Skip to Main Content

You’ve heard it many times before from your doctor: If you’re taking antibiotics, don’t stop taking them until the pill vial is empty, even if you feel better.

The rationale behind this commandment has always been that stopping treatment too soon would fuel the development of antibiotic resistance — the ability of bugs to evade these drugs. Information campaigns aimed at getting the public to take antibiotics properly have been driving home this message for decades.


But the warning, a growing number of experts say, is misguided and may actually be exacerbating antibiotic resistance.

The reasoning is simple: Exposure to antibiotics is what drives bacteria to develop resistance. Taking drugs when you aren’t sick anymore simply gives the hordes of bacteria in and on your body more incentive to evolve to evade the drugs, so the next time you have an infection, they may not work.

The traditional reasoning from doctors “never made any sense. It doesn’t make any sense today,” Dr. Louis Rice, chairman of the department of medicine at the Warren Alpert Medical School at Brown University, told STAT.


Some colleagues credit Rice with being the first person to declare the emperor was wearing no clothes, and it is true that he challenged the dogma in lectures at major meetings of infectious diseases physicians and researchers in 2007 and 2008. A number of researchers now share his skepticism of health guidance that has been previously universally accepted.

The question of whether this advice is still appropriate will be raised at a World Health Organization meeting next month in Geneva. A report prepared for that meeting — the agency’s expert committee on the selection and use of essential medicine — already notes that the recommendation isn’t backed by science.

In many cases “an argument can be made for stopping a course of antibiotics immediately after a bacterial infection has been ruled out … or when the signs and symptoms of a mild infection have disappeared,” suggests the report, which analyzed information campaigns designed to get the public on board with efforts to fight antibiotic resistance.

No one is doubting the lifesaving importance of antibiotics. They kill bacteria. But the more the bugs are exposed to the drugs, the more survival tricks the bacteria acquire. And the more resistant the bacteria become, the harder they are to treat.

The concern is that the growing number of bacteria that are resistant to multiple antibiotics will lead to more incurable infections that will threaten medicine’s ability to conduct routine procedures like hip replacements or open heart surgery without endangering lives.

So how did this faulty paradigm become entrenched in medical practice? The answer lies back in the 1940s, the dawn of antibiotic use.

A Petri dish of penicillin showing its inhibitory effect on some bacteria but not on others. Keystone Features/Getty Images

At the time, resistance wasn’t a concern. After the first antibiotic, penicillin, was discovered, more and more gushed out of the pharmaceutical product pipeline.

Doctors were focused only on figuring out how to use the drugs effectively to save lives. An ethos emerged: Treat patients until they get better, and then for a little bit longer to be on the safe side. Around the same time, research on how to cure tuberculosis suggested that under-dosing patients was dangerous — the infection would come back.

The idea that stopping antibiotic treatment too quickly after symptoms went away might fuel resistance took hold.

“The problem is once it gets baked into culture, it’s really hard to excise it,” said Dr. Brad Spellberg, who is also an advocate for changing this advice. Spellberg is an infectious diseases specialist and chief medical officer at the Los Angeles County-University of Southern California Medical Center in Los Angeles.

We think of medicine as a science, guided by mountains of research. But doctors sometimes prescribe antibiotics more based on their experience and intuition than anything else. There are treatment guidelines for different infections, but some provide scant advice on how long to continue treatment, Rice acknowledged. And response to treatment will differ from patient to patient, depending on, among other things, how old they are, how strong their immune systems are, or how well they metabolize drugs.

There’s little incentive for pharmaceutical companies to conduct expensive studies aimed at finding the shortest duration of treatment for various conditions. But in the years since Rice first raised his concerns, the National Institutes of Health has been funding such research and almost invariably the ensuing studies have found that many infections can be cured more quickly than had been thought. Treatments that were once two weeks have been cut to one, 10 days have been reduced to seven and so on.

There have been occasional exceptions. Just before Christmas, scientists at the University of Pittsburgh reported that 10 days of treatment for otitis media — middle ear infections — was better than five days for children under 2 years of age.

The superbugs are growing in number and strength. Hyacinth Empinado/STAT

It was a surprise, said Spellberg, who noted that studies looking at the same condition in children 2 and older show the shorter treatment works.

More of this work is needed, Rice said. “I’m not here saying that every infection can be treated for two days or three days. I’m just saying: Let’s figure it out.”

In the meantime, doctors and public health agencies are in a quandary. How do you put the new thinking into practice? And how do you advise the public? Doctors know full well some portion of people unilaterally decide to stop taking their antibiotics because they feel better. But that approach is not safe in all circumstances — for instance tuberculosis or bone infections. And it’s not an approach many physicians feel comfortable endorsing.

“This is a very tricky question. It’s not easy to make a blanket statement about this, and there isn’t a simple answer,” Dr. Lauri Hicks, director of the Centers for Disease Control and Prevention’s office of antibiotic stewardship, told STAT in an email.

“There are certain diagnoses for which shortening the course of antibiotic therapy is not recommended and/or potentially dangerous. … On the other hand, there are probably many situations for which antibiotic therapy is often prescribed for longer than necessary and the optimal duration is likely ‘until the patient gets better.’”

CDC’S Get Smart campaign, on appropriate antibiotic use, urges people never to skip doses or stop the drugs because they’re feeling better. But Hicks noted the CDC recently revised it to add “unless your healthcare professional tells you to do so” to that advice.

And that’s one way to deal with the situation, said Dr. James Johnson, a professor of infectious diseases medicine at the University of Minnesota and a specialist at the Minnesota VA Medical Center.

“In fact sometimes some of us give that instruction to patients. ‘Here, I’m going to prescribe you a week. My guess is you won’t need it more than, say, three days. If you’re all well in three days, stop then. If you’re not completely well, take it a little longer. But as soon as you feel fine, stop.’ And we can give them permission to do that.”

Spellberg is more comfortable with the idea of people checking back with their doctor before stopping their drugs — an approach that requires doctors to be willing to have that conversation. “You should call your doc and say ‘Hey, can I stop?’ … If your doctor won’t get on the phone with you for 20 seconds, you need to find another doctor.”

An earlier version of this story incorrectly described otitis media.

  • I have one tablet of amoxicillan for a reoccuring sinus infection leaving me with jaw/eye/ear pain and clear phlegm but my acid reflux can also trigger sinus infections. I find the antibiotics worked quite quickly yet gave me some nausea and red itchy skin on my neck which is normal, then that went away. I find antibiotics like these work well for dental problems though but not so keen on the side effects which come and go, just try to drink more water.

  • I had 8 weeks of 2 strong antibiotics and it’s left my throat feeling like phlegm in it all the time and constantly have to clear it but to no avail.
    If I had of known I would never had consented to the two antibiotics.
    I’m very very weak and can’t sleep at night

  • This also goes for the patient asking if they can get a refill to double the length of course. Doctors don’t listen to their patients enough. Then it might be weeks before they get back in and the bug has mutated enough that the particular drug is less effective.

  • I received a scholarship award back in high school AP Honors Biology but never pursued my doctorate. However my understanding banal as it may be is that both theories are correct. Overdosing the bacteria with antibiotics trains the bacterium to build resistance to the drug, but stopping antibiotics immediately when starting to feel good doesn’t mean the bacterium is no longer present. Therefore it may return with a strengthened resistance to the potency of the level of antibiotic used. Personal recommendation is take antibiotics 3 more days after feeling good. This doesn’t allow time for cells to build antigens against the drug and decimates the bacterium completely.

  • Yes Finally the truth has come out! I have always noted in the 1990s why I had to complete 40 tablets of apocillin when I was already feeling ok after 5 days (20 tablets). And I felt very bad in the last 5 days. Now glad that the truth is finally out. Once you feel better the body will have the power to ward off the rest of the bacteria. This means that for most of the people (depending upon also on the antibiotics) the medicine can be stopped once you feel perfectly Ok – ie after 5 days or worst 7 days.

  • Another problem is the overprescribing of too harsh antibiotics. Unnecessarily strong drugs or too long a dose might wreck a person’s microbiome, especially if he or she has a compromised digestive tract.

  • For new drug development, different doses are in fact studied; pharmacokinetics of these new drugs are carefully taken into consideration before deciding on Phase 2 study durations. Then, Phase 2 study results are used to figure out Phase 3 trial dosing.

    And that’s what Cempra did, but the mistake of having a loading dose to try to improve the results for seriously ill patients led to an ill-considered CRL from the FDA.

  • My experience as a dentist over 45 years was that a small percentage of folk would stop taking a Rx antibiotic once the pain or swelling subsided but would then store the remains and perhaps some months later self-administer because “I had a cold” or some other flimsy reason.

    • No excuse to self administer antibiotics, especially for a cold – will not make a difference. Maybe you could better educate your patients. After all, it only comprises a “small percentage” of your practice so the time you take may very well save humanity from the next superbug.

Comments are closed.