Drug problem

When a used car salesman tells you all of a vehicle’s positive aspects and none of the negatives, you might call him dishonest but shrewd. When pharmaceutical companies test new drugs and publish positive results without publishing negative ones, it’s called “publication bias.” And it happens a lot.

In a study published in the online journal PLOS Medicine in December, French researchers described a striking example of this bias. They looked at medical trials reported on ClinicalTrials.gov—a website the U.S. government set up to track trial data—and searched to see how often the results were published in scientific journals.

Their investigation was startling: Just half of the trials reported on the ClinicalTrials.gov database, judging by a random sampling, had ever been published in a science journal. When they were reported in journals, the resulting articles were less likely to completely reveal trial results or fully disclose side effects from the drug or intervention in question than the original ClinicalTrials.gov reports had. While 99 percent of the original reports listed “serious” observed side effects, only 63 percent of their corresponding journal articles listed them.

Such bias may be understandable from a human perspective. No pharmaceutical company or researcher wants to publish a bad or disappointing outcome from months or years of work. But bias in research can mislead doctors who are searching medical journals for evidence that the drugs they prescribe will work and not harm patients. The moral of the PLOS Medicine report? Doctors should search ClinicalTrials.gov, too.

Ancient kindred

Native Americans may be able to trace up to two-fifths of their genome to Europeans, say researchers reporting on ancient DNA in Nature in November. The team sampled the arm bone of a young boy buried with an ivory diadem near Mal’ta in south-central Siberia (a grave they claim is 24,000 years old), and found DNA markers related to populations in early America and Western Europe. If true, that means Native Americans may not be solely descended from East Asians as previously believed. —D.J.D.

By the numbers

$1 million

The fine Duke Energy Renewables agreed to pay Nov. 22 after its Wyoming wind farms killed 14 federally protected golden eagles and other birds. The case was the first such criminal prosecution by the Justice Department. Although the deaths of protected birds at traditional electric facilities can elicit large fines, the Obama administration has gone easy on wind farm casualties.

6th century

Possible date of the dawn of Buddhism, based on the age of an apparent tree shrine archaeologists discovered beneath Lumbini, the traditional birthplace of Gautama Buddha. The researchers who conducted the dig called it “the first archaeological evidence regarding the date of the life of Buddha,” a date scholars have disputed. Not all archaeologists are convinced by the discovery, though. (Antiquity)

1 millisecond

Speed of a seahorse nabbing its prey. Scientists found that seahorses, clumsy as they may appear, are stealthy predators: The shape of their head allows them to sneak up on copepods without disturbing the water. Copepods—tiny crustaceans—can escape at a velocity of 2 or 3 milliseconds, but that’s a little too slow. (Nature Communications) —D.J.D.