Pitfalls

Hospital formularies tend to overestimate the occurrence of antibiotic resistance in a community, as uncomplicated UTIs are often not cultured and complicated UTIs are more likely to be infected with resistant organisms [8]. This inherent selection bias often leads to clinicians avoiding standard antibiotics for uncomplicated UTIs.

‘Subclinical’ pyelonephritis has been described in patients who present with predominantly cystitis symptoms without characteristic pyelonephritis symptoms. It is often a diagnosis made in retrospect after a failure of initial cystitis treatment. Patients at higher risk of subclinical pyelonephritis include those with symptoms greater than 1 week, males, diabetic patients, pregnancy and if immunosuppression is present.

Although bacterial counts of >10⁵ are indicative of significant bacteriuria in uncomplicated pyelonephritis (80-95%), a significant proportion of patients may present with counts as low as 10² bacteria per ml in the setting of fever, loin pain and pyuria in catheterised patients.

Bacterial counts of >10⁵ are occasionally seen in asymptomatic patients, especially young women. Asymptomatic bacteriuria does not require treatment except in pregnant women.

Learning bite

Hospital formulary for antibiotic choice for UTI may have inherent selection bias.