- Email: [email protected]
Effect of Bacteremia in Elderly Patients With Urinary Tract Infection
CLINICAL INVESTIGATION
Effect of Bacteremia in Elderly Patients With Urinary Tract Infection Arturo Artero, MD, PhD, Ana Esparcia, MD, PhD, José M. Eiros, MD, PhD, Manuel Madrazo, MD, Juan Alberola, MD, PhD and José M. Nogueira, MD, PhD ABSTRACT Introduction: The clinical effect of bacteremia on outcomes in urinary tract infection (UTI) is still debated. This study aims to examine the clinical effect of bacteremia in elderly patients with UTI requiring hospital admission. Methods: This retrospective observational study recorded the clinical features, microbiology and outcomes in a Spanish cohort of patients aged Z65 years hospitalized for UTI in whom blood cultures were performed in the emergency department. The primary outcome of the study was in-hospital mortality. Results: Of 333 patients, with a mean age of 81.6 years, 137 (41.1%) had positive blood cultures. Escherichia coli, with 223 (66.9%) cases, was the most common microorganism isolated. Independent risk factors of bacteremia were temperature 4381C, heart rate 490 bpm and inversely both Enterococcus faecalis and Pseudomonas aeruginosa. Bacteremia was not associated with the length of stay in hospital (6.96 ⫾ 3.50 days versus 7.33 ⫾ 5.54 days, P ¼ 0.456). Mortality rate was 9.3% with no significant difference between bacteremic and nonbacteremic cases (8.8% and 9.7%, respectively, P ¼ 0.773). In-hospital mortality analyzed by logistic regression was associated with McCabe index 42 (20.5% survival versus 66.7% death, adjusted odds ratio ¼ 6.31, 95% CI: 2.71-14.67; P o 0.001) but not with bacteremia (41.4% survival versus 38.7% death, adjusted odds ratio ¼ 0.99, 95% CI: 0.43-2.29; P ¼ 0.992). Conclusions: Our study suggests that the presence or absence of bacteremia in elderly people with UTI requiring hospitalization does not have an influence on outcomes such as in-hospital mortality or length of stay. Key Indexing Terms: Bacteremia; Urinary tract infection; Elderly. [Am J Med Sci 2016;](]):]]]–]]].]
INTRODUCTION
U
rinary tract infection (UTI) is the second most common cause of infections in the general population and is the most common bacterial infection occurring in elderly people.1,2 In young people, UTIs are usually noncomplicated and mainly occur in healthy women as cystitis or pyelonephritis.3 The UTIs in elderly people, however, can present with unspecific manifestations such as acute confusion. Also, the spectrum of the infection is more extensive, ranging from mild urinary symptoms to severe sepsis or septic shock and with a mortality rate of 7-33%.1,4-8 The UTI is the most common cause of communityacquired bacteremia and sepsis.5,9 In all, 40-57% of community-acquired bacteremias in elderly people are of urinary origin.10,11 On the contrary, bacteremia is confirmed in 15-42% of patients with UTI.6,11 Bacteremia has commonly been considered as a marker of severe infection,12,13 however, the clinical effect of bacteremia on outcomes in UTI is still debated.5-7,13-15 In young women with uncomplicated UTI, bacteremia was not associated with the severity of the infection or with the outcomes.9,16 A total of 2 studies conducted on the clinical effect of bacteremia on prognosis in complicated UTI in adults showed contradictory results.6,16 Nevertheless, to our knowledge, no study of this subject has been conducted to date with elderly people. The aim
of this study is to assess the effect of bacteremia on the outcomes of community-acquired severe UTI in elderly people.
METHODS Design and Setting This is a retrospective cohort study of patients with community-acquired UTI or community-onset healthcare-associated UTI (HA-UTI) admitted to a university hospital over a 6-year period (from January 2008January 2014). Only those cases that met the following criteria were included: all patients were admitted directly from the emergency department of the hospital; age Z65 years; 1 or 2 microorganisms were isolated from urinary samples; blood cultures were performed in the emergency department at admission before the administration of antibiotics, a discharge diagnosis of acute pyelonephritis or sepsis, severe sepsis or septic shock of urinary origin were established and there was no other apparent source of infection. This study was conducted in the department of internal medicine of the University Hospital Dr. Peset in Valencia, Spain. The department of internal medicine accounts for 50 beds of the total of 549 beds in the hospital.
Copyright © 2016 Southern Society for Clinical Investigation. Published by Elsevier Inc. All rights reserved. www.amjmedsci.com www.ssciweb.org
1
Artero et al
Data Collection Clinical, epidemiological and microbiological data were retrospectively obtained from the review of electronic medical records. In-hospital mortality rate and length of hospital stay were the clinical outcomes evaluated. Definitions Pyelonephritis was diagnosed if a patient had the following symptoms: temperature of 381C, pyuria, flank pain or costovertebral angle tenderness or all of these. Sepsis, severe sepsis and septic shock were defined following the criteria of the American College of Chest Physicians and Society of Critical Care Medicine Consensus Conference.17 Bacteremic UTI was defined if the same organism was isolated from both urine and blood specimens of the same patient within 48 hours of each other.18 Isolation of coagulase-negative Staphylococcus in blood cultures was considered as a contaminant and these cases were included in the negative blood culture group. Definitions of community-acquired UTI, HA-UTI, a modified McCabe and Jackson’s classification and inadequate empirical antimicrobial treatment (IEAT) have been described in previous publications of our group.15,19 Empirical antimicrobial therapy was considered inappropriate when the causative microorganism was reported as nonsusceptible (intermediate susceptibility or resistant) to the antimicrobial agent.20 Statistical Analysis Continuous variables were expressed as means ⫾ standard deviation and Student’s t test was used for comparing means. For the comparison of categorical variables, we used the chi-squared test or Fisher’s exact test when appropriate. A P o 0.05 was considered significant. Logistic regression analysis was used to determine risk factors of in-hospital mortality. All statistical analyses were performed by using the statistical pack SPSS v22.0.
RESULTS From a total of 736 patients admitted to the internal medicine department with UTI as the principal hospital discharge diagnosis during the period of the study, 333 (45.2%) patients aged Z65 years in whom blood cultures were performed were included in this study. A total of 184 patients (55.3%) were women and the average age of the patients was 81.6 ⫾ 6.8 years. Blood cultures were positive in 137 (41.1%) cases. Epidemiological, clinical and microbiological characteristics of the patients according to whether blood cultures were positive or negative are shown in Table 1. The bacteremic group presented a higher proportion of women (62.0% versus 50.5%, P ¼ 0.037), a higher proportion of patients with temperature at admission greater than
2
381C (48.9% versus 30.6%, P ¼ 0.001) and heart rate at admission higher than 90 bpm (71.5% versus 58.2%, P ¼ 0.013). Urinary catheter was more common in patients in the nonbacteriemic group (20.4% versus 31.6%, P ¼ 0.024). E. coli, with 223 (66.9%) cases, was the most common microorganism isolated, followed by E. faecalis (n ¼ 47, 15%), Klebsiella pneumoniae (n ¼ 33, 9.9%), P. aeruginosa (n ¼ 29, 8.7%) and Proteus mirabilis (n ¼ 17, 5.7%). In all, 28 (8.4%) of the UTI cases were polymicrobial. Microorganisms isolated in the urine and blood samples coincided in every case. The proportion of E. coli and K. pneumoniae isolates that were extended-spectrum beta-lactamase–producing were 20.1% and 36.3%, respectively. None of the cases of E. faecalis was vancomycin-resistant. P. aeruginosa and K. pneumoniae were both more common in patients who had a urinary catheter (18.9% versus 12.3%, P o 0.001 and 18.9% versus 6.6%, P ¼ 0.001, respectively). The UTI caused by E. coli was statistically associated with bacteremia (P ¼ 0.004), whereas E. faecalis (P ¼ 0.001) and P. aeruginosa (P ¼ 0.017) were more common in nonbacteremic cases (Table 1). Independent risk factors of bacteremia analyzed by logistic regression were temperature 4381C, heart rate 490 bpm and inversely both E. faecalis and P. aeruginosa (Table 2). Bacteremia was not associated with the length of stay in hospital (6.96 ⫾ 3.50 days versus 7.33 ⫾ 5.54 days, P ¼ 0.456). Mortality rate was 9.3% with no significant difference between bacteremic and nonbacteremic cases (8.8% and 9.7%, respectively, P ¼ 0.773). McCabe index 4 2, septic shock, APACHE II Z 15 and inappropriate empirical antimicrobial therapy were associated with mortality, but only McCabe index 4 2 was associated with mortality by multivariate analysis (odds ratio ¼ 6.31, 95% CI: 2.71-14.67; P o 0.001) as is shown in Table 3. In all, 86 (25.8%) cases received inappropriate empirical antimicrobial therapy, and this was associated with chronic pulmonary obstructive disease (30.2% versus 12.1%, P o 0.001), solid neoplasia (22.1% versus 9.3%, P ¼ 0.002), polymicrobial infections (18.6% versus 4.9%. P o 0.001), HA-UTI (80.2% versus 53.0%, P o 0.001), E. faecalis (36.0% versus 6.5%, P o 0.001) and P. aeruginosa (15.1% versus 6.5%, P ¼ 0.014).
DISCUSSION In our study, elderly patients with communityacquired UTI and HA-UTI admitted to hospital had a high percentage of bacteremia and when evaluated by mortality and length of stay in hospital they did not have a poorer prognosis than those patients without bacteremia. Slightly more than 40% of patients with UTI in this series had bacteremia on admission, a percentage in the high range compared with other studies.6,11,21 THE AMERICAN JOURNAL
OF THE
MEDICAL SCIENCES
VOLUME ] NUMBER ]
] 2016
Bacteremic Urinary Tract Infection in the Elderly
TABLE 1. Epidemiological, clinical and microbiological characteristics of 333 cases of UTI according to positive or negative blood culture.
Age, mean ⫾ SD Age 4 85 years, no. (%) Female sex, no. (%) Healthcare-associated UTI, no. (%) Hospitalization for Z2 days in the past 90 days Nursing home residence Previous antibiotic use (3 months) Comorbidities, no. (%) Diabetes mellitus COPD Previous stroke Dementia Solid neoplasia Chronic kidney disease Recurrent urinary infection (Z3 episodes/year) Urinary catheter Mc Cabe index 4 2, no. (%) Clinical features Temperature 4 381C, no. (%) Heart rate 4 90 bpm, no. (%) Mean blood pressure, in mmHg, mean ⫾ SD Breath rate 4 20, no. (%) Severity of illness at admission APACHE II score Z 15, no. (%) APACHE II, mean ⫾ SD Septic shock, no. (%) Etiology, no. (%) Escherichia coli Klebsiella pneumoniae Enterococcus faecalis Pseudomonas aeruginosa Polymicrobial infection IEAT, no. (%) Mortality, no. (%) Length of stay in hospital, days, mean ⫾ SD
Bacteremic UTI
Nonbacteremic UTI
N ¼ 137
N ¼ 196
81.52 ⫾ 7.07 39 (28.5) 85 (62.0)
81.74 ⫾ 6.62 69 (35.2) 99 (50.5)
0.515 0.196 0.037
51 (37.2) 24 (17.5) 67 (48.9)
86 (43.9) 20 (10.2) 95 (48.5)
0.225 0.052 0.938
43 (31.4) 25 (18.2) 30 (21.9) 44 (32.1) 17 (12.4) 11 (8.0) 54 (39.4) 28 (20.4) 28 (20.4)
50 (25.5) 31 (15.8) 47 (24.0) 85 (43.4) 25 (12.8) 14 (7.4) 76 (38.8) 62 (31.6) 55 (28.1)
0.240 0.559 0.657 0.063 0.925 0.763 0.906 0.024 0.114
67 (48.9) 98 (71.5) 77.05 ⫾ 26.27 71 (51.8)
60 (30.6) 114 (58.2) 77.46 ⫾ 26.82 91 (46.4)
0.001 0.013 0.891 0.332
76 (55.5) 15.85 ⫾ 5.80 26 (19.0)
107 (54.6) 15.74 ⫾ 6.07 26 (13.3)
0.873 0.864 0.158
104 (75.9) 16 (11.7) 9 (6.6) 6 (4.4) 20 (10.2) 32 (23.4) 12 (8.8) 6.96 ⫾ 3.50
119 (60.7) 17 (8.7) 38 (19.4) 23 (11.7) 20 (10.2) 54 (27.6) 19 (9.7) 7.33 ⫾ 5.54
0.004 0.366 0.001 0.019 0.158 0.390 0.773 0.456
P Value
APACHE II, Acute Physiology and Chronic Health Evaluation classification system; COPD, chronic obstructive pulmonary disease; IEAT, inadequate empirical antibiotic treatment; UTI, urinary tract infection. Bold text indicates statistically significant P value.
This fact could be because of the age of our patients, as advanced patient age was the only consistent clinical risk factor for bacteremia in UTI in previous studies.3 Several clinical and epidemiological factors were associated with bacteremia, but after analysis by logistic regression the only independent risk factors were temperature 4381C and heart rate 490 bpm. These results are in accordance with other previous studies,6,8,11 which also showed the difficulty in predicting bacteremia from clinical and epidemiological features. Although E. coli was initially associated with a higher proportion of bacteremia, in accordance with published data,6,11,13,22 after multivariate analysis only E. faecalis
and P. aeruginosa were predictors of not developing bacteremia. The low prevalence of bacteremia in UTI caused by these 2 microorganisms was the reason for the initial apparent false relationship between E. coli and bacteremia. Urinary catheter was not associated with bacteremia in multivariate analysis. This might be explained by the fact that P. aeruginosa, which was inversely associated with bacteremia, was more common in the group of patients who had a urinary catheter. In our study, taking blood cultures does not offer any advantage in the identification of the causal microorganisms as the microorganisms isolated were concordant in urine and blood samples in all cases. Blood cultures, therefore, had no influence on the therapeutic
Copyright © 2016 Southern Society for Clinical Investigation. Published by Elsevier Inc. All rights reserved. www.amjmedsci.com www.ssciweb.org
3
Artero et al
TABLE 2. Risk factors of bacteremia in elderly patients with UTI. Multivariate analysis
Univariate analysis Bacteremia
No bacteremia
Variables
N ¼ 137
N ¼ 196
Female sex, no. (%) Urinary catheter, no. (%) Temperature 4381C, no. (%) Heart rate 490 bpm, no. (%) Escherichia coli, no. (%) Enterococcus faecalis, no. (%) Pseudomonas aeruginosa, no. (%)
85 28 67 98 104 9 6
99 62 60 114 119 38 23
(62.0) (20.4) (48.9) (71.5) (75.9) (6.6) (4.4)
OR (95% CI); P Value P Value
(50.5) (31.6) (30.6) (58.5) (60.7) (19.4) (11.7)
0.037 0.024 0.001 0.013 0.004 0.001 0.019
0.75 0.77 2.03 1.70 1.02 0.31 0.32
(0.44-1.28); (0.41-1.44); (1.25-3.31); (1.03-2.80); (0.56-1.88); (0.13-0.94); (0.11-0.94);
0.296 0.420 0.004 0.036 0.934 0.005 0.030
OR, odds ratio; UTI, urinary tract infection. Bold text indicates statistically significant P value.
patients than those included in this series. Our most significant finding is that bacteremia is not related to inhospital mortality. In this series, the mortality rate was quite low (9.3%) in spite of the relatively high prevalence of APACHE II score Z15 and septic shock. This low prevalence of mortality, although expected in UTI, represents the most serious limitation of the study. However, our findings are supported by the fact that when we analyze data of length of stay in hospital we find the same absence of influence of bacteremia, thus backing up the findings related to mortality. Another limitation is that this is a retrospective study and clinical data were obtained from patient’s electronic medical records, making it difficult to identify all possible confounding factors. Lastly, this study was conducted in a single center and as such the generalizability of our findings cannot be assured. Future multicenter studies including a higher number of patients are necessary to draw a definitive conclusion on the influence of bacteremia on the outcomes of UTI in elderly people.
decision-making. In the same way, previous studies in young women with pyelonephritis23,24 and children25 with UTI showed a percentage of discordant result between blood and urine samples o5%. Nevertheless, blood cultures may be clinically useful in cases such as severe infection in which the source of the infection is unknown when the patient is being evaluated as well as in cases of bacteremic UTI in which urine cultures were not adequately taken or processed. Our study found that McCabe index 42 was the only predictive factor of mortality. McCabe index has been associated with mortality in severe infections.26 The IEAT has been found to be associated with severe UTI in elderly patients.19 In this study although IEAT was initially associated with mortality, this association was not statistically significant in the multivariable logistic regression analysis, probably because of the sample size. However, taking into consideration the high percentage of IEAT in this series, we consider that enhancing practices to improve appropriate use of antibiotics is a necessary part of the care of elderly patients with UTI in our setting. Interestingly, bacteremia was not related to the severity of the infection as determined by APACHE II nor was the presence of sepsis and septic shock; this is in contrast to the results obtained in a previous study6 but in accordance with the study of Chen et al.16 However, both of these studies included younger
CONCLUSIONS According to our results, bacteremia is not associated with poorer prognosis in community-acquired severe UTI in elderly people. We propose that severe UTI in elderly people admitted to hospital, with or
TABLE 3. Risk factors of mortality in elderly patients with UTI. Multivariate analysis
Univariate analysis
OR (95% CI); P Value
Survival
Death
Variables
N ¼ 302
N ¼ 31
P Value
Bacteremic UTI, no. (%) McCabe index 4 2, no. (%) Septic shock, no. (%) APACHE Z 15, no. (%) IEAT, no. (%)
125 62 40 159 72
12 21 12 24 14
0.773 o0.001 o0.001 0.008 0.010
(41.4) (20.5) (13.2) (52.6) (23.8)
(38.7) (66.7) (38.7) (77.4) (45.2)
0.99 6.31 2.47 1.51 2.03
(0.43-2.29); 0.992 (2.71-14.67); o0.001 (0.94-6.43); 0.064 (0.54-4.21); 0.427 (0.88-4.67); 0.093
APACHE II, Acute Physiology and Chronic Health Evaluation classification system; IEAT, inadequate empirical antibiotic treatment; OR, odds ratio; UTI, urinary tract infection. Bold text indicates statistically significant P value.
4
THE AMERICAN JOURNAL
OF THE
MEDICAL SCIENCES
VOLUME ] NUMBER ]
] 2016
Bacteremic Urinary Tract Infection in the Elderly
without bacteremia, should be considered as a disease as potentially serious as urinary sepsis with bacteremia.
ETHICAL CONSIDERATIONS This study was approved by the hospital’s Clinical Investigation Ethics Committee and complies with ethical standards. The Committee waived the need to obtain informed consent because of the fact that it was a retrospective observational study and that all data were kept confidential and information identifying patients were removed. This project has not been funded.
REFERENCES 1. Nicolle LE. Urinary tract infections in the elderly. Clin Geriatr Med 2009; 25:423–36. 2. Liang SY, Mackowiack PA. Infections in the elderly. Clin Geriatr Med 2007; 23:441–56. 3. Foxman B, Brown P. Epidemiology of urinary tract infections: transmission and risk factors, incidence, and costs. Infect Dis Clin North Am 2003;17:227–41. 4. Nicolle LE, Bradley S, Colgan R, et al. IDSA guidelines for the diagnosis and treatment of asyntomatic bacteriuria in adults. Clin Infect Dis 2005;40:643–54. 5. Tal S, Guller V, Levi S, et al. Profile and prognosis of febrile elderly patients with bacteremic urinary tract infection. J Infect 2005;50:296–305. 6. Hsu CY, Fang HC, Chou KJ, et al. The clinical impact of bacteremia in complicated acute pyelonephritis. Am J Med Sci 2006;332:175–80. 7. Chen CY, Chen YH, Lu PL, et al. Proteus mirabilis urinary tract infection and bacteremia: risk factors, clinical presentation, and outcomes. J Microbiol Immunol Infect 2011;45:228–36. 8. Chin BS, Kim MS, Han SH, et al. Risk factors of the all-cause in-hospital mortality among Korean elderly bacteremic urinary tract infection patients. Arch Gerontol Geriatr 2011;52:50–5. 9. Ramos JM, Masiá M, Elía M, et al. Epidemiological and clinical characteristic of occult bacteremia in an adult emergency department in Spain: influence of blood culture results on changes in initial diagnosis and empiric antibiotic treatment. Eur J Clin Microbiol Infect Dis 2004;23:881–7. 10. Rebelo M, Pereira B, Lima J, et al. Predictors of in-hospital mortality in elderly patients with bacteraemia admitted to an Internal Medicina ward. Int Arch Med 2011;4:33. 11. Bahagon Y, Raveh D, Schlesinger Y, et al. Prevalence and predictive features of bacteremic urinary tract infection in emergency department patients. Eur J Clin Microbiol Infect Dis 2007;26:349–52. 12. Pien BC, Sundaram P, Raoof N, et al. The clinical and prognostic importance of positive cultures in adults. Am J Med 2010;123:819–28. 13. Jerkeman M, Braconier JH. Bacteremic and non-bacteremic febrile urinary tract infection—a review of 168 hospital-treated patients. Infection 1992;20:143–5.
14. Velasco M, Martínez JA, Moreno-Martínez A, et al. Blood cultures for women with uncomplicated acute pyelonephritis: are they neccessary? Clin Infect Dis 2003;37:1127–30. 15. Artero A, Zaragoza R, Camarena JJ, et al. Prognostic factors of mortality in patients with community-acquired bloodstream infection with severe sepsis and septic shock. J Crit Care 2010;25:276–81. 16. Chen Y, Nitzan O, Saliba W, et al. Are blood cultures necessary in the management of women with complicated pyelonephritis? J Infect 2006; 53:235–40. 17. Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992;101: 1644–55. 18. Lee H, Lee YS, Jeong R, et al. Predictive factors of bacteremia in patients with febrile urinary tract infection: an experience at a tertiary care center. Infection 2014;42:669–74. 19. Esparcia A, Artero A, Eiros JM, et al. Influence of inadequate antimicrobial therapy on prognosis in elderly patients with severe urinary tract infections. Eur J Intern Med 2014;25:523–7. 20. Shaw E, Benito N, Rodríguez-Baño J, et al. Risk factors for severe sepsis in community-onset bacteraemic urinary tract infection: impact of antimicrobial resistance in a large hospitalised cohort. J Infect 2015;70: 247–54. 21. Marschall J, Zhang L, Foxman B, et al. Both host and pathogen factors predispose to Eschirichia coli urinary-source bacteremia in hospitalized patients. Clin Infect Dis 2012;54:1692–8. 22. Leibovici L, Greenshtain S, Cohen O, et al. Toward improved empiric management of moderate to severe urinary tract infections. Arch Intern Med 1992;152:2481–6. 23. Smith WR, McClish DK, Poses RM, et al. Bacteremia in young urban women admitted with pyelonephritis. Am J Med Sci 1997;313:50–7. 24. Velasco M, Martínez JA, Moreno-Martínez A, et al. Blood cultures for women with uncomplicated acute pyelonephritis: are they necessary? Clin Infec Dis 2003;37:1127–30. 25. Piteti RD, Choi S. Utility of blood cultures in febrile children with UTI. Am J Emerg Med 2002;4:271–4. 26. Blanco J, Muriel-Bombín A, Sagredo V, et al. Incidence, organ dysfunction and mortality in severe sepsis: a Spanish multicentre study. Crit Care 2008;12(6):R158.
From the Department of Internal Medicine (AA, AE, MM) and Department of Microbiology (JA, JMN), Hospital Universitario Dr. Peset, Universitat de València, Valencia, Spain; School of Medicine (JME), Universidad de Valladolid, Valladolid, Spain. Submitted March 8, 2016; accepted May 31, 2016. The authors have no financial or other conflicts of interest to disclose. Correspondence: Arturo Artero, MD, PhD, Department of Internal Medicine, Hospital Universitario Dr. Peset, Av. Gaspar Aguilar 90, 46017 Valencia, Spain (E-mail: [email protected]).
Copyright © 2016 Southern Society for Clinical Investigation. Published by Elsevier Inc. All rights reserved. www.amjmedsci.com www.ssciweb.org
5
Effect of Bacteremia in Elderly Patients With Urinary Tract Infection Arturo Artero, MD, PhD, Ana Esparcia, MD, PhD, José M. Eiros, MD, PhD, Manuel Madrazo, MD, Juan Alberola, MD, PhD and José M. Nogueira, MD, PhD ABSTRACT Introduction: The clinical effect of bacteremia on outcomes in urinary tract infection (UTI) is still debated. This study aims to examine the clinical effect of bacteremia in elderly patients with UTI requiring hospital admission. Methods: This retrospective observational study recorded the clinical features, microbiology and outcomes in a Spanish cohort of patients aged Z65 years hospitalized for UTI in whom blood cultures were performed in the emergency department. The primary outcome of the study was in-hospital mortality. Results: Of 333 patients, with a mean age of 81.6 years, 137 (41.1%) had positive blood cultures. Escherichia coli, with 223 (66.9%) cases, was the most common microorganism isolated. Independent risk factors of bacteremia were temperature 4381C, heart rate 490 bpm and inversely both Enterococcus faecalis and Pseudomonas aeruginosa. Bacteremia was not associated with the length of stay in hospital (6.96 ⫾ 3.50 days versus 7.33 ⫾ 5.54 days, P ¼ 0.456). Mortality rate was 9.3% with no significant difference between bacteremic and nonbacteremic cases (8.8% and 9.7%, respectively, P ¼ 0.773). In-hospital mortality analyzed by logistic regression was associated with McCabe index 42 (20.5% survival versus 66.7% death, adjusted odds ratio ¼ 6.31, 95% CI: 2.71-14.67; P o 0.001) but not with bacteremia (41.4% survival versus 38.7% death, adjusted odds ratio ¼ 0.99, 95% CI: 0.43-2.29; P ¼ 0.992). Conclusions: Our study suggests that the presence or absence of bacteremia in elderly people with UTI requiring hospitalization does not have an influence on outcomes such as in-hospital mortality or length of stay. Key Indexing Terms: Bacteremia; Urinary tract infection; Elderly. [Am J Med Sci 2016;](]):]]]–]]].]
INTRODUCTION
U
rinary tract infection (UTI) is the second most common cause of infections in the general population and is the most common bacterial infection occurring in elderly people.1,2 In young people, UTIs are usually noncomplicated and mainly occur in healthy women as cystitis or pyelonephritis.3 The UTIs in elderly people, however, can present with unspecific manifestations such as acute confusion. Also, the spectrum of the infection is more extensive, ranging from mild urinary symptoms to severe sepsis or septic shock and with a mortality rate of 7-33%.1,4-8 The UTI is the most common cause of communityacquired bacteremia and sepsis.5,9 In all, 40-57% of community-acquired bacteremias in elderly people are of urinary origin.10,11 On the contrary, bacteremia is confirmed in 15-42% of patients with UTI.6,11 Bacteremia has commonly been considered as a marker of severe infection,12,13 however, the clinical effect of bacteremia on outcomes in UTI is still debated.5-7,13-15 In young women with uncomplicated UTI, bacteremia was not associated with the severity of the infection or with the outcomes.9,16 A total of 2 studies conducted on the clinical effect of bacteremia on prognosis in complicated UTI in adults showed contradictory results.6,16 Nevertheless, to our knowledge, no study of this subject has been conducted to date with elderly people. The aim
of this study is to assess the effect of bacteremia on the outcomes of community-acquired severe UTI in elderly people.
METHODS Design and Setting This is a retrospective cohort study of patients with community-acquired UTI or community-onset healthcare-associated UTI (HA-UTI) admitted to a university hospital over a 6-year period (from January 2008January 2014). Only those cases that met the following criteria were included: all patients were admitted directly from the emergency department of the hospital; age Z65 years; 1 or 2 microorganisms were isolated from urinary samples; blood cultures were performed in the emergency department at admission before the administration of antibiotics, a discharge diagnosis of acute pyelonephritis or sepsis, severe sepsis or septic shock of urinary origin were established and there was no other apparent source of infection. This study was conducted in the department of internal medicine of the University Hospital Dr. Peset in Valencia, Spain. The department of internal medicine accounts for 50 beds of the total of 549 beds in the hospital.
Copyright © 2016 Southern Society for Clinical Investigation. Published by Elsevier Inc. All rights reserved. www.amjmedsci.com www.ssciweb.org
1
Artero et al
Data Collection Clinical, epidemiological and microbiological data were retrospectively obtained from the review of electronic medical records. In-hospital mortality rate and length of hospital stay were the clinical outcomes evaluated. Definitions Pyelonephritis was diagnosed if a patient had the following symptoms: temperature of 381C, pyuria, flank pain or costovertebral angle tenderness or all of these. Sepsis, severe sepsis and septic shock were defined following the criteria of the American College of Chest Physicians and Society of Critical Care Medicine Consensus Conference.17 Bacteremic UTI was defined if the same organism was isolated from both urine and blood specimens of the same patient within 48 hours of each other.18 Isolation of coagulase-negative Staphylococcus in blood cultures was considered as a contaminant and these cases were included in the negative blood culture group. Definitions of community-acquired UTI, HA-UTI, a modified McCabe and Jackson’s classification and inadequate empirical antimicrobial treatment (IEAT) have been described in previous publications of our group.15,19 Empirical antimicrobial therapy was considered inappropriate when the causative microorganism was reported as nonsusceptible (intermediate susceptibility or resistant) to the antimicrobial agent.20 Statistical Analysis Continuous variables were expressed as means ⫾ standard deviation and Student’s t test was used for comparing means. For the comparison of categorical variables, we used the chi-squared test or Fisher’s exact test when appropriate. A P o 0.05 was considered significant. Logistic regression analysis was used to determine risk factors of in-hospital mortality. All statistical analyses were performed by using the statistical pack SPSS v22.0.
RESULTS From a total of 736 patients admitted to the internal medicine department with UTI as the principal hospital discharge diagnosis during the period of the study, 333 (45.2%) patients aged Z65 years in whom blood cultures were performed were included in this study. A total of 184 patients (55.3%) were women and the average age of the patients was 81.6 ⫾ 6.8 years. Blood cultures were positive in 137 (41.1%) cases. Epidemiological, clinical and microbiological characteristics of the patients according to whether blood cultures were positive or negative are shown in Table 1. The bacteremic group presented a higher proportion of women (62.0% versus 50.5%, P ¼ 0.037), a higher proportion of patients with temperature at admission greater than
2
381C (48.9% versus 30.6%, P ¼ 0.001) and heart rate at admission higher than 90 bpm (71.5% versus 58.2%, P ¼ 0.013). Urinary catheter was more common in patients in the nonbacteriemic group (20.4% versus 31.6%, P ¼ 0.024). E. coli, with 223 (66.9%) cases, was the most common microorganism isolated, followed by E. faecalis (n ¼ 47, 15%), Klebsiella pneumoniae (n ¼ 33, 9.9%), P. aeruginosa (n ¼ 29, 8.7%) and Proteus mirabilis (n ¼ 17, 5.7%). In all, 28 (8.4%) of the UTI cases were polymicrobial. Microorganisms isolated in the urine and blood samples coincided in every case. The proportion of E. coli and K. pneumoniae isolates that were extended-spectrum beta-lactamase–producing were 20.1% and 36.3%, respectively. None of the cases of E. faecalis was vancomycin-resistant. P. aeruginosa and K. pneumoniae were both more common in patients who had a urinary catheter (18.9% versus 12.3%, P o 0.001 and 18.9% versus 6.6%, P ¼ 0.001, respectively). The UTI caused by E. coli was statistically associated with bacteremia (P ¼ 0.004), whereas E. faecalis (P ¼ 0.001) and P. aeruginosa (P ¼ 0.017) were more common in nonbacteremic cases (Table 1). Independent risk factors of bacteremia analyzed by logistic regression were temperature 4381C, heart rate 490 bpm and inversely both E. faecalis and P. aeruginosa (Table 2). Bacteremia was not associated with the length of stay in hospital (6.96 ⫾ 3.50 days versus 7.33 ⫾ 5.54 days, P ¼ 0.456). Mortality rate was 9.3% with no significant difference between bacteremic and nonbacteremic cases (8.8% and 9.7%, respectively, P ¼ 0.773). McCabe index 4 2, septic shock, APACHE II Z 15 and inappropriate empirical antimicrobial therapy were associated with mortality, but only McCabe index 4 2 was associated with mortality by multivariate analysis (odds ratio ¼ 6.31, 95% CI: 2.71-14.67; P o 0.001) as is shown in Table 3. In all, 86 (25.8%) cases received inappropriate empirical antimicrobial therapy, and this was associated with chronic pulmonary obstructive disease (30.2% versus 12.1%, P o 0.001), solid neoplasia (22.1% versus 9.3%, P ¼ 0.002), polymicrobial infections (18.6% versus 4.9%. P o 0.001), HA-UTI (80.2% versus 53.0%, P o 0.001), E. faecalis (36.0% versus 6.5%, P o 0.001) and P. aeruginosa (15.1% versus 6.5%, P ¼ 0.014).
DISCUSSION In our study, elderly patients with communityacquired UTI and HA-UTI admitted to hospital had a high percentage of bacteremia and when evaluated by mortality and length of stay in hospital they did not have a poorer prognosis than those patients without bacteremia. Slightly more than 40% of patients with UTI in this series had bacteremia on admission, a percentage in the high range compared with other studies.6,11,21 THE AMERICAN JOURNAL
OF THE
MEDICAL SCIENCES
VOLUME ] NUMBER ]
] 2016
Bacteremic Urinary Tract Infection in the Elderly
TABLE 1. Epidemiological, clinical and microbiological characteristics of 333 cases of UTI according to positive or negative blood culture.
Age, mean ⫾ SD Age 4 85 years, no. (%) Female sex, no. (%) Healthcare-associated UTI, no. (%) Hospitalization for Z2 days in the past 90 days Nursing home residence Previous antibiotic use (3 months) Comorbidities, no. (%) Diabetes mellitus COPD Previous stroke Dementia Solid neoplasia Chronic kidney disease Recurrent urinary infection (Z3 episodes/year) Urinary catheter Mc Cabe index 4 2, no. (%) Clinical features Temperature 4 381C, no. (%) Heart rate 4 90 bpm, no. (%) Mean blood pressure, in mmHg, mean ⫾ SD Breath rate 4 20, no. (%) Severity of illness at admission APACHE II score Z 15, no. (%) APACHE II, mean ⫾ SD Septic shock, no. (%) Etiology, no. (%) Escherichia coli Klebsiella pneumoniae Enterococcus faecalis Pseudomonas aeruginosa Polymicrobial infection IEAT, no. (%) Mortality, no. (%) Length of stay in hospital, days, mean ⫾ SD
Bacteremic UTI
Nonbacteremic UTI
N ¼ 137
N ¼ 196
81.52 ⫾ 7.07 39 (28.5) 85 (62.0)
81.74 ⫾ 6.62 69 (35.2) 99 (50.5)
0.515 0.196 0.037
51 (37.2) 24 (17.5) 67 (48.9)
86 (43.9) 20 (10.2) 95 (48.5)
0.225 0.052 0.938
43 (31.4) 25 (18.2) 30 (21.9) 44 (32.1) 17 (12.4) 11 (8.0) 54 (39.4) 28 (20.4) 28 (20.4)
50 (25.5) 31 (15.8) 47 (24.0) 85 (43.4) 25 (12.8) 14 (7.4) 76 (38.8) 62 (31.6) 55 (28.1)
0.240 0.559 0.657 0.063 0.925 0.763 0.906 0.024 0.114
67 (48.9) 98 (71.5) 77.05 ⫾ 26.27 71 (51.8)
60 (30.6) 114 (58.2) 77.46 ⫾ 26.82 91 (46.4)
0.001 0.013 0.891 0.332
76 (55.5) 15.85 ⫾ 5.80 26 (19.0)
107 (54.6) 15.74 ⫾ 6.07 26 (13.3)
0.873 0.864 0.158
104 (75.9) 16 (11.7) 9 (6.6) 6 (4.4) 20 (10.2) 32 (23.4) 12 (8.8) 6.96 ⫾ 3.50
119 (60.7) 17 (8.7) 38 (19.4) 23 (11.7) 20 (10.2) 54 (27.6) 19 (9.7) 7.33 ⫾ 5.54
0.004 0.366 0.001 0.019 0.158 0.390 0.773 0.456
P Value
APACHE II, Acute Physiology and Chronic Health Evaluation classification system; COPD, chronic obstructive pulmonary disease; IEAT, inadequate empirical antibiotic treatment; UTI, urinary tract infection. Bold text indicates statistically significant P value.
This fact could be because of the age of our patients, as advanced patient age was the only consistent clinical risk factor for bacteremia in UTI in previous studies.3 Several clinical and epidemiological factors were associated with bacteremia, but after analysis by logistic regression the only independent risk factors were temperature 4381C and heart rate 490 bpm. These results are in accordance with other previous studies,6,8,11 which also showed the difficulty in predicting bacteremia from clinical and epidemiological features. Although E. coli was initially associated with a higher proportion of bacteremia, in accordance with published data,6,11,13,22 after multivariate analysis only E. faecalis
and P. aeruginosa were predictors of not developing bacteremia. The low prevalence of bacteremia in UTI caused by these 2 microorganisms was the reason for the initial apparent false relationship between E. coli and bacteremia. Urinary catheter was not associated with bacteremia in multivariate analysis. This might be explained by the fact that P. aeruginosa, which was inversely associated with bacteremia, was more common in the group of patients who had a urinary catheter. In our study, taking blood cultures does not offer any advantage in the identification of the causal microorganisms as the microorganisms isolated were concordant in urine and blood samples in all cases. Blood cultures, therefore, had no influence on the therapeutic
Copyright © 2016 Southern Society for Clinical Investigation. Published by Elsevier Inc. All rights reserved. www.amjmedsci.com www.ssciweb.org
3
Artero et al
TABLE 2. Risk factors of bacteremia in elderly patients with UTI. Multivariate analysis
Univariate analysis Bacteremia
No bacteremia
Variables
N ¼ 137
N ¼ 196
Female sex, no. (%) Urinary catheter, no. (%) Temperature 4381C, no. (%) Heart rate 490 bpm, no. (%) Escherichia coli, no. (%) Enterococcus faecalis, no. (%) Pseudomonas aeruginosa, no. (%)
85 28 67 98 104 9 6
99 62 60 114 119 38 23
(62.0) (20.4) (48.9) (71.5) (75.9) (6.6) (4.4)
OR (95% CI); P Value P Value
(50.5) (31.6) (30.6) (58.5) (60.7) (19.4) (11.7)
0.037 0.024 0.001 0.013 0.004 0.001 0.019
0.75 0.77 2.03 1.70 1.02 0.31 0.32
(0.44-1.28); (0.41-1.44); (1.25-3.31); (1.03-2.80); (0.56-1.88); (0.13-0.94); (0.11-0.94);
0.296 0.420 0.004 0.036 0.934 0.005 0.030
OR, odds ratio; UTI, urinary tract infection. Bold text indicates statistically significant P value.
patients than those included in this series. Our most significant finding is that bacteremia is not related to inhospital mortality. In this series, the mortality rate was quite low (9.3%) in spite of the relatively high prevalence of APACHE II score Z15 and septic shock. This low prevalence of mortality, although expected in UTI, represents the most serious limitation of the study. However, our findings are supported by the fact that when we analyze data of length of stay in hospital we find the same absence of influence of bacteremia, thus backing up the findings related to mortality. Another limitation is that this is a retrospective study and clinical data were obtained from patient’s electronic medical records, making it difficult to identify all possible confounding factors. Lastly, this study was conducted in a single center and as such the generalizability of our findings cannot be assured. Future multicenter studies including a higher number of patients are necessary to draw a definitive conclusion on the influence of bacteremia on the outcomes of UTI in elderly people.
decision-making. In the same way, previous studies in young women with pyelonephritis23,24 and children25 with UTI showed a percentage of discordant result between blood and urine samples o5%. Nevertheless, blood cultures may be clinically useful in cases such as severe infection in which the source of the infection is unknown when the patient is being evaluated as well as in cases of bacteremic UTI in which urine cultures were not adequately taken or processed. Our study found that McCabe index 42 was the only predictive factor of mortality. McCabe index has been associated with mortality in severe infections.26 The IEAT has been found to be associated with severe UTI in elderly patients.19 In this study although IEAT was initially associated with mortality, this association was not statistically significant in the multivariable logistic regression analysis, probably because of the sample size. However, taking into consideration the high percentage of IEAT in this series, we consider that enhancing practices to improve appropriate use of antibiotics is a necessary part of the care of elderly patients with UTI in our setting. Interestingly, bacteremia was not related to the severity of the infection as determined by APACHE II nor was the presence of sepsis and septic shock; this is in contrast to the results obtained in a previous study6 but in accordance with the study of Chen et al.16 However, both of these studies included younger
CONCLUSIONS According to our results, bacteremia is not associated with poorer prognosis in community-acquired severe UTI in elderly people. We propose that severe UTI in elderly people admitted to hospital, with or
TABLE 3. Risk factors of mortality in elderly patients with UTI. Multivariate analysis
Univariate analysis
OR (95% CI); P Value
Survival
Death
Variables
N ¼ 302
N ¼ 31
P Value
Bacteremic UTI, no. (%) McCabe index 4 2, no. (%) Septic shock, no. (%) APACHE Z 15, no. (%) IEAT, no. (%)
125 62 40 159 72
12 21 12 24 14
0.773 o0.001 o0.001 0.008 0.010
(41.4) (20.5) (13.2) (52.6) (23.8)
(38.7) (66.7) (38.7) (77.4) (45.2)
0.99 6.31 2.47 1.51 2.03
(0.43-2.29); 0.992 (2.71-14.67); o0.001 (0.94-6.43); 0.064 (0.54-4.21); 0.427 (0.88-4.67); 0.093
APACHE II, Acute Physiology and Chronic Health Evaluation classification system; IEAT, inadequate empirical antibiotic treatment; OR, odds ratio; UTI, urinary tract infection. Bold text indicates statistically significant P value.
4
THE AMERICAN JOURNAL
OF THE
MEDICAL SCIENCES
VOLUME ] NUMBER ]
] 2016
Bacteremic Urinary Tract Infection in the Elderly
without bacteremia, should be considered as a disease as potentially serious as urinary sepsis with bacteremia.
ETHICAL CONSIDERATIONS This study was approved by the hospital’s Clinical Investigation Ethics Committee and complies with ethical standards. The Committee waived the need to obtain informed consent because of the fact that it was a retrospective observational study and that all data were kept confidential and information identifying patients were removed. This project has not been funded.
REFERENCES 1. Nicolle LE. Urinary tract infections in the elderly. Clin Geriatr Med 2009; 25:423–36. 2. Liang SY, Mackowiack PA. Infections in the elderly. Clin Geriatr Med 2007; 23:441–56. 3. Foxman B, Brown P. Epidemiology of urinary tract infections: transmission and risk factors, incidence, and costs. Infect Dis Clin North Am 2003;17:227–41. 4. Nicolle LE, Bradley S, Colgan R, et al. IDSA guidelines for the diagnosis and treatment of asyntomatic bacteriuria in adults. Clin Infect Dis 2005;40:643–54. 5. Tal S, Guller V, Levi S, et al. Profile and prognosis of febrile elderly patients with bacteremic urinary tract infection. J Infect 2005;50:296–305. 6. Hsu CY, Fang HC, Chou KJ, et al. The clinical impact of bacteremia in complicated acute pyelonephritis. Am J Med Sci 2006;332:175–80. 7. Chen CY, Chen YH, Lu PL, et al. Proteus mirabilis urinary tract infection and bacteremia: risk factors, clinical presentation, and outcomes. J Microbiol Immunol Infect 2011;45:228–36. 8. Chin BS, Kim MS, Han SH, et al. Risk factors of the all-cause in-hospital mortality among Korean elderly bacteremic urinary tract infection patients. Arch Gerontol Geriatr 2011;52:50–5. 9. Ramos JM, Masiá M, Elía M, et al. Epidemiological and clinical characteristic of occult bacteremia in an adult emergency department in Spain: influence of blood culture results on changes in initial diagnosis and empiric antibiotic treatment. Eur J Clin Microbiol Infect Dis 2004;23:881–7. 10. Rebelo M, Pereira B, Lima J, et al. Predictors of in-hospital mortality in elderly patients with bacteraemia admitted to an Internal Medicina ward. Int Arch Med 2011;4:33. 11. Bahagon Y, Raveh D, Schlesinger Y, et al. Prevalence and predictive features of bacteremic urinary tract infection in emergency department patients. Eur J Clin Microbiol Infect Dis 2007;26:349–52. 12. Pien BC, Sundaram P, Raoof N, et al. The clinical and prognostic importance of positive cultures in adults. Am J Med 2010;123:819–28. 13. Jerkeman M, Braconier JH. Bacteremic and non-bacteremic febrile urinary tract infection—a review of 168 hospital-treated patients. Infection 1992;20:143–5.
14. Velasco M, Martínez JA, Moreno-Martínez A, et al. Blood cultures for women with uncomplicated acute pyelonephritis: are they neccessary? Clin Infect Dis 2003;37:1127–30. 15. Artero A, Zaragoza R, Camarena JJ, et al. Prognostic factors of mortality in patients with community-acquired bloodstream infection with severe sepsis and septic shock. J Crit Care 2010;25:276–81. 16. Chen Y, Nitzan O, Saliba W, et al. Are blood cultures necessary in the management of women with complicated pyelonephritis? J Infect 2006; 53:235–40. 17. Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992;101: 1644–55. 18. Lee H, Lee YS, Jeong R, et al. Predictive factors of bacteremia in patients with febrile urinary tract infection: an experience at a tertiary care center. Infection 2014;42:669–74. 19. Esparcia A, Artero A, Eiros JM, et al. Influence of inadequate antimicrobial therapy on prognosis in elderly patients with severe urinary tract infections. Eur J Intern Med 2014;25:523–7. 20. Shaw E, Benito N, Rodríguez-Baño J, et al. Risk factors for severe sepsis in community-onset bacteraemic urinary tract infection: impact of antimicrobial resistance in a large hospitalised cohort. J Infect 2015;70: 247–54. 21. Marschall J, Zhang L, Foxman B, et al. Both host and pathogen factors predispose to Eschirichia coli urinary-source bacteremia in hospitalized patients. Clin Infect Dis 2012;54:1692–8. 22. Leibovici L, Greenshtain S, Cohen O, et al. Toward improved empiric management of moderate to severe urinary tract infections. Arch Intern Med 1992;152:2481–6. 23. Smith WR, McClish DK, Poses RM, et al. Bacteremia in young urban women admitted with pyelonephritis. Am J Med Sci 1997;313:50–7. 24. Velasco M, Martínez JA, Moreno-Martínez A, et al. Blood cultures for women with uncomplicated acute pyelonephritis: are they necessary? Clin Infec Dis 2003;37:1127–30. 25. Piteti RD, Choi S. Utility of blood cultures in febrile children with UTI. Am J Emerg Med 2002;4:271–4. 26. Blanco J, Muriel-Bombín A, Sagredo V, et al. Incidence, organ dysfunction and mortality in severe sepsis: a Spanish multicentre study. Crit Care 2008;12(6):R158.
From the Department of Internal Medicine (AA, AE, MM) and Department of Microbiology (JA, JMN), Hospital Universitario Dr. Peset, Universitat de València, Valencia, Spain; School of Medicine (JME), Universidad de Valladolid, Valladolid, Spain. Submitted March 8, 2016; accepted May 31, 2016. The authors have no financial or other conflicts of interest to disclose. Correspondence: Arturo Artero, MD, PhD, Department of Internal Medicine, Hospital Universitario Dr. Peset, Av. Gaspar Aguilar 90, 46017 Valencia, Spain (E-mail: [email protected]).
Copyright © 2016 Southern Society for Clinical Investigation. Published by Elsevier Inc. All rights reserved. www.amjmedsci.com www.ssciweb.org
5