A Culture of Too Many Blood Cultures

A 59-day-old girl, born at 33 weeks gestation for maternal indications, received her routine 2-month immunizations at a well-child visit. She was diagnosed with anemia of prematurity but was otherwise healthy. Several hours following immunizations, she was noted to be fussy, and had 1 episode of nonbloody, nonbilious emesis following a feed. Her axillary temperature at the time was 100°F (37.8°C). Her parents notified her primary care pediatrician’s (PCP) office and was instructed to bring her to the emergency department (ED) given her age and temperature.

In the ED, the infant was well-appearing. Vital signs included a temperature of 37.6°C, a pulse of 151, a respiratory rate of 32, and an oxygen saturation of 99% in room air. Physical examination was unremarkable. Laboratory evaluation showed a white blood cell count of 14.1 k/uL, absolute neutrophil count of 7.46 k/uL, hemoglobin of 9.8 g/dL, hematocrit of 28.9%, platelets of 529 k/uL, C-reactive protein of 0.9 mg/dL, and procalcitonin of 0.29 ng/mL. The complete metabolic panel and urinalysis were within normal limits; blood and urine cultures were pending. The infant was discharged from the hospital given well-appearance and reassuring workup, with recent immunizations as the likely source of her symptoms.

The following morning, the infant continued to be well without fevers or emesis. The blood culture obtained in the ED resulted positive for gram-positive cocci at 19 hours following collection. The family was notified, and the infant was subsequently admitted to the hospital. Admission laboratory evaluation included a repeat blood culture and complete blood count, complete metabolic panel, C-reactive protein, and procalcitonin, which were reassuring and unchanged from the prior blood culture. The infant was well-appearing with normal vital signs. She was started on empirical intravenous antibiotics for potential bacteremia. The original blood culture speciated to methicillin-susceptible Staphylococcus aureus, and the repeat blood culture remained negative.

Following shared decision-making with the infant’s parents, antibiotics were discontinued at 48 hours of no growth on repeat blood culture, and the infant was discharged from the hospital. The positive blood culture was postulated to be a contaminant, and the infant’s original symptoms to be because of recent immunizations. The infant did well following discharge without return to the hospital.

Fever in the newborn period is typically defined by a temperature >38.0°C and is a common cause of evaluation in the inpatient and outpatient settings.¹  Infants who develop fever, or temperatures near 38.0°C, at <3 months of age are of particular concern to clinicians for their risk of invasive bacterial infection (IBI), defined as bacteremia and meningitis, and oftentimes undergo diagnostic testing.^1–4  Fever is frequently reported following immunization and occurs somewhere between 10% to 70% of infants within the first 3 days of immunization.^2,5–8  Existing literature on infants with postvaccination fever have found no cases of IBI in this population.^9–12  The recently published febrile infant guidelines currently excludes those immunized within the prior 48 hours since the risk of IBI is low in this subset of patients because of vaccine eligibility age and postimmunization status.^2,12–14 

Despite this, approximately 8% of infants with postimmunization fever sought medical care¹⁵  with >50% having at least 1 investigation performed^4,9–11  and 30% to 50% admitted to the hospital for observation.^4,9,10  Whereas the event of postimmunization fever is well known, there is no clear guidance on how providers should approach these infants. In this article, we look at 3 major points of decision making (the decision to pursue medical evaluation, the variation in practice interventions, and the implications of a positive blood culture) that may contribute to low-value care.

In a survey of >1600 families of 1 to 8 month old infants receiving routine immunizations, fever was the symptom most strongly associated with medical utilization.¹⁶  In a survey of more than 400 US pediatricians, hospital admission or ED visits were more likely recommended for younger infants (0–2 months) or a higher body temperature (>39.0°C).¹⁷  ED evaluation was less likely recommended for infants who developed fever within 48 hours of immunization,¹⁷  with potential exceptions for infants who are born prematurely, immunocompromised, or have known medical complexity.¹⁸  Prematurity has been associated with increased risk of infection and may have prompted further evaluation of our patient.^19,20 

There is significant variation across hospitals in the evaluation of febrile infants from 1 to 3 months, highlighting an opportunity to improve resource utilization.²¹  The recently published national guidelines aim to standardize the evaluation for febrile infants <60 days, though these guidelines exclude infants who are premature or have recently received immunizations.¹³ 

Multiple algorithms have been proposed to risk stratify the 1 to 3-month-old febrile infant. One recent algorithm, the sequential “step-by-step” laboratory evaluation, does not exclude infants with prematurity or recent immunization and has been shown to have higher sensitivity for identifying infants at low risk of IBI compared with older algorithms.³  This is because of the incorporation of newer blood tests, changes in bacterial pathogen epidemiology, and decreased rates of IBI with the advent of childhood vaccines.^1,18,22–26  Physician assessment of an ill-appearing child has also been found to be a valuable clinical predictor of IBI, with nearly 75% sensitivity and specificity for serious illness.^6,18,24  With her reassuring exam and laboratory findings, our patient is at low risk of IBI and suitable for outpatient management based on the step-by-step algorithm.³  Generally, providers can be reassured by a well-appearing infant with postimmunization fever and routine investigations are not warranted. The use of a validated risk stratification tool may help guide the need for further intervention. However, if there are concerns, well-appearing infants with postimmunization fever may be safely observed in the hospital without the need for invasive interventions.^10,14 

When a report of a positive blood culture result is received, physicians are faced with the dilemma of how to best assess the risk of bacteremia in a well-appearing child.²⁷  Peripheral blood culture contamination rates can be up to 10% of all collected cultures.^28–30  For children with positive blood culture results, 30% to 50% are found to be contaminants rather than true pathogens.^30–34  Contaminants are even more likely to be found in children <12 months old and in blood cultures collected in the ED.³⁵  These false positive blood cultures account for significant burden on additional hospital days, resource utilization, and unnecessary antibiotic therapy.^29,34 

The classification of true bacteremia from a positive blood culture varies depending on the type of pathogen isolated, number of colonies grown, time to positivity, and clinical context of the patient.^29,30,32  Whereas some organisms are typically deemed pathogens and others contaminants, there is a lack of uniformity with Staphylococcus aureus and the context (clinical appearance, presence of fevers, subsequent cultures, response to antibiotics) must be considered.^29,30,32  Our patient’s well appearance, lack of fevers, and subsequent negative culture led us to reason that her bacterial growth was likely a contaminant.

When a positive blood culture results, families are typically instructed to see a medical provider.²⁷  In a study of 79 febrile children aged 3 to 36 months with false positive blood cultures, 60% were seen exclusively by their PCP and 40% were seen in the ED. Of the patients seen by a pediatrician alone, none had additional laboratory studies performed and the majority (67%) had no antibiotics prescribed.²⁷  In contrast, those evaluated in the ED commonly had additional diagnostic testing, hospital admission, and parenteral antibiotic treatment.^27,32,34  Given that clinical history and physical examination can guide need for further intervention, 1 potential solution to improve the care of these infants is for primary evaluation in the PCP clinic.

Pursuing an extensive work up in a well-appearing infant with postimmunization fever may contribute to low value care, and any laboratory testing obtained has the potential to cause harm to the patient and be financially costly.¹⁸  There was >3-fold increase in work-pay losses among families seeking medical care for their child with fever after immunization.¹⁶  When reviewing febrile infants at low risk for IBI, patients who presented for outpatient evaluation saved >$3000 per patient compared with an inpatient observation, with no serious adverse outcomes in the outpatient group.³⁶  Moreover, each contaminated blood culture incurred an excess of $2800 in hospital charges.^27,28 

Many patients seek medical care for postimmunization fever without comprehension of the potential downstream consequences: from pain caused to the child, unnecessary interventions, hospital admission, to significant financial liability. With multiple points in the decision-making process, it is critical that we as providers include our families into conversations regarding diagnostic testing, hospital admission, and disposition. These are important opportunities to practice shared decision-making and high value care to reduce harm and costs to the patient.

It is important to recognize the low risk of IBI in otherwise well-appearing infants ages 1 to 3 months with postimmunization fever. We should consider the risks and benefits of evaluation, particularly as we shift to evidence-based approaches to perform fewer interventions for even younger febrile infants in the era of the new national guidelines. Whereas the value of early recognition and treatment of IBI cannot be discounted,³⁷  the sensitivity of our risk stratification tools and clinical judgement allows us to carefully consider how to minimize low-value care in the evaluation of well-appearing febrile infants with recent immunization.