Toward A Simple Diagnostic Index for Acute Uncomplicated Urinary Tract Infections

[Our journal club] unanimously agreed the topic is interesting, relevant, and unique. While the paper reports 60% lifetime prevalence of UTIs in women, students voiced potentially high numbers of unreported cases of UTI due to lack of access to medical care or opting for self management for this common disease. Only one student did not encounter a patient with a possible UTI diagnosis, while the rest had multiple encounters making this topic particularly relevant. Although UTIs is a well researched topic, this study is uniquely designed to do what has not been accomplished previously. Examining the combined effectiveness of different diagnostic tools is an unexplored territory because earlier studies evaluated efficacy only in isolation looking at only one test and it's effect on outcome.

The aim of this study was three fold: 1) to determine the value of diagnostic tests in predicting UTIs; 2) determine the effectiveness of diagnostic tool independently and when used in combination with one another; and 3) establish new guidelines for diagnosing UTIs. The implications of the findings promise not only to advance clinical knowledge, but also decrease costs and time for more efficient patient management.

Reflecting on our own experiences in working up a possible UTI case, the basic diagnostic tools we use are history taking and laboratory tests. Many red flags such as burning with urination, frequency, urgency, abdominal pain, and lack of vaginal discharge can be identified during history taking. The laboratory tests commonly ordered are urine analysis (UA) to confirm UTI, and urine culture to rule out false negatives in the presence of a strong history, but negative UA. Without experience with other medical systems, students were not familiar with the use of urine sediment and dipslide tests.

With sophisticated methods and statistical analysis, it is evident the researchers had a focused statistical approach. After exclusion criteria, this cross sectional study involved 196 patients seen in 20 general practitioners in. Students agreed that the researchers' decision to exclude UTIs in males and girls under 12 years of age is appropriate to ensure clear correlations between variables. UTIs in males and children are more likely to be complicated with anatomical defects or history of sexual abuse thus not fitting with the goal to examine acute uncomplicated cases. Further ensuring only uncomplicated UTIs were represented, symptoms lasting longer than 7 days were excluded since prolonged infection can result in plyonephritis. The exclusionary criteria further appropriately eliminated pregnant and lactating women since UTIs are more likely to occur in these populations and some symptoms of UTIs such as frequency of urination is considered normal in pregnancy. Recent use of antibiotics was also a disqualifier because it can affect the urine analysis. However, one interesting subpopulation, diabetics, was not excluded from the study. Students discussed pathophysiology of diabetes which results in increased risk of UTIs and frequency of urination, perhaps making this another appropriate disqualifier for the study. Especially considering that chronic conditions often result in more complicated infections with unusual flora requiring different diagnostic criteria and treatments. Similarly, some women who reported history of more than 3 UTIs raises high concern for complicated infection or anatomical derangement that would warrant advanced testing such as ultrasound or VCUG and thus could be excluded as well.

The assessment involved a questionnaire, dipstick, dipslide, and sediment. Analysis used regression to predict how variables individually affected rates of UTI identification. Variables without a strong correlation to UTI risk were eliminated. Using only the variables with the strongest predictive value, combinations of variables were examined.

Results of the study after using data from 196 women showed 61% positive cultures overall, which is consistent with literature values for lifetime prevalence of UTIs in women. Some counterintuitive findings were discovered. For example, more sexually active women had fewer numbers of positive cultures (50); whereas, less sexual activity was associated with a higher number of positive cultures (83). Explanations for this finding were few since it's against medical teaching that sexual activity increases the risk of UTI due to bacteria entering the urethra during intercourse. Another interesting finding was that women who are experiencing symptoms of UTI for the first time can still accurately self diagnose and were more likely to have a positive culture. The discussion pointed out that perhaps women with frequent UTIs were more familiar symptoms and treatment, thus more likely to self manage the condition, unreported. Compared to the significant alarm and anxiety associated a sudden change in body functioning not previously encountered, which would prompt immediate and thorough investigation. There seems to also be overlap between some categories of women such as those who had a UTI diagnosed by a physician and those who believed they had by self-report.

The study found three characteristics to be particularly predictive in the history of the patient with possible UTI: patient belief, presence of pain with urination, and presence of vaginal irritation. By asking these 3 questions, clinicians had a 50% chance of accurately diagnosis a UTI. This reaffirms the importance of a good history and reminds physicians that there is considerable value in subjective data. The accuracy of these three symptoms in predicting risk of UTI identifies the potential for effective public education and empowering women to watch for signs of possible UTIs.

Combining the history with a dipstick test increased the ability to predict UTIs to 73%. Dipstick tests are fast and inexpensive tests easily performed in office. Adding the more expensive and time-consuming sediment or dipslide test did not change the predictive value. Thus, the simple two-step process of history and dipstick was the strongest in predictive value. Without doing a urine culture, physicians can be confident in their diagnosis. False- negatives do occur where a strongly positive history and false dipstick is followed by a positive culture, but the since most cases of cystitis do not evolve into pyelonephritis and will self resolve, the implications of failure to routinely order cultures in uncomplicated suspected UTIs are minimal.

Every study has limitations and areas for improvement. First, considering that this is a study on uncomplicated UTIs, a definition of uncomplicated UTI should be made explicit. Second, a brief description of sediment and dipslide tests would have helped international readers unfamiliar with these diagnostic tools. Third, readers would have better understanding regarding significance of diagnostic tools if p values were given. Fourth, readers felt the figures were not intuitive to readers. Fifth, a larger sample size could have been obtained to increase generalizability for a study on this common disease. Lastly, there was a lack of cost-benefit ratios to thoroughly examine the possibility of decreased costs by utilizing only history and dipstick. Furthermore, its' impact on time spent in diagnosing and treating UTIs could have been further investigated in discussion.

In conclusion, there were unanimously favorable opinions about this study examining uncomplicated UTIs. A strength of the study included the emphasis on history empowers patients and validates their voice in medicine. This study found predictive symptoms that can become educational topics to help women correctly identify UTIs and when to seek treatment. In the end, the results confirm our approach used in our clinics placing importance with history taking and eliminating unnecessary tests.

Competing interests: None declared