You don't need to be signed in to read BMJ Blogs, but you can register here to receive updates about other BMJ products and services via our site.

Primary Care Corner with Geoffrey Modest MD: Take the full course of antibiotics???

31 Jul, 17 | by

A recent BMJ analysis article argued that taking the “full course of antibiotics” is often likely counterproductive (see doi: 10.1136/bmj.j3418 )​.


–international health organizations and the WHO have pushed for completing antibiotic regimens: a 2016 WHO advisory to patients stated “always complete the full prescription, even if you feel better, because stopping treatment early promotes the growth of drug-resistant bacteria”. The CDC has a similar message

–the authors note that there is impressive evidence that some micro-organisms (eg TB, gonorrhea, HIV, S. typhi) can create spontaneous resistant mutations on treatment, and these mutants subsequently can be transmitted as resistant strains. [and there are good data supporting longer term therapies]

–but many of the organisms with growing resistance worldwide are normal commensal flora (E coli, enterococci, staph, klebsiella, acinetobacter, pseudomonas, enterobacter), which actually develop resistance by longer antibiotic exposure



–this article does bring up several important points

​–though some infectious agents can develop resistance with inadequate antimicrobial therapy (eg HIV etc), all 3 of the “critical” resistant species per the WHO and 3 of the 6 listed as “high priority” bugs fall into the group of commensal flora, where more antibiotics leads to more resistance (see blog )

–we really do have a dearth of studies looking at what the optimal course should be, both by infection site and by antibiotic used

— the article mentions the lack of efficacy of shorter courses of antibiotics for otitis media.  When this study came out, I was surprised that the shorter course did not work, in part because otitis media is one of those bacterial infections that often seems to resolve spontaneously without antibiotics, depending on the organism, and in part because there was a similar situation with a few other infections where shorter courses worked well (see below).  In fact, this otitis study was limited to kids <2yo and treated only with amoxicillin-clavulanate. Those on the longer course had a shorter duration of symptoms, but there was no difference in recurrence rates, and no difference in nasopharyngeal colonization with penicillin-nonsususceptible pathogens (N Engl J Med 2016; 375: 2446).

–In the past, uncomplicated UTIs in women were treated with 7-10 day courses of antibiotics. Then there were studies showing efficacy of single-dose therapy. Now, because of somewhat improved efficacy, the recommendations are usually for longer courses, such as TMP/SMX DS bid for 3 days, fluoroquinolones for 3 days, nitrofurantoin for 5 days, or (as a last holdout) fosfomycin as a single dose. However, a study which I think was pivotal in promoting the longer antibiotic course found that although single dose amoxacillin 3 grams was only 69% effective vs multi-dose regimens being 84% effective, the data on TMP/SMX was no different between these lengths of treatment, with a single dose of 2-3 tablets of DS leading to the insignificant difference of 87% cure rate vs 90% with multi-dose regimens (see Philbrick JT. Arch Intern Med 1985;145:1672-1678). It would be interesting to know if single-dose therapy is still as effective, given changes in flora and antibiotic sensitivity over these decades, but the single dose is certainly easier, likely to have less effect on the microbiome or create resistance in commensals, and apparently fewer adverse reactions (though the quality of the data here is questionable). Single-dose TMP/SMX may still be reasonable???

–and there are some trials showing that shorter courses of therapy for community-acquired  pneumonia do well, including one using fever as a guideline to stop antibiotics; some other studies in patients with mild to moderate clinical disease have found that clinical outcomes were similar with 3-7 day treatments vs 7-10 day ones (eg, see Dimopoulos G. Drugs. 2008; 68:1841).

–one concern here is that there is not much incentive nowadays to do the necessary trials. Drug companies, the dominant funder of clinical studies, have no interest: many of the commonly used antibiotics are generic, and even for those expensive ones that are only available as brand-names, documenting efficacy by taking fewer days of meds is not in the interest of their stockholders



–as a perspective here for the above study on “finish the complete course”:

–90% of antibiotics go to animals, largely to improve their growth and not to treat infections, and this practice seriously augments the incidence of antibiotic-resistant bacteria (see WHO report of the scary increase in antibiotic resistance, targeting the 12 bacterial families with the greatest threat to humans; and on the emergence of untreatable superbugs such as colistin-resistant E coli

​–studies suggest that >70% of antibiotics prescribed in humans are for inappropriate indications (eg, acute bronchitis, non-strep pharyngitis, acute rhinosinusitis, upper respiratory infections), and 85% of antibiotic prescriptions are written in primary care.

–one option to consider is holding off on antibiotics for infections that seem to do reasonably well left untreated, as long as there can be good followup. for example, in patients with uncomplicated skin abscesses, antibiotics were efficacious only for staph infections (2/3 of the infections), though even then the actual differences were not great (80+% do well without them): see blog  . I have been trying to decrease antibiotics in not-so-sick kids with otitis media, where for decades in Europe clinicians have been reluctant to use antibiotics, and the suggestion here was that it seemed reasonable to follow patients closely off antibiotics. Another suggested solution, the one I have been using, is to offer a prescription but suggest to the parents to hold off on giving the antibiotics unless the child does not improve in a few days or if the kid’s symptoms are getting worse, pointing out that antibiotics can unusually have very severe adverse reactions and their use should be minimized. (though I’m not sure exactly what parents are doing.) And, parenthetically, one of my kids had otitis at a young age, i held off on antibiotics, then he developed a generalized rash (all fitting with a viral etiology). but if i had given him amoxacillin, he would likely have been labeled as penicillin-allergic…

So, to me the overwhelming issue is the necessity to dramatically decrease antibiotic use in livestock and in humans for inappropriate indications (and there has been some movement in both of these areas, but needs to be increased lots), but that we really do need to have good studies looking at the shortest courses which work. And the old addendum “and make sure you complete the full course of antibiotics” is likely wrong in many if not most cases. In my experience, many patients are ahead of the curve on this one: they often just stop the antibiotics when they are feeling better, and it seems that they often do well……

other relevant past blogs:​

—  current disturbing increase in untreatable gonorrhea;

blog noting that women who had taken antibiotics for more than 2 months between the ages of 40-59 had a 69% increased incidence of colorerctal adenomas

By submitting your comment you agree to adhere to these terms and conditions
You can follow any responses to this entry through the RSS 2.0 feed.
EBM blog homepage

Evidence-Based Medicine blog

Analysis and discussion of developments in Evidence-Based Medicine Visit site

Creative Comms logo

Latest from Evidence-Based Medicine

Latest from EBM