The cystitis antibiotic that worked last year

A woman in her forties comes in with burning, frequency and urgency for the third time in eighteen months. She has had the same three-day antibiotic each time, from a different clinic each time. The first course cleared it in a day. The second took most of a week. This time she has finished the course and nothing has changed, and she wants to know whether she needs a stronger antibiotic.

Stronger is the wrong axis. The question is whether it is the right drug for the organism she actually has, and that is a question nobody has yet asked, because nobody has cultured her urine.

Why Korea writes its own guideline#

There is no shortage of international guidance on urinary tract infection, and it would be cheaper to adopt one. Korea does not, and the reason is given in the first lines of the Korean guideline: clinical characteristics, causative organisms and antimicrobial susceptibility patterns may differ from country to country. An antibiotic that is a reasonable empirical first choice where local resistance to it is five percent is an unreasonable one where it is thirty.

The current Korean guideline on the antibiotic treatment of community-acquired urinary tract infection, published in Infection and Chemotherapy, is an update of the 2011 Korean guideline. It covers five situations that get treated as though they were one: asymptomatic bacteriuria, acute uncomplicated cystitis, acute uncomplicated pyelonephritis, complicated pyelonephritis related to urinary tract obstruction, and acute bacterial prostatitis. It is written for community-acquired infection in adults, and it explicitly leaves out healthcare-associated infection, catheter-associated infection and infection in immunocompromised patients.

What the culture is for#

A first episode of uncomplicated cystitis in an otherwise well woman is usually treated empirically, and that is reasonable. The organism is Escherichia coli in the great majority of cases and the empirical choice is made from local susceptibility data rather than from the individual patient.

A third episode in eighteen months is a different situation, and a course that has failed is a different situation again. At that point the useful information is which organism and what it is susceptible to, and the only way to get it is a culture taken before the next antibiotic starts. Extending a failing course does not generate that information. Escalating to a broader agent without it sometimes works and teaches nothing, and the next episode arrives with the same question unanswered and a narrower set of options.

Shorter, not longer#

The instinct when a short course fails is to give a longer one. The evidence on duration points the other way, although it answers a slightly different question — not what to do after a failure, but whether length is where the benefit lives at all. A network meta-analysis published in The Lancet Infectious Diseases pooled 61 randomised trials of antibiotic treatment for acute uncomplicated cystitis in adult women, covering 20,780 patients, and asked whether the regimen lengths in current guidelines could be shortened. For the third and fourth generation fluoroquinolones, a three-day regimen had a similar effect to a single dose on clinical response.

Duration is not the lever most patients think it is. If the organism is susceptible, a short course works. If it is not, no duration rescues it. Length is the wrong dial to turn when the drug is wrong.

And when it is a man#

A urinary tract infection in a man with an enlarged prostate is not uncomplicated, and treating it as though it were is how it comes back. The obstruction and the residual urine are part of the problem, and an antibiotic that clears the organism without the obstruction being addressed is a treatment for this episode rather than for the condition. That is why the infection and the prostate get assessed together rather than in sequence by two different clinicians.

What I would want a patient on her third course to take away is narrow and practical. Ask whether a culture has been sent. Ask what the empirical choice was based on. And treat a failed course as information rather than as a reason to ask for something stronger.