Table 1.

Common shortcomings of media reports

POTENTIAL SHORTCOMING	EXAMPLE OF BAD REPORTING	EXAMPLE OF GOOD REPORTING
Using a relative risk instead of (or without) an absolute risk	Press release by the NIH Screening for lung cancer decreases your risk of dying by 20%.27 The press release states the number of deaths in each group, but not the absolute risk reduction	Canadian Task Force on Preventive Health Care A discussion tool gives absolute numbers: 3 fewer individuals per 1000 will not die of lung cancer out of 1000 heavy smokers screened28
Confusing diagnostic procedures with screening	Fox News story Posts a story and a video about a 32-year-old woman who got a colonoscopy because of abdominal discomfort and weight loss. She encourages all young people to get screened29	HealthNewsReviews.org Explains the distinction between a diagnostic procedure and a screening procedure. In a screened population (without symptoms), the prevalence of disease is much less, so the benefits are less likely, but the harms are still present30
Only talking about benefits	Websites 98% of reports on lung cancer screening mention benefits, while only 48% present any harms31	Article about communicating risk Reports benefits and harms of screening tests using absolute (or natural) numbers with the same denominator32
Reporting benefits in relative terms and harms in absolute terms	Time article about SPRINT SPRINT reported a 38% reduction in heart failure and a 43% reduction in deaths from heart problems (relative risk) with a 1% to 2% increase of side effects apart from falls (absolute risk)33	The published SPRINT results The absolute reduction for heart failure was 0.8% and for cardiovascular mortality was 0.6%34
Equating increased 5-y survival with benefits	Radio advertisement, New Hampshire, 2007 Oct 29 “I had prostate cancer 5, 6 years ago. My chances of surviving prostate cancer, and thank God I was cured of it, in the United States, 82%. My chances of surviving prostate cancer in England, only 44% under socialized medicine”35	HealthNewsReview.org Explains how 5-y survival statistics should not be used to report benefits of screening. Lead-time bias and overdiagnosis will increase the 5-y survival rate in the screened group, even if there is no true benefit36
Assuming benefit when screening leads to more detection of disease	Press release from a company making 3D mammography technology “We remain resolute in our commitment to developing innovative new technologies ... which [detect] more invasive cancers than conventional mammography, improving a woman’s chance of survival”37	Testing Treatments “Finding more disease is not evidence of effectiveness. The possibility of overdiagnosis (which increases survival statistics) is always present. In the absence of studies confirming a benefit, we shouldn’t imply one”2
Thinking screening is the only reasonable choice	A video on Facebook Promotes prostate cancer screening by digital rectal examination or PSA.38 No explanations are provided about the total lack of evidence on the effectiveness of the digital rectal examination. There is no discussion about the option of not being screened, or about the possibility of being overdiagnosed	Globe and Mail column After reviewing the facts, the column clearly states that men should be informed before deciding to be screened or not39

NIH—National Institutes of Health, PSA—prostate-specific antigen, SPRINT—Systolic Blood Pressure Intervention Trial.