When Merck (NYSE: MRK) received bad news last year about its blood thinner vorapaxar, it stopped one trial and modified another because it was clear that patients who had previously had strokes were more likely to experience internal bleeding.

The drug is supposed to stop heart problems, at which, it turns out, it's pretty good. The modified trial, dubbed TRA-2P, passed its primary endpoint of reducing a combination of cardiovascular death, heart attack, stroke, or urgent coronary revascularization.

But the bleeding issue is still there, albeit at a lower level in patients without a history of stroke than were allowed to continue in the study. The company highlighted intracranial hemorrhage -- bleeding in the brain -- as one of the issues, which sounds like a deal-breaker to me.

For approval, the question is whether the benefit justifies the side effects. Merck is talking with regulators to see what they think of the data. Investors will get a better look during the American College of Cardiology meeting in March.

The problem is vorapaxar won't be a big seller with a bleeding side effect even if the benefits outweigh the risk. Let's say the drug saves two deaths from heart attacks for every death it creates from bleeding. Logically, you'd still want to take the medication because you're more likely to live taking the drug than not.

But from the perspective of a patients' family, those deaths are very different. Dying from a heart attack while on a drug that's supposed to treat it is unfortunate, but drugs won't help everyone. Dying from a bleeding event caused by the drug that supposed to help will be perceived by some patients' families as caused by the doctor who prescribed the drug. The fear of the wrath of patients' families will drive doctors away, even if the drug is approved.

Bleeding is always a problem with these types of heart drugs because reducing blood clots, which reduces heart attacks and strokes, also reduces clotting when bleeding occurs. Bleeding has been seen in other new heart drugs --- Xarelto from Bayer and Johnson & Johnson (NYSE: JNJ) and Eli Lilly's (NYSE: LLY) Effient -- depending on the patient population they're tested in, but not to the problematic extent that vorapaxar seems to be producing. At this point, Pfizer (NYSE: PFE) and Bristol-Myers Squibb's (NYSE: BMY) Eliquis appears to be the winner at balancing efficacy and safety and will likely be approved in May.

Retirement account losing a little blood? Put a Band-Aid on it with the Fool's new free report, "3 Stocks That Will Help You Retire Rich." You can get your copy for free for a limited time by clicking here.