Isolated Bandemia: What Should We Do with It?

Christine DeFranco DO and Terrance McGovern DO MPH
St. Joseph’s Regional Medical Center, Paterson, NJ

Almost immediately, emergency physicians know what to do with most abnormal lab results.  If a patient’s hemoglobin is 4.5 g/dL, they need a transfusion.  If their troponin is 12.3 ng/ml, they are having cardiac infarction.  If their lactic acid is 7.5 mmol/L, something terrible is going on.  What if the only abnormality they have on their blood work is bandemia?  Does it matter if their bands are 6%, 15%, or 32%? Or, are they all treated equally?

We recently had a case that challenged us. The patient was a 48-year-old male discharged from a nearby hospital two days prior for pancreatitis.  After experiencing continuous pain, the patient decided to come to the emergency department (ED) for evaluation. The patient’s physical exam was not very impressive with only mild epigastric abdominal tenderness.  He was afebrile, normotensive, and not tachycardic.  His blood work revealed only elevated bands of 24% with a normal WBC at 9.8K/mm3, normal complete metabolic profile, and a steadily decreasing lipase now at 102 U/L.  After receiving intravenous fluids and toradol for pain, the patient was feeling well and eager to be discharged. At this juncture, we were presented with the question: do we admit the patient or discharge him with close outpatient follow-up? We decided to discharge the patient.

In the past, there was a notion that band counts were not very useful in identifying infections.1 Bands are not the most specific indicator for infection because they can be elevated for many different reasons: seizures, toxic ingestions, metabolic abnormalities, inflammatory processes, and tissue damage.2 Drees et al. sought to determine the utility of bandemia in patients with normal white blood cell counts in a retrospective cohort study in 2012.  They separated their sample of patients who were admitted with normal white blood cell counts (3,800-10,800 per mm3) into either normal (≤10%), moderate (11%-19%) or high (≥20%) bands.2 After adjusting for age and vital signs, they found a significant association of increased bands with the rate of positive blood cultures and in-hospital death when compared to patients with normal band counts, (Table 1).

Table 1. Adjusted odds ratio (OR) for bandemia and positive blood cultures and in-hospital death2

Adjusted OR (95% Confidence Interval)
Positive blood culture
Moderate bands (11-19%)3.8 (2.0-7.2)
High bands (≥20%)6.2 (3.2-11.8)
In-hospital death
Moderate bands (11-19%)3.2 (1.7-6.1)
High bands (≥20%)4.7 (2.4-9.0)


Identifying and treating sepsis early in its course is essential for decreasing its morbidity and mortality.3 There is a desire to determine an early marker that we could use to identify septic patients as early as possible, hopefully even in the ED setting.  Could this marker be bandemia?  In a post hoc analysis of previously collected data on 289 patients that presented to the ED that had positive blood cultures, 80% of the patients had bandemia with a normal temperature and 79% had bandemia with a normal WBC.4 They suggest that bandemia (>5%) may be an early indicator of patients with an occult infection.4

While identifying patients that may end up having positive blood cultures is important, arguably more important are the patient related outcomes associated with bandemia.  In 2015, Shi et al looked at patients who were discharged from the ED that had bandemia (>10%).5 They evaluated whether the patients that were discharged from the ED had any negative clinical outcomes, defined as a return to the ED within seven days or death within 30 days.  For patients that had band counts >30% there was a five-fold increase in the rate of death at 30 days but no significant change in revisits to the ED within seven days when compared to patients that had band counts ≤ 30%.5

In retrospect, our disposition of the patient should have been admission. The patient returned to the ED less than 12 hours later with hemorrhagic pancreatitis and was intubated during his inpatient stay. Fortunately, the patient had a good prognosis.  This is an example of only one case, but it highlights the importance of bandemia, whether it is the only abnormality or not.  Having bands >10% should make all Emergency Physicians hesitant to discharge these patients, despite their clinical appearance, and lean towards an inpatient admission with early initiation of empiric antibiotics.


  1. Andron MJ, et al. Band neutrophil counts are unnecessary for the diagnosis of infection in patients with normal total leukocyte counts. Am J Clin Path. 1994;102:646-649.
  2. Drees M, et al. Bandemia with normal white blood cell counts associated with infection. The American Journal of Medicine. 2012;125:1124.e15-114.e15
  3. Gaieski DF, et al. Impact of time to antibiotics on survival in patients with severe sepsis or septic shock in whom early goal-directed therapy was initiated in the emergency department. Crit Care Med. 2010 Apr;38(4):1045-53.
  4. Seigel T, et al. Inadequacy of temperature and white blood cell count in predicting bacteremia in patients with suspected infection. The Journal of Emergency Medicine. 2012;42(3):254-259.
  5. Shi E, et al. Clinical outcomes in ED patients with bandemia. American Journal of Emergency Medicine. 2015;33: 876-881.
2017-07-13T13:29:21+00:00 October 13th, 2016|Critical Care, Fall 2016|

Leave A Comment