I once had a patient in whom I found a small breast lump. She was only thirty-two, and the lump was soft, non-tender, and mobile. But it was new. She examined her breasts monthly and was certain that she hadn’t felt it the month before. And she had a family history of breast cancer.

So, I asked myself, what to do? Her age—as well as the lump’s characteristics on exam—made the likelihood that it was cancer quite low. She had a lot of fibrocystic changes in her breasts (meaning they were lumpy to start with), and this was most likely a benign lump that had enlarged from a smaller one that had been present before but that had simply gone unnoticed in a field of lumps. But the family history of breast cancer was in a first-degree relative.

I was midway through my deliberations about how to proceed when I realized I’d failed to include her in the dialogue I was having with myself. Which made me think about how often doctors fail not only to make their reasoning transparent to patients but also to reason at all. Often, doctors, like everyone else, work out of habit. If you go to a surgeon, for example, you’re far more likely to end up having surgery than if you go first to an internist simply because—to state the obvious—surgeons perform surgery. What’s not obvious, however, is how easy it is for doctors to offer what lies in their own armamentarium without bothering to ask themselves if they should. You have a cough? Fine, take these antibiotics. You have a lump? Great, let’s take it out. But these answers aren’t always correct.

Luckily, patients have a recourse. They can force their doctors to reason through problems by asking them questions. Here, then, are five questions every patient should ask when a doctor offers them a treatment of any kind:

1. What’s the likelihood it will help me? For many treatments, this answer is known. Which doesn’t, of course, mean your doctor will know it. Doctors love to quote numbers based on nothing more than their experience or even their gut feelings. And, in fact, this is sometimes the best method available; not every question we can ask in medicine, after all, has been studied. But if it has been studied, that’s where your doctor’s answer should come from. So ask specifically where his answer comes from. And if your doctor is answering you based on a study, ask him how similar you are to the patients who were studied. If they were the same age, gender, ethnicity and had the same condition, generalizing the results of the study to your own circumstance is something you can do confidently.  If they’re different from you (and they mostly will be to some degree) ask your doctor how confident he is generalizing the results of the study he’s quoting to someone like you. Try to get an actual number out of him, a percentage chance that you will improve if you have the treatment done.
2. If it does help, how much will it help me? Just because a treatment makes you better doesn’t mean you should automatically have it done.  What if you’re in horrible pain from a herniated disc and your internist offers you morphine. Certainly, you think, that will make you better. But what if it only makes you slightly better? What if you still can’t function? Then morphine isn’t the answer.
3. What is the likelihood it will harm me? No intervention—absolutely none—is without risk. Some risks are tiny (e.g., the risk of a blood draw) while others are large (e.g., the risk of open heart surgery in a ninety-year-old with heart failure). Knowing the baseline risk of any intervention is a must. How else can you weigh it against the benefit you learned about when you asked questions #1 and #2?
4. If it harms me, how much will it harm me? And how bad will that harm be? If you have a 1% chance of being harmed by an intervention but the harm we’re talking about is death (as opposed to paralysis or just transient post-operative pain), you’re going to think about the likelihood of that harm very differently.
5. What’s likely to happen if I don’t do it? Never, never forget to ask this one. Doctors, again, are just like everyone else peddling their wares: they do what they’re trained to do. And what they’re expected to do. Which means when you come to a doctor with a problem he’s going to try to solve it, and solve it with the tools at his disposal. It’s the rare doctor who will ask himself, “Do I need to do anything at all?” Yet many problems, it turns out, fix themselves. Then again, many don’t. A good doctor knows the difference. Or at least, the likelihood that a problem might fix itself. So always ask what your doctor thinks is the natural history of your particular problem. Most back pain, for example—even from a herniated disc—eventually gets better on its own and doesn’t require surgery (if you can adequately control the pain with medication).

And most breast lumps turn out to be benign. But though my patient lacked most of the risk factors that would have made me suspect her lump was malignant (i.e., it wasn’t firm, fixed, or large), she did have that family history. So I raised and answered the five questions above for her, and we discussed the answers. And she decided to have the biopsy. And it was, in fact, benign. Unfortunately, she developed a rare complication from the biopsy, an infection. Which we treated with antibiotics. Which she turned out to be allergic to. In the end she was fine, but she ended up illustrating the very point I made to her: outcomes are often in doubt and which course is best is often difficult to say. If you have a suboptimal outcome from a treatment, at least if you and your doctor thought through all the risks and benefits you can content yourself you made the best decision with the data you had available at the time.

Alex Lickerman is an internal medicine physician at the University of Chicago who blogs at Happiness in this World.  He is the author of The Undefeated Mind: On the Science of Constructing an Indestructible Self.