|
Facts
of Life
Facts of Life:
Issue Briefings for Health Reporters
Vol. 10, No. 1
January 2005
Risky Business: Conveying
Disease Statistics
The Issue
The Facts
Catching the problem early
Expert Sources
References
The
Issue:
In health, risk is
the chance that a person will develop a disease over a certain period of
time. It is a mathematical description, a statistical
glimpse at what might be. The same “risk” attached to a new drug,
a surgical procedure or a behavior like unprotected sex can be described in
a variety of ways, from odds ratios to relative and absolute risk. Despite
recent news about pain relief drugs, common fears about breast cancer and prostate
cancer and the potential outbreak of a bird-flu pandemic, risk remains a poorly
understood concept by consumers, patients and even some doctors.
Risk as Rollercoaster
For many
patients, risk is meaningful beyond the numbers. In 2001, the odds of dying
in a pedestrian
accident were one in 46,960, while
the odds of dying
in an earthquake were one in 10,181,922. [1] Yet most people would call living
in an earthquake-prone area “riskier” than walking down the street.
Risk is an emotion-laden term, and personal perceptions can color how people
view the risk inherent in an
illness or a behavior.
What, Me Worry?
Risk is a population-based measure, the chance of something
happening as determined by its occurrence among a large group of people over
time. An
individual's risk varies considerably within the numerical boundaries of
a population's risk, due to variations in personal genetic, environmental and
behavioral factors. Individuals may alter their health behaviors after becoming
aware of population risk for a certain condition or they may dismiss population
risk as something that happens to everyone except themselves.
The
Facts:
- An analysis of randomized
controlled trials and systematic reviews suggest that health programs that
present their benefits in terms of relative risk are funded more often
than those that use absolute risk or other statistical descriptions. [2]
- Researchers found that
readers of breast cancer risk pamphlets were more likely to understand
risk with the help of graphics that contained human
stick figures rather than bar graphs. [3]
- In a study of 207 news
articles about three major medications, only 60 percent reported the benefits
of the medication in a quantitative form.
Of
those that did, only 2 percent mentioned the absolute, rather than
relative, benefits of the drug. [4]
- A survey of nearly
1,000 people found that individuals were less likely to recommend a hypothetical
medication when the risk of taking the drug was
presented in relative, rather than absolute, terms. [5]
- A systematic review
of studies examining individuals' beliefs about family history and the
risk of developing a disease found that individuals
base their belief on some medical factors, such as number of relatives
affected, as well as less objective criteria like a relative's suffering
during the disease. [6]
- In a recent Web questionnaire
that included several types of physician statements about cancer risk,
participants
who overestimated their risk were likely to
believe that the physician had minimized the risk to reduce the “patient’s” worry.
[7]
- When considering the risks of lung cancer surgery, possible death and
disability from the cancer seem to figure more prominently in a surgery
candidate's risk calculations than possible surgery-related death or postoperative
complications. [8]
- Money paid to students
for their participation in mock clinical trials made it more likely that
the students would agree to volunteer for the
trials, regardless of the level of risk associated with each trial, according
to a recent study. [9]
- Nearly 18 percent
of individuals who are older than 85 are considered to have a poor diet,
compared to 16 percent of all Americans. [9]
- Patients tend to overestimate
the risk of drug side effects when those side effects are presented verbally,
using terms like “rare” or “common,” compared
to numerical (percentage) resentations of side effect risks, a U.K. study
finds. [10]
- A systematic review of studies on communicating clinical evidence to patients
indicates that less-educated and older patients prefer proportions to percentages
when discussing risk with their physicians. [11]
Catching
the Problem Early
The numbers are everywhere: A drug ad touting a 40 percent
reduction in the risk of high cholesterol. A study that says taking another
drug can increase the risk of breast cancer by 20 percent. The advice that
a daily walk might cut the risk of future obesity by 10 percent. Although
the percentages sound authoritative, they're actually only the first step
in understanding risk, according to Dartmouth Medical School researchers
Lisa Schwartz, M.D., and Steven Woloshin, M.D.
Schwartz and Woloshin
say there are several questions that need answers before a patient or physician
can fully comprehend risk reported
for a drug or procedure.
For instance, what outcome does the “risk” measure — the
chance of developing a disease or the chance of dying from it? Is the disease
dangerous, in the sense that developing the disease necessarily leads to
death? Is risk calculated in an unusual group — elderly cancer patients,
for example — that might make the risk calculation less applicable
to the general public?
Risk may also be
expressed several ways mathematically, which can lead to confusion. One
of the most common ways of expressing risk is
by presenting
the absolute or relative difference in a condition between a treated
group and a non-treated group. Absolute and relative risks are calculated
from
the same numbers, but the two can “feel” very different,
Schwartz says.
For instance, if the risk of developing heart disease is four
in 100 patients who do not take a cholesterol-lowering drug and two in 100
patients who do,
the absolute difference in risk between the two groups is a simple subtraction:
2 percent.
Relative risk, on the other hand, is expressed as the ratio
of the risk percentages in each group. Using the same data, the
relative risk of heart disease is determined by dividing the 2 percent
of those with heart
disease who used the drug by the 4 percent of those with heart disease
who didn't. In this case, the reduction in risk would be
50 percent.
Both numbers are
technically correct, but make for very different headlines in a major health
news story. The key to understanding such a finding,
say Schwartz and Woloshin, is providing more background information to
evaluate
the numbers, including a full accounting how many heart disease events
occurred in the first place “
Just knowing a relative risk reduction like '50 percent fewer' without knowing
the base rate — 50 percent fewer than what? — is insufficient,” Woloshin
says.
The way that risk
is expressed can make a big difference in how prominent a risk seems, as
illustrated by reports on the recent Women's
Health
Initiative study on the effects of estrogen and progestin hormone
treatment. A Washington
Post article on the study reported that women taking the
hormones had “a 41 percent higher
incidence of strokes” compared with women taking a placebo.
A few paragraphs later in the article, the absolute numbers of the
study
came out: the 41
percent increase corresponded to eight more strokes per 10,000 women
for a year.[13]
Expert
Sources:
Lisa Schwartz, M.D.
Dartmouth Medical School
(802) 296-5178
lisa.schwartz@dartmouth.edu
Steven Woloshin, M.D.
Dartmouth Medical School
(802) 296-5178
steven.woloshin@dartmouth.edu
Allen Sanderson, Ph.D.
University of Chicago
(773) 256-6269
arsx@uchicago.edu
David Ropeik, M.A.
Harvard Center for Risk Analysis
(617) 432-6011
dropeik@hsph.harvard.edu
References
1. National Security
Council (2001) What are the Odds of Dying? Resource Sheet. Last accessed
12-10-04 at http://www.nsc.org/lrs/statinfo/odds.htm.
2. T. Fahey et al. (1995) Evidence based purchasing: understanding results
of clinical trials and systematic reviews. British Medical Journal, 311, 1056-1059.
3. M. M. Schapira et al. (2001) Frequency or probability? A qualitative study
of risk communication formats used in health care. Medical Decision Making,
21, 459-467.
4. R. Moynihan et al. (2000) Coverage by the news media of the benefits and
risks of medications. New England Journal of Medicine, 342, 1645-1650.
5. L.A. Hembroff et al. (2004) Treatment decision-making and the form of risk
communication: results of a factorial survey. BMC Medical Informatics and
Decision Making, 4, 20.
6. F.M. Walter et al. (2004) Lay understanding of familial risk of common
chronic diseases: a systematic review and synthesis of qualitative research.
Annals of Family Medicine, 2, 583-594.
7. A.D. Gurmankin et al. (2004) Intended message versus message received in
hypothetical physician risk communications: exploring the gap. Risk Analysis,
24, 1337-1347.
8. S. Cykert (2004) Risk acceptance and risk aversion: patients' perspectives
on lung surgery. Thoracic Surgery Clinics, 14, 287-293.
9. J.P. Bentley and P.G. Thacker (2004) The influence of risk and monetary
payment on the research participation decision making process. Journal
of Medical Ethics, 30, 293-298.
10. P. Knapp et al. (2004) Comparison of two methods of presenting risk information
to patients about the side effects of medicines. Quality and Safety in
Health Care,13, 176-180.
11. R.M. Epstein et al. (2004) Communicating evidence for participatory decision
making. Journal of the American Medical Association, 291, 2359-2366.
12. S. Okie. Hormone Treatment Is Called Harmful; Menopause Study Cites Health
Risks. Washington Post, July 10, 2002, A1.
13. Writing Group for the Women's
Health Initiative Investigators (2004) Risks and benefits of estrogen plus
progestin in healthy postmenopausal women: principal
results from the Women’s Health Initiative Randomized Controlled Trial.
Journal of the American Medical Association, 288,321-333.
The
Center for the Advancement of Health is an independent nonprofit organization
that promotes greater recognition of how psychological, social, behavioral,
economic and environmental factors influence health and illness. The Center
advocates the highest quality research and communicates it to the medical
community and the public. The fundamental aim of the Center is to translate
into policy and practice the growing body of evidence that can lead to the
improvement and maintenance of the health of individuals and the public.
The Center was founded by the John D. and Catherine T. MacArthur Foundation
and the Nathan Cummings Foundation, which continue to provide core funding.
Funding for this series was provided by the Robert Wood Johnson Foundation.
For Information Contact:
Kristina Campbell
Editor, Health Behavior News Service
Center for the Advancement of Health
2000 Florida Ave., NW, Suite 210
Washington, DC 20009
p. 202.387.2829 / f. 202.387-2857
press@cfah.org
http://www.cfah.org
© Copyright 2004, Center
for the Advancement of Health
PDF
Version |
