26.2 Infection, CSF Leak, and Pneumocephalus (04.B, E, F)

Air, fluid, and bacteria use the same hole. Independent OpenExamPrep teaching for outline items 04.B, 04.E, and 04.F is to separate wound and bone-flap infection, meningitis and abscess, CSF leak, and pneumocephalus so you do not treat Mount Fuji air with a BiPAP humidifier or a late infected flap with “another week of cephalexin.” Healthcare-associated meningitis drug choices overlap chapter 23.4; the perioperative move is source control and not making the air collection larger.

Wound infection, bone-flap infection, meningitis, and abscess

A surgical-site wound infection in the first 1–2 weeks is erythema, drainage, and fever at the incision. Superficial cellulitis can be antibiotics plus local care. Deeper infection tracking under the galea, through a wet suture line, or into a subgaleal collection needs a surgeon, not only a culture swab left on the dressing.

Bone-flap osteomyelitis is a slower clock. The avascular flap sits as a foreign body. Patients return weeks to months later with a boggy scalp, wound breakdown, low-grade fever, or a sunken-then-swollen flap. CT may show peri-flap collections; MRI helps when you need to know whether pus is epidural or in the bone. The usual definitive move is remove the infected flap, debride, culture bone and fluid, give a prolonged intravenous course (often on the order of 6 weeks, then delayed cranioplasty once the wound is clean). Leaving a frankly infected flap in and “covering” it is the board wrong answer.

Postoperative meningitis after craniotomy, transsphenoidal surgery, or a persistent leak presents with fever, headache, declining examination, and a CSF profile that is hard to read because postoperative blood already inflames the fluid. Still sample: Gram stain, culture, cell count, glucose, protein. Empiric therapy follows IDSA 2017 healthcare-associated ventriculitis and meningitis guidance: vancomycin plus cefepime, ceftazidime, or meropenem. Community ceftriaxone-only meningitis doses are the wrong gram-negative coverage after a craniotomy. Dexamethasone timed for community pneumococcus is not the first move in a day-8 post-craniotomy fever.

Abscess and subdural empyema after surgery are loculated pus. A ring-enhancing cavity in a tumor bed can be abscess or recurrent tumor; diffusion restriction and a fever help, but culture from an operative sample is better than 8 weeks of guessing. Subdural empyema after a bone-flap infection or sinus breach still needs emergency drainage (chapter 23.4). Antibiotics do not layer themselves over the hemisphere.

Tension pneumocephalus and the Mount Fuji sign

Some intracranial air after craniotomy is expected and shrinks over days. Tension pneumocephalus is air under pressure that behaves like a hematoma: declining consciousness, new deficit, sometimes Cushing physiology. CT classically shows bilateral frontal subdural air prying the frontal lobes apart and widening the interhemispheric fissure — the Mount Fuji sign. Scattered cisternal bubbles (air-bubble sign) support tension but are not required. Mount Fuji can appear in a still-talking patient; the examination decides urgency, not the eponym alone.

Two mechanisms fill the skull with air:

• Inverted soda-bottle (or pop-bottle): CSF leaks out, negative pressure sucks air in through the same or another defect. Sitting or semi-sitting posterior-fossa craniotomy, aggressive mannitol, hyperventilation, and overdrainage from an EVD or lumbar drain all empty the cisterns.
• Ball-valve: air enters a skull-base or wound fistula on a Valsalva or a positive-pressure breath and cannot leave.

Risk factors you should be able to list: sitting or semi-sitting craniotomy (symptomatic tension in roughly 0.5–3% of sitting posterior-fossa series), nitrous oxide expanding an existing air pocket, continuous lumbar drainage, an open frontal or sphenoid sinus, and a VP shunt that keeps dumping CSF. Small frontal air can take up to 3 weeks to resorb; it should not grow between scans.

Do not give high-flow noninvasive positive-pressure ventilation (NIPPV) naively. CPAP or BiPAP through a nasal or full-face mask can force air through a cribriform, ethmoid, or sphenoid fistula and convert simple pneumocephalus into tension. The same warning applies to bag-mask ventilation over an unsecured airway when a fresh skull-base leak is known — secure the airway with a tube if the patient needs pressure. High inspired oxygen (often 100% on a controlled airway, or high FiO2 without forcing nasal pressure) speeds nitrogen washout and is the conservative adjunct in a stable patient. Head-of-bed decisions are the next heading; they are not “sit them up and add BiPAP for atelectasis.” Neurologic collapse with Mount Fuji mass effect is urgent decompression (burr hole, twist-drill, or reopening) plus repair of the fistula and stop the overdrainage. A pressure bolt that vents air is a bridge, not a cure, if the soda-bottle is still open.

CSF leak: how to prove it, how not to position, repair versus drain

CSF rhinorrhea is a salty, metallic drip that increases when the patient leans forward. Otorrhea follows a temporal bone or middle-fossa breach. Wound leaks soak the dressing with a halo that is not specific (blood and serum also ring). Reservoir signs and glucose dipsticks are folklore you must be able to unlearn.

Glucose in nasal fluid is a weak screen. Nasal mucus is usually low in glucose, and a value >30 mg/dL is sometimes quoted as “more like CSF,” but blood, tears, and inflamed mucosa all contaminate the number. A negative glucose does not exclude a leak; a positive glucose does not prove one. The confirmatory tests are beta-2 transferrin (asialotransferrin, nearly specific for CSF and inner-ear perilymph) and beta-trace protein (prostaglandin D synthase). Both take laboratory time. Do not delay meningitis treatment or airway protection waiting for the send-out if the story and the CT fistula are already obvious.

Reverse Trendelenburg caution. Modest head-of-bed elevation is standard for a cranial CSF leak because it lowers the hydrostatic driving pressure at a skull-base defect and is more comfortable than a flat, dripping patient. Aggressive reverse Trendelenburg — sitting the patient bolt upright — plus a wide-open lumbar drain after a sitting craniotomy is how you recreate the soda-bottle and get Mount Fuji. Reverse Trendelenburg also increases venous air embolism risk in an open venous sinus in the operating room; that is an intraoperative lecture, but the ICU echo is the same geometry: head up plus an open cranial-to-air path plus CSF loss is not a free lunch. Do not combine a known anterior-fossa fistula with nasal high-flow oxygen or NIPPV just because the oxygen saturation is 91%.

Repair versus drain:

Lumbar drainage for a leak is legal only when CSF pathways communicate (chapter 23.3, chapter 26.3). A posterior-fossa mass or a casted aqueduct plus a lumbar catheter is how you buy herniation. If diversion is needed and the pathway is obstructive, use an EVD, not a lumbar drain.

Meningitis after a leak

An open fistula is a dirty door. Organisms are often pneumococcus and other upper-airway flora after a skull-base leak, and staphylococci or hospital gram-negatives after a craniotomy wound leak. Fever and a stiff neck 4 days into untreated rhinorrhea is meningitis until CSF says otherwise. Treat immediately, then close the leak; antibiotics alone while the patient keeps dripping will fail. Recurrence of pneumococcal meningitis months after a “repaired” skull-base fracture means you missed a fistula.

Exam traps

Diagnosing CSF by a halo on a pillow. Ruling out a leak because the glucose dipstick was low. Treating Mount Fuji with BiPAP “for the lungs.” Leaving an infected bone flap in because cranioplasty is “cosmetic.” Using community meningitis dexamethasone and ceftriaxone 1 g daily on postoperative day 9. Dropping a lumbar drain to 0 cm H2O and sitting a fresh sitting-craniotomy patient in a chair. Calling all intracranial air tension because a radiologist typed Mount Fuji on a talking patient — still stop NIPPV and stop overdrainage, then examine them.

In practice

Day 10 wet craniotomy wound, fever, cloudy CSF: vancomycin plus cefepime, look under the flap, do not wait for beta-2 transferrin. Month 3 boggy flap and a draining sinus: the flap comes out. Hour 6 after a semi-sitting vestibular-schwannoma resection, sudden coma, Mount Fuji: 100% oxygen on a tube if they need an airway, no nasal BiPAP, reverse the drain, decompress if they do not wake. Clear rhinorrhea after transsphenoidal surgery: send beta-2 transferrin, modest head-up, controlled lumbar drain only if the cisterns communicate, repair if it does not stop, and treat meningitis the hour the neck gets stiff.