Acute urinary tract infection (UTI) and pyelonephritis are common infections in children presenting to hospital emergency departments. Although there is no clear consensus internationally about which patients require intravenous (IV) antibiotic therapy, patients who are treated IV include those who are febrile or otherwise unwell, have underlying urologic conditions or who are very young.
The potential for treating children with IV antibiotics outside the hospital ward environment is being realized with increasing support for outpatient parenteral antimicrobial therapy (OPAT) programs. Ambulatory treatment in an outpatient clinic or in the home offers improved psychologic outcomes, is acceptable to families, reduces pressure on inpatient beds and is cost effective.1–4 UTI requiring IV antibiotics is amenable to ambulatory treatment as most UTIs can be treated with once-daily antibiotics (either an aminoglycoside5 or ceftriaxone6). Complete admission avoidance with the total IV course administered via OPAT directly from the emergency department (ED) is, therefore, an attractive option but is rarely practiced in children as there is little evidence to support its use in acute infections.7 This is reflected in the lack of reference to this pathway in international guidelines including the American Academy of Pediatrics and the National Institute of Health and Care guidelines.8,9
The few studies of admission avoidance using OPAT for UTI in children have been limited by the exclusion of patients deemed unwell or complicated and in some studies only including patients who could have been treated with oral antibiotics.10–14 At our institution, the guidelines for use of IV antibiotics in acute febrile UTI indistinguishable from pyelonephritis (henceforth termed UTI/pyelonephritis) are for patients who are unwell or complicated or are younger than 6 months of age. A previous study at our institution showed that despite having a hospital-in-the-home (HITH) service where a nurse visits the home to administer IV antibiotics under medical oversight, the management of UTI/pyelonephritis in the ambulatory setting remained the exception.15 Since then, we have developed a direct-to-home from the ED pathway for IV antibiotics via HITH for some acute infections.
The aim of this study was to determine the outcomes of treating patients with UTI/pyelonephritis with IV antibiotics at home directly from the ED. To provide context for this management pathway, we also determined how many patients with UTI are treated with IV versus oral antibiotics at our institution and compared the number of patients with UTI/pyelonephritis admitted to an inpatient ward with the patients treated at home via HITH.
MATERIALS AND METHODS
This was a retrospective study from August 2012 to July 2016 of all children with a UTI treated with IV antibiotics directly from ED to HITH. In addition, all patients who attended ED and had a discharge diagnosis of UTI/pyelonephritis from August 2012 to December 2015 had their clinical presentation and management reviewed.
The Royal Children’s Hospital Emergency Department where patients are assessed at presentation using the Australasian Triage Scale.16 To be eligible for direct transfer from ED to HITH, the following criteria needed to be met: (1) the patient should be considered medically (including hemodynamically) stable by ED senior medical staff with minimal concern about acute deterioration; (2) a maximum of twice daily interventions/home visits were required and (3) consent was needed from patient/parent. The RCH HITH team can visit within a 40 km radius of the hospital. This pathway includes patients with UTI/pyelonephritis. Patients are reviewed at home the following day by a physician from the HITH team who also determines antibiotic duration.
All patients with UTI/pyelonephritis diagnosed in ED and then transferred directly to HITH for IV antibiotics. The decision to treat a child for UTI is based on presenting clinical features and urine dipstick (including positive nitrites and/or leucocyte esterase) in ED. Our hospital guideline recommends gentamicin as first-line IV antibiotic to treat Gram-negative uropathogens for either hospital or treatment. UTI is confirmed on microscopy and culture and if the resistance pattern necessitates antibiotic change, this is made by the HITH medical team. If the culture is negative, a judgment is made on continuation of treatment based on the likelihood of the diagnosis, for example, flank pain, rigors and sterile pyuria with prior antibiotics. Both of these medical decisions are made by the HITH physicians in the same way as care is provided for children who are inpatients by ward-based physicians.
The data regarding these admissions were prospectively entered into a HITH database. Data were verified retrospectively through patient medical record review. This HITH database was cross-referenced with the ED database to ensure all patients were captured. The ED database was also used to find the number of overall presentations with UTI/pyelonephritis and inpatient admissions. Because of a change in hospital systems, these data were only accessible for the period from August 1, 2012, to the December 31, 2015.
Outcomes included representation to ED, readmission to our hospital, length of stay under HITH, complications and cost. Cost was calculated using medical and nursing costs, consumables and indirect overheads such as use of vehicles and information technology.
For comparisons between groups, for categorical data 2-tailed Fisher exact test was used, and for continuous data, Student’s t test was used. Odds ratios and 95% confidence intervals were calculated and for hypothesis testing. P < 0.05 was considered significant.
This study had ethics approval from the RCH Human Research Ethics Committee (32291A).
From August 2012 to July 2016, there were 62 episodes of patients with UTI/pyelonephritis admitted directly from ED to HITH, comprising 53 different patients. All patients fulfilled National Institute of Health and Care criteria for the diagnosis of UTI/pyelonephritis with fever, pyuria, plus or minus flank pain, and we have, therefore, presented the results of the entire cohort. However, several patients had received prior oral antibiotics and had negative urine cultures so in each table we have also provided the results for the patient group who had definitive bacteria identified on urine culture to ensure that these were not different.
Demographics and Clinical Features
Patients were a median age of 4.9 years, with the youngest patient 2.5 months of age (Table, Supplemental Digital Content 1, https://links.lww.com/INF/D140). Eighteen (29%) patients had an underlying condition, with the majority (14, 78%) of these being surgical malformations such as vesicoureteric reflux and neurogenic bladder with a Mitrofanoff conduit.
The majority of patients (58, 94%) had systemic features including fever, vomiting and/or tachycardia (Table, Supplemental Digital Content 1, https://links.lww.com/INF/D140). In the 15 patients with tachycardia, only 8 patients had documented resolution before transfer to HITH. One 12-year-old patient was hypotensive with a lowest documented blood pressure of 88/56 mm Hg, which normalized with IV fluids in the ED.
It was not always documented why patients were treated with IV versus oral antibiotics. Of the 52 episodes where the patient was febrile, 26 (50%) had recurrent UTI and 9 (17%) had underlying malformations. Of the 10 episodes without fever, 8 (80%) had a history of or current infection with a multidrug resistant bacteria with no oral antibiotic option, 1 patient was 10 weeks old and 1 patient had multiple courses of antibiotics before renal tuberculosis was diagnosed.
Investigations Including Microbiology
Blood was taken for investigation in 50 (81%) patients (Table 1). Neutrophils were high in 20 (48%) and C-reactive protein was greater than 50 mg/L in 14 (38%) and greater than 200 mg/L in 4 (11%). Eleven (23%) patients had mild electrolyte disturbances. Blood cultures were taken in 33 patients, and 32 (97%) were negative. One was positive for Escherichia coli in a 4-month-old infant, who was subsequently admitted to hospital for 5 days IV antibiotics and monitoring and then switched to oral antibiotics. Urine culture results were available for all 62 patients, although in 6, the general practitioner had started antibiotics before a urine sample. One patient was subsequently diagnosed with renal tuberculosis. Of the 61 considered to have a bacterial UTI, 46 (75%) had a pure positive culture, 6 (10) had mixed growth 106–107 cfu/L and 9 had no significant growth.
Of these 15 patients without a clear cultured organism, 6 had fever and pyuria but had received prior oral antibiotics from the general practitioner, 8 had fever and pyuria ± flank pain or rigors (4/8 had mixed growth) and 1 patient 12 weeks of age presented with irritability and vomiting, and an in-out catheter specimen with >999 × 106/L leucocytes and mixed growth.
Nine (15%) patients received IV fluids, and 8 (13%) received an antiemetic in ED. Parenteral antibiotics were IV via a peripheral cannula for 55 (89%) patients and intramuscular for 7 (11%; Table 2). The majority of patients (59, 95%) were treated with 1 antibiotic. Only 3 different antibiotics were used, with 50 (81%) episodes being empirically treated with an aminoglycoside. A clear duration of IV antibiotics planned in the ED was documented in 47 (76%) patients, with 39 (83%) planned for 2–3 days and the remainder 4–7 days. Two patients had a planned duration of greater than 5 days, both of whom had known resistant organisms with no oral antibiotic options.
Of the 62 episodes, 56 (90%) had successful completion of the IV antibiotic course via HITH (Table 3). Of the 6 (10%) patients who were readmitted during treatment, for 2, this was for less than 24 hours and they were discharged home. The reasons for readmission were inadequate improvement in symptoms (4 patients), underlying epilepsy (1) and misdiagnosis (1; see table, Supplemental Digital Content 5). None of the patients readmitted required treatment for septic shock. Regarding specific complications related to OPAT: no patient had antibiotic side effects necessitating change, while 2 (3%) had blocked IV access which was replaced in ED before returning home. The median length of stay under HITH for parenteral antibiotics of the 56 patients who were not readmitted was 2 days (range: 1–6 days; Fig., Supplemental Digital Content 3, https://links.lww.com/INF/D142), followed by oral antibiotics for the remainder of treatment. Three (5%) patients were readmitted with a further episode of UTI/pyelonephritis within 30 days of discharge from HITH, each again being treated via HITH.
The 62 patients treated at home were under HITH for a combined total of 142 days. The cost of being under HITH care per day for treatment of UTI/pyelonephritis was Australian dollar (AUD) 530 [US dollar (USD) 403], compared with an inpatient medical bed for the same condition, which was AUD 1297 (USD 986) per day in our institution. Based on this bed cost difference of AUD 767 (USD 583) per patient day, treating these 62 patients at home led to a real cost saving of AUD 108,914 (USD 82,775).
Comparison Between Patients Treated via HITH and in Hospital
From August 2012 to December 2015, 2949 patients 3 months to 18 years of age were diagnosed in the ED with UTI/pyelonephritis. 2231 (76%) were discharged on oral antibiotics, 664 (22%) were admitted to hospital for IV antibiotics and 54 (2%) were transferred directly to HITH (these are 54 of the 62 patients in this study). These 54 represented only 8% of the patients treated with IV antibiotics.
Compared with patients admitted to hospital, patients admitted to HITH were older (median age: 5.3 years via HITH versus 2.8 years in hospital), with a lower proportion of children 1 year of age or younger (13% versus 33%; odds ratio: 0.3; 95% confidence interval: 0.1–0.7; P < 0.01; Table 4). However, patients admitted to HITH were more likely to be febrile than inpatients: 44/54 (81%) versus 376/664 (57%; odds ratio: 3.4; 95% confidence interval: 1.7–6.8; P < 0.01). Of note though, 1011/2231 (45%) patients discharged from ED with oral antibiotics were also febrile. In other words, of all febrile patients presenting to ED and treated for UTI/pyelonephritis, only 413/1424 (29%) were admitted to hospital or HITH for IV antibiotics. For patients transferred to HITH from the ED, the median length of stay in ED was 4 hours (range: 0–14 hours) compared with transfer to an inpatient ward with an average of 5 hours (range: 1–21 hours; P = 0.02).
The main finding from this study is that patients with UTI/pyelonephritis requiring IV antibiotics selected for treatment under HITH directly from ED had good outcomes. The results for the group of patients with a definitive positive bacterial urine culture did not differ from the whole cohort. Of patients treated via this management pathway, 90% successfully completed the treatment course without readmission to hospital. For those who were readmitted, none had become severely unwell at home although numbers are too small to make a comprehensive statement about safety.
There have been a few previous studies where an OPAT model has been used to treat UTI/pyelonephritis in children, but our patient population has some major differences to the patients in those studies. In the 5 previously published studies (from 3 groups), outcomes for OPAT were good, the outpatient model was reported as safe and costs were lower.10–14 However, in all 5 studies, 96%–100% of patients presenting with a febrile UTI were treated with IV antibiotics, with almost no patients were started on oral antibiotics. This is in contrast to our study where only 29% of those presenting with febrile UTI were treated with IV antibiotics, indicating a higher threshold for IV treatment. This suggests that many patients in the previously published studies may have been able to be treated with oral antibiotics, and their potential overtreatment with IV antibiotics would skew the safety and efficacy data. In one study, the authors even acknowledged that the patients could have been treated with oral antibiotics.13 The National Institute of Health and Care guidelines for acute management of UTI/pyelonephritis in children do not give clear recommendations about which patients should be treated with IV antibiotics.9 They do, however, suggest that if IV treatment is initiated, it should be for 2–4 days; in our study the, median duration was 2 days. The American Academy of Pediatrics guidelines (2011) recommend that the route of administration should be based on practical considerations, while stating that initiating treatment orally or parenterally is equally efficacious.8,17 While these guidelines permit a wide range of treatment practices, it is important that IV antibiotics are not used where oral antibiotics would be as effective as this increases the likelihood of IV access complications and cost.
A recent Cochrane review of antibiotic use in children with febrile UTI/pyelonephritis suggested that oral antibiotics are as effective as iv.18 However, this was based on studies that excluded patients with previous UTI,19 prior oral antibiotics,20 urologic malformations,19,21 vomiting,20,22 dehydration22 and renal dysfunction,21,22 because these patients are considered to require IV antibiotics. In our study, there were patients in all of these groups receiving IV antibiotics, but importantly and for the first time, they were receiving IV antibiotics at home. This is again in contrast to the 5 previously published studies where specific exclusions from home treatment included previous UTI,11 prior oral antibiotics,11 urologic malformations or surgery,10–14 dehydration or reduced oral intake,10–13 serious comorbidities10,12,13 and abnormal serum creatinine.13
If a child is severely unwell at presentation (eg, with ongoing cardiovascular compromise), they are not considered for the home treatment pathway. However, we have shown that other factors relating to severity need not necessarily be used as exclusions for ambulatory care. We treated children via our direct-from-ED to home pathway that had vomited or were treated in the ED with IV fluids, with urologic malformations or other comorbidities, who had had previous UTI including multidrug resistant UTI and those with a variety of abnormal laboratory results including electrolytes and a high C-reactive protein. As patients with urologic malformations are particular predisposed to recurrent UTI and resistant organisms, having the option to treat them at home avoids frequent hospital admissions with the associated risk of resistant hospital-acquired infections. Our home pathway practice appears to be unique in that (1) IV antibiotics are limited to those needing them for clinical reasons and (2) patients are treated at home who have been excluded from previous studies as being too complicated.
Despite the good outcomes for patients treated at home, including complicated patients excluded from previous studies, only 8% of all patients treated with IV antibiotics were treated via HITH. This is despite the clear cost benefits of this pathway. This raises the possibility that more patients clinically requiring IV antibiotics for UTI could be treated with OPAT. Although a higher proportion of under 1-year-olds were admitted to hospital, this does not necessarily preclude these patients from ambulatory treatment as we also treated children in this age group at home. That a high proportion of children treated via HITH were febrile also indicates that fever alone should not be a prerequisite for requiring admission to hospital. Only 2 previous studies by the same authors have compared outcomes between home and hospital IV treatment for UTI. The first study compared OPAT versus inpatient IV treatment for all children with febrile UTI 2–24 months of age and found no difference between groups in duration of IV antibiotics, complications and success rate.11 OPAT was associated with a 73% cost-reduction, but use of IV antibiotics for all febrile children would not necessarily be considered usual practice. In the second study of children 2 months to 5 years of age, there was no difference between home versus hospital in treatment success, but the adverse event rate was substantially higher in hospital than at home (76% versus 16%).10 This is likely because of the fact that all children with complicated UTI were admitted to hospital so the groups were not comparable.
As this is the first study of an ambulatory pathway that includes complicated UTI/pyelonephritis, it provides valuable information regarding potential patient groups and challenges the dogma that certain clinical features necessitate hospital admission. However, numbers are low compared with the number treated in hospital, so firm conclusions cannot be drawn about safety or overall efficacy. While we have provided some institutional data on overall presentations with UTI/pyelonephritis to contextualize the current practice, the data available for direct comparison was limited. To allow generalizability, a larger comparative study will be required to compare outcomes prospectively.
In conclusion, we found that selected patients presenting to ED with febrile or complicated UTI/pyelonephritis may be treated directly via HITH. Ambulatory IV treatment directly from the ED is still considered novel and is not included in current guidelines. Our study adds weight to developing and studying OPAT models for UTI/pyelonephritis more widely with clear criteria for hospital versus home management. However, to ensure this model is not inappropriately applied to patients who could be treated with oral antibiotics, we recommend that studies reporting on OPAT practice should provide a description of numbers of patients treated with IV and oral antibiotics.
1. Small F, Alderdice F, McCusker C, et al. A prospective cohort study comparing hospital admission for gastro-enteritis with home management. Child Care Health Dev. 2005;31:555562.
2. Tiberg I, Katarina SC, Carlsson A, et al. Children diagnosed with type 1 diabetes: a randomized controlled trial comparing hospital versus home-based care. Acta Paediatr. 2012;101:10691073.
3. Balaguer A, Gonzalez de Dios J. Home versus hospital intravenous antibiotic therapy for cystic fibrosis. Cochrane Database Syst Rev. 2012;(3):CD001917.
4. Svahn BM, Remberger M, Heijbel M, et al. Case-control comparison of at-home and hospital care for allogeneic hematopoietic stem-cell transplantation: the role of oral nutrition. Transplantation. 2008;85:10001007.
5. Poey N, Madhi F, Biscardi S, et al. Aminoglycosides monotherapy as first-line treatment for febrile urinary tract infection in children. Pediatr Infect Dis J. 2017;36:11041107.
6. Smith JK, Alexander S, Abrahamson E. Ambulatory intravenous ceftriaxone in paediatric A&E: a useful alternative to hospital admission? Emerg Med J. 2011;28:877881.
7. Bryant PA, Katz NT. Inpatient versus outpatient parenteral antibiotic therapy at home for acute infections in children: a systematic review. Lancet Infect Dis. 2018;18:e45e54.
8. Roberts KB; Subcommittee on Urinary Tract Infection, Steering Committee on Quality Improvement and Management, Urinary tract infection: clinical practice guideline for the diagnosis and management of the initial UTI in febrile infants and children 2 to 24 months. Pediatrics 2011;128:595610.
9. Baumer JH, Jones RW. Urinary tract infection in children, National Institute for Health and Clinical Excellence. Arch Dis Child Educ Pract Ed 2007;92:189192.
10. Peña A, Zambrano A, Alvarado M, et al. [Evaluation of the effectiveness, safety and costs of outpatient intravenous antimicrobial treatment (OPAT) vs hospitalized in urinary infection in pediatrics]. Rev Chilena Infectol. 2013;30:426434.
11. Pena DA, Viviani ST, Le Corre PN, Morales MV, Montecinos BC, Gajardo SC. [Treatment of urinary tract infections in febrile infants: experience of outpatient intravenous antibiotic treatment]. Rev Chilena Infectol 2009;26:350354.
12. Doré-Bergeron MJ, Gauthier M, Chevalier I, et al. Urinary tract infections in 1- to 3-month-old infants: ambulatory treatment with intravenous antibiotics. Pediatrics. 2009;124:1622.
13. Gauthier M, Chevalier I, Sterescu A, et al. Treatment of urinary tract infections among febrile young children with daily intravenous antibiotic therapy at a day treatment center. Pediatrics. 2004;114:e469e476.
14. Sannier N, Le Masne A, Sayegh N, et al. Ambulatory management of acute pyelonephritis in children. Acta Paediatr. 2000;89:372373.
15. Hensey CC, Sett A, Connell TG, et al. A comparison of hospital versus outpatient parenteral antibiotic therapy at home for pyelonephritis and meningitis. Pediatr Infect Dis J. 2017;36:827832.
16. Australasian College for Emergency Medicine. The Australasian Triage Scale. Emerg Med (Fremantle). 2002;14:335336.
17. Subcommittee On Urinary Tract Infection. Reaffirmation of AAP clinical practice guideline: the diagnosis and management of the initial urinary tract infection in febrile infants and young children 2–24 months of age. Pediatrics 2016;138.
18. Strohmeier Y, Hodson EM, Willis NS, et al. Antibiotics for acute pyelonephritis in children. Cochrane Database Syst Rev. 2014;(7):CD003772.
19. Hoberman A, Wald ER, Hickey RW, et al. Oral versus initial intravenous therapy for urinary tract infections in young febrile children. Pediatrics. 1999;104(1 pt 1):7986.
20. Bocquet N, Sergent Alaoui A, Jais JP, et al. Randomized trial of oral versus sequential IV/oral antibiotic for acute pyelonephritis in children. Pediatrics. 2012;129:e269e275.
21. Neuhaus TJ, Berger C, Buechner K, et al. Randomised trial of oral versus sequential intravenous/oral cephalosporins in children with pyelonephritis. Eur J Pediatr. 2008;167:10371047.
22. Montini G, Toffolo A, Zucchetta P, et al. Antibiotic treatment for pyelonephritis in children: multicentre randomised controlled non-inferiority trial. BMJ. 2007;335:386.