The Diagnosis and Management of UTI in >65s: To Dipstick or Not? The Argument For Dipsticks

Sir Robert Hutchinson used to say that the ghosts of dead patients that haunt us do not ask why we did not employ the latest fad of clinical investigation; they ask “Why did you not test my urine?” [1] Examination of the urine is a fundamental part of the physical examination of the patient. The urine dipstick is quick, convenient, cheap and easy to perform. With minimal training, the presence or absence of significant disease can be indicated in a matter of seconds.

With any pathological test, however, the challenge concerns the interpretation of the results. This is especially the case with urine dipsticks, which should be viewed as a screening test to indicate whether further, more specific testing is advisable. Who would deny the utility value of urine dipsticks in the detection of glycosuria, proteinuria, haematuria etc.? No-one, however, would tell a patient they are diabetic and start them on treatment on the basis of a single dipstick indicating glycosuria. The absence of glycosuria, however, in a patient who is not on insulin or other anti-hyperglycaemia drugs, is a good indication that the patient does not have diabetes, which in itself is an extremely useful piece of information when assessing a patient who is unwell, or who requires surgery or other active management.

The issue with the over-diagnosis of urinary tract infections (UTIs) through the use of dipsticks is purely down to the interpretation of a positive result. Therefore anyone who is responsible for acting on the result of the test needs to understand its limitations, and what a positive and what a negative result actually means. It does not mean that dipsticks are useless or should be banned. A dipstick has a perfectly respectable negative predictive value for infection in patients of both sexes and all ages [2]. Indeed, up until the advent of automated cell counters in Microbiology laboratories, it was not unusual for dipsticks to be used as the sole screening tool in the laboratory to decide which urine specimens required microscopy and culture, and which could be reported as ‘No evidence of a urinary tract infection’ with no further work undertaken on them. Are we now saying that a large proportion of urine reports over many years were totally invalid?

Increasing antimicrobial resistance and the overuse of antibiotics are major issues. It is undeniable that the over-diagnosis of UTIs based on positive urine dipsticks is fuelling both of these problems, especially in the elderly population in which asymptomatic bacteruria is common. Banning the dipstick, even in this group of patients, is unlikely to make a significant impact on the problem. It could, in fact, make things worse. If an elderly patient develops an acute confusional state and is unable to detail whether they have any symptoms that could indicate a UTI, a negative dipstick will quickly rule this out as a likely cause and diagnostic efforts can be diverted to the other potential causes of this common condition. Without that negative dipstick result, how often will empirical antibiotics be prescribed in the acute situation? While there will be no evidence that it is the correct course of action, there will be none to show this is not the case. If the only alternative is to admit the patient to an already overwhelmed hospital, I suspect this will not be a rare occurrence in busy emergency departments or care homes being covered by over-stretched GPs. Even if it is decided that a urine sample should be sent to the Microbiology laboratory for culture, this is not going to produce any more evidence of infection than a positive dipstick would. The sample will grow bacteria if they are present in significant numbers, but is still not telling anyone whether this is indicative of infection or is merely a reflection of asymptomatic bacteruria, in exactly the same way that a positive dipstick result would have done (only several days sooner). What is critical is the education of the people who are interpreting the results of the tests, whether these are done at the point of care or in the laboratory.

England has a Commissioning for Quality and Innovation (CQUIN) target for 2019–20 which penalises dipstick use in the diagnosis of urinary tract infections in the over 65s. Unfortunately this is designed to reward poor practice (e.g. if a dipstick result is not recorded, this is counted as being a ‘pass’ in the same way as if one had not been done, or if the result has not influenced the diagnosis, so poor documentation is an asset), and totally ignores the significant number of patients who are not prescribed antibiotics on the basis of a negative dipstick. I am yet to work out, as I dredge through the case notes of randomly selected patients who have been coded as having had a UTI, how I am supposed to discern the mental machinations of whoever has come up with that diagnosis, and what influence the positive dipstick result had on that decision? Yet this is the crux of the CQUIN; if the positive dipstick influenced the diagnosis, then that is a fail. For the reasons I have outlined above, however, I refuse to instruct our Emergency Department to ban the use of dipsticks. If the patient has no signs or symptoms suggestive of a UTI and the only factor used is the positive dipstick, then that is wrong. The confused patient with a negative dipstick, who can give no history, and who is diagnosed (almost certainly incorrectly) with a UTI, is apparently a diagnostic success, though, as the result of the dipstick test was not taken into account in the diagnostic process!

I think we should keep the dipstick, but try to ensure it is used correctly and appropriately, and any result is only acted upon by people who have a full understanding of the limitations of the test. Surely this is the case with other investigations carried out in the practice of medicine, so why is this particular example so contentious?