One must complete the course of antibiotics as prescribed by physician, as written above one must not stop the course as minimal number of infectious microbe may reside and resist. It would survive and through clonal propagation would spread to environment and recurrence in host may also occur. This is the one of most important reason for spread of resistant and hard to treat infections.

The important thing is this Mantra can not applied to severe infections. Next question arise is "Do we able to make it sure that "clinical cure" occured relatively simultaneous with "pathogen clearance"? If we are not able to measure it, it means shorter periode is not a choice. 

Hi!

That’s a great question.

Using shorter courses of antibiotics is a very interesting perspective. The STOP.IT trial (RG Sawyer, NEJM ‎2015), concerning intraabdominal infectious, was quite impressive in that regard. But as has been stated above, it depends on the situation.

When treating an infection (and let’s assume that we are just talking about bacterial infections), one must take into account 3 main factors: the antibiotic; the immune status; the control of the infectious focus.

Sometimes, you don’t even need an antibiotic to treat the infection – if a patient has a focalized infections such as a gangrenous foot ulcer with osteomyelitis due to diabetes mellitus, and amputation is preformed, you can control the focus without further antibiotic being required.

In other situations, if you have an antibiotic that can easily reach the infectious focus in a patient with a preserved immune status, control of the focus is not necessary. That's what happens in most patients with a pneumonia (in which a lobectomy is obviously not performed…). In these patients, it is assumed that the immune system is initially overwhelmed by the infectious agent, which leads to a harmful and unadjusted inflammatory response. Theoretically, if you use antibiotics for some time to reduce the bacterial load, you could stop it earlier and allow the immune system to do the rest. But the question remains – when can you stop it? Can we trust the clinical symptoms alone as surrogates? Should we also use some lab analysis? And what would be the cut-offs used? I don’t think that these questions have been clearly elucidated.

There are other infections, such as meningitis, in which focal inflammatory response is impaired (eg, no complement activity in the CSF) and the clearance mainly relies on the antibiotic. So, a short course of antibiotic would probably be insufficient.

There are also really difficult to treat infections (such as brain abscesses or spondylodiscitis) in which short courses of antibiotics are very likely to fail, because both the antibiotic penetration and immune responses are impaired at the focus of infection. In these situations, you frequently need both focus control and prolonged antibiotic therapy to have a favourable outcome.

And finally, you have the case of slow-growing agents that require long courses of treatment (because the antimicrobials can’t easily eliminate quiescent organisms), such as M. tuberculosis, Coxiella burnetii or Brucella spp.

In these last two situations (difficult to treat infections and slow-growing microorganisms) the patient may start to feel better before you have a cure, so in these cases symptoms can hardly be used as surrogates.

All in all, I think that shorter courses of antibiotics could become the next “hot” recommendation in antibiotic stewardship programs; I guess that it should be considered mainly in infections in which the antibiotic has a good penetration and the immune system is preserved. But frankly, I wouldn’t change my practice until solid recommendations are made.

Best regards.