Diagnostic Performance of Cerebrospinal Fluid Cell Ratio and Cell Index for Extra-Ventricular Drain-Associated Ventriculitis

When a patient suffers a severe brain injury or a bleed inside the skull, doctors often insert a thin, flexible tube called an external ventricular drain. This tube acts as a safety valve, siphoning off excess fluid to relieve dangerous pressure on the brain. However, because this tube creates a direct path from the outside world into the deepest part of the brain, it carries a risk of infection. This infection, known as ventriculitis, is a serious complication that can lead to permanent damage or death. The challenge for medical teams is that the symptoms of this infection are often vague and can be mistaken for the patient's original brain injury or the effects of sedation. Furthermore, the standard test to confirm an infection—growing bacteria from a fluid sample in a lab—can take days, and sometimes the test comes back negative even when an infection is present. In the meantime, doctors are left guessing whether to start powerful antibiotics, which carry their own risks if used unnecessarily.

To solve this puzzle, researchers at the Prince of Wales Hospital in Hong Kong looked back at nearly six hundred patients who had received these drains over a fifteen-year period. They focused on a specific problem: when a patient has a brain bleed, their spinal fluid naturally contains red blood cells. This blood can also bring white blood cells, the body's infection fighters, into the fluid. Consequently, a high count of white blood cells in the fluid does not always mean an infection is present; it might just be a reaction to the blood or the trauma of the surgery itself. The medical team wanted to find a way to tell the difference between white blood cells caused by a simple injury and those caused by a dangerous bacterial infection. They tested two new ways of looking at the data: one method compared the number of white blood cells to the number of red blood cells in the fluid, and the other compared that ratio to what was found in the patient's own blood.

The study found that out of the 598 patients, 44 had a confirmed bacterial infection in their brain fluid. These infected patients faced a significantly higher risk of dying in the hospital compared to those who did not have an infection. When the researchers analyzed the fluid samples, they discovered that simply counting the white blood cells was not a reliable way to spot the infection. The count was often high in patients who were not infected, simply because of the blood in their system. However, the two new methods the team tested performed better than the raw white blood cell count alone. By calculating the relationship between the white and red blood cells, and by adjusting that relationship against the patient's blood levels, the doctors could distinguish between a sterile reaction to injury and a true bacterial invasion with greater accuracy than the traditional count.

The researchers determined that these adjusted calculations were superior to the traditional white blood cell count alone, though the two new methods performed similarly to each other. They identified specific thresholds where the numbers shifted, suggesting that if the ratio of white to red blood cells in the fluid was very high, an infection was likely. However, the study noted that there is currently no consensus on the exact thresholds that best predict the presence of ventriculitis, and the overall performance of these indices was only moderate. While these new methods were not perfect and still missed some cases, they offered a clearer picture than the old standard. The study suggests that doctors should stop relying on the raw number of white blood cells when checking for this type of infection and instead use these adjusted ratios. This approach could help medical teams make faster, more informed decisions about when to start antibiotics, potentially saving lives by treating the infection early without exposing patients to unnecessary medication when no infection exists.