Double Lens 185
affecting perceptions of statistics have been explored in economics and
other fields, but have been less examined in medicine. The psychologist
Daniel Kahneman received the 2002 Nobel Prize in economics for de-
scribing how individuals commonly use ‘‘biases and heuristics’’—for
example, overvaluing the odds of bad events and undervaluing the prob-
abilities of good outcomes (3). Several psychological studies have probed
such misperceptions of statistics, generally among college students and
in regard to hypothetical situations. Interpretations of statistical infor-
mation are affected by how it is framed, either positively or negatively
(4). Among medical students, numerical estimates have been found to be
more effective than verbal ones. Rather than being wholly disadvanta-
geous, these seemingly illogical heuristics and biases may offer certain ad-
vantages. Gerd Gigerenzer’s book Simple Heuristics That Make Us Smart
argues how evolutionarily, unconscious cognitive strategies allow hu-
mans to very rapidly draw intuitive conclusions about complex situations
(e.g., the likelihood of escaping predators in one direction or the other,
given the distance to forest, etc.) (5).
Yet how these intuitive, but subjective and ‘‘illogical,’’ factors affect
physicians and patients in their assessments of risks and benefits remains
much less clear. Donald Redelmeier, an internist, outlined categories of
biases and heuristics that distort patients’ perceptions of risks in medi-
cine (6). Health care professionals present risks in varying ways, for
example, using a wide range of verbal descriptions (7). Though differ-
ent groups of medical personnel assign similar quantitative probabilities
to qualitative descriptions of odds, providers may differ from patients
in interpreting the words used (e.g., ‘‘ ‘likely’ ’’ versus ‘‘ ‘unlikely’ ’’) (8).
Even when numbers rather than verbal descriptions are used, patients
may assess risks differently if presented with rates rather than propor-
tions (9). The sequential order in which risks and benefits are presented
can also influence responses (10). Many questions persist as to the im-
plications of these issues among patients and doctors. Though aspects of
these issues have been explored, particularly in genetics, these issues of
interpretation of statistics have been probed less in other areas of doctor-
patient communication. Psychological experiments have been conducted
using hypothetical scenarios with students. But it is not clear how these
results apply in the real world. Hence, understanding how doctors who
have become patients view these issues and change their perceptions can
be helpful.