The Failure of 'Empirical Antibiotic Therapy' For Urinary Tract Infections

At a recent conference (pdf), Dr. Rebecca Roberts described how medical residents and doctors treat urinary tract infections. But first, a bit about 'empirical therapy.'

When a patient is sick it can take anywhere from 24-72 hours at a hospital with good facilities to identify what organism is causing the disease and which antibiotics will be effective against it. Empirical therapy uses the patterns of resistance (and infections) of previous cases at the hospital in conjunction with other guidelines to choose the most effective antibiotic therapy. Think of it as a highly educated guess--and it is a guess. For example, if you knew that ciprofloxacin only killed half of all urinary tract infections, you might choose another antibiotic.* That's the theory; let's see how the reality looks.

Here are the percentages of antibiotics prescribed at Roberts' hospital:

Quinolones (e.g., ciprofloxacin) 361 (55%)
Beta-Lactams 32 (5%)
Trimethiprim/Sulfamethaxole** 214 (33%)
Nitrofurantoin 6 (1%)

Now here's what urinary tract infections (E. coli) are resistant to at that same hospital:

Resistant to Quinolones (e.g., ciprofloxacin) 19%
Resistant to Beta-lactams 4%
Resistant to Trimethiprim/Sulfamethaxole** 12%
Resistant to Nitrofurantoin 4%

These patterns are similar to studies and surveys from other hospitals--the numerical frequencies will change, but, for the most part, the rank order of resistance is similar. If something seems kind of funny, you're right: the drugs that are most prescribed will be the least effective. So much for empirical therapy.

Now, this does support that whole natural selection thingee, but one would like to think that we're using the most effective antibiotics. After all, all the bacteria in the patient are being bombarded with the antibiotic, including the E. coli living in the intestinal tract which can be different from the E. coli infecting the urinary tract.

What's even more terrifying is that many of the residents had never heard of nitrofurantoin, the most effective antibiotic against urinary tract infections...

*There are, of course, other considerations. Antibiotics of 'last resort' that are used in critical cases (e.g., daptomycin) would probably be effective, but should not be used. However, nitrofurantoin is only used to treat urinary tract infections. It can have side effects, particularly in the elderly, and should be taken with food, but then again ciprofloxacin can also have side effects.

**also known as cotrimoxazole.