Empirical treatment of uncomplicated cystitis

Acute uncomplicated cystitis in healthy non-pregnant women is a common problem, affecting 10-15% of women each year.¹ It is readily managed by treatment for up to three days, and courses of three days or less seldom give rise to the development of bacterial resistance in the gut.²,³

Uncomplicated cystitis accounts for a substantial part of all prescribed antimicrobials. Results of standard treatment in general practice have until recently been satisfying: cure rates for most antibiotics have varied between 85% and 95%.⁴ Rising levels of resistance cause concern, because the increase may lead to the liberal use of newer antimicrobials with broader spectrums, resulting in increased resistance to these drugs too. Local variation in levels of community based prescribing may be reflected in changes in local levels of antibiotic resistance in uropathogens.⁵

A recent national survey has described susceptibility among 103 223 isolates of uropathogens from microbiological laboratories in nine geographical regions of the United States.⁶ The resistance of Escherichia coli to co-trimoxazole ranged from 22% in the western United States to 10% in the northeast. The corresponding figures were 7% to 1% for nitrofurantoin and 1% to 0% for fluoroquinolones. A parallel review article presents similar ranges of resistance in some north European countries.⁷

Should these results lead to changes in the present policy for empirical treatment of uncomplicated cystitis in women? Not necessarily, because data on resistance from microbiological laboratories should not be interpreted as representative of actual levels of resistance in general practice. Laboratory data are based on selected urinary samples from general practice, mostly from patients with suspected complicating factors. These patients are only a small fraction of all patients with cystitis, and they often harbour more resistant bacteria.⁸ A recent Norwegian study supports this: antimicrobial resistance found in a microbiological laboratory in urine samples from general practice was substantially higher than the actual resistance found in consecutive urine samples from women with uncomplicated cystitis in general practice.⁹ The different levels in the samples from women with uncomplicated cystitis and in the laboratory samples were 13% and 18%, respectively, for trimethoprim, 1% and 5% for mecillinam, and 3% and 10% for nitrofurantoin. These differences are fairly representative of other settings.¹⁰

Further, the antimicrobials usually prescribed for cystitis achieve far higher concentrations in the urine than the cut off points used in vitro for labelling of bacteria as susceptible or not.⁷ Consequently antimicrobials will in practice eradicate uropathogens from the urine in a fair number of cases where the bacteria were labelled as resistant. A practical cut-off point has been proposed for co-trimoxazole. An alternative agent should be prescribed when its resistance level, as reported by local microbiological laboratories, is above 20%.⁷ Similar calculations could be made for trimethoprim alone, as its antibacterial activity is not too different from that of co-trimoxazole.⁹

Which antimicrobials should then be used? As uncomplicated cystitis is so common, antimicrobials that are currently seldom prescribed for other infections should be preferred. Also, because increased prescription of one antimicrobial may result in higher resistance locally, the doctor's choice of antimicrobial should vary from time to time and from one patient to another.⁵ The actual local bacterial resistance levels should also be considered.⁵,⁷

Nitrofurantoin is one of the oldest antimicrobials in the market, and curiously enough it still has rather less than 10% bacterial resistance in most geographical settings.⁷,⁹,¹⁰ Mecillinam is not in general use in all Western countries, and bacterial resistance is still surprisingly low, considering that it is a derivative of penicillin.⁹,¹⁰ Most coliforms are susceptible to it; Gram positive cocci are not. Still, mecillinam will in most settings be effective against about 90% of most uropathogens causing uncomplicated cystitis in women, and it is a fair treatment choice.

Trimethoprim is roughly as effective as co-trimoxazole but has fewer adverse effects.⁹ Its antibacterial spectrum favours its use in the treatment of urinary tract infections, and its high concentration in urine makes it effective against most uropathogens. Because of limited use in many countries, its resistance level is still favourable in many areas of western Europe, often well under 20%. Local variations occur, though.⁵

Quinolones are still among our most effective antimicrobials, though the high level of resistance found in several countries is alarming.⁷ Current levels of use in many countries are equally alarming. To preserve their effectiveness they should be reserved for serious or life threatening infections and not used for simple, self limiting conditions such as uncomplicated cystitis in women, as long as more basic antimicrobials are still effective.

Empirical treatment of uncomplicated acute cystitis in adult women should stay simple. The disease can still be managed using short courses of traditional antimicrobials such as nitrofurantoin, mecillinam, or trimethoprim, according to levels of resistance in the local area. Microbiological laboratories overestimate this bacterial resistance.