Acute urinary tract infection (UTI) is one of the most common acute bacterial infections among adult females.1,2
Empiric antibiotic treatment has been advocated as cost-effective,3
but unselective antibiotic use will result in a growing problem of antibiotic resistance, which has been identified in 20% of laboratory specimens.4,5
There are current proposals to make courses of trimethoprim available over the counter. This potential overuse of antibiotics for UTI creates an urgent need to address the question: can we use history and physical examination, or near-patient tests, for better diagnosis and targeting of antibiotics?
A systematic review of the role of symptoms in diagnosis identified few high-grade studies (those with ≥50 consecutive patients, and independent blind comparison of symptoms and signs with a gold standard among patients with suspected UTI), none of which were based in primary care.6
Only one study, which had poor methodology, assessed the predictive value of combining symptoms; and all previous studies used a very insensitive cut-off for laboratory diagnosis, whereas recent laboratory diagnosis standards suggest colony counts down to 103
cfu/ml (colony-forming units per millilitre) are classified as UTI.7
How this fits in
This is one of the few adequately powered studies in primary care to confirm which clinical variables and range of dipstick variables independently predict rigorous laboratory diagnosis of urinary tract infection (UTI), and to validate clinical decision rules based on these independent predictors. Among female patients presenting with suspected uncomplicated UTI in primary care, those with dysuria, cloudy urine, and nocturia are very likely to have UTI, but even when all these features are absent, 33% of patients presenting with suspected uncomplicated UTI have UTI. A clinical decision rule based on either nitrite or both leucocytes and blood could also be used to target investigations or treatment, but the negative predictive values are poor: even when all these dipstick results are negative, 24% of female patients still have UTI.
Clinicians commonly use dipsticks to rule out infection, and they are the most widely used simple near-patient test in primary care.8–11
Although summary data are available for studies that assessed nitrite and leucocyte esterase separately, primary data are needed to assess the independent predictive value of all dipstick results.12
The evidence base for dipstick use in primary care is poor, due to the paucity of studies and ‘spectrum bias’.6,11,13
Studies from primary care have a range of one or more limitations:9,11,14–18
they have either not assessed the independent value of dipstick results and symptoms (hence potentially over-complicating clinical decision rules); and/or not used the range of dipstick variables (most include nitrite and leucocyte but not blood); and/or failed to develop and then test algorithms in separate samples (McIsaac et al
being the exception16
); and/or had low power. Only the most recent dipstick studies have used the recent more rigorous laboratory guidelines for diagnosis.16,18
The current authors have reported a study where a clinical score and a dipstick score were developed for women presenting with suspected UTI.19
The independent predictive values of symptoms and of dipsticks results were assessed. Based on accumulating evidence and recent international consensus, the latter study used more sensitive laboratory gold standards to include lower colony counts.19
However, the predictive value of any scoring system that is tested in the same sample used to calculate the scoring system is likely to have artificially inflated predictive values: both a training and a validation set are needed. To estimate the more realistic predictive values of these scores, this study assessed the predictive value of the scores and the component variables of the scores in a new validation sample.