But the meta-analysis was seriously flawed by what epidemiologists call confounding by indication
, which is the potential for bias to be introduced when a group of patients on medication is compared with another group off medication.
This confounding by indication
is probably one of the main challenges for the internal validity of observational studies.
This selective use of treatment leads to confounding by indication
, a well-recognized limitation of observational studies of the effectiveness of therapeutic interventions.
They found associations between each of the selected outcomes and PPIs, suggesting that there was possible confounding by indication
or disease severity that was unaccounted for in their study.
Still, the final results aren't in yet, and this study is limited by a small number of events to date, its retrospective design, and the potential for confounding by indication
We restricted our study population to patients eligible for either treatment and we used an intent-to-treat approach combined with multivariate regression to control for confounding by indication
Various rheumatic disease characteristics are associated with both the outcomes (23) and the choice of drug therapy (24) in rheumatic diseases, introducing significant confounding by indication
in examining drug effects.
That is, channeling of antimicrobial drugs toward such patients may result in confounding by indication
there is confounding by indication
[Greenland and Neutra, 1980]).
Kotecha and coworkers assessed each study in terms of its degree of bias due to confounding by indication
and other limitations using two standardized bias scoring systems: the Cochrane Collaboration's risk of bias tool for randomized controlled trials and the Risk of Bias Assessment Tool for Non-randomized Studies (RoBANS).
The association between broad-spectrum agents and higher readmission rates could be a result of confounding by indication
, wherein patients are given broad-spectrum drugs by physicians because the patients present as very sick, Dr.
These innovative applications include the use of pre/post comparison series designs, dynamic controls for confounding by indication
, visual time plots for evaluating drug indications and study outcomes, and explicit controls for healthy adherer bias.