Omphalitis Hospitalizations at a US Children’s Hospital Free

This descriptive study was conducted at a tertiary children’s health care organization in the Midwestern United States with 2 freestanding hospitals and ∼14,000 annual admissions. We included children <12 months admitted to the children’s hospital with an International Classification of Disease, Ninth or 10th Revision (ICD-9, ICD-19), discharge diagnosis of omphalitis (ICD-9 [771.3–771.4] and ICD-10 [P38.1, P38.9, P38.-, L08.82]) between January 1, 2006, and December 31, 2020. Patients were excluded if they did not give general consent for research or did not receive antibiotics. Data were obtained from electronic health records and chart reviews by 4 trained physician reviewers.⁶ 

Descriptive statistics were used to characterize patient demographics and relevant clinical outcomes, including frequency with percentage for categorical variables and median with interquartile range (IQR) for continuous variables. We limited our description of presenting signs and symptoms, laboratories, and hospital length of stay to patients with omphalitis as the primary reason for admission, because we would be unable to determine if findings were related to omphalitis or an alternate diagnosis among patients who were admitted for a different reason (eg, birth hospitalization for extreme prematurity).

Presenting signs and symptoms were collected from admission history. Fever was defined as caregiver report of elevated temperature or documented temperature of ≥38.0°C. Antibiotics were categorized as antimethicillin resistant Staphylococcus aureus (MRSA) if they included trimethoprim-sulfamethoxazole, clindamycin, vancomycin, or linezolid. Days of antibiotic therapy for omphalitis were identified by chart review to disentangle antibiotic days for other indications.

Outcomes included complications of omphalitis, 30-day related cause revisit, and death. Complications of omphalitis included diagnosis of necrotizing fasciitis, peritonitis, intestinal gangrene, liver abscess, portal vein thrombosis, or septic umbilical arteritis on the basis of literature review.^5,7,8 

Stata version 16.0 (Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC; 2019) was used for all analyses. This study was approved by the organization’s institutional review board.

Ninety-one patients with ICD-9/ICD-10 codes for omphalitis were identified, 78 were included (47 omphalitis as primary reason for admission). There was an average of 5.2 omphalitis encounters per year. Most patients (96.2%) were 28 days of age or less (Table 1).

The majority of patients admitted primarily for omphalitis (n = 47) presented with rash (93.6%). Fussiness/irritability was present in 40.4%, and fever was seen in 12.8% of patients. The median white blood cell count was 12.5 × 10³/uL and C-reactive protein was minimally elevated with a median of 0.4 mg/dL (IQR 0.29–0.85; Table 2). Median length of stay was 4 days (IQR 3–10) for patients admitted primarily for omphalitis (Table 3).

Among all patients, blood cultures were positive in 3 (3 of 78, 3.8%): Streptococcus viridans, Group B Streptococcus, and 1 patient who had multiple organisms (Group A Streptococcus, E-Coli, and coagulase-negative staphylococci [CoNS], which was deemed a contaminant). One additional patient had a positive blood culture for CoNS, deemed a contaminant by the medical team. Two patients had positive urine cultures (Klebsiella and E-Coli). There were no positive cerebral spinal fluid cultures.

The majority of patients had positive wound cultures (70 of 78, 89.7%) with primarily gram-positive organisms. Organisms included methicillin-sensitive Staphylococcus aureus (21), Enterococcus (8), corynebacterium (7), CoNS deemed a contaminant by the medical team (5), E-Coli (5), CoNS deemed a pathogen by the medical team (4), Klebsiella (3), MRSA (2), Group B Strep (2), Group A Strep (2), bacillus (2), organism not specified (2), and 1 each of actinomyces, citrobacter, Haemophilus parainfluenza, peptostreptococcus, Proteus mirabilis, morganelle morganii, and usual skin flora.

Imaging was obtained in 16 (20.5%) patients, identifying urachal remnants in 5 patients (5 of 78, 6.4%) and abscess in 4 patients (4 of 74, 5.4%) (Table 3). Anti-MRSA antibiotics were used in 67.9% of patients. Median IV antibiotic duration in hospital was 5 days (IQR 3–7). Oral antibiotics were initiated in 32.1% of patients before discharge.

No patients had complications of omphalitis or death. Five patients (5 of 78, 6.4%) had a 30-day revisit for a related cause.

In our descriptive study, we found that most patients with omphalitis presented with rash, had unremarkable laboratory studies, and experienced good outcomes. Bacteremia was uncommon and wound cultures identified methicillin-sensitive Staphylococcus aureus, enterococcus, and Corynebacterium as the most-common organisms. There was variability in duration of intravenous (IV) antibiotics. No patients had an omphalitis complication or died.

The majority of omphalitis literature focuses on infants in low-income economy countries where risk factors and resources for management differs from the United States, making comparison with our study challenging. Most US studies are >20 years old and have focused on patients with complications (eg, necrotizing fasciitis), thus may not be applicable to patients with less-severe presentations.^5,7–10 

The majority of patients with omphalitis are young, with 96.2% <28 days of age in our study. Infants 28 days of age or less who present with concerns for an infectious process often undergo laboratory testing for invasive bacterial infections.^11–16  In our study, white blood cells and C-reactive protein were normal or minimally elevated, and bacteremia was uncommon. No patients had meningitis. Our findings, combined with previous reports of a very low rate of invasive bacterial infections in infants with uncomplicated skin and soft tissue infections, suggest a potential opportunity to forego blood and cerebral spinal fluid cultures in patients with omphalitis who are without signs of critical illness.^11,17,18 

The microbiology findings of our study are consistent with previous US studies showing a predominance of gram-positive organisms,⁸  with gram-negative organisms seen in a smaller percentage of patients. MRSA was uncommon in our study; MRSA rates varied regionally and fluctuated over our study period.^19–24  More-recent studies of omphalitis bacteriology in low-income economy countries have identified organisms such as Neisseria spp. or Pseudomonas spp.; however, these findings may not be applicable to high-income economy countries where risk factors (such as peripartum-skilled birth attendants, infection control practices, and cord care) may differ.^3,25  Wound culture results and local antibiotic resistance patterns should be considered when selecting appropriate antimicrobial therapy and antimicrobial stewardship strategies used for inappropriate antibiotic selection.

IV antibiotic duration varied in our study, and approximately one-third of patients were continued on oral antibiotics after discharge. The small sample size in our study precludes the ability to compare outcomes with short (eg, ≤3 days) versus long IV antibiotic courses. Additionally, there were no patients who received exclusively oral antibiotics. Future multicenter, comparative effectiveness studies would be helpful to inform antibiotic route and duration recommendations.

No patients experienced a poor outcome, including complications associated with omphalitis, or death. This differs from previous US literature, which showed up to a 7% mortality rate, primarily from complications such as necrotizing fasciitis.⁸  The lower mortality seen in our study may reflect earlier recognition and management, less complications (eg, necrotizing fasciitis), updated recommendations for cord care, and overclassification and/or increasing antimicrobial options.²⁶  For example, previous US case series included patients from as early as 1967, when current antimicrobial agents with strong gram-positive activity, such as Clindamycin (introduced in 1966), may not have been widely available.⁸ 

There were several limitations to our study. This study at a tertiary children’s hospital may not be generalizable to other settings. Birth records were unavailable and thus variables with potential associations with omphalitis, such as birth weight, membrane rupture, or cord care, were not reliably captured. Patients with readmissions at other health systems or deaths outside the hospital were not captured.

We identified patients on the basis of ICD-9/10 diagnosis codes for omphalitis, and patients may have been misclassified. To limit underclassification, we reviewed cases with diagnosis codes for necrotizing fasciitis but no omphalitis code; however, none of the 3 identified cases had indication of abdominal or umbilical involvement. Furthermore, because omphalitis is a clinical diagnosis, there was the potential for overclassification of cases by providers. For example, 3 patients with a diagnosis code for omphalitis did not receive antibiotics and were excluded from our study. It is possible that some patients who were included in our study did not have infection necessitating hospitalization and/or treatment. Future studies should examine the rate and impact of overclassification of omphalitis.

In our retrospective descriptive study of patients hospitalized at a US tertiary pediatric health care organization over a 15-year period, we described presenting features, resource utilization, bacteriology, and outcomes for patients with omphalitis. Although blood testing, including inflammatory markers or cultures, did not inform management for most patients, wound cultures identified organisms, commonly gram-positive, in the majority of cases. Future studies should use a multicenter or research network approach given low incidence, and explore the possibility for reduction of blood testing, antibiotic exposure, and lengths of hospital stay for infants with omphalitis admissions.

We thank Alicen Spaulding, PhD, MPH, for her contribution in study design and conceptualization.