Pediatric Blood Culture Results for Paenibacillus urinalis at a Tertiary Hospital; 5.5-Year Evaluation

1. Introduction

The name Paenibacillus comes from the Latin Paene (meaning immediately, completely), bacillus; meaning almost completely rod-like bacteria (1). Paenibacillus was originally considered within the genus Bacillus and was defined as a separate genus from Bacillus in 1993 (1,2). Paenibacillus is generally a rod-like, aerobic or facultative anaerobic, endospore-forming, gram-variable staining bacterium. In culture, Paenibacillus can be identified using Matrix-Assisted Laser Desorption Ionization-Time-of-Flight Mass Spectrometry (MALDI-TOF MS) or polymerase chain reaction (PCR) (16s rRNA metagenome sequencing). Before these methods were used, they were defined as gram-positive rod, bacillus, or bacillus-like bacteria.

Paenibacillus is abundant in nature and environmental samples and generally does not contain invasion factors that can lead to invasive infection. Therefore, Paenibacillus growth in sterile areas such as blood is generally considered a contamination. However, in children and adults, it can lead to infections in risk groups (such as prematurity, hydrocephalus, chronic kidney disease, sickle cell anemia, malignancy, and other immunosuppressive conditions). In many of these cases, it may not be clear whether Paenibacillus is the actual cause of the infection (1,2).

There are over 200 identified species of the Paenibacillus genus (3). Over 20 of these species have been identified in human clinical samples (such as Paenibacillus thiaminolyticus, P. alvei, P. dendritiformis, and P. urinalis). While species such as P. thiaminolyticus, P. alvei, and P. dendritiformis have been reported as infectious agents in newborns and infants, a wide variety of Paenibacillus species have been reported as infectious agents in adults (1-6). In the literature, a limited number of cases of P. urinalis in adults have been reported (4). However, no information has been found regarding P. urinalis infection in children and infants. This retrospective study aimed to clinically evaluate the growth of P. urinalis in sterile site cultures (blood, catheter, cerebrospinal fluid (CSF) in a tertiary reference hospital over a 5-year period.

2. Materials and Methods

In this study, a total of 170 patients under the age of 18 who were hospitalized in the Pediatric Clinics of Bursa Uludağ University (BUU) Faculty of Medicine Hospital, a tertiary referral hospital, between January 2020 and July 2025, and who showed growth of P. urinalis in sterile site cultures were retrospectively evaluated. Approval for the study was obtained from the BUU Health Research Ethics Committee (ethics committee decision dated 03.12.2025 and numbered 2025/993/21-8). Only sterile site (blood, catheter, and CSF) culture growths were considered. Blood samples were inoculated into BACTEC™ Peds Plus/F culture bottles (Becton Dickinson, USA) under aseptic conditions and incubated for 5 days in the BD BACTEC™ FX200 (Becton Dickinson, USA) automated microbiological monitoring system. During the incubation period, the system automatically assessed the presence of growth by continuously monitoring the carbon dioxide levels resulting from microbial metabolism in the vials. After five days of incubation, passages were made from vials showing growth (positive) onto blood agar and eosin methylene blue agar under sterile conditions. These media were then incubated at 37 °C for 24 hours under atmospheric conditions. Following incubation, the developing colonies were identified using a Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry (Bruker Daltonics, Germany) system. Since P. urinalis growth is generally considered contamination, standard culture antibiograms were not performed. A second culture was taken from cases where culture growth was reported. Five cases with second growth were considered significant, and antibiotic treatment was appropriately determined empirically. In four of these five cases, no P. urinalis growth was observed in the third control cultures. In cases with Paenibacillus growth, simultaneous growth of other bacteria was also recorded. The culturing time (time to positivity) of the first and second positive cases (5 cases) was extracted from the records. To determine the source of Paenibacillus growth, an outbreak analysis study was conducted by the Hospital Infection Control Committee in 2024 and 2025. Within this framework, skin swab cultures, environmental cultures, sterile syringe and antiseptic solution culture samples were taken from hospitalized patients in clinics where growth was observed and evaluated.

Descriptive statistics were used to analyze P. urinalis growth rates based on the ward of admission and patient age. To evaluate potential differences in Time to Positivity between the initial positive cultures and the subsequent control cultures obtained from the same patient at different time points, paired comparisons were performed. Given the small sample size and the inability to robustly assume normality, the nonparametric Wilcoxon signed-rank test was employed instead of the parametric paired-samples t-test. A p-value of less than 0.05 was considered statistically significant. All analyses were conducted using IBM SPSS Statistics for Windows, Version 22.0

3. Results

Over a five-year period, a total of 170 hospitalized children showed growth of P. urinalis in sterile area cultures (blood; 162, catheter; 7, CSF; 1) (Table 1). Additionally, one case (13 days old) showed growth in conjunctival culture, which was not included in this study. The distribution of cases with P. urinalis growth according to age groups and wards is presented in Table 1 and Table 2, respectively. More than one-third of the cases (34.1%) were ≤3 months old, and more than half (51.1%) were <2 years old. More than a quarter of the cases (28.8%) were detected in the neonatal intensive care unit. 40% of all P. urinalis growths occurred in only 3 wards (neonatal intensive care unit, hematology, oncology) (Table 2). Except for one case, all growths were blood-derived (95.2% blood, 4.1% catheter). No simultaneous blood culture growth (P. urinalis) was detected in catheter cultures. In 7 of the cases with initial Paenibacillus growth (7/170; 4.1%), simultaneous co-growths (3 Staphylococcus capitis, 1 Staphylococcus hominis, 1 Staphylococcus haemolyticus, 1 Staphylococcus cohni, 1 Corynebacterium striatum) were detected. In 5 of the 170 cultures (2.9%), P. urinalis growth was again detected in the control culture. In these cases with a second P. urinalis growth, no accompanying co-growth was detected in the culture. In four of these 5 cases, P. urinalis growth was not detected in the third control culture, while Streptococcus mitis grew in one case. One case was discharged before a third culture could be taken.

In cases with recurrent growth (5 cases); the mean time to positivity in the culture of the first P. urinalis growing cases was 28+18 hours (mean + SD, min-max; 12-58 hours), while in cases with a second growth, it was 24+7 hours (mean + SD, min-max; 17-33 hours) (p=0.4).

91.1% (155/170) of the cases before the first culture, were taking antibiotics. Of the cases taking antibiotics, 126 (74.1%) were taking antibiotics expected to be effective against Paenibacilli (such as cefotaxim, ceftriaxone, cefepim, meropenem, ciprofloxacin). 15 (8.8%) patients were not taking antibiotics during the period of the first culture growth; and no growth was detected in the control culture of any of these patients. Antibiogram was not performed because P. urinalis growth was reported as contamination. The initial P. urinalis growth was not considered as an indication in the treatment change for the patients. However, in 5 cases where a second growth occurred and which could be clinically significant, the treatments were changed to cefotaxim or meropenem, taking into account the literature. No patients were lost.

The 23-day-old infant with CSF growth, had clinical findings consistent with sepsis/meningitis, with CSF values of glucose 46 mg/dl, protein 109 mg/dl, rare leukocytes and no bacteria on direct Gram staining. To determine the source of P. urinalis infection, assessments were conducted under the responsibility of the Hospital Infection Control Committee as part of an outbreak analysis, two times one year apart. For this purpose, environmental cultures were taken from different pediatric wards (in July 2024 and July 2025). In this context, P. urinalis growth was detected in 33% (15/45 samples) of samples taken from clean, packaged linens before use from different pediatric wards (neonatal intensive care unit, pediatric intensive care unit, pediatric hematology, pediatric oncology, pediatric health clinic, laundry, etc.). Considering room environmental samples, P. urinalis was not detected in any of the environmental room samples taken from 4 different wards (neonatal intensive care unit, pediatric intensive care unit, pediatric hematology, pediatric oncology). In ambient air cultures (neonatal intensive care unit, pediatric clinic, laundry, operating room, etc.), P. urinalis growth was observed in 53.8% (7/13 air samples). No growth was detected in a total of 15 different sterile syringes in the neonatal intensive care unit, pediatric oncology, and pediatric clinics. No growth was found in detergents and additives used in laundry. No growth was detected in antiseptics containing alcohol, iodine, and chlorhexidine used in the area. Of 20 skin swab cultures taken from 10 patients in the neonatal intensive care unit and pediatric clinics, 65% (13/20) showed growth of P. urinalis. While 60% of swabs were positive after disinfection with 2% chlorhexidine/70% isopropyl alcohol before collection, growth was detected in 70% (7/10) of swab samples taken without disinfection.

4. Discussion

The Paenibacillus genus is very widespread in nature (soil, plant roots, humans, animals, environmental samples, etc.) (1). Many Paenibacillus species can secrete substances such as antimicrobials, pesticides, plant growth-promoting factors, and probiotics. Therefore, it is a type of bacteria that may also be important from an agricultural point of view (1,2,3). Paenibacillus can contain enzymes that cause spoilage of milk and dairy products and can cause their deterioration. In raw milk, the vast majority (>95%) of bacteria that grow during its natural shelf life in the refrigerator (e.g., after 10 days) are Paenibacillus. Paenibacillus endospores can survive in harsh environmental conditions (high heat, pressure, biocides, ultraviolet radiation, pasteurization, etc.). Therefore, they can be detected in raw and pasteurized milk (7). Their presence in environmental samples may increase the risk of contamination in a hospital setting. Paenibacillus infections are generally opportunistic infections and tend to occur in high-risk groups and immunocompromised individuals.

Paenibacillus infections in infants have been primarily identified in recent years due to advances in bacterial identification diagnostic tests. With classical methods where bacterial identification was insufficient, the Paenibacillus genus was identified as bacillus species. MALDI-TOF MS and rRNA genetic evaluations have made Paenibacillus identification clearer and allowed for species-level identification.

Growth of Paenibacillus spp. in sterile site cultures may indicate contamination. However, it is known that many non-invasive bacterial agents can be the actual cause of infection in risk groups and immunocompromised individuals. Paenibacillus, particularly P. thiaminolyticus, can cause sepsis and meningitis in newborns and can lead to adverse neurological sequelae, including mortality (1,2,5,6). In a prospective study of term neonatal sepsis in Uganda, PCR detected P. thiaminolyticus (70%) or Paenibacillus spp (30%) in 6% (37/631) of clinical neonatal sepsis cases, 30% of which showed poor prognosis, and 14% died within the first year. Approximately 20% of survivors had neurological sequelae, including post-infectious hydrocephalus (6). However, there is no literature data yet indicating that P. urinalis is a serious clinical infection agent in children (including neonates).

A recent large-scale study evaluated 177 Paenibacillus infections in children and adults across 40 articles/reports (2). The literature review in this study identified a total of 38 infections caused by 23 Paenibacillus species in adults. Adult infections were generally seen in risk groups, clinical findings varied, and the prognosis was generally good (only 2 mortality, approximately 4%). No particular Paenibacillus type was observed in adults (2). In infants, infections generally presented with sepsis, meningitis, and hydrocephalus. Almost all cases in infants developed before the age of 3 months. In infant infections, some types such as P. thiaminolyticus (79%), P. alvei (1%), P. dendritiformis (1%), were particularly prevalent. And other species that could not be distinguished but were only detected at the genus level (19%). All newborns and young infants with Paenibacillus infection presented with a severe infection picture (meningitis, sepsis, hydrocephalus, severe cerebral destruction, etc.) and a high mortality rate (17%) (2). The authors’ interpretation is that Paenibacillus infections in newborns and infants have a different etiology and a worse prognosis than in adults (2). Furthermore, these data suggest that some Paenibacillus species may present with different clinical manifestations and carry a higher risk of being more invasive, depending on the age group.

There are over 200 identified species of the Paenibacillus genus. Over 20 of these species have been identified in human clinical samples (3). While P. thiaminolyticus, P. alvei, and P. dendritiformis have been reported in children and infants, various species have been reported in adults. In the literature review, data on P. urinalis infections are very limited. It was first isolated and identified in 2008 from the urine of a woman without signs of urinary tract infection (8). In a tertiary hospital in Korea, P. urinalis growth was detected in blood cultures of adults aged 21-76 years, who are generally in the risk group (osteosarcoma, intracerebral hemorrhage, sinusitis treated with acupuncture, subarachnoid hemorrhage, aspiration pneumonia, etc.). With antibiotic treatment (such as moxifloxacin, levofloxacin, ceftriaxone, piperacillin-tazobactam), fever subsided within days and clinical improvement was observed. The authors considered most cases as contamination (4). No cases of clinical P. urinalis infection in children were found in the literature review. Generally, the presence of positive blood and CSF cultures provides evidence in determining the etiology. Even if blood cultures are positive, in some cases, agents thought to be non-invasive may be considered as contamination. However, in risk groups and immunocompromised individuals, the decision of contamination should be made with caution. Positive double-arm blood cultures and/or simultaneous growth or early growth signal (<24-48 hours) in peripheral blood-catheter samples suggest a true infectious agent rather than contamination.

The vast majority of our cases were hospitalized infants and children at risk. Initial growths were generally considered contamination, and no antibiotic change was made. There are currently no species specific EUCAST clinical susceptibility breakpoints for P. urinalis. This organism is generally not listed in the breakpoint tables (9). For Paenibacillus species, susceptibility testing is usually performed by broth microdilution MIC methods, or sometimes, generally interpreted using breakpoints of some similar bacteriae (such as bacillus spp.). However, in 5 cases where the same agent grew a second time, despite the absence of an antibiogram, it was considered a possible agent, and regarding the literatüre recommendations, empirically susceptible antibiotics (such as cefotaxim/ceftriaxone, cefepim, meropenem) were started. In cases with a second bacterial growth (approximately 3%), the time to positivity was found to be an average of 24 hours (17-33 hours). No patient experienced a third bacterial growth, and no mortality was observed. Similarly, bacterial growth in one of our cases, where CNS (central nervous system) infection was suspected and pleocytosis was detected in the CSF, may indicate a possible infection, but it had a good prognosis. In this context, it can be said that even if P. urinalis is the causative agent in the infection, it is not invasive and has a good prognosis.

In Paenibacillus infections, the source of the causative agent may not be clear. One study suggested that it may originate from non-sterile practices involving the umbilical cord of newborns (6). As a risk factor, one study in US, noted that neonatal Paenibacillus infections were seen in premature babies or babies of mothers using narcotics (2). Furthermore, it has been suggested that preparations containing Paenibacillus as probiotics should not be used in newborns until safety data is proven (2). In a study conducted in Uganda, Paenibacillus was not detected by PCR in maternal blood, vaginal, placental, and cord blood samples of infants who developed Paenibacillus infection. These data suggest that neonatal Paenibacillus infections may not be maternally related (5).

The genus Paenibacillus is very common in nature (soil, plant roots, humans, animals, environmental samples, food, etc.) (1). Paenibacillus endospores can survive in harsh environmental conditions (high heat, pressure, biocides, ultraviolet radiation, pasteurization, etc.) (7). They can be detected in food (plant roots, milk and dairy products, whether pasteurized or not) (7). Their easy presence in environmental samples can increase the risk of contamination in a hospital setting. In our study, under the guidance of the Hospital Infection Control Committee, two-stage environmental and screening cultures were performed to investigate the source of possible P. urinalis contamination. P. urinalis growth was not detected in sterile syringes, antiseptics such as alcohol, iodine, and chlorhexidine, detergents used in room cleaning and laundry, or environmental samples taken from rooms. On the contrary, P. urinalis growth was found in 33% of clean packaged sheet samples, 53% of room air samples, and 65% of skin swab cultures from hospitalized patients. These data suggest that the primary factor in blood culture contamination is infected clean linens and associated airborne transmission. This can be due to the Paenibacillus surviving capasity in hard conditions. The fact that more than half of the skin swab cultures from hospitalized patients showed growth supports the possibility of blood culture contamination via skin contact. Pre-treatment of skin swabs with 2% chlorhexidine/70% isopropyl alcohol (60% growth) did not significantly affect P. urinalis growth compared to pre-treatment (70% growth). This supports the idea that blood cultures can be contaminated despite chlorhexidine/alcohol disinfection due to skin colonization. P. urinalis is not considered among the commensal bacteria of the skin. However, its detection at high levels in our cases (even after routine skin disinfection) suggests that P. urinalis is a bacterium that can easily adapt to the skin. Although skin colonization is thought to be secondary and transient due to linen or airborne contamination, no studies have been conducted on the duration of colonization.

The optimal treatment for Paenibacillus infections is unclear. They are generally resistant to penicillin, ampicillin, vancomycin, and macrolides. In most human cases, Paenibacillus isolates are considered susceptible to third-generation cephalosporins (such as ceftriaxone, cefotaxim), aminoglycosides such as meropenem and gentamicin, and new quinolones (such as moxifloxacin, levofloxacin) (2,6,4). In antibiograms performed for Bacillus spp., meropenem is considered susceptible if its minimal inhibitory concentration (MIC) value is below 0.25, ciprofloxacin if its MIC value is below 0.001, and vancomycin if its MIC value is below 2 (9). Since no antibiograms were performed in any of our cases, regarding the EUCAST recommendations, we could not determine susceptibility. However, the administration of antibiotics reported to be potentially susceptible (such as cefotaxim/ceftriaxone, cefepim, meropenem) after the second culture may explain the absence of growth in the third cultures.

Limitations of the study: The fact that it is a single-center, retrospective study and that no antibiogram was performed can be considered among the limitations of the study. In cases with a second culture, the recurrence of the same pathogen, the shorter TTP (Term-Time Transition) supporting true growth, and the finding of improvement with empirical treatment in cases with a second culture, as well as the absence of a negative outcome in a case discharged without treatment, may also support the possibility of a second contamination.

Strengths of the study: This study is, to our knowledge, the largest study evaluating positive blood culture growth using the MALDI-TOF MS method, conducted only in pediatric clinics over a relatively long period (approximately 5.5 years). The distribution of cases according to age group and wards, and the determination of average TTP times can be considered advantages of the study. Furthermore, within the framework of a two-stage outbreak analysis, evaluating/demonstrating the causes of blood contamination (such as the correlation between the detection of the pathogen in bed linens and room air, consistent with the characteristics of P. urinalis) can contribute to the literature in terms of preventing possible contaminations. Additionally, the fact that P. urinalis is frequently detected on the skin of hospitalized patients (not mentioned before), despite not being a commensal bacterium, and possibly its long-term presence/adaption in the skin flora, can be considered among the strengths of this study in terms of contributing to the literature.

In conclusion, P. urinalis blood culture growth is primarily considered a contamination. However, these growths should be carefully evaluated in risk groups. In clustered growths in hospitalized patients, environmental and skin colonization may play a role. So, preventive measurements should be taken taking into account of the P. urinalis specialities.