We can eliminate that last cause first. The duration of antibiotic therapy is usually determined by educated guess, only occasionally by experiment, and almost never with a view to minimizing resistance.
The “finish your antibiotics or you’ll encourage resistance” paradigm is based on a faulty premise – that antibiotic resistance develops stepwise during the course of treatment. Instead, there is usually a small pre-existing population of fully-resistant bacteria[1] . Prolonged courses of antibiotics actually give these bugs a selective advantage[2], and are more likely to create resistance. Shorter-duration therapies are being encouraged with the specific goal of reducing resistance[3] .
There are exceptions, of course. Resistance in TB and in some Gram-negative non-fermenters (Pseudomonas, Acinetobacter) can arise step-wise during the course of treatment, and failure to complete therapy can contribute to resistance development. But combination therapy, not longer duration, is the most effective method of suppressing resistance.
It may not be possible to determine whether over-prescription or agricultural use of antibiotics contributes more to resistance. To do so empirically, we’d have to identify a representative cohort of resistant infections, and trace the history and origins of resistance in each individual case. I know of only one limited instance in which this has been done[4] .
But from a public health perspective, there is no need to choose one source or prioritize its elimination over the other. Both livestock use and overprescription contribute to the emergence and spread of resistance.
Turning the tide of resistance is not easy or simple, but it can be done. MRSA rates have dropped by half or more in the UK and France since 2000. Smaller, but significant drops in MRSA rates are also beginning to show up in the US.
As with the emergence of resistance, its decline can not be attributed to any single factor or intervention. Given the diversity and ubiquity of resistance mechanisms, it’s likely that many interventions are required – particularly better stewardship, hygiene, and reduced antibiotic use in food animals[5] .
Unfortunately none of these interventions are money-makers. Everyone benefits but no one profits. Until the US healthcare system becomes far more socialized, I expect us to lag far behind other countries in effective control of antibiotic resistance. Infectious diseases are more a political than a medical problem.
Footnotes
[1] The human microbiome harbors a diverse reservoir of antibiotic resistance genes.
[2] Suppression of Emergence of Resistance in Pathogenic Bacteria: Keeping Our Powder Dry, Part 1
[3] The New Antibiotic Mantra—“Shorter Is Better”
[4] Intermingled Klebsiella pneumoniae Populations Between Retail Meats and Human Urinary Tract Infections.