[ Pobierz całość w formacie PDF ]

ment numbers (for a total number of 50 patients the probability
for an imbalance as large as 18:32 is _ 0.05); it is often preferable
to stratify patients prior to grouping them in order to avoid imbal-
ances (r. within strata; ’! see stratification); for balancing num-
bers, esp. in case of small groups, it is often suitable to restrict ran-
domization e.g. in random permuted blocks (restricted r., block r.);
’! see also block size, minimization; in variable block r., a modi-
fied version of block r., the investigator does not know the num-
ber of patients to be recruited before balance is achieved (variable
rac
block sizes); it is advisable not to inform the investigator of block
sizes; in the biased coin method one observes continuously which
treatment has the least patients so far; that treatment is then as-
signed with a probability >1/2 (e.g. ¾) to the next patient; if little
is known about a new treatment in contrast to a control treatment,
esp. if this is placebo, then unequal r. may be an attractive, case
saving alternative (e.g. in phase ii or rare diseases), whereby for
every patient e.g. in the control group two patients are allocated to
the new treatment; such a 2:1 allocation would be equivalent (in
terms of power) to perform a 1:1 allocation and eliminating about
10% of the patients from the trial; unequal r. should however not
exceed a 3:1 ratio in order to avoid a considerable loss of power; a
similar r. strategy is followed in the play-the-winner allocation;
unequal r. might also be desirable when more than one treatment
155
group is to be compared with a standard control, increasing the
relative number receiving the control treatment; ’! see also ran-
domized consent design, square-root rule.
randomization code Code according to which treatments are allo-
cated to patients in a controlled clinical trial; under blinded
conditions the trialist must be able to break the code in emer-
gency cases (serious adverse events) in order to identify the
treatment; usually codes for each patient are contained in separate
envelopes; ’! see also disclosure procedures.
randomized consent design Here, in contrast to the common proce-
dure, randomization takes place before seeking informed con-
sent of patients to treatment; this results apparently in three, rath-
er than two groups: a standard treatment group as control (without
consent) and the study group which is asked for consent to the new
treatment; those patients not giving consent to the new treatment
are ultimately combined with the control group mentioned
previously; a prerequisite for the successful implementation of a
r.c.d. is that the percentage of patients in the seek consent group
and who accept the study treatment will be close to 100%; such a
design may be considered in surgical trials when it would be dif-
ficult to assign a patient at random to a more radical operation in
comparison with e.g. a standard chemotherapy; ethical problems
concerning the group  without consent may however arise when
protocols require e.g. invasive diagnostic or other procedures be-
ing not necessarily part of a  standard treatment.
randomized controlled clinical trial ’! see controlled clinical
trial.
range Interval between the lowest and the highest value within a dis-
tribution of data; ’! see percentile range, standard devia-
tion.
rap
rapporteur Original member state (reference member state or
expert of a member state) within the EC, in which a marketing
authorization for a medicinal product has been obtained accord-
ing to the criteria laid down by the EC directives or first member
state to which a high-tech procedure application (decentra-
lised procedure application) has been addressed by a company
or expert selected by the committee for proprietary medicinal
products(CPMP)/committee for veterinary medicinal prod-
ucts (CVMP) or expert designated by national authorities; the r.
notifies the CPMP of the application, prepares an evaluation re-
port with questions, circulates it to all member states and the com-
pany, makes a compilation of all eventual objections (which are
discussed/filtered by the appropriate working party and CPMP),
and sends the resulting list of objections to the company and to
156
all member states; after the answer of the company to all member
states the r. collects again the conclusions on the answers to the
questions raised and applies for the opinion of the CPMP; then the
r. as well as the member states concerned notify the Commission
of the European Community of their decision on the action to be
taken following the opinion of the CPMP; if there were no serious
objections the Commission would adopt a decision to implement
the opinion of the CPMP, if there were objections the Council
would reach a decision; if however no decision is reached by the
Council after 3 months the application would be considered to
have been rejected; ’! see also decentralised procedure.
rare diseases ’! see orphan diseases.
rate A numerical statement of the frequency of an event, expressed in
terms of person-time; differences and rates are common summary
measures; ’! see also proportion.
rating scale scale with a set of numerical categories; ’! see scale.
raw data Records or certified copies of the original clinical and labora-
tory findings from the trial; term is sometimes used as a synonym for
data in case record forms; ’! see also source data verification.
reaction ’! see adverse reaction.
Read clinical classification (RCC) System using five character al-
phanumeric codes for codifying diseases, diagnoses, diagnostic
procedures, examination findings, signs, symptoms, patients his-
tory, drugs, treatment, laboratory results, environmental and so-
cial conditions, administrative procedures, outcome and severity
measurements within a hierarchical dictionary containing more
than 30,000 terms.
rebound effect Reappearance of a sign or symptom after abrupt
withdrawal of a drug e.g. after stopping antihypertensive treatment
with clonidine blood pressure may  overshoot in rare cases.
rar
recall ’! see product recall, GMP, withdrawal.
rechallenge Reappearance of an adverse reaction on repeated expo-
sure (ethically justified only when benefits outweigh the risks); to
avoid false positive r. tests due to placebo effects or a flare-up of
the disease immediately before, the r. must be carefully planned
and performed; ’! see dechallenge, single case experiment.
rechallenge trial ’! see design.
recommended daily allowances (USRDA) values for vitamins and
minerals, established by the FDA (US) for labeling purposes; ’! see
recommended dietary allowances.
recommended dietary allowances (RDA) values for vitamins and
minerals, determined by the Food and Nutrition Board of the
National Research Council (US); intake of the RDA will provide
adequate nutrition in most healthy persons under usual environ-
157
mental stresses; they are not minimum requirements; ’! see rec-
ommended daily allowances.
reconciliation EC (IV):  a comparison, making due allowance for
normal variation, between the amount of product or materials
theoretically and actually produced or used
.
recordkeeping In USA records of a clinical trial have to be retained
for a period of 2 years following the date on which: a) the test article
is approved by the FDA for marketing for the purposes which were
the subject of the trial, b) the entire trial is discontinued or termi-
nated; records of institutional review boards must be kept for a
minimum of 3 years after completion of the research; EC: retention
of patient identification codes, patient files and other source data
by investigator for at least 15 years, all relevant documentation [ Pobierz całość w formacie PDF ]