The epiphany came in a dentist's chair. That's where Charles Kelman, a former Wills Eye resident, dreamed up the idea of modern-day cataract surgery. Why not use the same ultrasound technology for teeth cleaning to take out cloudy lenses in the eyes?

At the time, in the late 1960s, everyone thought cataract surgery had peaked and there was nothing left to discover: A patient endured a bloody, painful operation, got hospitalized for a week, and emerged with decent vision but thick Coke-bottle glasses.

The majority of eye surgeons deemed Kelman a reckless cowboy and pooh-poohed his ideas. He wanted to break up the lens with ultrasound and then vacuum it out through a small hole instead of yanking it out in one piece. Thanks to a small group of believers who persevered with Kelman, his ideas became the foundation for swift, painless contemporary cataract surgery.

Now consider this: What if back in the 1970s, Obama had been president? Obama recently signed into law the stimulus package, which included $1.1 billion to jump-start research on the effectiveness of drugs and technologies. The goal is to answer such questions as "Does one therapy work better than another?" and "Is it worth paying for?"

In the beginning, Kelman's way of doing cataract surgery was less safe, more expensive, and less effective than the established method. Compared to the old method, in the first several years - or even the first decade of its evolution - cataract surgery as it is today likely would have been institutionally and financially stymied.

I wonder whether the Obama plan will hamper such innovations and wind up as a Trojan horse for cutting costs and rationing care. While the plans won't be made public until later this month, Obama seems intent on creating a government body that picks winners and losers, approving some innovations but not others.

This worries Bernadine Healy, former director of the National Institutes of Health: "Even though your loved one's cancer very likely would be helped by a $50,000 drug treatment, his doctor might well say no," she wrote recently on her blog Heart to Heart. "That would happen if comparative-effectiveness research has ruled that the benefits of the drug to the average patient don't justify its price when compared with yesterday's medicines."

I worry too that this strategy could pull the physician further from the patient. Is it a thinly disguised version of a large HMO - the government - determining how doctors should think and treat their patients individually?

We already suffer from insurance companies directing care. I cringe every time I get a call from a pharmacist, in the middle of a busy clinic, saying "Sorry, doctor, that eye drop is not covered by the patient's insurance. Could you use this drop?" More often than not, the suggested drop is subpar.

The drops I prescribe take some thinking and are personalized to the patient's condition: What will give the best antibiotic coverage for the type of infection this could be? Or which glaucoma drop hasn't worked so far - and which one would I predict next could work? Recipelike algorithms don't take these details into account.

I'm not saying doctors should spend health-care dollars for patients willy-nilly. We should determine care based on the best medical evidence and fully consider cost as well. Certainly the more evidence, the better.

One big problem, though, is that the price of care is hidden.

So here's my two cents on how to cut costs and keep innovations coming. Physicians often recommend care without knowing how much patients will have to pay.

When I order an eye drop, say, I would like to get the cost at one centralized Web site instead of having to go to each insurer's Web site. That way, instead of learning later that they couldn't afford the drop and have had no treatment, I could discuss the matter preemptively.

And that's the ultimate question: What are we willing to spend for new treatments and for old ones? The doctor, the patient, and now the government will have to work on that answer together - or fight about it separately.

Readers respond. My last column - "Waiting to see the doc: We truly do apologize" - struck a nerve.

The verdict from more than 100 e-mails was mixed on whether I was sufficiently sympathetic. Some readers felt I validated their frustrations, others were adamant that a late appointment is unacceptable no matter what.

I received some constructive suggestions. Doctors' offices should help patients understand how long the expected backup will be. They should encourage patients to call ahead to see if there's a delay and then be transparent about the wait time at the office. These are good ideas that I think many doctors, including myself, should take to heart.

Contact Rachel Sobel at rachelkimsobel@gmail.com