (e.g., age, sex, race, weight) 
and other (e.g., renal or hepatic function) subgroups so that 
possible differences in efficacy or safety can be identified. 
Usually, however, subgroup responses should be examined in the 
larger data base used in the overall analysis.
    The data listings requested as part of the report (usually in an 
appendix) are those needed to support critical analyses, not the 
more comprehensive archival data listings required by some 
regulatory authorities. Data listings which are part of the report 
should be readily usable by the reviewer. Thus, although it may be 
desirable to include many variables in a single listing to limit 
size, this should not be at the expense of clarity. An excess of 
data should not be allowed to lead, for example, to overuse of 
symbols instead of words or easily understood abbreviations or to 
too small displays. In this case, it is preferable to produce 
several listings.
    Data should be presented in the report at different levels of 
detail: overall summary figures, and tables for important 
demographic, efficacy, and safety variables may be placed in the 
text to illustrate important points; other summary figures, tables, 
and listings for demographic, efficacy, and safety variables should 
be provided in section 14; individual patient data for specified 
groups of patients should be provided as listings in Appendix 16.2; 
and all individual patient data (U.S. archival listings) should be 
provided in Appendix 16.4.
    In any table, figure, or data listing, estimated or derived 
values, if used, should be identified in a conspicuous fashion. 
Detailed explanations should be provided as to how such values were 
estimated or derived and what underlying assumptions were made.
    The guidance provided below is detailed and is intended to 
notify the applicant of virtually all of the information that should 
routinely be provided so that postsubmission requests for further 
data clarification and analyses can be reduced as much as possible. 
Nonetheless, specific requirements for data presentation and/or 
analysis may depend on specific situations, may evolve over time, 
may vary from drug class to drug class, may differ among regions, 
and cannot be described in general terms. It is therefore important 
to refer to specific clinical guidelines and to discuss data 
presentation and analyses with the reviewing authority, whenever 
possible. Detailed written guidance on statistical approaches is 
available from some authorities.
    Each report should consider all of the topics described (unless 
clearly not relevant) although the specific sequence and grouping of 
topics may be changed if alternatives are more logical for a 
particular study. Some data in the appendices are specific 
requirements of individual regulatory authorities and should be 
submitted as appropriate. The numbering should then be adapted 
accordingly.
    In the case of very large trials, some of the provisions of this 
guideline may be impractical or inappropriate. When planning and 
when reporting such trials, contact with regulatory authorities to 
discuss an appropriate report format is encouraged.
    The provisions of this guideline should be used in conjunction 
with other ICH guidelines.

Structure and Content of Clinical Study Reports

1. Title Page

    The title page should contain the following information:
    - Report title
    - Name of drug
     - Indication studied
    - If not apparent from the title, a brief (one to two sentences) 
description giving design (parallel, cross-over, blinding, 
randomized) comparison (placebo, active, dose/response), duration, 
dose, and patient population
    - Name of the sponsor
    - Protocol identification
    - Drug development phase
    - Study initiation date (first patient enrolled, or any other 
verifiable definition)
    - Date of early study termination, if any
    - Study completion date (last patient completed)
    - Name and affiliation of signatory (principal) investigator(s) 
or responsible medical officer
    - Name of company signatory (the person responsible for the 
study report within the company.) The name, telephone number, and 
fax number of the company contact persons for questions arising 
during review of the study report should be indicated on this page 
or in the letter of application.
    - Statement indicating whether the study was performed in 
compliance with good clinical practices (GCP's), including the 
archiving of essential documents
    - Date of the report (identify any earlier reports from the same 
study by title and date)

2. Synopsis

    A brief synopsis (maximum: three pages) that summarizes the 
study should be provided (see Annex I of the guideline for an 
example). The synopsis should include numerical data to illustrate 
results, not just text or p-values.

3. Table of Contents for the Individual Clinical Study Report

    The table of contents should include:
    - The page number or other locating information of each section, 
including summary tables, figures, and graphs;
    - A list and the locations of appendices, tabulations, and any 
case report forms provided.

4. List of Abbreviations and Definitions of Terms

    A list of the abbreviations and definitions of unusual or 
specialized terms or measurements units used in the report should be 
provided. In addition, abbreviated terms should be spelled out and 
the abbreviation indicated at first appearance in the text.

5. Ethics

5.1 Independent Ethics Committee (IEC) or Institutional Review Board 
(IRB)
    It should be confirmed that the study and any amendments were 
reviewed by an independent ethics committee or institutional review 
board. A list of all IEC's or IRB's consulted should be given in 
Appendix 16.1.3 and, if required by the regulatory authority, the 
name of the committee Chair should be provided.

[[Page 43912]]

5.2 Ethical Conduct of the Study
    It should be confirmed that the study was conducted in 
accordance with the ethical principles that have their origins in 
the Declaration of Helsinki.
5.3 Patient Information and Consent
     How and when informed consent was obtained in relation to 
patient enrollment (e.g., at allocation, prescreening) should be 
described.
    Representative written information for the patient (if any) and 
a sample patient consent form should be provided in Appendix 16.1.3.

6. Investigators and Study Administrative Structure

    The administrative structure of the study (e.g., principal 
investigator, coordinating investigator, controller, steering 
committee, administration, monitoring and evaluation committees, 
institutions, statistician, central laboratory facilities, clinical 
research organization, clinical trial supply management) should be 
described briefly.
    A list of the investigators with their affiliations, their role 
in the study, and their qualifications (curriculum vitae or 
equivalent), and a similar list for other persons whose 
participation materially affected the conduct of the study should be 
provided in Appendix 16.1.4. In the case of very large trials with 
many investigators, the above information may be abbreviated to 
consist of general statements of qualifications for persons carrying 
out particular roles in the study with only the name, degree, and 
institutional affiliation and roles of each investigator or other 
participant.
     The listing should include:
     (a) Investigators.
    (b) Any person carrying out important study observations, such 
as a nurse, physician's assistant, clinical psychologist, clinical 
pharmacist, or house staff physician. It is not necessary to include 
in this list a person with only an occasional role, e.g., an on-call 
physician who dealt with a possible adverse effect or a temporary 
substitute for any of the above.
    (c) The author(s) of the report, including the responsible 
biostatistician(s).
    Where signatures of the principal investigators are required by 
regulatory authorities, these should be included in Appendix 16.1.5 
(see Annex II for a sample form). Where these are not required, the 
signature of the responsible medical officer should be provided in 
Appendix 16.1.5.

7. Introduction

    The introduction should contain a brief statement (maximum: one 
page) placing the study in the context of the drug's development, 
relating the critical features of the study (e.g., rationale and 
aims, target population, treatment, duration, primary endpoints) to 
that development. Any guidelines which were followed in the 
development of the protocol or any other agreements/meetings between 
the Company and regulatory authorities which are relevant to the 
particular study should be identified or described.

8. Study Objectives

    A statement describing the overall purpose(s) of the study 
should be provided.

9. Investigational Plan

9.1 Overall Study Design and Plan - Description
    The overall study plan and design (configuration) of the study 
(e.g., parallel, cross-over) should be described briefly but clearly 
using charts and diagram as needed. If other studies used a very 
similar protocol, it may be useful to note this and describe any 
important differences. The actual protocol and any changes should be 
included as Appendix 16.1.1 and a sample case report form (unique 
pages only; i.e., it is not necessary to include identical pages 
from forms for different evaluations or visits) as Appendix 16.1.2. 
If any of the information in this section comes from sources other 
than the protocol, these should be identified.
    The information provided should include:
    - Treatments studied (specific drugs, doses, and procedures).
    - Patient population studied and the number of patients to be 
included.
    - Level and method of blinding/masking (e.g., open, double-
blind, single-blind, blinded evaluators, and unblinded patients and/
or investigators).
    - Kind of control(s) (e.g., placebo, no treatment, active drug, 
dose-response, historical) and study configuration (parallel, cross-
over).
    - Method of assignment to treatment (randomization, 
stratification).
    - Sequence and duration of all study periods, including 
prerandomization and post-treatment periods, therapy withdrawal 
periods, and single- and double-blind treatment periods. The timing 
of randomization should be specified. It is usually helpful to 
display the design graphically with a flow chart which includes 
timing of assessments (see Annexes IIIa and IIIb for an example).
    - Any safety, data monitoring, or special steering or evaluation 
committees.
    - Any interim analyses.
9.2 Discussion of Study Design, Including the Choice of Control 
Groups
    The specific control chosen and the study design used should be 
discussed, as necessary. Examples of design issues meriting 
discussion follow.
    Generally, the control (comparison) groups that are recognized 
are placebo concurrent control, no treatment concurrent control, 
active treatment concurrent control, dose comparison concurrent 
control, and historical control. In addition to the type of control, 
other critical design features that may need discussion are use of a 
cross-over design and selection of patients with particular prior 
history, such as response or nonresponse to a specific drug or 
member of a drug class. If randomization was not used, it is 
important to explain how other techniques, if any, guarded against 
systematic selection bias.
    Known or potential problems associated with the study design or 
control group chosen should be discussed in light of the specific 
disease and therapies being studied. For a cross-over design, for 
example, there should be consideration, among other things, of the 
likelihood of spontaneous change in the disease and of carry-over 
effects of treatment during the study.
    If efficacy was to be demonstrated by showing equivalence, i.e., 
the absence of a specified degree of inferiority of the new 
treatment compared to an established treatment, problems associated 
with such study designs should be addressed. Specifically there 
should be provided a basis for considering the study capable of 
distinguishing active from inactive therapy. Support may be provided 
by an analysis of previous studies similar to the present study with 
respect to important design characteristics (e.g., patient 
selection, study endpoints, duration, dose of active control, 
concomitant therapy) showing a consistent ability to demonstrate 
superiority of the active control to placebo. How to assess the 
ability of the present study to distinguish effective from 
ineffective therapy should also be discussed. For example, it may be 
possible to identify a treatment response (based on past studies) 
that would clearly distinguish between the treated population and an 
untreated group. Such a response could be the change of a measure 
from baseline or some other specified outcome like healing rate or 
survival rate. Attainment of such a response would support the 
expectation that the study could have distinguished the active drug 
from an inactive drug. There should also be a discussion of the 
degree of inferiority of the therapy (often referred to as the delta 
value) the study was intended to show was not exceeded.
    The limitations of historical controls are well known (e.g., 
difficulty of assuring comparability of treated groups, inability to 
blind investigators to treatment, change in therapy/disease, 
difference due to placebo effect) and deserve particular attention.
    Other specific features of the design may also deserve 
discussion, including presence or absence of washout periods and the 
duration of the treatment period, especially for a chronic illness. 
The rationale for dose and dose-interval selection should be 
explained, if it is not obvious. For example, once daily dosing with 
a short half-life drug whose effect is closely related in time to 
blood level is not usually effective; if the study design uses such 
dosing, this should be explained, e.g., by pointing to 
pharmacodynamic evidence that effect is prolonged compared to blood 
levels. The procedures used to seek evidence of ``escape'' from drug 
effect at the end of the dose-interval, such as measurements of 
effect just prior to dosing, should be described. Similarly, in a 
parallel design dose-response study, the choice of doses should be 
explained.
9.3 Selection of Study Population
9.3.1 Inclusion Criteria
    The patient population and the selection criteria used to enter 
the patients into the study should be described, and the suitability 
of the population for the purposes of the study discussed. Specific 
diagnostic criteria used, as well as specific disease requirements 
(e.g., disease of a particular severity or duration, results of a 
particular test or physical examination, particular features of 
clinical history, such as failure or success on prior therapy, or 
other potential prognostic factors, and any age, sex, or ethnic 
factors) should be presented.

[[Page 43913]]

    Screening criteria and any additional criteria for randomization 
or entry into the drug treatment part of the trial should be 
described. If there is reason to believe that there were additional 
entry criteria, not defined in the protocol, the implications of 
these should be discussed. For example, some investigators may have 
excluded, or entered into other studies, patients who were 
particularly ill or who had particular baseline characteristics.
9.3.2 Exclusion Criteria
    The criteria for exclusion at entry into the study should be 
specified and the rationale (e.g., safety concerns, administrative 
reasons, or lack of suitability for the trial) provided. The impact 
of exclusions on the generalisability of the study should be 
discussed in the study report or in an overview of safety and 
effectiveness.
9.3.3 Removal of Patients From Therapy or Assessment
    The predetermined reasons for removing patients from therapy or 
assessment observation, if any, should be described, as should the 
nature and duration of any planned followup observations in those 
patients.
9.4 Treatments
9.4.1 Treatments Administered
    The precise treatments or diagnostic agents to be administered 
(active drug(s), placebo(s), procedures) in each arm of the study, 
and for each period of the study, should be fully described by name 
of drug(s), marketing formulation, route and mode of administration, 
dose, and dosage schedule.
9.4.2 Identity of Investigational Product(s)
    In the text of the report, a brief description of the 
investigational product(s) (formulation, batch number) should be 
given.
    In Appendix 16.1.6 of the report, the batch/serial number(s) of 
treatment(s) administered and a description of their formulation, 
content, external appearance (manufacturing and packaging batch 
records) should be provided. If more than one batch of test drug was 
used, patients receiving each batch should be identified.
    The source of placebos and active control drugs should be 
provided. Any modification of active control drugs from their usual 
commercial state should be noted, and the steps taken to assure that 
their bioavailability was unaltered should be described.
    For long-duration trials of investigational products with 
limited shelf-lives or incomplete stability data, the logistics of 
resupply of the materials should be described. Any use of test 
materials past their expiry date should be noted, and patients 
receiving them identified. If there were specific storage 
requirements, these should also be described.
9.4.3 Method of Assigning Patients to Treatment Groups
    The specific methods used to assign patients to treatment 
groups, e.g., centralized allocation, allocation within sites, 
adaptive allocation (that is, assignment on the basis of earlier 
assignment or outcome), should be described in the text of the 
report, including any stratification or blocking procedures. Any 
unusual features should be explained.
    A detailed description of the randomization method, including 
how it was executed, should be given in Appendix 16.1.7 with 
references cited if necessary. A table exhibiting the randomization 
codes, patient identifier, and treatment assigned should also be 
presented in the appendix. For a multicenter study, the information 
should be given by center. The method of generating random numbers 
should be explained.
    For a historically controlled trial, it is important to explain 
how the particular control was selected and what other historical 
experiences were examined, if any, and how their results compared to 
the control used.
9.4.4 Selection of Doses in the Study
    The doses or dose ranges used in the study should be given and 
the basis for choosing them described (e.g., prior experience in 
humans, animal data).
9.4.5 Selection and Timing of Dose for Each Patient
    Procedures for selecting each patient's dose of test drug and 
control agent should be described. These procedures can vary from 
simple random assignment to a selected fixed drug/dose regimen, to 
some specified titration procedure, to more elaborate response-
determined selection procedures, e.g., where dose is titrated upward 
at intervals until intolerance or some specified endpoint is 
achieved. Procedures for back-titration, if any, should also be 
described.
    The timing (time of day, interval) of dosing and the relation of 
dosing to meals should be described, and if it was not specified, 
this should be noted.
    Any specific instructions to patients about when or how to take 
the dose(s) should be described.
9.4.6 Blinding
    A description of the specific procedures used to carry out 
blinding should be provided (e.g., how bottles were labeled, double 
dummy techniques), including the circumstances in which the blind 
would be broken for an individual or all patients, the procedures 
used, and who had access to patient codes. If the study allowed for 
some investigators to remain unblinded (e.g., to allow them to 
adjust medication), the means of shielding other investigators 
should be explained. Measures taken to ensure that drug and placebo 
were indistinguishable and evidence that they were 
indistinguishable, should be described, as should the appearance, 
shape, smell, and taste of the test material. Measures to prevent 
unblinding by laboratory measurements, if used, should be described.
    If blinding was considered unnecessary to reduce bias for some 
or all of the observations, this should be explained; e.g., use of a 
random-zero sphygmomanometer eliminates possible observer bias in 
reading blood pressure and Holter tapes are often read by automated 
systems that are presumably immune to observer bias. If blinding was 
considered desirable, but not feasible, the reasons and implications 
should be discussed. Sometimes blinding is attempted but is known to 
be imperfect because of obvious drug effects in at least some 
patients (dry mouth, bradycardia, fever, injection site reactions, 
changes in laboratory data). Such problems or potential problems 
should be identified and if there were any attempts to assess the 
magnitude of the problem or manage it (e.g., by having some end 
point measurements carried out by people shielded from information 
that might reveal treatment assignment), they should be described.
    A description of any packaging and labeling techniques used for 
blinding, e.g., double dummy techniques, special labels that reveal 
blind-breakage, sealed code lists/envelopes, etc., should be 
provided in Appendix 16.1.7.
9.4.7 Prior and Concomitant Therapy
    Which drugs or procedures were allowed before and during the 
study, whether and how their use was recorded, and any other 
specific rules and procedures related to permitted or forbidden 
concomitant therapy should be described. How allowed concomitant 
therapy might affect the outcome due either to drug-drug interaction 
or to direct effects on the study endpoints should be discussed, and 
how the independent effects of concomitant and study therapies could 
be ascertained should be explained.
9.4.8 Treatment Compliance
    The measures taken to ensure and document treatment compliance 
should be described, e.g., drug accountability, diary cards, blood, 
urine, or other body fluid drug level measurements, or medication 
event monitoring.
9.5 Efficacy and Safety Variables
9.5.1 Efficacy and Safety Measurements Assessed and Flow Chart
    The specific efficacy and safety variables to be assessed and 
laboratory tests to be conducted, their schedule (days of study, 
time of day, relation to meals, and the timing of critical measures 
in relation to test drug administration, e.g., just prior to next 
dose, 2 hours after dose), the methods for measuring them, and the 
persons responsible for the measurements should be described. If 
there were changes in personnel carrying out critical measurements, 
these should be reported.
    It is usually helpful to display graphically in a flow chart 
(see Annex III of the guideline) the frequency and timing of 
efficacy and safety measurements; visit numbers and times should be 
shown, or, alternatively, times alone can be used (visit numbers 
alone are more difficult to interpret). Any specific instructions 
(e.g., guidance or use of a diary) to the patients should also be 
noted.
    Any definitions used to characterize outcome (e.g., criteria for 
determining occurrence of acute myocardial infarction, designation 
of the location of the infarction, characterization of a stroke as 
thrombotic or hemorrhagic, distinction between TIA and stroke, 
assignment of cause of death) should be explained in full. Any 
techniques used to standardize or compare results of laboratory 
tests or other clinical measurements (e.g., ECG, chest X-ray) should 
also be described. This is particularly important in multicenter 
studies.
    If anyone other than the investigator was responsible for 
evaluation of clinical outcomes (e.g., the sponsor or an external 
committee to review X-rays or ECG's or to 

[[Page 43914]]
determine whether the patient had a stroke, acute infarction, or sudden 
death) the person or group should be identified. The procedures, 
including means of maintaining blindness, and centralizing readings 
and measurements, should be described fully.
    The means of obtaining adverse event data should be described 
(volunteered, checklist, questioning), as should any specifically 
planned followup procedures for adverse events or any planned 
rechallenge procedure.
    Any rating of adverse events by the investigator, sponsor, or 
external group (e.g., severity rating, likelihood of drug causation) 
should be described and criteria for such ratings, if any, given. If 
efficacy or safety was to be assessed in terms of categorical 
ratings or numerical scores, the criteria used for point assignment 
(e.g., definitions of point scores) should be provided. For 
multicenter studies, how methods were standardized should be 
indicated.
9.5.2 Appropriateness of Measurements
    If any of the efficacy or safety assessments were not standard, 
i.e., widely used and generally recognized as reliable, accurate, 
and relevant (able to discriminate between effective and ineffective 
agents), its reliability, accuracy, and relevance should be 
documented. It may be helpful to describe alternatives considered 
but rejected.
    If a surrogate end point (a laboratory measurement or physical 
measurement or sign that is not a direct measure of clinical 
benefit) was used as a study end point, this should be justified, 
e.g., by reference to clinical data publications, guidelines, or 
previous actions by regulatory authorities.
9.5.3 Primary Efficacy Variable(s)
    The primary measurements and end points used to determine 
efficacy should be clearly specified. Although the critical efficacy 
measurements may seem obvious, when there are multiple variables, or 
when variables are measured repeatedly, the protocol should identify 
the primary ones, with an explanation of why they were chosen, or 
designate the pattern of significant findings or other method of 
combining information that would be interpreted as supporting 
efficacy. If the protocol did not identify the primary variables, 
the study report should explain how these critical variables were 
selected (e.g., by reference to publications, guidelines, or 
previous actions by regulatory authorities) and when they were 
identified (i.e., before or after the study was completed and 
unblinded). If an efficacy threshold was defined in the protocol, 
this should be described.
9.5.4 Drug Concentration Measurements
    Any drug concentrations to be measured, and the sample 
collection times and periods in relation to the timing of drug 
administration, should be described. Any relation of drug 
administration and sampling to ingestion of food, posture, and the 
possible effects of concomitant medication/alcohol/caffeine/nicotine 
should also be addressed. The biological sample measured and the 
method of measurement used should be described, referring to 
published and/or internal assay validation documentation for 
methodological details. Where other factors are believed important 
in assessing pharmacokinetics (e.g., soluble circulating receptors, 
renal or hepatic function), the timing and plans to measure these 
factors should also be specified.
9.6 Data Quality Assurance - Audit Certificate
    The quality assurance and quality control systems implemented to 
assure the quality of the data should be described in brief. If none 
were used, this should be stated.
    Any steps taken at the investigation site or centrally to ensure 
the use of standard terminology and the collection of accurate, 
consistent, complete, and reliable data, such as training sessions, 
monitoring of investigators by sponsor personnel, instruction 
manuals, data verification, cross-checking, use of a central 
laboratory for certain tests, centralized ECG reading, or data 
audits, should be described. It should be noted whether investigator 
meetings or other steps were taken to prepare investigators and 
standardize performance.
    If the sponsor used an independent internal or external auditing 
procedure, it should be mentioned here and described in Appendix 
16.1.8 and audit certificates (if available; see Annexes IVa and IVb 
for descriptions) provided in the same appendix.
9.7 Statistical Methods Planned in The Protocol and Determination of 
Sample Size
9.7.1 Statistical and Analytical Plans
    The statistical analyses planned in the protocol and any changes 
made before outcome results were available should be described. In 
this section emphasis should be on which analyses, comparisons, and 
statistical tests were planned, not on which ones were actually 
used. If critical measurements were made more than once, the 
particular measurements (e.g., average of several measurements over 
the entire study, values at particular times, values only from study 
completers, or last on-therapy value) planned as the basis for 
comparison of drug and control should be specified. Similarly, if 
more than one analytical approach is plausible, e.g., changes from 
baseline response, slope analysis, life table analysis, the planned 
approach should be identified. Also, whether the primary analysis is 
to include adjustment for covariates should be specified.
    If there were any planned reasons for excluding from analysis 
patients for whom data are available, these should be described. If 
there were any subgroups whose results were to be examined 
separately, these should be identified. If categorical responses 
(global scales, severity scores, responses of a certain size) were 
to be used in analyzing responses, they should be clearly defined.
    Planned monitoring of the results of the study should be 
described. If there was a data monitoring committee, either within 
or outside the sponsor's control, its composition and operating 
procedures should be described and procedures to maintain study 
blinding should be given. The frequency and nature of any planned 
interim analysis, any specified circumstances in which the study 
would be terminated, and any statistical adjustments to be employed 
because of interim analyses should be described.
9.7.2 Determination of Sample Size
    The planned sample size and the basis for it, such as 
statistical considerations or practical limitations, should be 
provided. Formulae for sample size and power calculation should be 
given together with their derivations or source of reference. 
Estimates used in the formulae should be given and explanations 
provided as to how they were obtained. For a positive control study 
intended to show that a new therapy is at least as effective as the 
standard therapy, the sample size determination should specify a 
``delta value'' (a difference between treatments that would be 
considered unacceptably large) and therefore the difference the 
study is designed to be able to exclude.
9.8 Changes in the Conduct of the Study or Planned Analyses
    Any change in the conduct of the study or planned analyses 
(e.g., dropping a treatment group, changing the entry criteria or 
drug dosages, adjusting the sample size, etc.) instituted after the 
start of the study should be described. The time(s) and reason(s) 
for the change(s), the procedure used to decide on the change(s), 
the person(s) or group(s) responsible for the change(s), and the 
nature and content of the data available (and to whom they were 
available) when the change was made should also be described, 
whether the change was documented as a formal protocol amendment or 
not. (Personnel changes need not be included.) Any possible 
implications of the change(s) for the interpretation of the study 
should be discussed briefly in this section and more fully in other 
appropriate sections of the report. In every section of the report, 
a clear distinction between conditions (procedures) planned in the 
protocol and amendments or additions should be made.

10. Study Patients

10.1 Disposition of Patients
    There should be a clear accounting of all patients who entered 
the study, using figures or tables in the text of the report. The 
numbers of patients who were screened, were randomized, and who 
entered and completed each phase of the study (or each week/month of 
the study) should be provided, as well as the reasons for all 
postrandomization discontinuations, grouped by treatment and by 
major reason (e.g., lost to followup, adverse event, poor 
compliance). In some cases, it may also be relevant to provide a 
breakdown of the reasons for excluding patients during screening, if 
this could help clarify the appropriate patient population for 
eventual drug use. A flow chart is often helpful (see Annexes Va and 
Vb of the guideline). Whether patients are followed for the duration 
of the study, even if drug is discontinued, should be made clear.
    In Appendix 16.2.1, there should also be a listing of all 
patients discontinued from the study after enrollment, broken down 
by center and treatment group, giving a patient identifier, the 
specific reason for discontinuation, the treatment (drug and dose), 
cumulative dose (where appropriate), and the duration of treatment 
before discontinuation. Whether or not the blind for the patient was 
broken at the time of discontinuation should be noted. It may also 
be useful to include other information, such as critical demographic 
data (e.g., age, sex, race), concomitant medication, and the major 

[[Page 43915]]
response variable(s) at termination. See Annex VI for an example of 
such a listing.
10.2 Protocol Deviations
    All important deviations related to study inclusion or exclusion 
criteria, conduct of the trial, patient management, or patient 
assessment should be described.
    In the body of the text, protocol deviations should be 
appropriately summarized by center and grouped into different 
categories, such as:
    - Those who entered the study even though they did not satisfy 
the entry criteria.
    - Those who developed withdrawal criteria during the study but 
were not withdrawn.
    - Those who received the wrong treatment or incorrect dose.
    - Those who received an excluded concomitant treatment.
    In Appendix 16.2.2, individual patients with these protocol 
deviations should be listed, broken down by center for multicenter 
studies.
11. Efficacy Evaluation
11.1 Data Sets Analyzed
    Exactly which patients were included in each efficacy analysis 
should be precisely defined, e.g., all patients receiving any drug, 
all patients with any efficacy observation or with a certain minimum 
number of observations, only patients completing the trial, all 
patients with an observation during a particular time window, only 
patients with a specified degree of compliance, etc. It should be 
clear, if not defined in the study protocol, when (relative to study 
unblinding), and how inclusion/exclusion criteria for the data sets 
analyzed were developed. As a general rule, even if the applicant's 
proposed primary analysis is based on a reduced subset of the 
patients with data, there should also be for any trial intended to 
establish efficacy an additional analysis using all randomized (or 
otherwise entered) patients with any on-treatment data.
    There should be a tabular listing of all patients, visits, and 
observations excluded from the efficacy analysis provided in 
Appendix 16.2.3 (see Annex VII of the guideline for an example). The 
reasons for exclusions should also be analyzed for the whole 
treatment group over time (see Annex VIII of the guideline for an 
example).
11.2 Demographic and Other Baseline Characteristics
    Group data for the critical demographic and baseline 
characteristics of the patients, as well as other factors arising 
during the study that could affect response, should be presented in 
this section and comparability of the treatment groups for all 
relevant characteristics should be displayed by use of tables or 
graphs in section 14.1. The data for the patient sample included in 
the ``all patients with data'' analysis should be given first. This 
can then be followed by data on other groups used in principal 
analyses, such as the ``per-protocol'' analysis or other analyses, 
e.g., groups defined by compliance, concomitant disease/therapy, or 
demographic/baseline characteristics. When such groups are used, 
data for the complementary excluded group should also be shown. In a 
multicenter study, comparability should be assessed by center, and 
centers should be compared.
    A diagram showing the relationship between the entire sample and 
any other analysis groups should be provided.
    The critical variables will depend on the specific nature of the 
disease but will usually include:
     Demographic variables
    - Age
    - Sex
    - Race
     Disease factors
    - Specific entry criteria (if not uniform), duration, stage, and 
severity of disease and other clinical classifications and 
subgroupings in common usage or of known prognostic significance.
    - Baseline values for critical clinical measurements carried out 
during the study or identified as important indicators of prognosis 
or response to therapy.
    - Concomitant illness at trial initiation such as renal disease, 
diabetes, heart failure.
    - Relevant previous illness.
    - Relevant previous treatment for illness treated in the study.
    - Concomitant treatment maintained, even if the dose was changed 
during the study, including oral contraceptive and hormone 
replacement therapy; treatments stopped at entry into the study 
period (or changed at study initiation).
     Other factors that might affect response to therapy 
(e.g., weight, renin status, antibody levels, metabolic status).
     Other possibly relevant variables (e.g., smoking, 
alcohol intake, special diets) and, for women, menstrual status and 
date of last menstrual period, if pertinent for the study.
    In addition to tables and graphs giving group data for these 
baseline variables, relevant individual patient demographic and 
baseline data, including laboratory values, and all concomitant 
medication for all individual patients randomized (broken down by 
treatment and by center for multicenter studies) should be presented 
in by-patient tabular listings in Appendix 16.2.4. Although some 
regulatory authorities will require all baseline data to be 
presented elsewhere in tabular listings, the appendix to the study 
report should be limited to only the most relevant data, generally 
the variables listed above.
11.3 Measurements of Treatment Compliance
    Any measurements of compliance of individual patients with the 
treatment regimen under study and drug concentrations in body fluids 
should be summarized, analyzed by treatment group and time interval, 
and tabulated in Appendix 16.2.5.
11.4 Efficacy Results and Tabulations of Individual Patient Data
11.4.1 Analysis of Efficacy
    Treatment groups should be compared for all critical measures of 
efficacy (primary and secondary end points; pharmacodynamics), as 
well as benefit/risk assessment measured in each patient where these 
are utilized. In general, the results of all analyses contemplated 
in the protocol and an analysis including all patients with on-study 
data should be performed in studies intended to establish efficacy.
    Analyses based on continuous variables (e.g., mean blood 
pressure or depression scale score) and categorical responses (e.g., 
cure of an infection) can be equally valid; ordinarily both should 
be presented if both were planned and are available. If categories 
are newly created (i.e., not in the statistical plan), the basis for 
them should be explained. Even if one variable receives primary 
attention (e.g., in a blood pressure study, supine blood pressure at 
week x), other reasonable measures (e.g., standing blood pressure 
and blood pressures at other particular times) should be assessed, 
at least briefly. In addition, the time course of response should be 
described, if possible. For a multicenter study, data display and 
analysis of individual centers should be included to give a clear 
picture of the results at each site, especially the larger sites.
    If any critical measurements or assessments of efficacy or 
safety outcomes were made by more than one party (e.g., both the 
investigator and an expert committee may offer an opinion on whether 
a patient had an acute infarction), overall differences between the 
ratings should be shown, and each patient having disparate 
assessments should be identified. The assessments used should be 
clear in all analyses.
    In many cases, efficacy and safety end points are difficult to 
distinguish (e.g., stroke in a thrombolytic trial, deaths in a fatal 
disease study). Many of the principles addressed below should be 
adopted for critical safety measures as well.
11.4.2 Statistical/Analytical Issues
    The statistical analysis used should be described for clinical 
and statistical reviewers in the text of the report, with detailed 
documentation of statistical methods (see section Annex IX) 
presented in Appendix 16.1.9. Important features of the analysis, 
including the particular tests used, adjustments made for 
demographic or baseline measurements or concomitant therapy, 
handling of dropouts and missing data, adjustments for multiple 
comparisons, special analyses of multicenter studies, and 
adjustments for interim analyses, should be discussed. Any changes 
in the analysis made after blind-breaking should be identified.
    In addition to the general discussion, the following specific 
issues should be addressed (unless not applicable):
11.4.2.1 Adjustments for Covariates
    Selection of, and adjustments for, demographic or baseline 
measurements, concomitant therapy, or any other covariate or 
prognostic factor should be explained in the report, and methods of 
adjustment, results of analyses, and supportive information (e.g., 
ANCOVA or Cox regression output) should be included in the detailed 
documentation of statistical methods. If the covariates or methods 
used in these analyses differed from those planned in the protocol, 
the differences should be explained and where possible and relevant, 
the results of planned analyses should also be presented. Although 
not part of the individual study report, comparisons of covariate 
adjustments and prognostic factors across individual studies may be 
an informative analysis in a summary of clinical efficacy data.
11.4.2.2 Handling of Dropouts or Missing Data
    There are several factors that may affect dropout rates. These 
include the duration of 

[[Page 43916]]
the study, the nature of the disease, the efficacy and toxicity of the 
drug under study, and other factors that are not therapy related. 
Ignoring the patients who dropped out of the study and drawing 
conclusions based only on patients who completed the study can be 
misleading. A large number of dropouts, however, even if included in 
an analysis, may introduce bias, particularly if there are more 
early dropouts in one treatment group or the reasons for dropping 
out are treatment or outcome related. Although the effects of early 
dropouts, and sometimes even the direction of bias, can be difficult 
to determine, possible effects should be explored as fully as 
possible. It may be helpful to examine the observed cases at various 
time points or, if dropouts were very frequent, to concentrate on 
analyses at time points when most of the patients were still under 
observation and when the full effect of the drug was realized. It 
may also be helpful to examine modeling approaches to the evaluation 
of such incomplete data sets.
    The results of a clinical trial should be assessed not only for 
the subset of patients who completed the study, but also for the 
entire patient population as randomized (the intent-to-treat 
analysis) or at least for all those with any on-study measurements. 
Several factors should be considered and compared for the treatment 
groups in analyzing the effects of dropouts: the reasons for the 
dropouts, the time to dropout, and the proportion of dropouts among 
treatment groups at various time points.
    Procedures for dealing with missing data, e.g., use of estimated 
or derived data, should be described. Detailed explanation should be 
provided as to how such estimations or derivations were done and 
what underlying assumptions were made.
11.4.2.3 Interim Analyses and Data Monitoring
    The process of examining and analyzing data accumulating in a 
clinical trial, either formally or informally, can introduce bias 
and/or increase type I error. Therefore, all interim analyses, 
formal or informal, preplanned or ad hoc, by any study participant, 
sponsor staff member, or data monitoring group should be described 
in full, even if the treatment groups were not identified. The need 
for statistical adjustment because of such analyses should be 
addressed. Any operating instructions or procedures used for such 
analyses should be described. The minutes of meetings of any data 
monitoring group and any data reports reviewed at those meetings, 
particularly a meeting that led to a change in the protocol or early 
termination of the study, may be helpful and should be provided in 
Appendix 16.1.9. Data monitoring without code-breaking should also 
be described, even if this kind of monitoring is considered to cause 
no increase in type I error.
11.4.2.4 Multicenter Studies
    A multicenter study is a single study under a common protocol, 
involving several centers (e.g., clinics, practices, hospitals) 
where the data collected are intended to be analyzed as a whole (as 
opposed to a post-hoc decision to combine data or results from 
separate studies). Individual center results should be presented, 
however, and statistical tests for homogeneity across centers, i.e., 
for detecting treatment-by center interaction, should be provided, 
if feasible. The significance level used to declare the significance 
of a given test for treatment-by-center interaction should be 
considered in light of the sample sizes involved. Any extreme or 
opposite results among centers should be noted and discussed, 
considering such possibilities as differences in study conduct, 
patient characteristics, or clinical settings. As mentioned in 
previous sections, demographic, baseline, and postbaseline data, as 
well as efficacy data, should be presented by center, even though 
the combined analysis is the primary one.
11.4.2.5 Multiple Comparisons
    False positive findings increase in number as the number of 
significance tests (number of comparisons) performed increases. If 
there was more than one primary endpoint (outcome variable), or if 
there were multiple treatment groups, or subsets of the patient 
population being examined, the statistical analysis should reflect 
awareness of this and either explain the statistical adjustment used 
for type I error criteria or give reasons why it was considered 
unnecessary.
11.4.2.6 Use of an ``Efficacy Subset'' of Patients
    Particular attention should be devoted to the effects of 
dropping patients with available data from analyses because of poor 
compliance, missed visits, ineligibility, or any other reason. As 
noted above, an analysis using all available data should be carried 
out for all studies intended to establish efficacy, even if it is 
not the analysis proposed as the primary analysis by the applicant.
11.4.2.7 Active-Control Studies Intended to Show Equivalence
    If an active control study is intended to show equivalence 
(i.e., lack of a difference greater than a specified size) between a 
new drug and a control, the analysis should show the confidence 
interval for the comparison between the two agents for critical end 
points and the relation of that interval to the prespecified degree 
of inferiority that would be considered unacceptable. (See 9.2, for 
important considerations when using the active control equivalence 
design.)
11.4.2.8 Examination of Subgroups
    If the size of the study permits, important demographic or 
baseline value-defined subgroups should be examined for unusually 
large or small responses and the results presented, e.g., comparison 
of effects by age, sex, or race, by severity or prognostic groups, 
by history of prior treatment with a drug of the same class. If 
these analyses were not carried out because the study was too small, 
it should be noted. These analyses are not intended to ``salvage'' 
an otherwise nonsupportive study but may suggest hypotheses worth 
examining in other studies or be helpful in refining, for example, 
labeling information, patient selection, or dose selection. Where 
there is a prior hypothesis of a differential effect in a particular 
subgroup, this hypothesis and its assessment should be part of the 
planned statistical analysis.
11.4.3 Tabulation of Individual Response Data
    In addition to tables and graphs representing group data, 
individual response data and other relevant study information should 
be presented in tables. Some regulatory authorities may require all 
individual data in archival case report tabulations. What needs to 
be included in the report will vary from study to study and from one 
drug class to another and the applicant should decide, if possible 
after consultation with the regulatory authority, what to include in 
an appendix to the study report. The study report should indicate 
what material is included as an appendix, what is in the more 
extensive archival case report tabulations, if required by the 
regulatory authority, and what is available on request.
    For a controlled study in which critical efficacy measurements 
or assessments (e.g., blood or urine cultures, pulmonary function 
tests, angina frequency, or global evaluations) are repeated at 
intervals, the data listings accompanying the report should include, 
for each patient, a patient identifier, all measured or observed 
values of critical measurements, including baseline measurements, 
with notation of the time during the study (e.g., days on therapy 
and time of day, if relevant) when the measurements were made, the 
drug/dose at the time (if useful, given as milligrams per kilogram 
(mg/kg)), any measurements of compliance, and any concomitant 
medications at the time of, or close to the time of, measurement or 
assessment. If, aside from repeated assessments, the study included 
some overall responder versus nonresponder evaluation(s) 
(bacteriologic cure or failure), it should also be included. In 
addition to critical measurements, the tabulation should note 
whether the patient was included in the efficacy evaluation (and 
which evaluation, if more than one), provide patient compliance 
information, if collected, and a reference to the location of the 
case report form, if included. Critical baseline information such as 
age, sex, weight, disease being treated (if more than one in study), 
and disease stage or severity, is also helpful. The baseline values 
for critical measurements would ordinarily be included as zero time 
values for each efficacy measurement.
    The tabulation described should usually be included in Appendix 
16.2.6 of the study report, rather than in the more extensive case 
report tabulations required by some regulatory authorities, because 
it represents the basic efficacy data supporting summary tables. 
Such a thorough tabulation can be unwieldy for review purposes, 
however, and more targeted displays should be developed as well. For 
example, if there are many measurements reported, tabulations of the 
most critical measurements for each patient (e.g., the blood 
pressure value at certain visits might be more important than 
others) will be useful in providing an overview of each individual's 
results in a study, with each patient's response summarized on a 
single line or small number of lines.
11.4.4 Drug Dose, Drug Concentration, and Relationships to Response
    When the dose in each patient can vary, the actual doses 
received by patients should be shown and individual patient's doses 
should be tabulated. Although studies not designed as dose-response 
studies may have limited ability to contribute dose-response 

[[Page 43917]]
information, the available data should be examined for whatever 
information they can yield. In examining the dose response, it may 
be helpful to calculate dose as mg/kg body weight or milligrams per 
meter squared (mg/m2) body surface.
    Drug concentration information, if available, should also be 
tabulated (Appendix 16.2.5), analyzed in pharmacokinetic terms and, 
if possible, related to response.
    Further guidance on the design and analysis of studies exploring 
dose-response or concentration response can be found in the ICH 
Guideline ``Dose-Response Information to Support Drug 
Registration.''
11.4.5 Drug-Drug and Drug Disease Interactions
    Any apparent relationship between response and concomitant 
therapy and between response and past and/or concurrent illness 
should be described.
11.4.6 By-Patient Displays
    While individual patient data ordinarily can be displayed in 
tabular listings, it has on occasion been helpful to construct 
individual patient profiles in other formats, such as graphic 
displays. These might, for example, show the value of (a) particular 
parameter(s) over time, the drug dose over the same period, and the 
times of particular events (e.g., an adverse event or change in 
concomitant therapy). Where group mean data represent the principal 
analyses, this kind of ``case report extract'' may offer little 
advantage; it may be helpful, however, if overall evaluation of 
individual responses is a critical part of the analysis.
11.4.7 Efficacy Conclusions
    The important conclusions concerning efficacy should be 
concisely described, considering primary and secondary endpoints, 
prespecified and alternative statistical approaches, and results of 
exploratory analyses.

12. Safety Evaluation

    Analysis of safety-related data can be considered at three 
levels. First, the amount of exposure (dose, duration) should be 
examined to determine the extent to which safety can be assessed 
from the study. Second, the more common adverse events, e.g., 
laboratory test changes, should be identified, classified in some 
reasonable way, compared for treatment groups, and analyzed, as 
appropriate, for that may affect the frequency of adverse reactions, 
such as time dependence, relation to demographic characteristics, 
relation to dose or drug concentration. Finally, serious adverse 
events and other significant adverse events should be identified, 
usually by close examination of patients who left the study 
prematurely because of an adverse event, whether or not identified 
as drug related, or who died.
    The ICH Guideline on Clinical Safety Data Management, 
Definitions and Standards for Expedited Reporting defines serious 
adverse events as follows: A ``serious adverse event'' (experience) 
or reaction is any untoward medical occurrence that at any dose: 
results in death, is life-threatening, requires inpatient or 
prolongation of existing hospitalization, results in persistent or 
significant disability/incapacity, or is a congenital anomaly/birth 
defect.
    For the purpose of this guideline, ``other significant adverse 
events'' are marked hematological and other laboratory abnormalities 
and any medical events that led to an intervention, including 
withdrawal of drug treatment, dose reduction, or significant 
additional concomitant therapy.
    In the following sections, three kinds of analysis and display 
are suggested:
    (1) Summarized data, often using tables and graphical 
presentations presented in the main body of the report;
    (2) Listings of individual patient data; and
    (3) Narrative statements of events of particular interest.
    In all tabulations and analyses, events associated with both 
test drug and control treatment should be displayed.
12.1 Extent of Exposure
    The extent of exposure to study drugs (and to active control and 
placebo) should be characterized according to the number of patients 
exposed, the duration of exposure, and the dose to which they were 
exposed.
     Duration: Duration of exposure to any dose can be 
expressed as a median or mean, but it is also helpful to describe 
the number of patients exposed for specified periods of time, such 
as for 1 day or less, 2 days to 1 week, more than 1 week to 1 month, 
more than 1 month to 6 months. The numbers exposed to drug for the 
various durations should also be broken down into age, sex, and 
racial subgroups, and any other pertinent subgroups, such as disease 
(if more than one is represented), disease severity, concurrent 
illness.
     Dose: The mean or median dose used and the number of 
patients exposed to specified daily dose levels should be given; the 
daily dose levels used could be the maximum dose for each patient, 
the dose with longest exposure for each patient, or the mean daily 
dose. It is often useful to provide combined dose-duration 
information, such as the numbers exposed for a given duration (e.g., 
at least 1 month) to the most common dose, the highest dose, the 
maximum recommended dose. In some cases, cumulative dose might be 
pertinent. Dosage may be given as the actual daily dose or on a mg/
kg or mg/m2 basis as appropriate. The numbers of patients 
exposed to various doses should be broken down into age, sex, and 
racial subgroups, and any other pertinent subgroups.
     Drug concentration: If available, drug concentration 
data (e.g., concentration at the time of an event, maximum plasma 
concentration, area under curve) may be helpful in individual 
patients for correlation with adverse events or changes in 
laboratory variables. (Appendix 16.2.5.)
    It is assumed that all patients entered into treatment who 
received at least one dose of the treatment are included in the 
safety analysis; if that is not so, an explanation should be 
provided.
12.2 Adverse Events (AE's)
12.2.1 Brief Summary of Adverse Events
    The overall adverse event experience in the study should be 
described in a brief narrative, supported by the following more 
detailed tabulations and analyses. In these tabulations and 
analyses, events associated with both the test drug and control 
treatment should be displayed.
12.2.2 Display of Adverse Events
    All adverse events (including events likely to be related to the 
underlying disease or likely to represent concomitant illness, 
unless there is a prior agreement with the regulatory authority to 
consider specified events as disease related) should be displayed in 
summary tables (section 14.3.1). The tables should include changes 
in vital signs and any laboratory changes that were considered 
serious adverse events or that resulted in withdrawal of treatment.
    In most cases, it will also be useful to describe in such 
tables, ``treatment emergent signs and symptoms'' (TESS; those not 
seen at baseline, and those which worsened even if present at 
baseline).
    The tables should list each adverse event, the number of 
patients in each treatment group in whom the event occurred, and the 
rate of occurrence. Adverse events should be grouped by body system. 
Each event may then be divided into defined severity categories 
(e.g., mild, moderate, severe) if these were used. The tables may 
also divide the adverse events into those considered at least 
possibly related to drug use and those considered not related, or 
use some other causality (e.g., unrelated or possibly, probably, or 
definitely related). For any such categorization, the categories 
should be defined and the person(s) responsible for classification 
of each event identified. Even when such a causality assessment is 
used, the tables should include all adverse events, whether or not 
considered drug related, including events thought to represent 
intercurrent illnesses. Subsequent analyses of the study or of the 
overall safety data base may help to distinguish between adverse 
events that are, or are not, considered drug related. So that it is 
possible to analyze and evaluate the data in these tables, it is 
important to identify each patient having each adverse event. An 
example of such a tabular presentation is shown below.

BILLING CODE 4160-01-F

[[Page 43918]]
[GRAPHIC][TIFF OMITTED]TN23AU95.010



BILLING CODE 4160-01-C

[[Page 43919]]

    In addition to these complete tables provided in 14.3.1, an 
additional summary table comparing treatment and control groups, 
without the patient identifying numbers limited to relatively common 
adverse events (e.g., those in at least 1 percent of the treated 
group), should be provided in the body of the report.
    In presenting adverse events, it is important both to display 
the original terms used by the investigator and to attempt to group 
related events (i.e., events that probably represent the same 
phenomena) so that the true occurrence rate is not obscured. One way 
to do this is with a standard adverse reaction dictionary.
12.2.3 Analysis of Adverse Events
    The basic display of adverse event rates described in section 
12.2.2 (and located in section 14.3.1) of the report should be used 
to compare rates in treatment and control groups. For this analysis 
it may be helpful to combine the event severity categories and the 
causality categories, leading to a simpler side-by-side comparison 
of treatment groups. In addition, if study size and design permit, 
it may be useful to examine the more common adverse events that seem 
to be drug related for relationship to dosage and to mg/kg or mg/
m2 dose, to dose regimen, to duration of treatment, to total 
dose, to demographic characteristics, such as age, sex, race, to 
other baseline features, such as renal status to effectiveness 
outcomes, and to drug concentration, if data are available. It may 
also be useful to examine time of onset and duration of adverse 
events. A variety of additional analyses may be suggested by the 
study results or by the pharmacology of the drug.
    It is not intended that every adverse event be subjected to 
rigorous statistical evaluation. It may be apparent from initial 
display and inspection of the data that a significant relation to 
demographic or other baseline features is not present. If the 
studies are small and if the number of events is relatively small, 
it may be sufficient to limit analyses to a comparison of treatment 
and control.
    Under certain circumstances, life table or similar analyses may 
be more useful than reporting of crude adverse event rates.
12.2.4 Listing of Adverse Events by Patient
    All adverse events for each patient, including the same event on 
several occasions, should be listed in Appendix 16.2.7, giving both 
preferred term and the original term used by the investigator. The 
listing should be by investigator and by treatment group and should 
include:
    - Patient identifier
    - Age, race, sex, weight (height, if relevant)
    - Location of case report forms (CRF's), if provided.
     - preferred term
    - The adverse event -
     - reported term
    - Duration of the adverse event
    - Severity (e.g., mild, moderate, severe)
    - Seriousness (serious/nonserious)
    - Action taken (none, dose reduced, treatment stopped, specific 
treatment instituted, etc.)
    - Outcome (e.g., CIOMS format)
    - Causality assessment (e.g., related/not related). How this was 
determined should be described in the table or elsewhere.
    - Date of onset or date of clinic visit at which the event was 
discovered
    - Timing of onset of the adverse event in relation to last dose 
of study drug (when applicable)
    - Study treatment at time of event or most recent study 
treatment taken
    - Study drug dose in absolute amount, mg/kg or mg/m2 at 
time of event
    - Drug concentration (if known)
    - Duration of study drug treatment
    - Other drug treatment during study
    Any abbreviations and codes should be clearly explained at the 
beginning of the listing or, preferably, on each page.
12.3 Deaths, Other Serious Adverse Events, and Other Significant 
Adverse Events
    Deaths, other serious adverse events, and other significant 
adverse events (see definition in section 12) deserve special 
attention.
12.3.1 Listing of Deaths, Other Serious Adverse Events, and Other 
Significant Adverse Events
    Listings, containing the same information as called for in 
section 12.2.4 above, should be provided for the following events.
12.3.1.1 Deaths
    All deaths during the study, including the posttreatment 
followup period, and deaths that resulted from a process that began 
during the study, should be listed by patient in section 14.3.2.
21.3.1.2 Other Serious Adverse Events
    All serious adverse events (other than death but including the 
serious adverse events temporally associated with or preceding the 
deaths) should be listed in section 14.3.2. The listing should 
include laboratory abnormalities, abnormal vital signs, and abnormal 
physical observations that were considered serious adverse events.
12.3.1.3 Other Significant Adverse Events
    Marked hematological and other laboratory abnormalities (other 
than those meeting the definition of serious) and any events that 
led to an intervention, including withdrawal of drug treatment, dose 
reduction, or significant additional concomitant therapy, other than 
those reported as serious adverse events, should be listed in 
section 14.3.2.
12.3.2 Narratives of Deaths, Other Serious Adverse Events, and 
Certain Other Significant Adverse Events
    There should be brief narratives describing each death, each 
other serious adverse event, and those of the other significant 
adverse events that are of special interest because of clinical 
importance. These narratives can be placed either in the text of the 
report or in section 14.3.3, depending on their number. Events that 
were clearly unrelated to the study drug may be omitted or described 
very briefly. In general, the narrative should describe:
    - The nature and intensity of event, the clinical course leading 
up to the event, with an indication of timing relevant to drug 
administration; relevant laboratory measurements whether the drug 
was stopped, and when; countermeasures; post mortem findings; 
investigator's opinion on causality, and sponsor's opinion on 
causality, if appropriate.
    In addition, the following information should be included:
    - Patient identifier
    - Age and sex of patient; general clinical condition of patient, 
if appropriate
    - Disease being treated (if the same for all patients this is 
not required) with duration of current episode of illness
    - Relevant concomitant/previous illnesses with details of 
occurrence/duration
    - Relevant concomitant/previous medication with details of 
dosage
    - Study drug administered, drug dose, if this varied among 
patients, and length of time administered
12.3.3 Analysis and Discussion of Deaths, Other Serious Adverse 
Events, and Other Significant Adverse Events
    The significance of the deaths, other serious adverse events, 
and other significant adverse events leading to withdrawal, dose 
reduction, or institution of concomitant therapy should be assessed 
with respect to the safety of the drug. Particular attention should 
be paid to whether any of these events may represent a previously 
unsuspected important adverse effect of the drug. For serious 
adverse events that appear of particular importance, it may be 
useful to use life table or similar analyses to show their relation 
to time on drug and to assess their risk over time.
12.4 Clinical Laboratory Evaluation
12.4.1 Listing of Individual Laboratory Measurements by Patient 
(16.2.8) and Each Abnormal Laboratory Value (14.3.4)
    When required by regulatory authorities, the results of all 
safety-related laboratory tests should be available in tabular 
listings, using a display similar to the following, where each row 
represents a patient visit at which a laboratory study was done, 
with patients grouped by investigator (if more than one) and 
treatment group, and columns include critical demographic data, drug 
dose data, and the results of the laboratory tests. As not all tests 
can be displayed in a single table, they should be grouped logically 
(e.g., hematological tests, liver chemistries, electrolytes, 
urinalysis). Abnormal values should be identified, e.g., by 
underlining, bracketing. These listings should be submitted as part 
of the registration/marketing application, when this is required, or 
may be available on request.

                                                                        

[[Page 43920]]
                                                                                                                
----------------------------------------------------------------------------------------------------------------
                                                                                      Laboratory Tests          
  Patient      Time       Age       Sex      Race      Weight      Dose    -------------------------------------
                                                                               SGOT        SGPT    AP.........X 
----------------------------------------------------------------------------------------------------------------
#1          T0         70        M         W         70 kg      400 mg      V1*         V5        V9            
            T1         ........  ........  ........  .........  ..........  V2          V6        V10           
            T2         ........  ........  ........  .........  ..........  V3          V7        V11           
            T3         ........  ........  ........  .........  ..........  V4          V8        V12           
            .........  ........  ........  ........  .........  ..........  ..........  ........  ..............
#2          T10        65        F         B         50 kg      300 mg      V13         V16       V19           
            T21        ........  ........  ........  .........  ..........  V14         V17       V20           
            T32        ........  ........  ........  .........  ..........  V15         V18       V21           
----------------------------------------------------------------------------------------------------------------

                                                                                                  --------------
*Vn = value of a particular test
    For all regulatory authorities, there should be a by-patient 
listing of all abnormal laboratory values in section 14.3.4, using 
the format described above. For laboratory abnormalities of special 
interest (abnormal laboratory values of potential clinical 
importance), it may also be useful to provide additional data, such 
as normal values before and after the abnormal value, and values of 
related laboratory tests. In some cases, it may be desirable to 
exclude certain abnormal values from further analysis. For example, 
single, nonreplicated, small abnormalities of some tests (e.g., uric 
acid or electrolytes) or occasional low values of some tests (e.g., 
transaminase, alkaline phosphatase, blood urea nitrogen (BUN)) can 
probably be defined as clinically insignificant and excluded. Any 
such decisions should be clearly explained, however, and the 
complete list of values provided (or available to authorities on 
request) should identify every abnormal value.
12.4.2 Evaluation of Each Laboratory Parameter
    The necessary evaluation of laboratory values must in part be 
determined by the results seen, but, in general, the following 
analyses should be provided. For each analysis, comparison of the 
treatment and control groups should be carried out, as appropriate, 
and as compatible with study size. In addition, normal laboratory 
ranges should be given for each analysis.
12.4.2.1 Laboratory Values Over Time
    For each parameter at each time over the course of the study 
(e.g., at each visit) the following should be described: the group 
mean or median values, the range of values, and the number of 
patients with abnormal values, or with abnormal values that are of a 
certain size (e.g., twice the upper limit of normal, 5 times the 
upper limit; choices should be explained). Graphs may be used.
12.4.2.2 Individual Patient Changes
    An analysis of individual patient changes by treatment group 
should be given. A variety of approaches may be used, including:
    I. ``Shift tables'' - These tables show the number of patients 
who are low, normal, or high at baseline and then at selected time 
intervals.
    II. Tables showing the number or fraction of patients who had a 
change in parameter of a predetermined size at selected time 
intervals. For example, for BUN, it might be decided that a change 
of more than 10 mg per deciliter BUN should be noted. For this 
parameter, the number of patients having a change less than this or 
greater than this would be shown for one or more visits, usually 
grouping patients separately depending on baseline BUN (normal or 
elevated). The possible advantage of this display, compared to the 
usual shift table, is that changes of a certain size are noted, even 
if the value is not abnormal.
    III. A graph comparing the initial value and the on-treatment 
values of a laboratory measurement for each patient by locating the 
point defined by the initial value on the abscissa and a subsequent 
value on the ordinate. If no changes occur, the point representing 
each patient will be located on the 45 deg. line. A general shift to 
higher values will show a clustering of points above the 45 deg. 
line. As this display can show only a single time point for a single 
treatment, interpretation requires a time series of these plots for 
treatment and control groups. This kind of display identifies 
outliers readily (it is useful to include patient identifiers for 
the outliers).
12.4.2.3 Individual Clinically Significant Abnormalities
    Clinically significant changes (defined by the applicant) should 
be discussed. A narrative of each patient whose laboratory 
abnormality was considered a serious adverse event and, in certain 
cases, considered another significant event should be provided under 
section 12.3.2 or 14.3.3. When toxicity grading scales are used 
(e.g., World Health Organization, National Cancer Institute), 
changes graded as severe should be discussed regardless of 
seriousness. An analysis of the clinically significant changes, 
together with a recapitulation of discontinuations due to laboratory 
measurements, should be provided for each parameter. The 
significance of the changes and likely relation to the treatment 
should be assessed, e.g., by analysis of such features as 
relationship to dose, relationship to drug concentration, 
disappearance on continued therapy, positive dechallenge, positive 
rechallenge, and the nature of concomitant therapy.
12.5 Vital Signs, Physical Findings, and Other Observations Related 
to Safety
    Vital signs, other physical findings, and other observations 
related to safety should be analyzed and presented in a way similar 
to laboratory variables. If there is evidence of a drug effect, any 
dose-response or drug concentration-response relationship or 
relationship to patient variables (e.g., disease, demographics, 
concomitant therapy) should be identified and the clinical relevance 
of the observation described. Particular attention should be given 
to changes not evaluated as efficacy variables and to those 
considered to be adverse events.
12.6 Safety Conclusions
    The overall safety evaluation of the study drug(s) should be 
reviewed, with particular attention to events resulting in changes 
of dose or need for concomitant medication, serious adverse events, 
events resulting in withdrawal, and deaths. Any patients or patient 
groups at increased risk should be identified and particular 
attention paid to potentially vulnerable patients who may be present 
in small numbers, e.g., children, pregnant women, frail elderly, 
people with marked abnormalities of drug metabolism or excretion. 
The implication of the safety evaluation for the possible uses of 
the drug should be described.

13. Overall Conclusions

    The efficacy and safety results of the study and the 
relationship of risks and benefit should be briefly summarized, 
referring to the tables, figures, and sections above as needed. The 
presentation should not simply repeat the description of results nor 
introduce new results.
    The conclusions should clearly identify any new or unexpected 
findings, comment on their significance, and discuss any potential 
problems such as inconsistencies between related measures. The 
clinical relevance and importance of the results should also be 
discussed in the light of other existing data. Any specific benefits 
or special precautions required for individual subjects or at-risk 
groups and any implications for the conduct of future studies should 
be identified.

14. Tables, Figures, and Graphs Referred to but not Included in the 
Text

    Figures should be used to visually summarize the important 
results, or to clarify results that are not easily understood from 
tables.
    Important demographic, efficacy, and safety data should be 
presented in summary figures or tables in the text of the report. 
However, if these become obtrusive because of size or number they 
should be presented here, cross-referenced to the text, along with 
supportive, or additional, figures, tables, or listings.
    The following information may be incorporated in this section:
14.1 Demographic Data

[[Page 43921]]

    Summary figures and tables
14.2 Efficacy Data
    Summary figures and tables
14.3 Safety Data
    Summary figures and tables
14.3.1 Displays of Adverse Events
14.3.2 Listings of Deaths, Other Serious and Significant Adverse 
Events
14.3.3 Narratives of Deaths, Other Serious and Certain Other 
Significant Adverse Events
14.3.4 Abnormal Laboratory Value Listing (each patient)

15. Reference List

    A list of articles from the literature pertinent to the 
evaluation of the study should be provided. Copies of important 
publications should be attached in an appendix (16.1.10 and 
16.1.11). References should be given in accordance with the 
internationally accepted standards of the 1979 Vancouver Declaration 
on ``Uniform Requirements for Manuscripts Submitted to Biomedical 
Journals'' or the system used in ``Chemical Abstracts.''

16. Appendices

    This section should be prefaced by a full list of all appendices 
available for the study report. Where permitted by the regulatory 
authority, some of the following appendices should not be submitted 
with the report but should be provided only on request.
    The applicant should therefore clearly indicate those appendices 
that are submitted with the report.
    N.B. To have appendices available on request, they should be 
finalized by the time of filing of the submission.
16.1 Study Information
16.1.1 Protocol and protocol amendments
16.1.2 Sample case report form (unique pages only)
16.1 .3 List of IEC's or IRB's (plus the name of the committee Chair 
if required by the regulatory authority) - Representative written 
information for patient and sample consent forms
16.1.4 List and description of investigators and other important 
participants in the study, including brief (1 page) curriculum 
vitaes or equivalent summaries of training and experience relevant 
to the performance of the clinical study
16.1.5 Signatures of principal investigator(s) (or coordinating 
investigators) or responsible medical officer, depending on the 
regulatory authority's requirement
16.1.6 Analytical documentation--batch certificate for the 
investigational product(s)
16.1.7 Randomization scheme and codes (patient identification and 
treatment assigned )
16.1.8 Audit certificates (if available) (see Annex IVa and IVb of 
the guideline)
16.1.9 Documentation of statistical methods
16.1.10 Publications based on the study
16.1.11 Important publications referenced in the report
16.2. Patient Data Listings
16.2.1 Discontinued patients
16.2.2 Protocol deviations
16.2.3 Patients excluded from the efficacy analysis
16.2.4 Demographic data
16.2.5 Compliance and/or Drug Concentration Data (if available)
16.2.6 Individual Efficacy Response data
16.2.7 Adverse event listings (each patient)
16.2.8. Listing of individual laboratory measurements by patient, 
when required by regulatory authorities
16.3 Case Report Forms
16.3.1 CRF's of serious adverse events and withdrawals for adverse 
events
16.3.2 Other CRF's submitted
16.4. Individual Patient Data Listings (U.S. Archival Listings)
BILLING CODE 4160-01-F

[[Page 43922]]
[GRAPHIC][TIFF OMITTED]TN23AU95.011



[[Page 43923]]
[GRAPHIC][TIFF OMITTED]TN23AU95.012



[[Page 43924]]
[GRAPHIC][TIFF OMITTED]TN23AU95.013



[[Page 43925]]
[GRAPHIC][TIFF OMITTED]TN23AU95.014



[[Page 43926]]
[GRAPHIC][TIFF OMITTED]TN23AU95.015



BILLING CODE 4160-01-C

ANNEX IV a

    AUDIT CERTIFICATES (if available)
     The audit certificate is a document indicating:
     - The audit (type, identification number)
     - The audited system, clinical trial, or organization
     - The audit dates
     - The date of release of the audit report

[[Page 43927]]

    The Audit Certificate should be prepared and used in accordance 
with standard operating procedures. To be valid the Audit 
Certificate should be signed by the responsible auditor(s) or the 
head of the independent audit organization as defined in the 
standard operating procedures.
    Audit Certificates may be added to the clinical study report.
    Since it is recognized that not every clinical study will be 
subject to an independent audit, Audit Certificates for systems 
audits can demonstrate the independent assessment of the systems or 
part of the systems used to ensure that the trial is performed and 
the data are generated in compliance with good clinical practice.
    Audit Certificates pertaining to audits of systems implemented 
and existing during the planning, conduct, data analysis, and 
reporting of the reported clinical trial may be added to the 
clinical study report together with the certificate for the audit of 
the reported study, if such an audit was conducted.
    The Audit Certificate should be retained in accordance with any 
regulatory requirements.
BILLING CODE 4160-01-F

[[Page 43928]]
[GRAPHIC][TIFF OMITTED]TN23AU95.016



[[Page 43929]]
[GRAPHIC][TIFF OMITTED]TN23AU95.017



[[Page 43930]]
[GRAPHIC][TIFF OMITTED]TN23AU95.018



[[Page 43931]]
[GRAPHIC][TIFF OMITTED]TN23AU95.019



[[Page 43932]]
[GRAPHIC][TIFF OMITTED]TN23AU95.020



[[Page 43933]]
[GRAPHIC][TIFF OMITTED]TN23AU95.021



BILLING CODE 4160-01-C

[[Page 43934]]


ANNEX IX

Guidance for Section 11.3.2--Statistical/Analytical Issues

A. Statistical Considerations

    Details of the statistical analysis performed on each primary 
efficacy measure should be presented in an appendix. Details 
reported should include at least the following information:
    (a) The statistical model underlying the analysis. This should 
be presented precisely and completely, using references if 
necessary.
    (b) A statement of the clinical claim tested in precise 
statistical terms, e.g., in terms of null and alternative 
hypotheses.
    (c) When statistically reasonable and appropriate, the power 
against specific clinically meaningful alternatives for those tests 
that fail to reject the null hypothesis to indicate whether the 
study size was adequate.
    (d) The statistical methods applied to estimate effects, 
construct confidence intervals, etc. Literature references should be 
included where appropriate.
    (e) The assumptions underlying the statistical methods. It 
should be shown, insofar as statistically reasonable, that the data 
satisfy crucial assumptions, especially when necessary to confirm 
the validity of an inference. When extensive statistical analyses 
have been performed by the applicant, the extent to which the 
analyses were planned prior to the availability of data should be 
considered, and, if they were not, how bias was avoided in choosing 
the particular analysis used as a basis for conclusions should be 
described. This is particularly important in the case of any 
subgroup analyses, because if such analyses are not preplanned, they 
will ordinarily not provide an adequate basis for definitive 
conclusions.
    (i) In the event data transformation was performed, a rationale 
for the choice of data transformation along with interpretation of 
the estimates of treatment effects based on transformed data should 
be provided.
    (ii) A discussion of the appropriateness of the choice of 
statistical procedure and the validity of statistical conclusions 
will guide the regulatory authority's statistical reviewer in 
determining whether reanalysis of data is needed.
    (f) The test statistic, the sampling distribution of the test 
statistic under the null hypothesis, the value of the test 
statistic, significance level (i.e., p-value), and intermediate 
summary data, in a format that enables the regulatory authority's 
statistical reviewer to verify the results of the analysis quickly 
and easily. The p-values should be designated as one- or two-tailed. 
The rationale for using a one-tailed test should be provided.
    For example, the documentation of a two-sample t-test should 
consist of the value of the t-statistic, the associated degrees of 
freedom, the p-value, the two sample sizes, mean and variance for 
each of the samples, and the pooled estimate of variance. The 
documentation of multi-center studies analysed by analysis of 
variance techniques should include, at a minimum, an analysis of 
variance table with terms for centers, treatments, their 
interaction, error, and total. For crossover designs, the 
documentation should include information regarding sequences, 
patients with sequences, baselines at the start of each period, 
washouts and length of washouts, dropouts during each period, 
treatments, periods, treatment by period interaction error, and 
total. For each source of variation, aside from the total, the table 
should contain the degrees of freedom, the sum of squares, the mean 
square, the appropriate F-test, the p-value, and the expected mean 
square. Generally, it is recommended that regression type sums of 
squares be provided in addition to any other analyses.
    Intermediate summary data should display the demographic data 
and response data, averaged or otherwise summarized, for each 
center-by-treatment combination (or other design characteristic such 
as sequence) at each observation time.

B. Format and Specifications for Submission of Data Requested by 
Regulatory Authority's Statistical Reviewers

    In the report of each controlled clinical study, there should be 
data listings (tabulations) of patient data utilized by the sponsor 
for statistical analyses and tables supporting conclusions and major 
findings. These data listings are necessary for the regulatory 
authority's statistical review, and the sponsor may be asked to 
supply these patient data listings in a computer-readable form.

    Dated: August 15, 1995.
William K. Hubbard,
Acting Deputy Commissioner for Policy.
[FR Doc. 95-20813 Filed 8-22-95; 8:45 am]
BILLING CODE 4160-01-F