Herbert and Bo have addressed a crucial issue in clincial trials, and
in their
metanalyses. Clinical trial methodology is frequently inadequately
explicit, with
significant variation within and between institutions. Many clinical
cointerventions (confounders introduced after randomization by the
interaction
of the patient with clinicians) can influence the outcomes of studies.
Failure to
uniformly apply these cointerventions to all experimental groups, or as
Herbert and Bo have suggested, falure to adequately apply the experimental
intervention, can significantly alter outcomes and their interpretation.
Scientific rigor requires more attention to experimental methods, and
to
cointerventions in many clincial trials, than is commonly employed. I
applaud
the emphasis of Herbert and Bo on the quality of the interventions in
clinical
trials. This apears to be largely overlooked.
Competing interests:
None declared
Competing interests:
No competing interests
02 September 2005
Alan H. Morris
Professor of Medicine
Pulmonary Division, LDS Hospital, Salt Lake City, Utah, 84143, USA
Rapid Response:
Quality of clinical trial methods
Herbert and Bo have addressed a crucial issue in clincial trials, and
in their
metanalyses. Clinical trial methodology is frequently inadequately
explicit, with
significant variation within and between institutions. Many clinical
cointerventions (confounders introduced after randomization by the
interaction
of the patient with clinicians) can influence the outcomes of studies.
Failure to
uniformly apply these cointerventions to all experimental groups, or as
Herbert and Bo have suggested, falure to adequately apply the experimental
intervention, can significantly alter outcomes and their interpretation.
Scientific rigor requires more attention to experimental methods, and
to
cointerventions in many clincial trials, than is commonly employed. I
applaud
the emphasis of Herbert and Bo on the quality of the interventions in
clinical
trials. This apears to be largely overlooked.
Competing interests:
None declared
Competing interests: No competing interests