Deadly Bloodstream Infections Strike One in Four Children with Central Lines, Study Finds

Central venous catheters are among the most indispensable tools in modern pediatric medicine. These thin, flexible tubes, threaded into large veins that lead directly to the heart, deliver chemotherapy, nutrition, antibiotics, and fluids to children whose bodies cannot manage without them. Yet the same lifeline that sustains critically ill children also opens a direct corridor for bacteria and fungi to enter the bloodstream. A new retrospective cohort study from Iran’s largest pediatric referral hospital now offers one of the most detailed pictures to date of just how dangerous that corridor can be, finding that one in four hospitalized children with a central line developed a catheter-associated bloodstream infection, and that more than one in five of those infected children died before leaving the hospital.

The research, published in BMC Infectious Diseases by Hamed Tabasizadeh, Babak Pourakbari, Setareh Mamishi, and colleagues at Tehran University of Medical Sciences, followed 723 children younger than 18 years who carried central venous catheters at the Children’s Medical Center in Tehran between 2021 and 2024. The team combed through four years of routinely collected clinical records, laboratory results, and microbiology reports to answer several intertwined questions: which children were most likely to develop a central line-associated bloodstream infection, which infected children were most likely to die, which organisms were responsible, how quickly blood cultures turned negative after treatment began, and how often the culprit microbes resisted multiple antibiotics at once.

The headline numbers are stark. Of the 723 eligible children, 182, or 25.2 percent, experienced at least one central line-associated bloodstream infection during their admission. Overall, 128 children in the cohort, or 17.7 percent, died in the hospital. Among the children who did develop a catheter-associated infection, all-cause in-hospital mortality reached 22.0 percent. These figures place the burden of catheter-related sepsis in this setting among the most serious complications in pediatric hospital care, and they underscore why infection prevention teams treat every central line as a potential hazard that must be continuously justified and vigilantly maintained.

Where a child was treated mattered considerably. Using modified Poisson regression, a statistical technique that estimates risk ratios directly rather than odds ratios and performs better when outcomes are common, the researchers adjusted for sex, age group, and admission ward. Compared with children in the pediatric or general intensive care units, those in the cardiac ICU and open-heart ICU faced a 35 percent higher risk of catheter-associated bloodstream infection, with an adjusted risk ratio of 1.35 and a 95 percent confidence interval of 1.00 to 1.82. By contrast, children on transplant and oncology wards had a 43 percent lower risk, with an adjusted risk ratio of 0.57 and a confidence interval of 0.37 to 0.88. The authors suggest that ward-specific surveillance should be a priority, since the drivers of infection, from line-handling practices to the intensity of nursing care, differ substantially across hospital units.

To identify which factors were associated with death among infected children, the team turned to Firth penalized logistic regression, a method designed for situations where events are sparse and conventional models can produce unstable or inflated estimates. The model’s performance was internally assessed using bootstrap validation, in which the analysis is repeated across thousands of resampled datasets to check that the estimates hold up. Two factors stood out. Children whose catheters had been placed in the femoral vein, in the groin, had roughly three times the odds of dying in hospital compared with children whose lines were placed elsewhere, with an adjusted odds ratio of 3.04 and a confidence interval of 1.37 to 6.82. Femoral sites are often chosen in emergencies but are widely regarded as more vulnerable to contamination from the surrounding skin flora. The second mortality signal was thrombocytopenia: every halving of the platelet count raised the odds of death by 51 percent, with an adjusted odds ratio of 1.51 and a confidence interval of 1.20 to 1.94. Falling platelets are a classic laboratory marker of worsening sepsis, reflecting both consumption of clotting factors and the inflammatory storm that severe bloodstream infection unleashes.

The microbiology behind these infections revealed a familiar but troubling cast of characters. Klebsiella species, Gram-negative bacteria notorious for their ability to acquire resistance genes, and Staphylococcus epidermidis, a skin-dwelling organism that becomes a pathogen the moment it finds a plastic surface to colonize, predominated among the bloodstream isolates. This pairing captures the two faces of catheter infection: organisms that exploit the device itself and organisms that thrive in the hospital environment. The distribution of organisms also influenced the clinical course. The time to documented culture negativity, the interval between starting treatment and the first follow-up blood culture showing no growth, differed significantly across organism groups, with a P value of 0.002. In the subgroup analysis, 98 children had Gram-negative infections, 58 had Gram-positive infections, and 16 had fungal infections, and the pace at which their bloodstreams cleared varied accordingly. The authors note that follow-up cultures were not protocolized as part of the study, meaning the timing reflected routine clinical practice rather than a standardized schedule.

Perhaps the most alarming finding concerns antimicrobial resistance. Multidrug resistance, defined as non-susceptibility to at least one agent in three or more antimicrobial classes, was detected in 82.0 percent of assessable children with catheter-associated infections. In practical terms, for every five infected children whose isolates could be tested, four harbored organisms that shrugged off most of the standard antibiotic arsenal. Susceptibility analyses were conducted at the isolate level, with denominators varying by antibiotic because only tested isolates contributed to each percentage, while the primary multidrug-resistance prevalence was estimated at the patient level to avoid inflating the figure by counting repeated isolates from the same child. The supplementary matrices of organism-specific non-susceptibility paint a granular picture of which drugs retained activity against which pathogens, information that clinicians in similar settings can use to guide empiric therapy.

The methodological choices in the study deserve attention because they shape how the findings should be interpreted. Modified Poisson regression with heteroskedasticity-consistent standard errors allowed the team to report adjusted risk ratios that are easier to grasp clinically than odds ratios. Firth’s penalized approach tamed the small-sample instability inherent in mortality modeling among a subset of 182 children. Bootstrap validation provided a check on model performance, assessed through measures such as the area under the receiver operating characteristic curve. The researchers also handled a subtle confounding problem carefully: concurrent mechanical ventilation and inotropic support were recorded as secondary severity markers but kept out of the primary mortality model, because the timing of these interventions relative to infection onset was unavailable and they could just as easily be consequences of infection as causes of death.

As with any single-center retrospective study, there are limits. The data come from one tertiary referral hospital in Tehran, which concentrates the sickest children, including transplant recipients, oncology patients, and post-cardiac-surgery cases, so the infection and mortality proportions may not generalize to community hospitals or to settings with different resistance epidemiology. The requirement for informed consent was waived by the institutional review board because the analysis used routinely collected clinical data, and the study received no specific external funding. The authors declare no competing interests. Still, the consistency of the signals, higher risk in cardiac units, femoral catheters and thrombocytopenia marking mortality, Klebsiella and Staphylococcus epidermidis dominating the isolates, and multidrug resistance in more than four fifths of assessable cases, gives the findings weight.

The study’s conclusions translate into a clear agenda for pediatric hospitals everywhere. Ward-specific surveillance should identify the units where catheter infections cluster. Reliable insertion and maintenance practices, from sterile technique at placement to meticulous dressing changes and hub disinfection, remain the foundation of prevention. Regular review of line necessity, asking every day whether each catheter still earns its place in a child’s vein, shortens exposure and cuts risk. And antimicrobial stewardship programs should be locally informed, built on each hospital’s own susceptibility data rather than generic guidelines, because in a setting where 82 percent of infecting organisms are multidrug resistant, guessing wrong about empiric therapy can be fatal. For the children whose lives depend on central lines, these measures are not bureaucratic box-ticking; they are the difference between a lifeline and a liability.

Subject of Research: Risk factors, mortality, and antimicrobial resistance in central line-associated bloodstream infections among hospitalized children

Article Title: Central line-associated bloodstream infection in hospitalized children: a retrospective cohort study of risk, mortality, and antimicrobial resistance

Article References: Tabasizadeh, H., Pourakbari, B., Sotudeh, M., Sadeghi, R. H., Farahmand, M., & Mamishi, S. (2026). Central line-associated bloodstream infection in hospitalized children: a retrospective cohort study of risk, mortality, and antimicrobial resistance. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14528-4

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14528-4

Keywords: central line-associated bloodstream infection, central venous catheter, pediatrics, sepsis, antimicrobial resistance, multidrug resistance, Klebsiella, Staphylococcus epidermidis, mortality, risk factors, intensive care unit, Iran