In March of 2017, the Infectious Disease Society of America published novel guidelines to aid clinicians in the diagnosis and management of health- care associated meningitis, ventriculitis and neurosurgical site infections, including cerebrospinal fluid (CSF) shunts/drains, ventriculo-peritoneal (VP), atrial (VA) and pleural (VPL) shunts, intra-thecal pumps, deep brain simulators and head trauma. The guidelines were created with the help of 10 expert panelists, one of them being our own Dr. Karin Byers from the University of Pittsburgh.
CNS infections acquired in the nosocomial setting are associated with different pathogens and clinical/diagnostic presentations when compared to community acquired meningitis, and this can create clinical confusion and sub-optimal management if not addressed appropriately. It was because of these issues that the above guidelines were recently instituted.
So what are the key points in these new guidelines? Similar to the endocarditis guidelines that were presented in 2015, there is a strong emphasis in formulating a multi-disciplinary team, consisting of Infectious Disease specialists, neurosurgeons, neurologists and critical care specialists in order to optimize care for patients who suffer from these infections. It is acknowledged that health-care associated CNS infections can be challenging to diagnose clinically; although it was strongly recommended that symptoms of new altered mental status, headache, seizures, meningismus, lethargy or nausea should raise suspicion for infection. In addition, symptoms of peritonitis or pleuritis in individuals with VP and VPL shunts should also raise concerns for infection, along with any abnormalities associated with tubing sites, including redness, drainage and tenderness over these sites.
As mentioned, CSF analysis in these patients may vary considerably, which can raise confusion as to whether infection is present or not. The panelists felt that the CSF profile alone could not reliably include/exclude infection, and in particular, a negative CSF Gram stain was not enough to reliably rule out infection in these patients. They do emphasize the importance of CSF cultures however, stating that if these cultures are negative and suspicion for infection remains high, then cultures should be held for up to 10 days in order to detect slower growing pathogens such as P. acnes. Cultures growing non-virulent pathogens from a single sample or from broth alone were not thought to be definitive for infection, while the presence of any type of virulent pathogen (Staph aureus, fungi, Gram negatives) was felt to be enough to secure a diagnosis of infection. Overall, a combination of clinically relevant symptoms, CSF pleocytosis (particularly low CSF glucose or rising CSF white count) and positive CSF culture data was most suggestive of an underlying nosocomial infection, regardless of what type was present. For VA shunts, the addition of blood cultures were felt to be diagnostically important, while in other types of infections, routine blood cultures were not necessarily required unless signs of sepsis were noted.
Diagnostic imaging focused primarily on the use of MR imaging with gadolinium in order to most effectively detect changes of ventriculitis, meningitis and intra-cranial abscesses that could be associated with these infections. The additional of abdominal imaging (via ultrasound or CT scan) was recommended for any patients with VP shunts who presented with abdominal symptoms. The use of CSF pro-calcitonin and CSF lactate were thought to be helpful in aiding in the diagnosis of infection, along with serum procalcitonin to help determine if CSF changes noted in a patient who recently underwent CNS surgery were directly from the surgery itself or an underlying bacterial infection, however the data (especially for the utility of procalcitonin) was limited and therefore a definitive recommendation could not be made on the use of these tests. Nucleic acid amplification tests such as PCR, when available, were thought to help with more accurate diagnosis and reducing the time to diagnosis, with a good specificity (93.4%) but poor sensitivity (47.1%). Therefore when cultures remained negative and infection was still highly suspected, obtaining PCR testing was thought to be appropriate to help detect more fastidious organisms. Adding PCR testing to standard CSF cultures increased the overall yield of detection by 25% in one study (Deutch et al. Neurosurgery 2007). The use of CSF galactomannan and beta-D-glucan were thought to be very useful in aiding in the diagnosis of fungal CNS infections.
Treatment focused both on empiric and focused regimens, after culture data was available. Empiric therapy focused on using a combination of vancomycin (covering staphylococci and streptococci) along with an anti-pseudomonal beta-lactam with good CNS penetration (cefepime, ceftazidime, or meropenem); the decision on which drug to use is to be based on the local antibiogram of the institution where the patient is being cared for. For those with true beta-lactam allergies, the addition of aztreonam or ciprofloxacin to vancomycin was suggested, and with those who had vancomycin allergies, the substitution of thus drug with either daptomycin, linezolid or trimethoprim-sulfamethoxazole was suggested, again based on information from local antibiogram data. Targeted therapy was based on switching to the most narrow, effective agent that had good CNS penetration. The guidelines specifically address treatment of methicillin sensitive and resistant Staphylococcus aureus, coagulase-negative Staphylococci, P. acnes, enteric Gram-negative organisms, Pseudomonas, drug resistant Gram-negative organisms (ESBL pathogens and Acinetobacter species) and Candida.
In addition to appropriate antimicrobial therapy, it was strongly emphasized that removal of the infected hardware, drain, shunt or pump, whenever possible, was critical in the overall management of the infection and addressing source control. In general, treatment should continue for 10-14 days, and up to 21 days for certain Gram-negative infections. If a non-virulent pathogen, such as coagulase negative Staph or P. acnes was noted, and the patient had minimal presenting symptoms and a bland CSF profile, a shorter course of 10 days could be considered. If an external ventricular drain infection was being managed, then treatment was to continue for 14 days after clearance of cultures from the CSF.
Lastly, the panelists do make mention of perioperative interventions in order to minimize the risk of infection, focusing mostly on the use of appropriate perioperative antibiotic administration as well as the use of a standardized protocol for shunt and drain placement. The use of antibiotic impregnated CSF shunts and drains were also recommended. Prolonged use of antibiotic prophylaxis was discouraged, and the panelists specifically made mention of the fact that antibiotics do not need to be continued for the duration of extra-ventricular drain placement; in addition, they discouraged the use of ongoing prophylactic antibiotic use in individuals with CSF leaks, and instead recommended repair of a CSF leak that persisted for 7 days or more. The administration of pneumococcal vaccines however was recommended in these patients.
It is hoped that these new guidelines will help shed some light on a this specific sub-set of infections, the incidence of which will continue to be more and more prevalent in our hospitals as newer neurosurgical techniques and technologies develop. As more research is done, these guidelines will be modified accordingly to help optimize the management and care of patients who suffer from these infections as well.
Pubmed link for guidelines: