Prevalence of and factors associated with atypical presentation in bacteremic urinary tract infection

A delay in the diagnosis of urinary tract infection (UTI) is not uncommon. Atypical presentation is often cited as one of the causes of diagnostic delays. However, few studies have investigated the prevalence of atypical presentation and determined factors associated with atypical presentation at initial contact among patients with UTI. Therefore, a retrospective and prospective cohort study using chart review was conducted in two acute care hospitals. We included 285 consecutive patients hospitalized for bacteremic UTI. The primary outcome was atypical presentation, defined as the absence of any urinary tract symptom or sign at initial contact. Of all patients, the median age was 82 years, 186 (65.3%) were women, and 53 (18.6%) had dementia. Urinary tract symptoms and signs were absent at initial contact in 144 patients (50.5%; 95% CI 44.7–56.4%). The multivariable analysis revealed that older age, male sex, dementia, and early visit from symptom onset were significantly associated with an increased risk of atypical presentation. Patients with atypical presentation were less likely to receive a correct diagnosis at initial contact than patients with urinary tract symptoms and signs (OR 0.30; 95% CI 0.17–0.51). Atypical presentation in patients with bacteremic UTI is common and negatively affects the correct diagnosis of UTI.

www.nature.com/scientificreports/ Among all the cases, 82 (28.8%) and 169 (59.3%) were cases of complicated UTI and pyelonephritis, respectively. The most common pathogen was Escherichia coli (n = 211, 74.0%), followed by Klebsiella species (n = 27, 9.5%). For the primary outcome, any urinary tract symptom or sign was absent at initial presentation in 144 (50.5%) patients (Table 2). For the secondary outcomes, the proportion of patients without fever and any urinary tract symptoms or signs was 11.2%, and the proportion of patients with a correct UTI diagnosis at initial presentation was 70.9%. Antibiotic therapy was started at initial contact for any reason in 223 (78.3%). The most common initial diagnosis at first contact was UTI or urosepsis (n = 202, 70.9%), followed by unspecified fever (n = 27, 9.5%) and pneumonia (n = 14, 4.9%).

Discussion
Our findings showed that half of the patients with bacteremic UTI had no urinary tract symptoms or signs at initial presentation. Independent predictive factors for the absence of urinary tract symptoms and signs were older age, male sex, dementia, and early visit from symptom onset. Moreover, the absence of urinary tract symptoms and signs at initial presentation was associated with an increased risk of a diagnostic delay of UTI and initiation of antimicrobial therapy.
Our results are consistent with those of past studies 6,10-12 reporting that more than half of patients with UTIs had no urinary tract symptoms. However, the proportion of correct diagnoses for bacteremic UTI at initial contact in the present study was 70.9%, while that of a past study conducted more than two decades was 43% 6 . This implies that the diagnostic accuracy for bacteremic UTI at initial contact has been improved in the past two decades. Nonetheless, a substantial proportion of bacteremic UTI patients were not diagnosed correctly at initial presentation. Therefore, some strategies to improve the accuracy of bacteremic UTI diagnosis at initial presentation are needed.
In the present study, advanced age and dementia were independent predictive factors for the absence of urinary tract symptoms and signs. This result supports past 6,8,10 and recent studies 9 reporting that UTI patients who were older were less likely to have urinary tract symptoms and signs. In addition, our research revealed that the absence of urinary tract symptoms and signs at initial presentation was associated with an incorrect initial diagnosis for UTI and delayed initiation of antimicrobial therapy. Given that elderly UTI patients are more likely to die than younger UTI patients 9 and that diagnostic and treatment delays may result in the poor prognosis of UTI patients 4,5 , some efforts to diagnose UTI correctly among elderly patients with dementia will be warranted.
Male sex was another independent predictive factor for the absence of urinary tract symptoms and signs in our research. Given that UTI has been more frequently studied in women than in men 9 , it is no surprise that the presenting features of men with UTI are atypical. It may be similar to ischemic heart disease, for which symptoms in women with angina are atypical because classical symptoms of angina are mainly based on male patients 13 . Infection/inflammation of the male accessory gland is difficult to diagnose based on just the clinical history and physical examination 14 . Therefore, it is possible that atypical presentation of bacteremic UTI is more frequent in men. Our findings indicate that it is difficult to differentiate UTI from other diseases in men based on only clinical symptoms and signs because urinary tract symptoms and signs are often absent. Moreover, given the high prevalence of asymptomatic bacteriuria among elderly men [15][16][17][18] , an early correct diagnosis of UTI may be difficult, even if urine laboratory tests are performed. Given that there are few studies investigating the clinical features of men with UTI 9 , further studies are warranted to investigate the clinical features of men with UTI and develop strategies to diagnose UTI in men.

Strength and weakness.
This study was the first to investigate risk factors for atypical presentation and the effect of atypical presentation on a diagnostic delay in bacteremic UTI patients. To avoid the incorrect inclusion of infectious and noninfectious diseases other than UTI, we used a strict definition of UTI as a reference standard. Nonetheless, several limitations should be mentioned. First, the data collected in a real-world practice were used by using electronic medical records. Therefore, the information analyzed in the present study might not be accurate. Moreover, the inclusion of only two hospitals in the same area limits the generalizability of our findings. Second, lower abdominal tenderness and costovertebral angle tenderness were not documented or evaluated in some patients. Therefore, the presence of these physical signs was underestimated in the present study. However, given that physicians who suspect UTI generally perform these examinations, it is possible that no documentation reflects atypical presentation in UTI patients. Third, there might have been a substantial loss of targeted UTI patients due to the lack of submission of blood or urine cultures by principal physicians. Fourth, it is uncertain that our results can be generalized to UTIs without bacteremia, which is more common than UTIs with bacteremia. Fifth, acute prostatitis might be incorrectly included in the present study because tenderness of the prostatic gland was not evaluated in all cases. Sixth, overemphasis on avoidance of a delayed diagnosis for bacteremic UTI might increase the overuse of tests and overdiagnosis of UTI, which might result in the overuse Table 3. Results of the multivariable analysis a for factors associated with no urinary symptoms or signs at presentation among patients with bacteremic urinary tract infections. a Variables were removed one-by-one using a backward stepwise method until all remaining variables had a p value of less than 0.2. The following variables were considered: age, sex, hypertension, dementia, stroke, ischemic heart disease, benign prostatic hypertrophy, diabetes, and time to initial visit from symptom onset. The level of statistical significance was set at 5% CI, confidence interval.

Variables
Odds ratios (95% CI) P value www.nature.com/scientificreports/ of antimicrobial drugs 11,19 . Finally, the majority of patients included in the present study were more than 75 years old. Therefore, this limits the generalizability of our results and may cause bias because some of independent variables used in the multivariable regression analysis were related to age. Nonetheless, past studies have also reported that most patients with bacteremic UTI were elderly 4,10 .

Methods
Study setting and design. A retrospective and prospective multicenter cohort study was conducted by using medical electronic records to determine the prevalence of and risk factors associated with atypical presentation in UTI patients. Inclusion and exclusion criteria. All consecutive patients who were hospitalized due to UTI in Tochigi Medical Center (from September 2014 to March 2021) and Saiseikai Utsunomiya Hospital (from January 2019 to December 2020) were included. We included only UTI patients who had bacteremia and satisfied the definition of UTI. An accurate diagnosis of UTI is difficult in clinical practice. Moreover, the clinical distinction between asymptomatic bacteriuria and UTI is often difficult 15 . Therefore, to avoid the inclusion of infectious diseases other than UTI and noninfectious diseases, we included UTI patients with the same bacterial pathogen isolated from both urine and blood. Patients hospitalized due to renal abscess, proctitis, and cystitis were excluded. Based on past studies 20,21 and guidelines 22 , bacteremic UTI was defined if all of the following were satisfied: (1) the patient experienced fever or chills, any symptoms of cystitis (dysuria, urgency, frequency, suprapubic pain, or lower abdominal tenderness on physical examination), or any symptoms of pyelonephritis (flank pain, back pain, or costovertebral angle tenderness on physical examination) during the index episode; (2) the same bacterial pathogen was isolated in concurrent urinary and blood cultures, and there were no other sources of infection; and (3) the bacterial pathogen was present at ≥ 10 4 CFU/ml of urine. UTI with any indwelling catheter, neurogenic bladder, obstructive uropathy, or urinary retention due to benign prostatic hypertrophy was defined as a complicated bacteremic UTI. During the study period, a total of 285 patients who met the inclusion criteria were included in the final analysis.
Data collection and outcome measures. Physicians reviewed electronic medical records and retrieved information on patient age, sex, past medical history, medication use, time to initial hospital visit from symptom onset, symptoms, physical findings, initial diagnosis at the first visit, and prognosis. Chronic symptoms were not collected.
The primary outcome was the proportion of patients with atypical presentation at initial contact. Atypical presentation was defined as the absence of any new urinary tract symptoms and signs. Urinary tract symptoms and signs were defined as dysuria, urgency, frequency, suprapubic pain, flank pain, back pain, or lower abdominal tenderness or costovertebral angle tenderness on physical examination. The secondary outcome was the proportion of patients with the absence of new urinary tract symptoms and signs and no fever at initial contact. Fever was defined as a body temperature of 38.0 °C of greater at any site. Other secondary outcomes were the proportion of patients who were diagnosed correctly as having UTI at the initial visit and the proportion of patients who were prescribed any antimicrobial drugs at the initial visit.
Statistical analysis. Descriptive statistics were used to report the baseline characteristics of the study population. To determine the predictive factors associated with atypical presentation at initial contact, multivariable analysis with binary logistic regression was performed. The associations between the primary outcome and selected variables were examined. The following variables were considered: age, sex, hypertension, dementia, stroke, ischemic heart disease, benign prostatic hypertrophy, diabetes, and time to initial visit from symptom onset. The variables were removed one-by-one in the backward stepwise method until all remaining variables had a p value of less than 0.2. To investigate the hypothesis that atypical presentation can result in diagnostic and treatment delays for UTI, we also investigated the association between the absence of urinary tract symptoms and signs and a correct diagnosis of UTI and initiation of antimicrobial therapy at initial presentation. The level of statistical significance was set at 5%. These analyses were performed by using Stata version 15 (LightStone, Tokyo, Japan).

Conclusions
The absence of urinary tract symptoms and signs at initial presentation was common and associated with an increased risk of a delay in both diagnosis and therapy in patients with bacteremic UTI. Independent predictive factors for the absence of urinary tract symptoms and signs were older age, male sex, dementia, and early visit from symptom onset. Given that diagnostic and treatment delays may result in the poor prognosis of UTI patients, further studies are warranted to investigate risk factors associated with the absence of urinary tract www.nature.com/scientificreports/ symptoms and signs and develop strategies for a correct diagnosis in UTI patients without urinary tract symptoms and signs.