Diagnostic value of dipstick test in adult symptomatic urinary tract infections: results of a cross-sectional Tunisian study

Urinary tract infections (UTIs) are common. The diagnosis is confirmed by urine culture which is costly and takes at least 24 hours before results are known. The aim of this study was to determine the diagnostic accuracy of dipstick test for the diagnosis of UTI in symptomatic adult patients. We conducted a cross-sectional study in the department of Infectious Diseases, Sousse-Tunisia during a two-year period. We included all patients with clinical signs of UTI. Urine samples were tested for the presence of leukocyte esterase (LE) and nitrites. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of LE and nitrites were calculated against urine culture as gold standard. Four hundred thirty one patients, 139 men (32%) and 292 women (68%) were included. One hundred sixty six patients (39%) had UTI. The most frequently isolated microorganism was Escherichia coli (75%). LE had a high sensitivity (87%) but a low specificity (64%), while nitrites had a high specificity (95%) but a low sensitivity (48%). Combined positive LE and nitrites had a high PPV (85%) and combined negative LE and nitrites had a high NPV (92%), while positive LE combined with negative nitrites had a low PPV (47%) and a low NPV (53%). In conclusion, in adult patients with UTI symptoms, an alternate diagnosis should be considered if the LE is negative, while an early empirical antibiotic therapy against Enterobacteriaceae should be started if the nitrites are positive.


Introduction
Urinary tract infections (UTIs) are common. They occur in 60% of women at least once in their lifetime and account for 23% of hospitalacquired infections [1,2]. Clinical signs of UTIs lack of sensitivity and specificity and the definitive diagnosis is based on the isolation of a microorganism by urine culture. Since urine culture is costly and takes at least 24 hours before results are known, the use of simpler and faster diagnostic methods such as dipstick test should be useful to guide the initial diagnosis especially in an emergency context. In 2015, the Société de Pathologie Infectieuse de Langue Française recommended dipstick test as a diagnostic method in UTIs for its high negative predictive value (NPV) in women and its high positive predictive value (PPV) in men. Moreover, an English study has demonstrated that the use of urinary dipstick reduced the urine laboratory workload [3]. The aim of this study was to determine the diagnostic accuracy of dipstick test for the diagnosis of UTI in symptomatic adult patients [4].

Methods
Study design and population: we conducted a cross-sectional study in the department of Infectious Diseases in Farhat Hached hospital, Sousse-Central Tunisia, during a two-year period. We included all hospitalized patients and outpatients, aged ≥ 18 years, suspected of having a UTI based on the presence of one or more among the following clinical signs: urinary frequency or urgency; dysuria, urinary retention lumbar; flank, iliac fossa, pelvic, or perineal pain; hematuria, mental confusion and fever. If another diagnosis such as acute appendicitis, salpingitis or endometritis was made, the patient was excluded from the study. Ethical considerations: as our study didn't involve changes to the patients' usual medical management, no study protocol had been submitted to our hospital Ethics Committee approval. However, patients were provided with oral information and gave their verbal consent before being included in the study.
Thus, UTIs were more frequent in women than in men (p<0.001) ( Figure 1). The isolated microorganisms were Enterobacteriaceae in 154 cases (93%), Gram positive cocci in 11 cases (6.5%) and Pseudomonas aeruginosa in one case (0.5%). The most frequently isolated microorganisms were Escherichia coli (E. coli) (75%) and Klebsiella spp (10%) ( Table 1). Overall, the sensitivity of LE was 87% and the specificity was 64%, the sensitivity of nitrites was 48% and the specificity was 95%. The PPV and the NPV were 60% and 89% respectively for LE and 85% and 74% respectively for nitrites. LE was more specific in men (78%) compared to women (45%), and the NPV of nitrites was higher in men (85%) compared to women (68%) ( Table 2). Overall, LE positivity combined with nitrites positivity had a PPV of 85%, LE negativity combined with nitrites negativity had a NPV of 92%, LE positivity combined with nitrites negativity had a PPV of 47% and LE negativity combined with nitrites positivity, noted in only 1.6% of patients, had a PPV of 86% (Table 3).

Discussion
In the present study, LE had a high sensitivity (87%) and a high NPV (88%) but a low specificity especially in women (45%), while nitrites had a high specificity (95%) and a high PPV (85%) but a low sensitivity (48%). Combined positive LE and nitrites had a high PPV (85%) and combined negative LE and nitrites had a high NPV (92%), while positive LE combined with negative nitrites had a low PPV (47%) and a low NPV (53%). The diagnostic value of LE and of nitrites varied widely between the different studies. However, like in our study, a high NPV (87 to 93%) and a high specificity (94 to 99%) with low sensitivity (28 to 60%) were frequently noted for LE and nitrites respectively ( Table 4). The differences between these studies may be explained by the heterogeneity of the patients' inclusion criteria (men or/and women; symptomatic or asymptomatic UTIs, cut-off used to define a positive urine culture) and by the important variation in the interpretation of dipstick test between practitioners Negative nitrite test in patients with positive urine culture may be noted when a UTI is caused by a microorganism that does not contain nitrite reductase or when the urine has been in the bladder for insufficiently long period (less than four hours). Lack of dietary nitrate and dilution of nitrite in urine may also explain false negative nitrite [11,12,16]. In our study, clinically suspected UTIs were confirmed by positive urine culture in only 38% of patients. They were more common in women (46%) than in men (24%), and the most frequently isolated microorganism was E. coli. In other studies, positive urine cultures in patients with clinically suspected UTIs varied considerably from 17 to 66%. This may be explained by the differences in inclusion criteria (women, men, or women and men; clinical signs) and in the cut off value (≥ 10 3 colony forming units (CFU)/mL or ≥ 10 5 CFU/mL) to diagnose a UTI in these studies [5][6][7][8][9].
Like in our study, E. coli was the most frequently (48 to 73.5%) isolated microorganism in UTIs in different studies [7][8][9][10][11].  Nitrite test has a positive predictive value of 85% and a sensitivity of 48% for the diagnosis of UTIs.

Competing interests
The authors declare no competing interests.

Authors' contributions
Foued Bellazreg designed the study, collected and analyzed data and wrote the article, Maha Abid analyzed data, Nadia Ben Lasfar and Zouhour Hattab collected data, Wissem Hachfi and Amel Letaief designed the study and reviewed the article. All authors have read and agreed to the final manuscript. Table 1: microorganisms isolated from urinary tract infections in our study Table 2: diagnostic value of LE and nitrites in our study