Antibiotics for Sinusitis

You’re a patient. That cold just isn’t getting better and you have purulent drainage from your nose, and your face hurts and your teeth hurt. You probably have sinusitis, right? You go to a doctor to get an antibiotic.

You’re a doctor. Deep down, you know there’s a good chance the patient has a self-resolving condition.  You’d rather not do x-rays on every patient who presents with these symptoms, because x-rays are expensive, expose the patient to harmful radiation, and they are not always accurate.  You could puncture the sinuses and take a sample for bacteriological culture, but that’s expensive, painful, and the patient would NOT appreciate it. The patient may not really need treatment, but you want very much to do something to help. If you can find a reason to give the patient an antibiotic, you can feel that you have done something worthwhile.  Antibiotics don’t work for a viral infection, but you rationalize that you’re not 100% sure it’s not bacterial, and that sometimes a bacterial infection develops superimposed on a viral infection and mild bacterial infections can develop into severe ones with complications, and maybe you could ward that off.  You convince yourself that it really would be prudent to prescribe an antibiotic. Both you and your patient are happy. The patient gets better. You remember this pleasant experience and are reinforced to do the same next time.

Maybe that’s not such a good idea.

A recent study in JAMA showed that antibiotics were not superior to placebo for treating bacterial sinusitis diagnosed by the clinical criteria used by many primary care physicians. There is a growing concern that we have been overtreating sinus infections. Recent research has also shown that we were overtreating ear infections in children, that many of them resolve just as fast without treatment. That doesn’t mean antibiotics should never be used for ear infections. There are now guidelines for using age and clinical presentation to determine which children to treat and which ones can be safely observed without antibiotics. Most of these observed ear infections will resolve but some will eventually require antibiotics too. The situation with sinusitis is similar: most patients may not need antibiotics, and we’re trying to thrash out better criteria for identifying those who will.

Critics point to this kind of thing as a defect of conventional medicine. We use treatments that are not based on good evidence. We do things for years and then find out we were wrong. New studies are constantly contradicting older studies. We keep changing our minds.

In fact, this kind of thing illustrates the value of scientific medicine. Science never claims to have absolute truth. It is willing to change its practices as the evidence changes. More importantly, evidence-based doctors are willing to stop doing things that the evidence says don’t work. No amount of influence from “Big Pharma” will perpetuate the use of drugs for sinus infections if the evidence shows they are unnecessary. No amount of irrational belief will trump the evidence. No doctor can get away with saying, “In my experience, …” when good evidence contradicts that experience.

Remember when annual chest x-rays and urinalyses were part of the recommended annual physical? No more. And it would be easy to make a long list of similar relinquishments.

Contrast that with “alternative” medical systems like homeopathy or chiropractic. They don’t change very much, and they almost never give anything up. I searched the literature and consulted professors of chiropractic in an attempt to find anything that chiropractors have uniformly stopped doing, and I was only able to find one thing that fit the bill. I don’t think any chiropractor is still using Palmer’s “nerve tracing” technique where he claimed to be able to feel nerves unknown to anatomists.

Limiting antibiotic use to strict evidence-based guidelines has several benefits: it reduces medical expenses, it minimizes unnecessary side effects, and it reduces the risk that the bacteria will develop resistance to the antibiotic. 

So we’re going to bite the bullet, follow the science, and learn to be much more selective about using antibiotics. Now, the only problem with being scientific is that the patients are not going to be happy. They’re used to going to the doctor to get a “cure.” They’re used to seeing their sinusitis symptoms resolve after getting antibiotics, and they will be unwilling to forgo treatment long enough to find out if their symptoms resolve just as fast without antibiotics. They want their doctor to “do something.” And when he doesn’t, they may be attracted to alternative providers and quacks who will always “do something” and get the inevitable placebo effect.

The frustration science-based doctors feel was illustrated by a tongue-in-cheek quiz in the February 2008 issue of the medical humor magazine Placebo Journal:

 You diagnose a young child with acute otitis media. Which are the most common reactions by the mother when you recommend observation for the treatment instead of antibiotics?

A. Anger

B. Disbelief. Then anger.

C. Disbelief. Then anger. Then appreciation for the honesty and information after which she leaves the office to go to the ER for antibiotics.

D. None of the above as you stopped this bullsh#t approach to treatment years ago because you saw you were losing too many patients. 

And in the general course of human perversity, some patients will inevitably go to the other extreme and cite these studies to refuse treatment for cases where antibiotics really are indicated.

The sound bite may be that “antibiotics don’t work for sinusitis,” but the real story is a bit more complex. The challenge is twofold: getting doctors to apply the best evidence appropriately, and getting patients to understand why they may not be getting that magic pill they have learned to expect.

Posted in: Science and Medicine

Leave a Comment (22) ↓

22 thoughts on “Antibiotics for Sinusitis

  1. Michelle B says:

    So many important points excellently covered! Thanks.

    My mum-in-law would continue to go to her GP for antibiotics to treat her common cold despite my informing her that it was not necessary in order for her to get completely well, that it was an expense not required, and that it could build up her resistance to an antibiotic in the future when she really needed its help.

    Even after her GP stopped giving antibiotics she still went to him hoping that he would change his mind, and go against the evidence of which he presented to her why he is no longer prescribing antibiotics for the common cold–that is how ‘addicted’ she was to what she decided the optimal treatment was. Eventually she stopped going altogether when she got a cold, saying, the doctor will not give me anything to make me better so why go? (At least she did not go to a non-evidence based practitioner of medicine)

    I pointed out that since she was over eighty she needed to keep contact with her doctor when she got ill, and she would be more vulnerable to secondary infections, of which antibiotics would cure. I doubt she will heed my advice. I truly appreciate the frustration that a thoughtful, conscientious, and well informed doctor must go through each and every day.

  2. delaneypa says:

    This is something I have already had to address three times today – “my sinuses are back where’s my Z-pak”. The NEJM finding is nothing new, it seems every few years that ABX are found to be ineffective for sinusitis in the absence of certain red flags. The same story also appears in the newspapers every winter at least once (along with “doctors discover cough medicine ineffective for colds”). I save some of those articles and print them out for patients. Heavens knows they won’t believe me, a board-certified MD. But if it’s in the newspaper it must be true!

    Someday I would like to conduct a study of antibiotics use among children of physicians (or physicians themselves) versus the general population, see if we practice what we preach.

  3. Fission says:

    In cases like this I sometimes wonder if it would benefit patients and doctors to “prime” the patients first.

    Maybe as part of the paperwork a patient fills out before their appointment they could answer questions like “Do you believe that a drug of some kind is always necessary to help you overcome your health issue?” (clearly the wording could be better, but you get my point).

    I’m sure lots of patients (including me?) would see this as an annoyance, but forcing them to actually answer a question like that (rather than just read a pamphlet of facts) might make a difference.

  4. TurangaLeela says:

    I would love it if doctors would give me the ‘wait and see’ advice for ear infections, however, I have never once heard it.

  5. humgums says:

    “Someday I would like to conduct a study of antibiotics use among children of physicians (or physicians themselves) versus the general population, see if we practice what we preach.”

    That would be an interesting study.

    Anecdotally: I was discussing the revised AHA guidelines for antibiotic prophylaxis with my cardiologist when they first came out (antibiotic prophylaxis for infective endocarditis is no longer recommended before dental treatment in most cases). He was somewhat skeptical and his parting comment was “I hope you will continue to premedicate yourself when you get your teeth cleaned (I’ve got MVP with moderate regurgitation – for which the AHA no longer recommends prophylaxis).”

  6. McDoctor says:

    I think it’s worth pointing out that this particular JAMA study recruited just 240-250 patients over 3 years from 70+ primary care physicians. That’s little more than 1 patient per doctor per year. This is incredibly low given the volume of patients presenting with sinus symptoms in a given year in a typical practice.

    Either doctors were intentionally excluding many patients (very unlikely) or this study truly shows how difficult it is to convince patients to “wait and see” and potentially not prescribe an antibiotic. Just 1 patient per year per for any given doctor was willing to enroll in a study wherein they were only 50% less likely to receive an antibiotic.

    I advocate a wait and see approach in most cases. I do sometimes write a script and encourage parents to hold off on filling it, as a compromise. For my own daughter I would prefer she not take the antibiotic unless she clearly had an indication, and I’d wager my own family members use antibiotics far below the norm.

  7. BlazingDragon says:

    I do hope this wont’ be taken as a blanket by doctors who will then never prescribe antibiotics until the patient is well and truly sick. When I get a sinus infection, it trickles down into my lungs, aggravates my asthma, and gives me bronchitis if not treated aggressively (about 70-80% of the time). My ear infections don’t go away on their own either. Doctors seem to “throw the baby out with the bathwater” every time a study like this comes out.

    I agree that we shouldn’t over-use antibiotics, but is there any talk in human medicine about the awful over-use of antibiotics in the raising of animals? Massive amounts of antibiotics are used in animals every year, whether they get sick or not, whether they are needed or not, because, statistically, animals get to slaughter weight faster if all of the animals are given antibiotics. This practice is a major source for antibiotic-resistant bacteria.

    I strongly believe in using evidence to weed out useless therapies and woo-based crap, but I have often seen doctors use studies such as this one to make their lives easier (come back in a week if your ears/sinuses aren’t better). Things like this take the personal touch out of practicing medicine. Too many physicians I’ve witnessed short-circuit the decision tree completely (or completely ignore the fact that a patient may not fit on a decision tree designed for “average” populations). Having been a victim of this type of “treatment,” I’m very afraid that broad application of relatively simple-minded decision trees will leave those among us that don’t fit into these simple-minded decision trees out in the cold (even more than we are already left out in the cold).

    Disclaimer: This fear of the way in which evidence-based medicine is currently being practiced is NOT an excuse for woo-based therapies, nor is it a rejection of using good evidence to make medicine better. It is simply a statement of my experiences with the treatment I get when I would never qualify for any study because my biology is so odd when reacting to most treatments.

  8. PalMD says:

    It’s tough out there. There’s now evidence that nasal steroids are also ineffective in acute sinusitis. It’s hard being a doctor—people are messy machines and require a lot of talking to, and sometimes they stomp out mad.

    I always make sure they feel free to come back or call me and let me know how they are doing, and that usually gets me by, but honestly, even my own wife doesn’t believe me.

  9. humgums says:

    “but honestly, even my own wife doesn’t believe me.”

    LOL! I was completing my PhD in immunology when my wife and I started trying to have a baby. I did a test for HCG (which uses run of the mill immunology techniques) with a kit only available to research and clinical labs – they weren’t available OTC 32 years ago. I found that she was pregnant, but she didn’t believe it until her OB got the results from a commercial lab.

  10. BlazingDragon says:


    I don’t think it’s any wonder why nasal steroids aren’t useful for sinusitis. I am guessing that a lot of cases of sinusitis are bacterial… and steroids are powerful immunosuppressants in addition to being powerful anti-inflammatories. The benefit of shrinking of the nasal passages may be mitigated or wiped out by the suppression of the immune system. I dunno if this is actually true, but it never made much sense to me to hand out steroids for every little inflammatory condition (as seems to be the practice for at least the last 5-6 years) like asthma (associated with bronchitis/pneumonia) and sinusitis. Because of the immunosuppression inherent in steroids, this seems like it could backfire in many cases. Steroids can be used in cases where the sinus passages are extremely swollen and are can be damned helpful in cases of asthmatic bronchitis, but I think they are over-used these days.

    I wonder how many GPs handing out prednisolone dose packs are aware of the small but definitely non-zero risk of avascular necrosis (aware enough to not hand them out willy-nilly, which has been my observation). It’s one thing to be aware of a risk as a statistic on a piece of paper, it’s another thing to know that even 0.1% of patients can easily be one person whose life can be forever altered by a simple steroid dose pack.

  11. PalMD says:

    I don’t see internists handing out steroids like candy anymore. When I was a student, I saw a family practice doc who used to give his hay fever patients a shot of depo medrol every spring…not the greatest idea.

    But inhaled corticosteroids have a very low incidence of systemic side effects. They aren’t associated with significant osteopenia, avn, dm, etc.

    Very short courses of steroids for asthma, gout, etc are usually quite safe, and have little immune or adrenal suppression associated with them.

  12. McDoctor says:

    Inhaled steroids long term do seem to increase the risk of bacterial pneumonia (in COPD patients), evidenced in 2 recent RCT’s. It would make sense then that intranasal steroids could increase the risk for bacterial sinusitis, but probably only with long term use.

  13. Noni Mausa says:

    1971. Minneapolis. I had the Cough From Hell and after a month it showed no sign of resolving. It was the sort of hacking cough that made little old ladies, strangers on the bus, try to take me for medical help.

    So I finally went to the clinic. After a cursory inspection and a diagnosis of Viral Infection DKWK*, the resident said he would give me a prescription for antibiotics.

    I refused of course because if it really was a virus the antibiotic wouldn’t help the cough, but might educate little baby bacteria about developing immunity. After paying my fee, I went home, grumbling and coughing.

    37 years later we’re still at it? Oh dear.


    * Don’t Know What Kind

  14. HCN says:

    PalMD said “But inhaled corticosteroids have a very low incidence of systemic side effects. They aren’t associated with significant osteopenia, avn, dm, etc.

    Very short courses of steroids for asthma, gout, etc are usually quite safe, and have little immune or adrenal suppression associated with them.”

    My experience with them for hayfever (allergic rhinitis?) was that they were yucky. It may have been the solvent was irritating, but I could never ever get used to the steroid spray up my nose that was prescribed by the allergy doctor (this was in the late 1970s). I was also told I had to use them consistently over a week for them to work.


    Fortunately, my allergies were (and are) of the sort that responded to the weekly shots (yay!… or maybe the two years of getting shots was a placebo), avoiding allergens (sorry, dear spouse, I’m allergic to dusting), and more modern medications (hurray for the ones that don’t knock you out! … and hurray for benadryl for allowing me to sleep during itchy times). Well, for a week in the Spring I do live with puffy eyes, but I can live with that… as long as I don’t have to breathe in those awful steroid inhalers. Ugh.

  15. Laurel says:

    This has driven me nuts for decades now. Whenever I’m offered an antibiotic by a doctor I ask that a white cell count be performed. The doctors almost always back down and say they’re not sure I need an antibiotic, and I ask them why on earth they want to give me a yeast infection and make antibiotics less effective by giving them to people who don’t need them.

    The answer: “People don’t like to leave the doctor’s office without a prescription.”

    WTF?! I know it’s hard to deal with cranky people whose sinuses hurt, but isn’t this the kind of thing that brought us antibiotic-resistant staph?

    Couldn’t these cases be used as teaching moments?

  16. BlazingDragon says:


    I’m glad to hear GPs aren’t handing out steroid dose packs like candy any more. Avascular necrosis is a nasty thing. I would think even a 0.01% side effect, when pushed to a LOT of patients by GPs could result in a lot of damage (avascular necrosis often leads to total hip replacement).

    I guess I’m the flip side of the coin as far as antibiotics are concerned… I always seem to have sinus/throat infections turn into bacterial infections of the lungs without them. My track record (since childhood) is quite awful on this score. It doesn’t matter if I currently have a virus when I go to a doctor… within 1-3 days, I’ll have some nasty bacterial secondary infection and once it has a foothold, it can take 2-3 weeks to go away. I was once sick for a total of two months because a doctor refused to give me a prescription for antibiotics “because you only have a virus” (then the student health clinic shut down and I couldn’t get any antibiotics when I needed them two days later. By the time I got them, it was way too late).

    HCN: Newer steroid nasal inhalers use water-based suspensions/colloids and are MUCH easier on the nasal tissues (the solvent-based ones give me nosebleeds).

    As a general comment, steroids are powerful, immediate immunosupressives and anti-inflammatories. One of the best reasons to use them is that they are nearly instantaneous (when compared to other stuff like methotrexate, etc.). I have no doubt that steroid nasal sprays begin to suppress your immune system within the first one to two doses.

    This issue of antibiotic use for sinusitis is a bit sensitive for me because I had to get antibiotics for a sinus infection 3 weeks ago… Ten days of antibiotics cleared it up… but after being off them for about a week, it was back, with a vengeance. I just don’t seem to do well in clearing bacteria (and I never have). Ugh.

  17. HCN says:

    BlazingDragon said “Newer steroid nasal inhalers use water-based suspensions/colloids and are MUCH easier on the nasal tissues (the solvent-based ones give me nosebleeds). ”

    So, I’ve been told. Still, it is nicer that I’ve managed to control my allergies now with avoiding some allergens, and only go through one bottle of prescription allergy meds per year. Though it may have to do with age, and my immune system deciding to not freak out over every bit of pollen.

    Though I do use cortisone (sp?) cream on my hands when I react to nickel.

  18. BlazingDragon says:

    They do seem to work as advertised HCN, but I’m still not comfortable with putting immunosupressives in my nose long-term like that, so I don’t use them.

    My allergies are not well-controlled, but I’m only allergic to fungi (not cats, trees, dust mites, house dust, etc), so I’ve been told there is no data on whether allergy shots will work for me or not (apparently being only allergic to fungi is rare enough it’s hard to find enough people to do an RCT). I’m also kinda leery about allergy shots because some simple skin testing for various fungi kicked my asthma into overdrive (in addition to leaving me with near golf-ball sized lumps for several weeks for 6/21 fungi). I’m also put completely out by Allegra (much worse than benadryl for me). I guess I’m just a physicians nightmare … :)

    I have a friend who has a nickel allergy.. it sounds pretty nasty and makes it hard to find watches, among other things. Topical steroids, when used occasionally, seem to be pretty safe.

  19. Harriet, your article provides exactly the kind of information that the public wants and needs. I strongly urge you to start a webpage and post this and similar articles there since so many consumers turn first to the Internet for medical information and as others have pointed out are more apt to believe what they find there than to believe flesh and blood MDs.

    I also suggest that in order to draw in the greatest number of rational, intelligent consumers, the ones you have the best chance of helping, that you present the site as being that of one or several experienced doctors and in no way associate it with Skeptics or Quackbusters simply because many consumers have already been convinced by the Great American Alt Marketing Machine that such groups are terribly biased and unreliable.

  20. Harriet Hall says:


    Thanks for the vote of confidence. Start a webpage? We have already started a blog that is authored by 5 experienced MDs and that is in no way associated with any skeptical group. Its name says it all: science-based medicine. We are hoping that this blog will become the place to go for unbiased medical information.

  21. Chloe says:

    I would like to point out that after about ten years of on and off sinus pain, facial pain and tooth pain, all apparently minor, I turned out to have a rare sarcoma of the skull base.

    Sometimes it’s a zebra.

Comments are closed.