up’. Practice showing one slide before wrongly
loading all of them.(j) Relate the middle part of your talk to your
take-home message (e.g. if disease Yis type I
diabetes, then ‘As shown in this slide, the
patient population included 30% adolescents
because this group represents a relevant
fraction of the whole population with type I
diabetes’).
(k) At the end, repeat the scientific conclusions,
briefly review the data that you have presented
in their support, and then interpret these
conclusions, once again, into your take-home
message.
Most people are in an altered psychological state
shortly after giving a talk, whether or not it seemed
to go well. In this psychological state, they gladly
accept thanks and congratulations, but are incap-
able of hearing constructive feedback. Feedback is
essential to either improve the talk the next time
round or to improve one’s presentation skills in
general. Seek out this learning opportunity from
friends, and tell them in advance that you will be
asking for this feedback, probably a few days after
the event.
Newer forms of clinical trials
publication
Electronic publishing is relatively new and is not
yet in any standardized form. It is important to
understand, however, the main classes of electronic
publication, before taking the big step of commit-
ting your clinical trial report to it. Only then can the
central question be answered for that clinical trial:
Would electronic publication make these data
more easily available to the audience that can
best use them (Geddes, 1999)?
The CD-ROM versus the textbook is probably
the most primordial form of the digital versus
analog debate. This battle has probably now been
fought to a standstill, with winners and losers on
both sides. Example replacements include the
approximately two dozen annual volumes of
Index Medicus, or both 37 annual volumes of
Headacheand 17 annual volumes ofCephalalgia,
by single CD-ROM disks. This replacement saves
trees, speeds search times, and has lower produc-
tion and shipping expenses, but requires readers to
have access to a computer at the same place as the
disk. Clinical trial databases can be usefully placed
on CD-ROM, and this can facilitate explorations
beyond the prospective trial objectives. Epidemio-
logical studies, where huge numbers of patients are
often studied, may be especially suited to this form
of publication.
Many traditional journals have sprouted electro-
nic limbs. The most common form at present is
probably thedistributionof electronicfacsimiles of
printed papers, usually in pdf format which can be
read using Adobe^1 Acrobat^1 software that can be
downloaded without charge. Access to these facsi-
miles is usually restricted to those who also have a
subscription to the paper version of the journal and
thus represents a duplication of or extension to
paper publication, rather than its replacement.
In some cases, journals publish electronically a
wider selection of submitted papers than can be
accommodated in their paper versions, or restrict
new electronic material to correspondence that
does not appear in print (Chalmers, 1999;
Delamothe and Smith, 1999; McConnell and
Horton, 1999).
Songet al. (1999) have suggested that electronic
journals can reduce publication bias (see above)
principally by accommodating and providing
access to greater quantities of published materials.
Chalmers (1999, and see above) is an enthusiast, so
presumably this is correct. Chalmers and Altman
(1999) have even proposed that not only will pub-
lication bias be reduced but also that the intrinsic
quality of clinical trials themselves could be
improved as a result of electronic publication;
this remains to be proved. However, this enlarged
volume of publications also mandates a different
peer-reviewsystem, oreven nopeer-review at all. It
is possible that electronic publications may come
to be suspected as both providing higher quantities
of information but possibly with lower quality than
more orthodox publications.570 CH42 PUBLISHING CLINICAL STUDIES