STATISTICAL TRICKS CAN DISTORT HEALTH RISKS
Recently a 76-year-old bookseller who had been taking the blood thinner clopidogrel (Plavix) became understandably alarmed by headlines warning that the drug could cause a fatal—but rare—blood disease. He decided on his own to stop taking the drug. Just a week later, he suffered a transient ischemic attack, or near-stroke.
Fortunately, he recovered quickly and, needless to say, took my advice to go back on the blood-thinning medication. The drug’s ability to guard against life-threatening blood clots far outweighed the remote risk of blood disease. As this patient found out, you can’t always count on the news media to characterize health risks fully or accurately.
In a study that was recently published in The New England Journal of Medicine, researchers analyzing more than 200 news stories about three common medications reported that many of these stories provided statistics that could easily be misleading.
HOW ROSY IS THIS SCENARIO?
Which medication would you rather take to guard against heart attack—one that reduces heart-attack risk by 2 percent, or one that reduces it by 50 percent? The fact is, statistical sleight of hand can make the same study yield either result.
Suppose, for example, a new clinical trial reports that 2 out of 100 people (2 percent) placed on an experimental heart drug had heart attacks over a three-year period, compared with 4 out of 100 people (4 percent) who’d been given a placebo. Patients and physicians mainly need to know from that study whether the medication helped a significant percentage of people—or only a few. To find that out, you’d subtract the experimental result from the control result to determine the absolute difference (4%-2%=2%). In this example, only two percent of patients who took the medication were actually protected by it.
By contrast, many news reports commonly rely on a different statistical calculation, called relative difference, which divides one result by the other to create a ratio. (2%/4%=50%). According to that approach, the medication would appear to reduce the risk of heart attacks by an impressive 50 percent. But that number is highly misleading because it fails to indicate that only a small fraction of patients who took the drug actually benefited.
IS THIS PICTURE TOO BLEAK?
Along with exaggerating the benefits of an intervention, news stories often make health risks seem far graver than they actually are. That can happen when a health risk is expressed as what’s called a relative risk.
Relative risk assumes that unity (1.0) represents the normal risk posed by a disease. Factors that lower the risk are assigned values less than 1.0 (for instance, a relative risk of 0.5 represents half the normal risk), while factors that increase the risk are assigned values greater than 1.0. For example, one Harvard study found that the relative risk of a heart attack for a physically active person during a single vigorous workout is 2.4—or more than twice the risk for that same individual at rest. When looked at in isolation, that statistic might suggest that you should hang up your gym bag forever.
To judge the magnitude of a health risk, you need to know how commonly that risk occurs within a given population and time period. For example, your annual risk of dying from a motor-vehicle accident is about 160 in 1 million.
That statistic—called the absolute risk—may not mean much when looked at by itself, however. For added context, you need to compare it with similar data on other risks. Once you’re aware that 15 per million Americans die each year from drowning and that 11 per million die from house fires, you can better appreciate the risk posed by motor-vehicle accidents.
But the absolute risk that a physically active person will have a heart attack during a given workout is minuscule— roughly 1 in 2 million. Moreover, large studies have repeatedly concluded that active people in general have half as much risk of a heart attack or cardiac death as their sedentary counterparts do.
In other words, a single episode of exertion increases the relative risk of death during the activity, but this is balanced by a protective effect from regular exercise. When taken as a whole, it’s likely that the benefits of regular exercise outweigh the risks.
That same benefits-vs.-risks balancing act occurs with every medical decision. To make sense of the science, be sure to consider the overall evidence rather than a single statistic. And focus on absolute numbers—not just the relative ones.
ALARM OVER SMOKE DETECTORS
Question: The smoke detectors in my house have small print indicating they contain radioactive material Is there any cause for concern?
Answer: No. Ionization detectors use a tiny amount of americium 241, a radioactive element, to make the air in a small chamber conduct an electric current. Smoke particles entering the chamber disrupt the current, setting off the alarm. The risk from the minute amount of radiation emitted is negligible. Such exposure is roughly equivalent to moving from one apartment to another one on the floor above, and hence that much closer to the sun.
A far greater risk is relying on ionization smoke detectors alone to protect your family. Ionization devices respond quickly to open flames. But a slow, smoldering fire, the more common type of home fire, is better detected by photoelectric units, which rely on a beam of light and a light-sensitive photocell. Moreover, photoelectric detectors are almost as good as ionization detectors in responding to “fast” fires. When Consumer Reports last tested smoke detectors, the best performers were either ionization units or combination units with both a photoelectric and an ionization sensor.
FEAR OF FIBERGLASS
Question: The fiberglass insulation in my basement ceiling is exposed. Because my wife’s throat sometimes feels scratchy when she works in the basement, she won’t let the children play there for fear the fiberglass is harmful. Is it?
Answer: Probably not in this situation. Studies have shown a possible link between exposure to fiberglass and lung cancel; but only in workers who inhale huge amounts of the fibers for many years during manufacture or installation. Fiberglass insulation that is fixed in place usually doesn’t give off airborne particles.
HEAT, HUMIDITY, AND HEALTH
Question: I’m concerned about the unhealthy effects of heat and humidity in public places. What are the dangers?
Answer: Heatstroke and respiratory infection. The risk of heatstroke becomes significant when the ambient, or external, temperature rises above 93° F. The risk of respiratory infection increases as the relative humidity falls below 20 percent. For health and comfort, ambient temperatures of 65° to 75° F and humidity levels of 30 to 40 percent are ideal.
NORMAL BODY TEMPERATURE
Question: My temperature never seems to reach the “normal” of 98.6° F. In fact, I rarely get a reading much higher than 97.5° or so, unless I’m sick. Is this unusual?
Answer: Not at all. The time-honored “normal” oral temperature of 98.6° F represented the average for healthy people, and that number has been revised downward to 98.2°. Some perfectly healthy people never break 98.0°. In addition, your normal body temperature can vary, depending in part on the time of day: It’s consistently lowest in the morning and highest in the late afternoon or evening. That daily variation can range anywhere from about 0.7° to 2.6° F.
Health handbook introducing you to read the article: HAIR CARE: ESTROGEN AND HAIR LOSS
Copyright ownership rights: The Best of Health – Consumer Reports