Healthcare and Medicine Reference
Additional information about the diversity of individuals and their states is
given then by the analysis of the ranges of observations within and between
various sets. Beginners must be aware that confidence intervals and ranges
of observation are different subjects and that they have different meanings .
Both are relevant and useful and provide complementary information.
Our uncertainty about most quantitative parameters found, physiological
and pathological parameters like averages or quantiles, rates, ratios, pro-
portions, relative risks, etiological fractions, or protective efficacy rates are
then expressed by interval estimates of what we have seen with all pos-
sible implications for our decision making to solve the problem under study.
Overlapping confidence intervals of effectiveness of treatment in various
clinical trials make us think whether such results are similar or dissimilar.
British biostatisticians have presented an overall fundamental and detailed
overview of obtaining, interpreting, and using confidence intervals in
describing health phenomena and using them. 70
2.3 Fallacies in Medical Reasoning and
Scientific Thinking in General
Perhaps the most important source of fallacies in medical and scientific
thinking in general is to derive conclusions and make proposals about
causal relationships based on one fulfilled criterion of a cause-effect link
only. For example, if some relative risks or odds ratio are found to be unusu-
ally high, we can conclude that the causal link may be strong but we know
nothing else. A different situation may occur in a field trial of a new vac-
cine's effectiveness, indicating a high protective efficacy ratio (proportion of
the whole of cases prevented by intervention), let us say 80%. This is a good
indicator of causality provided that other causal criteria are fulfilled. A con-
trolled clinical trial is closer to this ideal than other observational analytical
or other type of research.
Covering the entire main body of fallacies in medicine might be over-
whelming for uninitiated readers and it goes well beyond the scope of this
introductory reading. We have devoted a book to the topic, titled Fallacy
Free Reasoning in Medicine 16 that may be welcomed by some readers as an
extension of this chapter and section.
Some of them are also of common importance in the risk and cause
assessment domains as covered in this chapter and in Section 3.1.6 of
Chapter 3 that follows.