The health-care reform debate has heated up again this week, with President Obama giving a speech to the American Medical Association and Republicans gearing up for a fight to keep the government from encroaching further into the industry.

There's a lot to debate -- a public insurance plan, follow-on biologics, and Medicare Advantage payments, for example. But the one that scares me as an investor, and should probably scare other health-care investors, is the government's plan to get into comparative medicine. The effect on companies is very unpredictable.

Compare and contrast
Drugs can be compared in a couple of different ways. The cleaner way to make a comparison is with a direct head-to-head clinical trial where patients are randomized in groups taking different medicines and followed until some clinical outcome occurs. But, as you might guess, this is costly.

So, instead, we're likely to see a lot of studies done the dirtier way, using meta-analysis, especially with an expected increased use of electronic medical records. This is where investigators comb through databases looking for patients on different drugs for the same condition to compare. The Food and Drug Administration already does something similar using patient databases from UnitedHealth Group (NYSE: UNH) and WellPoint to try and spot side effects caused by drugs. But these after-the-fact studies really need to be followed up with cleaner, hypothesis-driven, comparative clinical trials to make sure the findings are relevant.

Comparative studies aren't really new. In highly competitive therapeutic areas like diabetes, companies usually have to run head-to-head trials to get doctors to prescribe their drug. Both Novo Nordisk and Amylin Pharmaceuticals , which have drugs before the Food and Drug Administration, have tested their drug candidates against drugs that are currently available. When companies run comparison trials, they take on the risk of failure. I mean, if the new drug turns out to be worse ... oops.

But, when the government runs the clinical trials, the companies assume a level of risk they have no control over. For instance, last year the National Institutes of Health ran a study that found Gilead Sciences ' (Nasdaq: GILD) Truvada HIV combo drug outperformed GlaxoSmithKline 's Epzicom compound. The results seem to be helping Gilead capture new patients. Last quarter, sales of Truvada were up 23% with a slight currency headwind compared to just a 10% rise of Epzicom at constant currencies. Glaxo seems to be the loser here, from something outside its control.

The clear winner here will be generic-drug makers like Teva Pharmaceuticals (Nasdaq: TEVA) and Mylan . Their low margins would never let them run a head-to-head comparison trial against a branded drug, not to mention the fact that such a trial would help out competitors selling the same generic drug. But if the government is willing to pay for those trials, the generic-drug companies are in an almost no-lose situation.

Slippery slope
The question then becomes: what to do with the data from a comparison trial? Currently it's just out there -- perhaps with some touting from the winning company -- but doctors and patients usually get to make up their minds about whether to follow the conclusions of the study or not. Continued...