Efficacy and safety of individualised versus standard 10-day antibiotic treatment in children with febrile urinary tract infection (INDI-UTI): a pragmatic, open-label, multicentre, randomised, controlled, non-inferiority trial in Denmark
Naqash Javaid Sethi, Emma Louise Malchau Carlsen, Abdullah Tabassum, et al
Lancet Infect Dis 2025; 25: 925-935
Summary
Background
The optimal antibiotic duration for febrile urinary tract infection (UTI) in children remains uncertain. We aimed to assess whether individualised treatment was non-inferior to standard 10-day treatment in terms of recurrent UTI and superior in reducing overall antibiotic exposure.
Methods
INDI-UTI was a pragmatic, open-label, multicentre, randomised, controlled, non-inferiority trial conducted at eight Danish hospitals. Children aged 3 months to 12 years who were febrile (≥38°C), within 24 h of treatment start, and with significant growth of uropathogenic bacteria were randomly assigned (1:1) using a web-based module with randomly permuted blocks to individualised or standard 10-day treatment. Main exclusion criteria included known urinary tract abnormalities, complicated medical history, bacteraemia, and elevated serum creatinine. The individualised group stopped treatment 3 days after adequate clinical improvement (ie, absence of fever, flank pain, and dysuria), with a minimum treatment duration of 4 days. The primary outcomes were recurrent UTI within 28 days after treatment cessation (non-inferiority margin 7·5 percentage points) and total antibiotic days within 28 days of treatment initiation (superiority assessment). No sample size calculation was performed for the assessment of total antibiotic days. Safety was assessed in all included patients. Main analyses were done in the intention-to-treat population. This study is registered with ClinicalTrials.gov, NCT05301023.
Findings
Between March 28, 2022, and March 3, 2024, 694 patients were assessed for eligibility and 408 patients were randomly assigned to individualised (n=205; median antibiotic duration 5·3 days [IQR 4·8 to 6·5]) or standard 10-day treatment (n=203; 10·0 days [10·0 to 10·0]). Median age was 1·5 years (IQR 0·7 to 5·4), and there were 326 (80%) female and 82 (20%) male participants. Recurrent UTI within 28 days occurred in 23 (11%) of 205 patients in the individualised group and 12 (6%) of 203 patients in the standard 10-day group (difference 5·3 percentage points, one-sided 97·5% CI –∞ to 11·1, pnon_inferiority=0·24). Total antibiotic days within 28 days were 6·0 (IQR 5·3 to 7·5) in the individualised group and 10·0 (10·0 to 10·0) in the standard 10-day group (median difference –4·0 days [97·5% CI –4·5 to –3·7], p<0·0001). The incidence rate of antibiotic-related adverse events within 28 days was 6·8 per 100 patient-days in the individualised group and 11·1 per 100 patient-days in the standard 10-day group (rate ratio 0·61 [95% CI 0·47 to 0·80], p=0·0003). Serious adverse events occurred in 17 (8%) of 205 patients in the individualised group and 15 (7%) of 203 patients in the standard 10-day group (difference 0·9 percentage points [95% CI –4·6 to 6·5], p=0·79).
Table 1. Baseline characteristics of the intention-to-treat population
| Individualised group (n=205) | Standard 10-day group (n=203) | ||
|---|---|---|---|
| Age, years | 1·6 (0·6–5·4) | 1·4 (0·7–5·5) | |
| Age 3–23 months | 112 (55%) | 117 (58%) | |
| Age 2–5 years | 54 (26%) | 37 (18%) | |
| Age 6–12 years | 39 (19%) | 49 (24%) | |
| Female sex | 166 (81%) | 160 (79%) | |
| Male sex | 39 (19%) | 43 (21%) | |
| Previous UTI | 31 (15%) | 28 (14%) | |
| Previous febrile UTI | 22 (11%) | 18 (9%) | |
| Febrile UTI in the past year | 15 (7%) | 8 (4%) | |
| Constipation | 25 (12%) | 26 (13%) | |
| Temperature, °C | 39·5 (39·0–40·0) | 39·5 (38·9–40·0) | |
| Fever duration before treatment initiation, h | 72 (30–118) | 60 (24–104) | |
| Clinical symptoms | |||
| Any clinical symptom | 183 (89%) | 177 (87%) | |
| Reduced food intake | 150 (73%) | 145 (71%) | |
| Stomach pain | 77 (38%) | 65 (32%) | |
| Nausea | 74 (36%) | 70 (34%) | |
| Vomiting | 70 (34%) | 65 (32%) | |
| Dysuria | 58 (28%) | 58 (29%) | |
| Flank pain | 36 (18%) | 26 (13%) | |
| Pollakisuria | 19 (9%) | 21 (10%) | |
| Increased regurgitation | 9 (4%) | 10 (5%) | |
| Haematuria | 3 (1%) | 11 (5%) | |
| Urine collection method | |||
| Midstream collection | 148 (72%) | 138 (69%) | |
| Catheterisation | 49 (24%) | 56 (28%) | |
| Bladder puncture | 8 (4%) | 9 (4%) | |
| Urine dipstick result | |||
| Leucocyturia | 188 (92%) | 186 (92%) | |
| Nitrite | 85 (42%) | 76 (37%) | |
| Leucocyturia or nitrite | 196 (96%) | 191 (94%) | |
| C-reactive protein* | 73 (25–130) | 73 (30–120) | |
| Uropathogenic bacteria | |||
| Escherichia coli | 192 (94%) | 183 (90%) | |
| Non-Escherichia coli† | 13 (6%) | 20 (10%) | |
| Restrictive microbiological criteria‡ | 192 (94%) | 180 (89%) | |
| Abnormal urine dipstick and restrictive microbiological criteria | 187 (91%) | 172 (85%) | |
Data on C-reactive protein was missing for nine patients in the individualised group and eight in the standard 10-day group.†
Non-Escherichia coli pathogens in the individualised group included Klebsiella pneumoniae(n=2), Klebsiella oxytoca (n=2), Enterococcus faecalis (n=2), Staphylococcus saprophyticus (n=2), Aerococcus urinae (n=1), Proteus mirabilis (n=1), Staphylococcus aureus (n=1), Proteus vulgaris(n=1), and Raoultella planticola (n=1). In the standard 10-day group, non-Escherichia colipathogens included Klebsiella pneumoniae (n=7), Klebsiella oxytoc a (n=3), Aerococcus urinae(n=3), Enterococcus faecalis (n=1), Staphylococcus saprophyticus (n=1), Proteus mirabilis (n=1), Staphylococcus aureus (n=1), Citrobacter koseri (n=1), Pseudomonas aureginosa (n=1), and Streptococcus agalactiae (n=1).‡
Excluding patients with 103 CFU per mL using catheterisation and 104 CFU per mL in two cultures using midstream collection.
Table 2. Characteristics of the baseline antibiotic treatment in the intention-to-treat population
| Individualised group (n=205) | Standard 10-day group (n=203) | ||
|---|---|---|---|
| Amoxicillin–clavulanic acid1 | 141 (69%) | 140 (69%) | |
| Empirical treatment | 132/141 (94%) | 135/140 (96%) | |
| Resistance | 9/132 (7%) | 8/135 (6%) | |
| Mecillinam1 | 28 (14%) | 34 (17%) | |
| Empirical treatment | 27/28 (96%) | 30/34 (88%) | |
| Resistance | 1/27 (4%) | 2/30 (7%) | |
| Gentamicin and ampicillin* | 22 (11%) | 17 (8%) | |
| Empirical treatment | 22/22 (100%) | 17/17 (100%) | |
| Resistance | 0 | 0 | |
| Amoxicillin1 | 12 (6%) | 9 (4%) | |
| Empirical treatment | 6/12 (50%) | 7/9 (78%) | |
| Resistance | 3/6 (50%) | 3/7 (43%) | |
| Other antibiotics† | 2 (1%) | 3 (1%) | |
| Empirical treatment | 2/2 (100%) | 1/3 (33%) | |
| Resistance | 1/2 (50%) | 0 | |
39 patients received intravenous antibiotics empirically with combined gentamicin and ampicillin due to the following reasons: unstable appearance at treatment initiation (n=14), stable appearance but markedly increased C-reactive protein (median C-reactive protein 195 [IQR 123–257]) at treatment initiation (n=10), not tolerating oral antibiotics (n=6), no specific reason (n=5), and elevated serum creatinine at treatment initiation (n=4).†
In the individualised group, one patient was treated with oral trimethoprim (8 mg per kg per day in two doses) and one with oral sulfamethizol (50 mg per kg per day in three doses); in the standard 10-day group, one patient was treated with oral trimethoprim, one with oral dicloxacillin (50 mg per kg per day in three doses), and one, who was not empirically treated, with intravenous cefuroxime (60 mg per kg per day in three doses) due to the growth of a multiresistant pathogen with no suitable oral treatment option.
Table 3. Outcome results for the intention-to-treat population
| Individualised group (n=205) | Standard 10-day group (n=203) | Group difference | pnon-inferiority | psuperiority | |
|---|---|---|---|---|---|
| Primary outcomes | |||||
| Recurrent UTI within 28 days | 23 (11%) | 12 (6%) | 5·3 (−∞ to 11·1)* | 0·24 | NA |
| Antibiotic days within 28 days | 6·0 (5·3-7·5) | 10·0 (10·0–10·0) | −4·0 (−4·5 to −3·7)† | NA | <0·0001 |
| Secondary outcomes | |||||
| Recurrent UTI within 100 days | 33 (16%) | 29 (14%) | 1·8 (−∞ to 9·0)* | 0·012 | NA |
| Absence incidence rate within 28 days‡ | 348/1720 (20·2) | 442/1858 (23·8) | 0·84 (0·66 to 1·08)§ | NA | 0·18 |
| Exploratory outcomes | |||||
| Hospital contacts during treatment for the baseline infection | 76 (37%) | 31 (15%) | 21·8 (13·0 to 30·2)¶ | NA | <0·0001 |
| Hospital contacts for UTI symptoms within 28 days | 50 (24%) | 28 (14%) | 10·6 (2·3 to 18·3)¶ | NA | 0·0066 |
| Resistant recurrent infection within 100 days‖ | 3 (2%) | 5 (3%) | −1·0 (−4·4 to 2·1)¶ | NA | 0·51 |
Risk difference (one-sided 97·5% CI).†
Difference in median antibiotic days in 28 days (two-sided 97·5% CI).‡
98 (71%) of 138 patients attending school or daycare in the individualised group and 101 (70%) of 144 patients attending school or daycare in the standard 10-day group had available data; accordingly, 126 (67 in the individualised group and 59 in the standard 10-day group) patients not attending school or daycare were removed from the analysis and 83 (40 in the individualised group and 43 in the standard 10-day group) patients were lost to follow-up.§
Incidence rate ratio (95% CI).¶
Risk difference (95% CI).‖
Two (1%) of 205 patients in the individualised group and two (1%) of 203 patients in the standard 10-day group had a resistant recurrent infection within 28 days.
Table 4. Adverse events in the safety population
| Empty Cell | Individualised group | Standard 10-day group | Group difference | psuperiority |
|---|---|---|---|---|
| Antibiotic-related adverse event within 28 days* | ||||
| Number of patients | 173 | 168 | .. | .. |
| Any adverse event | 102 (59%) | 109 (65%) | −5·9 (−16·3 to 4·5)† | 0·27 |
| Loss of appetite or nausea | 65 (38%) | 72 (43%) | .. | .. |
| Diarrhoea | 55 (32%) | 61 (36%) | .. | .. |
| Abdominal pain | 46 (27%) | 40 (24%) | .. | .. |
| Vomiting | 27 (16%) | 32 (19%) | .. | .. |
| Adverse event incidence rate | 329/4844 (6·8) | 523/4704 (11·1) | 0·61 (0·47 to 0·80)‡ | 0·0003 |
| Serious adverse events within 100 days | ||||
| Number of patients | 205 | 203 | ||
| Any serious adverse event | 17 (8%) | 15 (7%) | 0·9 (−4·6 to 6·5)† | 0·79 |
| All-cause hospitalisation | 17 (8%) | 15 (7%) | .. | .. |
| Bacteraemia | 2 (1%) | 0 | .. | .. |
173 (84%) of 205 patients in the individualised group and 168 (83%) of 203 patients in the standard 10-day group had available data; accordingly, 67 patients (32 in the individualised group and 35 in the standard 10-day group) were lost to follow-up.†
Risk difference (95% CI).‡
Incidence rate ratio (95% CI).
Interpretation
Children with febrile UTI assigned to individualised treatment duration had an increased risk of recurrent UTI (by 5·3 percentage points) but reduced antibiotic use and fewer adverse event days within 28 days compared with those assigned to standard 10-day treatment. These findings highlight the potential of individualised treatment strategies to reduce antibiotic exposure and associated harms in most children with febrile UTI, supporting antimicrobial stewardship goals. Further research is needed to identify those requiring 10-day treatment to avoid compromising care for most children with febrile UTI who respond well to shorter durations.
Funding
Copenhagen University Hospital Rigshospitalet Research Fund, Innovation Fund Denmark, and Greater Copenhagen Health Science Partners.
