Intravenous Antibiotic Duration for Nontyphoidal Salmonella Bacteremia in Children Free

This is a retrospective medical record review performed within a tertiary care children’s health care system in the southeastern United States. This study included children 3 months to 18 years of age admitted to the hospital with NTSB from January 1, 2008 to December 31, 2018 who received at least 1 dose of IV antibiotics. Patients were identified utilizing International Classification of Diseases, Ninth Revision or International Classification of Diseases, Tenth Revision codes for Salmonella. Patients with a focal invasive Salmonella infection, malaria, a hemoglobinopathy, or a compromised immune system were excluded. Study data were collected and managed by using Research Electronic Data Capture, a secure data collection and management application.^10,11  This study was reviewed and approved by the hospital institutional review board.

Data collected included demographics, symptomatology, laboratory studies, and the clinical course, including length of stay (LOS), antibiotic duration, and susceptibilities. A positive blood culture result was defined as any calendar day with a positive blood culture result irrespective of the timing of antibiotic administration. Outcome variables were return to the emergency department (ED) and readmissions to the hospital within 90 days. Returns and readmissions were accessible if patients returned to any facility within our health care system as well as to select outside facilities participating in a shared electronic medical record system.

Outcomes of the SDG and LDG as well as demographic, clinical, and laboratory characteristics were compared by using χ² or Fisher’s exact test for categorical data and Wilcoxon rank test for continuous data. Two-sided P values were considered statistically significant if <.05. Statistical analysis was performed by using SAS 9.4.

A total of 957 children met the inclusion criteria, and 57 children were found to have NTSB (Fig 1). Of the 57 children, 24 (42%) were in the SDG and 33 (58%) were in the LDG. Baseline demographic characteristics of patients were similar in the SDG and LDG (Table 1). The median IV antibiotic duration was 3.0 days in the SDG and 5.0 days in the LDG. The 2 patients admitted to the PICU were both in the LDG, but they were not considered ill at presentation and were transferred to the floor the day after admission. There was no difference in calendar days with a positive blood culture result between the SDG and LDG. Median LOS in the SDG was significantly shorter at 3.0 days compared with 4.0 days in the LDG (P ≤ .001).

For IV antibiotic administration, 52 patients (91%) received ceftriaxone and 11 patients (19%) received ampicillin during their care encounter. Amoxicillin was the most common oral antibiotic choice prescribed at discharge (67%).

Ten nontyphoidal Salmonella isolates demonstrated resistance to ampicillin, ciprofloxacin, trimethoprim-sulfamethoxazole, and/or ceftriaxone (labeled number 1–10 in Table 2). Two patients with isolates resistant to ciprofloxacin reported international travel. One patient in the LDG with a multidrug resistance isolate had glutaric aciduria type 1 and had been exposed to numerous antibiotics in the past. The remaining 7 patients, including 1 with a multidrug resistance isolate in the SDG, did not have any identified risk factors for resistance.

Two patients in the SDG and 1 patient in the LDG were reevaluated in the emergency department for continued minor gastrointestinal symptoms but did not require readmission (P = .567). There was a 5% absolute risk difference of returning to the ED between the SDG (8%) and LDG (3%); 95% confidence interval −7% to 17%. The 2 patients in the SDG returned to the ED on days 9 and 32 postdischarge, and the patient in the LDG returned to the ED on day 22 postdischarge. The patient in the LDG was readmitted to the hospital for reasons unrelated to the sequelae of their salmonellosis. Otherwise, no patients in either group were readmitted to our health care system within 90 days of initial hospital discharge.

Considering that patients with persistent bacteremia may have been disproportionally represented in the LDG, we performed a subanalysis of 19 patients in the SDG and 18 patients in the LDG with only 1 day of bacteremia. The patients in the LDG had a longer LOS in the hospital (P < .001) with no difference in returns or readmissions.

This study showed that there was no difference in returns or readmissions of children with NTSB treated with ≤3 days of IV antibiotics compared with those treated with >3 days. Our findings are consistent with previous studies that have reported similar rates of complications and readmissions in patients with shorter courses of antibiotics.^7–9  Patients in the SDG also had a significantly shorter hospital LOS, which may translate to lower health care costs and risk of morbidities associated with hospitalization.

To our knowledge, this is the second study of this size to evaluate the implications of shorter IV antibiotic courses in otherwise healthy children with NTSB. A recent study defined short IV antibiotic duration as <7 days,⁹  so the majority of our patients would have been included in their SDG as the median IV antibiotic duration in our LDG was 5 days. An IV antibiotic duration of ≤3 days without an increase in complications supports a shift to shorter IV antibiotic courses. This may also reflect increasing antimicrobial stewardship cognizance.

In our study, 7% (64 of 957) of patients with nontyphoidal Salmonella infection had bacteremia, which is comparable to the 5% to 11% previously reported in the literature.^5,6  Among the 64 with NTSB, 5 patients (8%) experienced a focal invasive complication similar to a previous study.¹²  There were 18% of isolates with drug resistance, slightly less than the 20% to 30% in preceding studies.^13,14  We considered travel as a risk factor for resistance in our study because there are higher reports of quinolone and multidrug resistance to Salmonella species on other continents, especially Asia.¹⁵  All isolates in our study were susceptible to at least 1 oral and 1 IV antibiotic that was suitable to complete the treatment course.

There are several limitations of our study. First, it is a retrospective medical record review, which is susceptible to selection bias, accuracy in the extraction of patient information, and reporting of subjective variables such as duration of symptoms and indicators of ill appearance. Evaluating patients with bacteremia could also inherently include a population that is more ill. There also may have been unknown clinical factors that led to a longer duration of antibiotics. We had a modest sample size and are underpowered to detect small but potentially clinically meaningful differences in returns and readmissions, although adverse outcomes were rare occurrences. Lastly, this was a single-center study that only included previously healthy children, so the results may not be generalizable to all patient populations.

The duration of IV antibiotics varied for the treatment of NTSB, but returns and readmissions were rare in both the SDG and LDG. Our results prompt consideration to treat otherwise healthy children with NTSB >3 months of age with a short (≤3 days) course of IV antibiotics before transitioning to oral antibiotic therapy. Prospective studies that assess the optimal route and duration of antibiotics in previously healthy children with NTSB would reinforce our findings.