Think the news media is untrustworthy? Try reading the medical journals.

Major studies are being retracted almost daily for everything from lousy data to outright fraud -- and that's just scratching the surface, because the studies that haven't been retracted could actually be worse.

A leading medical journal says most of today's studies are published without key data -- omissions that could hurt or even kill patients as doctors unwittingly rely on an incomplete picture of a drug's risks when they make treatment decisions.

BMJ recently published no less than seven papers on the problems with today's research -- but the biggest are those sins of omission.

One paper shows how quickly incomplete data can turn a snowball into an avalanche as research that combines data from earlier studies -- called meta-analyses -- ends up magnifying all those flaws and omissions.

Researchers reviewed 41 meta-analyses on nine different meds and found that 38 of them would have had different results if the researchers had access to the missing data.

Think about that next time your doctor offers you a med and cites some important-sounding study in a major medical journal.

In another paper, researchers checked to see how many studies complied with a rule that requires summaries of every single study to be posted on a government Web site, ClinicalTrials.gov within a year of completion.

They found that just 22 percent complied with this "mandatory" rule.

Looks like "mandatory" just ain't what it used to be, but there's a reason for that: When a study shows a drug doesn't work or comes with too many risks, Big Pharma would rather sweep it under the rug -- and researchers are often willing to grab a broom and help out.

One study a couple years back found that 85 percent of published company-funded studies found a benefit for the drug -- an unbelievable success rate no doubt aided by the fact that only 32 percent of company-funded trials are even published in the first place.

You can probably find the other 68 percent in the nearest shredder. (Read more about why drug studies can't be trusted here.)

Put it all together, and you get what an editorial in BMJ calls a "culture of haphazard publication and incomplete data disclosure" that makes any true risks-benefits analysis "almost impossible."

I'd say we're way beyond "almost."