Abstract
Objective. Urinary tract infections (UTIs) are common among infants and toddlers. Children can be treated effectively with short courses (2–4 days) of intravenous (IV) therapy followed by oral therapy. If IV therapy is chosen, use of once-daily dosing may allow outpatient management instead of hospital admission. However, no description of ambulatory treatment with IV antibiotics of UTI among febrile children has been reported to date. We aimed to describe the feasibility and complications of outpatient management with IV antibiotics of UTI among febrile children, at the day treatment center (DTC) of a tertiary-care pediatric hospital.
Methods. Between April 1, 2002, and March 31, 2003, a prospective cohort of patients 3 months to 5 years of age who were examined in the emergency department (ED) and diagnosed as having presumed febrile UTI were treated according to a clinical protocol. Patients were treated at the DTC unless they met exclusion criteria, in which case they were hospitalized. The DTC was open 7 days per week, including holidays, from 8:30 am to 4:30 pm. At the DTC, patients were initially treated with a daily dose of IV gentamicin, until the child had been afebrile for at least 24 hours, and with oral amoxicillin, until preliminary urine culture results were available. Children allergic to penicillin received gentamicin only. IV antibiotics were administered through peripheral IV access; the IV catheter’s patency was maintained with injection of 50 U of heparin once daily throughout the treatment period. Parental satisfaction with the DTC experience was assessed with an anonymous, self-administered questionnaire.
Results. Two hundred ninety-one episodes of presumed febrile UTI were diagnosed in the ED, of which 212 (72.9%) were sent to the DTC. There were 71 hospital admissions (24.4%); in 9 of these instances, the child was admitted because parents refused or were unable to comply with DTC treatment. Adherence to the treatment protocol in the ED was excellent; in 92.1% of presumed febrile UTI episodes (268 of 291 episodes), the patient was referred to the appropriate setting for treatment. In 8 instances, patients who met an exclusion criterion were sent to the DTC. They should have been hospitalized, according to the protocol. At the DTC, a final diagnosis of UTI was made in 178 of the 212 episodes (84%). Patients treated at the DTC, with a final diagnosis of UTI, had a median age of 12.0 months (range: 3–68 months), and their mean initial temperature was 39.2°C (SD: 1.1°C). Patients were afebrile by 24 hours in 52% of UTI episodes and by 48 hours in 82%. Minor problems with IV access occurred in 9.0% of cases. The duration of IV antibiotic therapy at the DTC was 1.9 days (SD: 0.9 day). The mean number of visits to the DTC, including appointments for renal ultrasound and voiding cystourethrography evaluations, was 3.5 (SD: 0.9). Parents were present at all scheduled visits in 98.9% of cases. Four patients needed to be hospitalized from the DTC, but in only 1 case was hospital admission related to UTI treatment. Four patients with UTI treated in the DTC had positive blood cultures, 2 with Escherichia coli (both successfully treated at the DTC) and 2 with contaminants. For 4 children treated at the DTC, UTI was caused by gentamicin-resistant E coli. One patient became afebrile within 24 hours after treatment initiation with IV gentamicin; he was then treated with oral cefixime. A second patient was treated with IV ceftriaxone, administered at the DTC once culture results were available, and remained febrile for <72 hours. The last 2 patients were hospitalized; one, who was also allergic to cephalosporins, had been febrile for 72 hours at the time of hospitalization (once hospitalized, he was treated with IV amikacin), and the other was admitted to the hospital for an unrelated problem, namely, scalp cellulitis. None of these 4 patients was initially bacteremic or became bacteremic during the treatment period. Repeat urine culture was performed within 14 days after treatment initiation in 146 instances, and results were negative in all cases. At telephone follow-up assessments 14 days after discharge, no patient had been rehospitalized because of UTI. Successful treatment at the DTC (defined as attendance at all visits, normalization of temperature within 96 hours, negative control urine cultures, if performed, and absence of hospitalization from the DTC) was observed in 96.6% of the 178 UTI episodes. Overall adherence of physicians to the protocol at the DTC was 87.1% (95% confidence interval: 82.2-92.0%). One hundred seventy-two satisfaction questionnaires were returned and revealed good, very good, or excellent parental satisfaction in 98.8% of cases.
Conclusions. Our data show that ambulatory treatment with IV antibiotics, at a DTC, may be used for at least three-fourths of UTIs among febrile children 3 months to 5 years of age. It is safe and feasible and appears very satisfactory to parents. Although ambulatory treatment with IV antibiotics is more invasive than oral therapy during the initiation of UTI treatment, it ensures almost full compliance, allows close medical supervision, and facilitates investigations related to the UTI. It is an interesting alternative to hospitalization.
Urinary tract infection (UTI) is one of the most common bacterial infections among children. Acute pyelonephritis (APN) may result in significant acute morbidity and may cause permanent renal damage. Children can be treated effectively with short courses (2–4 days) of intravenous (IV) therapy followed by oral therapy.1 If IV therapy is chosen, then use of once-daily dosing (ODD) may allow outpatient management instead of hospital admission. However, no description of ambulatory treatment with IV antibiotics of UTI among febrile children has been reported to date. The purpose of this report is to describe the feasibility and complications of outpatient management with IV antibiotics of UTI among febrile children, at a day treatment center (DTC) within a tertiary-care pediatric hospital.
METHODS
A clinical protocol for the treatment of UTI was implemented at Hôpital Sainte-Justine, a tertiary-care pediatric center in Montreal, Canada, in January 2002. We prospectively evaluated its application from April 1, 2002, to March 31, 2003. All children 3 months to 5 years of age with a presumed diagnosis of febrile UTI in the emergency department (ED) were eligible for treatment according to the protocol. Children were considered febrile if their parents reported fever at home or if, in the ED, they had a body temperature of >38.5°C rectally or 38°C orally. The evaluation to be performed in the ED included a complete blood count, blood culture, measurement of serum blood urea nitrogen, creatinine, and electrolyte levels, urinalysis, and urine culture. Depending on the child’s age, the latter was to be performed through suprapubic aspiration, bladder catheterization, or clean voiding. Children who appeared in nontoxic condition, who had normal renal function and electrolyte levels, and who had no exclusion criteria received a single dose of IV gentamicin (5 or 2.5 mg/kg per dose, depending on the time of administration), a single dose of IV ampicillin, and 3 doses of oral amoxicillin to be taken until their visit at the DTC, scheduled within the next 24 hours. Children who were allergic to penicillin received gentamicin only. IV antibiotics were administered through peripheral IV access; the IV catheter’s patency was maintained with injection of 50 U of heparin once daily throughout the treatment period. Parents were asked to measure the child’s temperature every 4 to 6 hours during IV treatment. Exclusion criteria for treatment at the DTC included toxic appearance, >5% dehydration, dubious parental compliance, parental inability to comply with the physical requirements of treatment at the DTC or parental refusal, known significant uropathy (grade IV or V vesicoureteral reflux [VUR], hydronephrosis, or prior nephrostomy or vesicostomy), history of surgical intervention involving the urinary tract during the past 3 months, hearing deficit, abnormal serum creatinine levels, and other serious medical conditions. Children meeting any of these criteria were hospitalized.
The DTC was open 7 days per week, including holidays, from 8:30 am to 4:30 pm and was staffed by pediatricians working as hospitalists. At the DTC, treatment was continued with ODD of IV gentamicin (5 mg/kg per day) until the child had been afebrile for at least 24 hours and with oral amoxicillin until preliminary urine culture results were available (amoxicillin was discontinued when Gram-negative bacilli were identified in urine cultures; it was considered unnecessary to pursue oral amoxicillin therapy as empirical treatment at this stage, with IV gentamicin being sufficient to ensure good treatment of the isolated organism). If 4 doses of gentamicin were required, then a trough gentamicin level was sampled before the fourth dose. After gentamicin treatment was stopped, oral antibiotic therapy was continued for a total of 10 days; the choice of oral antibiotics was made by the treating physician, according to urine culture results. Children experiencing a first episode of UTI underwent renal ultrasound evaluation and voiding cystourethrography (VCUG) before being discharged from the DTC; VCUG was performed at least 3 days after admission to the DTC, after the child became afebrile. Repeat urine cultures were performed at the time of VCUG or at the treating physician’s request; blood cultures were also repeated at the physician’s request. Telephone follow-up monitoring was performed by a designated nurse 14 days after treatment initiation at the DTC.
Parental satisfaction with the DTC experience was assessed with an anonymous, self-administered questionnaire at the last DTC visit. A single questionnaire per episode of UTI was completed. The questionnaire was available in both English and French and consisted of 10 multiple-choice questions. Responses were either dichotomous (yes/no) or involved choosing an answer from a 5-point scale. Parents were asked to rate their overall experience at the DTC on such a 5-point scale. Other questions pertained to the clarity of explanations given in the ED, prior hospitalizations and comparisons between the DTC experience and those hospitalizations, parental worries about fever at home and about the presence of indwelling IV access during the course of therapy, and parental satisfaction when trying to reach someone from the DTC outside operating hours.
The diagnosis of UTI was deemed certain in the following instances: (1) any amount of Gram-negative bacteria grew in urine cultures obtained though suprapubic aspiration, (2) >5 × 104 colonies per mL of a single pathogen grew in urine cultures obtained through bladder catheterization (excluding lactobacilli, corynebacteria, and coagulase-negative staphylococci; urine cultures were also considered positive if they revealed >104 colonies per mL of Pseudomonas species), or (3) urine cultures obtained through clean voiding revealed >105 colonies per mL of a single pathogen (excluding lactobacilli, corynebacteria, and coagulase-negative staphylococci). The diagnosis of UTI was deemed probable if urine culture results were not consistent with the criteria listed above but were deemed diagnostic of UTI by the treating physician. All cases of certain or probable UTI had a final diagnosis of UTI for the purposes of this study.
Cost-effectiveness of treatment at the DTC was assessed by subtracting the daily costs for treatment at the DTC during the study period from the daily costs for hospitalization and multiplying the result by the number of UTI episode-days at the DTC during the 12 months of the study. Descriptive statistics were calculated for the entire cohort. Univariate and multivariate analyses using logistic regression were performed to identify factors predictive of successful implementation of the treatment protocol. A variable was created to reflect adherence of ED practitioners to patient referral to the appropriate treatment setting (DTC or hospital ward). The following covariates were included in the analysis involving this outcome: age, distance between home and hospital, years of experience of ED physician, whether this was the child’s first UTI, time of day, and type of day (weekday or weekend). Two other variables were created, to reflect (1) adherence to protocol guidelines for antibiotic administration at the DTC (defined as adequate choice and dose of initial antibiotics administered and cessation of IV gentamicin and oral amoxicillin therapy according to the criteria described previously) and (2) successful treatment at the DTC (defined as attendance at all visits, normalization of temperature within 96 hours, negative control urine cultures, if performed, and absence of hospitalization from the DTC). The following covariates were included in multivariate models involving these outcomes: age, distance between home and hospital, type of bacteria found in urine cultures, and whether this was the child’s first UTI. Maximum-likelihood estimates of regression coefficients were used to estimate crude and adjusted odds ratios (ORs) for each of the exposure variables. Ninety-five percent confidence intervals (CIs) were calculated for all estimates reported. All analyses were performed with SPSS statistical software, version 11.0.1 (SPSS Chicago, IL). This study was approved by the president of the hospital’s ethics committee.
RESULTS
Two hundred ninety-one episodes of presumed febrile UTI were diagnosed among 275 patients in the ED. Of these episodes, 212 (72.9%) were treated at the DTC, 71 (24.4%) required hospital admission, and 8 (2.7%) were treated at home with oral antibiotics. Patient flow throughout the study is summarized in Fig 1. The median age of the patients evaluated in the ED was 13 months (range: 3–68 months), and 70.2% of the patients were female.
Patient flow throughout the study.
Adherence to the treatment protocol in the ED was excellent, as shown in Table 1. In >90% of presumed febrile UTI episodes, the patient was referred to the appropriate setting to obtain treatment. In 8 instances, patients who met an exclusion criterion were sent to the DTC; they should have been hospitalized, according to the protocol. Of these patients, 4 had underlying renal anomalies (1 had recently undergone ureteral reimplantation, 1 had hydronephrosis and hydroureter, and 2 had grade V reflux; 1 of the latter also had a single kidney, a vesicostomy, and high serum creatinine levels but was sent to the DTC at his parents’ explicit request), 3 had a serious concurrent medical condition (namely significant dehydration, cystic fibrosis, and pneumonia), and 1 had an initial serum creatinine level of >54 mmol/L. Reasons for hospitalization are detailed in Table 2.
Adherence to Treatment Protocol in ED (Total N = 291)
Reasons for Hospitalization (n = 71 Episodes of Presumed Febrile UTI Referred to Hospital Ward)
A final diagnosis of UTI was made in 178 episodes (84%) of presumed febrile UTI treated in the DTC; it was excluded in 34 (16.0%). Characteristics of patients with a final diagnosis of UTI treated at the DTC or in the hospital are detailed in Table 3. Laboratory results obtained in the course of each episode of UTI treated in these settings are reported in Table 4. Urine culture was performed through bladder catheterization in 79.1% of UTI episodes treated at the DTC or in the hospital; samples were obtained through suprapubic aspiration in 0.8% of episodes, through clean voiding in 16.3%, and through bag collection in 3.3% (these percentages were nearly identical for the DTC and hospital settings).
Characteristics of Patients With a Final Diagnosis of UTI Treated at the DTC or in the Hospital (Total N = 227)
Laboratory Results in the Course of Episodes With a Final Diagnosis of UTI Treated at the DTC or in the Hospital
The clinical course at the DTC is described in Table 5. The mean number of days of treatment with IV antibiotics at the DTC was 1.9 days (SD: 0.9 day), and the mean number of visits per episode of UTI, including appointments for renal ultrasound and VCUG evaluations, was 3.5 (SD: 0.9). In 82% of UTI episodes, body temperature normalized within 48 hours after treatment initiation. Problems with IV access occurred in only 16 instances (9.0%); they included failure to establish IV access (antibiotics were administered intramuscularly in this case) and the need to reinstall IV access during the course of therapy. There were no major complications, such as hemorrhage, cellulitis, or mechanical breakage of the IV line. Trough levels of gentamicin were measured for 5 patients and were <1 mg/L in all cases. Parental compliance with scheduled visits was full for 98.9% (95% CI: 97.4-100.4%) of episodes. In both instances of incomplete compliance, parents came to all visits involving treatment and renal ultrasound evaluation but failed to come to the appointment scheduled for VCUG. Four patients were hospitalized during the course of treatment at the DTC, 3 for other medical problems (appearance of a rash, rotavirus gastroenteritis, and scalp cellulitis) and 1 because of the growth of a strain of Escherichia coli resistant to gentamicin in urine cultures (see below). No patient treated at the DTC was rehospitalized because of UTI within 14 days after diagnosis, as ascertained in the telephone follow-up assessment.
Clinical Course of UTI Episodes Treated at the DTC (Total N = 178)
Four patients with UTI treated in the DTC had positive blood cultures, 2 with E coli (their ages were 6 and 7 months, and they were both successfully treated at the DTC) and 2 with contaminants. The risk of bacteremia with a noncontaminant organism was almost 5 times higher for hospitalized children than for children treated at the DTC (crude OR: 4.8; 95% CI: 0.8-29.3); this result was not statistically significant, however, even after adjustment for age.
For 4 children treated at the DTC, UTI was caused by gentamicin-resistant E coli. One child became afebrile within 24 hours after treatment initiation with IV gentamicin and oral amoxicillin; he was then treated with oral cefixime. A second patient was treated with IV ceftriaxone administered at the DTC once culture results were available and remained febrile for <72 hours. The last 2 patients were hospitalized; one, who was also allergic to cephalosporins, had been febrile for 72 hours at the time of hospitalization (once hospitalized, he was treated with IV amikacin) and the other was admitted to the hospital for an unrelated problem, namely, scalp cellulitis. None of these 4 patients was initially bacteremic or became bacteremic during the treatment period.
At the DTC, repeat urine culture was performed for 146 patients, a mean of 5.6 days (SD: 2.4 days) after treatment initiation. Reasons for repeating the urine culture included VCUG in 93.8% of cases (routine urine culture was part of the procedure) and perceived lack of response to therapy in 4.1%. A single repeat urine culture was positive, with 104 colonies per mL of Klebsiella species (this patient had been treated for UTI caused by E coli). Five patients with UTI treated at the DTC had repeat blood cultures, a mean of 5.6 days (SD: 4.0 days) after treatment initiation (usually because of persistent fever and as a control for a bacteremic child in 1 instance); all results were negative. Treatment at the DTC was deemed successful, according to our predefined criterion, in 96.6% of episodes of confirmed UTI.
Renal ultrasound and VCUG evaluations were performed for patients treated for a first episode of UTI at the DTC, a mean of 0.9 day (SD: 0.3 day) and 6.4 days (SD: 4.0 days) after treatment initiation, respectively. Ultrasound results were normal for 75.7% of patients (109 of 144 patients); results were consistent with pyelonephritis for 11.1%, pyelectasis for 6.3%, ureteral duplication for 4.9%, and hydronephrosis, hydroureter, and nephromegaly for 0.7% each. VUR was documented for 26.4% of children (37 of 140 children) who underwent VCUG, including grade I reflux for 8.6%, grade II for 14.3%, grade III for 2.1%, and grade IV for 1.4%.
In univariate analysis, compared with children who had experienced UTI in the past, patients treated for their first episode of UTI were 3 times more likely to be sent to the right treatment location by the ED practitioner (crude OR: 3.1; 95% CI: 1.27-7.64). This result was no longer significant after adjustment for other covariates in multivariate modeling (adjusted OR: 2.3; 95% CI: 0.9-6.2). No other covariate was significantly associated with adequate treatment setting referral. None of the covariates we studied was significantly associated with adherence to protocol guidelines for antibiotic administration at the DTC or with successful treatment at the DTC. Overall adherence of physicians to the protocol at the DTC was 87.1% (95% CI: 82.2-92.0%).
One hundred seventy-two anonymous parental satisfaction questionnaires were returned. Seventy-five percent described the overall experience at the DTC as excellent, 20.3% as very good, and 3.5% as good. Only 2 parents (1.2%) considered their experience fair or poor. Of 43 parents whose child had been hospitalized in the past, 65.1% thought that their experience at the DTC was much better than their hospital experience, 20.9% that it was better, and 9.3% that it was similar. Thirty-two percent and 20.4% expressed worry at the thought of going home with a child who was still febrile and who had indwelling IV access, respectively.
At our institution, the average daily cost for hospitalization on the wards where children with UTI are usually admitted (general pediatric units) was CAN $400 (US $300) during the study period, whereas the daily cost for treatment at the DTC was CAN $52 (US $39); these rates did not include medication, investigation costs, and physician fees, which would be approximately the same in both settings. If the 178 episodes of UTI treated at the DTC had been managed on a hospital ward, with a mean of 3 days of IV therapy (534 UTI episode-days), then the additional cost attributable to hospitalization during the 12 months of our study would have been at least [CAN $(400 − 52) × 534], or CAN $185 832 (US $139 374). This is an approximate and minimal estimate that does not take into account the 34 patients with presumed UTI who were sent to the DTC for IV therapy and for whom a final diagnosis of UTI was not made (these patients would also have been hospitalized in the absence of a DTC).
DISCUSSION
Our study showed excellent adherence of both physicians and parents to a protocol for the treatment of febrile UTI among young children with daily IV gentamicin, administered in an ambulatory setting. Few patients needed to be hospitalized from the DTC; most of those who did required hospitalization for problems unrelated to their UTI. Parents described overwhelmingly their experience with ambulatory treatment at the DTC as satisfactory.
Some studies have reported that the susceptibility of the renal parenchyma to infection varies according to age, with infants <1 year of age being at greatest risk and those >5 years of age at lowest risk of developing renal scars.2 This view had a considerable effect on recommendations for the initial treatment of UTI among febrile children.3 However, according to a recent prospective study, the development of cortical scarring does not decrease with age (the study challenges the conventional view that the risk of renal scars after APN diminishes with age).4 A recent review of the use of antibiotics for APN in the pediatric population did not make distinctions between age groups, except for infants <2 to 3 months of age.5 In our study, we chose to treat children 3 months to 5 years of age with initial IV antibiotics, considering that the ideal age cutoff for less-invasive treatment is still undefined.
In 1999, Hoberman et al6 compared oral and initial IV therapy with a third-generation cephalosporin for the treatment of febrile UTI among young children and concluded that oral cefixime was as safe and effective as initial IV treatment with cefotaxime. The numbers of children with renal parenchymal damage on dimercaptosuccinic acid scans in each treatment group did not differ significantly, but the study’s power to detect such a difference was probably insufficient (CIs were wide).1 Optimal treatment of young children with APN remains a matter of controversy.1 Administering parenteral antibiotics on an outpatient basis is an interesting alternative for children who are not in toxic condition or dehydrated and whose renal function is normal.3,5 In order for outpatient parenteral management of an acute infection such as APN to be feasible, specific challenges must be overcome, ie (1) the IV antibiotics must have a prolonged effect, allowing ODD; (2) the antibiotics given must be as effective and appropriate as those that would be administered in the hospital (it would be debatable to use a wider-spectrum antibiotic simply to ensure outpatient management); (3) good coordination is required between the site where the initial diagnosis is made and treatment is started and that where ensuing treatment is to be administered; (4) families must accept coming back to the site where treatment is to be administered (the DTC in our study); (5) compliance of parents with visits is mandatory; and (6) overall treatment must be safe, without significant complications specifically related to outpatient management itself.
Aminoglycosides are among the antibiotics usually recommended for parenteral treatment of UTI.3 Their spectrum of action is more specific for Gram-negative bacteria, the most common cause of APN, than is that of third-generation cephalosporins. They achieve higher kidney tissue levels, relative to serum levels, than do β-lactams, and allergic reactions to aminoglycosides are exceedingly rare. Moreover, their cost is very low. Administration of IV gentamicin every 24 hours for the treatment of APN is considered appropriate for children with adequate renal function.1 Three trials provide data to support the safety and efficacy of ODD with aminoglycosides (gentamicin7,8 and netilmicin9), compared with 8-hourly or standard daily dosing, for children with APN. The effectiveness of ODD with 5 mg/kg IV gentamicin for the treatment of pediatric UTI has been demonstrated,8 and daily doses of 7.5 mg/kg gentamicin have been used in some studies.7 We chose the former dose to minimize potential adverse effects. Because the purpose of our study was not to specifically assess side effects of ODD of gentamicin, nephrotoxicity and ototoxicity were not systematically evaluated. The mean duration of treatment with gentamicin (1.9 days plus the initial dose in the ED) was comparable to that described by other authors.7,8
In 178 UTI episodes treated at our DTC, resistance to gentamicin was observed on only 4 occasions (1.9%). The overall sensitivity to gentamicin was thus high (98.1%) and compares with that (97%) reported by Chong et al.8 We decided to add 1 dose of IV ampicillin followed by oral amoxicillin, pending preliminary urine culture results, because of the possibility of infection by Enterococcus species (Gram-stained smears of urine are not routinely available in our ED).
The other requirements for outpatient management of an acute and invasive infection such as APN with IV antibiotics (eg, adherence of numerous physicians to a common treatment protocol and compliance of parents) have not been studied previously. The main pediatric experience with IV antibiotics out of the hospital setting concerns home care programs for children with serious infections requiring prolonged treatment with IV antibiotics (eg, acute osteomyelitis and septic arthritis).10 In these programs, each child has shown symptomatic improvement before home care is started, and the family and home situation are usually evaluated before discharge.10 Parents are trained to administer IV antibiotics, and primary IV access is a central venous catheter in the majority of cases.10 These situations differ greatly from the care required in the context of APN.
When our clinical protocol was first implemented, we were wary regarding the receptivity of ED physicians to this significant change in their practice. Use of the protocol was strongly promoted, but each physician had the liberty of applying it or not. In the ED, rapidity of intervention is critical; we knew that, if the protocol was considered to be too cumbersome, then its use would be minimal. The easiest treatment modalities for APN available to our ED physicians were admission to the hospital and oral treatment at home. Referral to the DTC meant prescription of gentamicin at different doses, depending on the time of day, and specific explanations to parents about follow-up visits to the DTC and about the fact that the child was leaving the ED with an indwelling IV access. Patients were referred to the appropriate treatment setting (DTC or hospital ward) in the majority of cases, however, and gentamicin was administered in 97% of 212 cases referred to the DTC. The choice of antibiotics was that proposed in the protocol in 84% of cases, with discrepancies mostly involving IV ampicillin or oral amoxicillin. Antibiotic doses were considered adequate in >90% of cases. Adherence to the clinical protocol by ED physicians was thus excellent. The same observation applies to the DTC, where overall compliance with protocol guidelines for IV antibiotic administration and cessation was 87%. In neither setting were the risk factors we studied significantly associated with observance of protocol guidelines. The exemplary adherence to the protocol in both the ED and the DTC makes it difficult to detect factors associated with compliance, given our sample size.
Parents refused referral to the DTC or were judged unable to comply with DTC treatment by ED physicians in only 9 instances (1 of these involved a 3-month-old child with 2 deaf parents); in these cases, children were hospitalized. Therefore, DTC treatment was usually feasible when indications to proceed were present. The rate of attendance of parents at all scheduled visits was 99%. This extremely high compliance was surprising to us. Indeed, parents had spent several hours in the ED the day before their first visit to the DTC, and >40% lived >20 km from the hospital (Table 3). Furthermore, the population in our ED is very cosmopolitan, with potential problems resulting from language or cultural barriers. This exceptional compliance with DTC appointments is probably attributable to several factors, although we cannot ascertain their relative importance, ie (1) parents were given written information about follow-up care at the DTC; (2) children were sent home with an indwelling peripheral IV access; (3) most parents were eager to see their child’s symptoms, which had usually been present for a few days, under control; and (4) some parents had been told in the ED that the child had a “kidney infection,” which they thought was serious. Good compliance with appointments ensures adequate antibiotic administration; it would be difficult to establish similar compliance with oral antibiotics. In the prospective study by Hoberman et al,6 cefixime, the oral antibiotic used, was detectable for only 85% of children for whom a urine specimen was available.
UTI can result in significant morbidity, including bacteremia and sepsis. In our study, UTI-related bacteremia was documented for 1.1% of 172 children treated at the DTC and for 6.4% of 59 hospitalized children (overall incidence: 2.5%). This incidence of UTI-related bacteremia is similar to rates of 4% to 9.3% reported in 3 recent studies.6,11,12 Young age is the risk factor most consistently associated with increased risk of UTI-associated bacteremia.11,12 After the first 1 month of life, however, the risk of bacteremia does not seem to differ among age subgroups of infants and toddlers,6 and defining a critical period after which bacteremia is nearly nonexistent is thus difficult. To date, no significant association has been found between the following variables and the risk of bacteremia: initial body temperature,11,13 irritability,11 duration of fever before antibiotic administration,6,13 and erythrocyte sedimentation rate.6,11 Hoberman et al6 found a significant increase in peripheral white blood cell counts among bacteremic patients, compared with nonbacteremic patients, but 4 other groups did not.11–14 The significance of C-reactive protein with respect to bacteremia is also controversial.6,13 Distinguishing bacteremic from nonbacteremic children in the ED is therefore difficult. In our opinion, this is an additional argument supporting initial IV treatment of young children with UTI.
Outpatient parenteral management of an acute infection such as APN can be achieved in different settings, such as in the ED itself, at home, or at a DTC, as in our study. Although daily IV antibiotics can be administered in an ED, this practice may increase the congestion of an already overburdened ED and is thus far from ideal. Home IV therapy with peripherally inserted central catheters has been described for the treatment of acute cellulitis among children.15 However, it can be difficult to organize in 12 to 24 hours, 7 days per week, after a first visit to the ED. For an infection that usually requires IV treatment of short duration, such as APN, the logistic complexity of organizing home IV therapy appears somewhat excessive. Treatment of patients at a DTC located within the hospital where the initial diagnosis of UTI was made (or in direct connection with it) seems a much better solution. If no well-defined DTC is available, then outpatient IV therapy can be administered in a circumscribed area adjacent to a hospital ward.
CONCLUSIONS
Our data show that ambulatory treatment with IV antibiotics, at a DTC, may be used for at least three-fourths of UTIs among febrile children 3 months to 5 years of age. It is safe and feasible and appears very satisfactory to parents. Although ambulatory treatment with IV antibiotics is more invasive than oral therapy during the initiation of UTI treatment, it ensures almost full compliance, allows close medical supervision, and facilitates investigations related to the UTI. It is an interesting alternative to hospitalization.
Acknowledgments
We thank Sylvie Lebrun and Josée Liberge for their collaboration in data collection and Stéphanie Tremblay for her help in establishing our treatment protocol.
REFERENCES
- Copyright © 2004 by the American Academy of Pediatrics