Switch Cefepime to Piperacillin-Tazobactam
The most appropriate intervention is to discontinue cefepime and switch to piperacillin-tazobactam. This patient's clinical presentation is classic for cefepime-induced neurotoxicity (CIN), which is causing her encephalopathy with the characteristic EEG findings of generalized periodic discharges with triphasic morphology.
Why This is Cefepime Neurotoxicity
This patient has the perfect storm for CIN:
- Stage IIIb CKD with acute worsening (creatinine jumped from 1.9 to 3.5) 1
- Advanced age (77 years old) 2
- Classic EEG findings: generalized periodic discharges with triphasic morphology are pathognomonic for beta-lactam neurotoxicity 3, 2
- Temporal relationship: neurological symptoms appeared on day 2 after cefepime initiation, which is typical 4, 5
- No alternative explanation: CT head unremarkable, no focal neurological deficits
The French Society guidelines explicitly state that cefepime has the second-highest pro-convulsive activity among beta-lactams (160, compared to penicillin G = 100), and that renal failure is the primary risk factor for beta-lactam neurotoxicity 1. Cefepime concentrations above 22 mg/L are associated with neurotoxicity in 50% of patients, and this patient's acute kidney injury makes toxic accumulation virtually certain 1.
Why Piperacillin-Tazobactam is the Right Choice
Piperacillin has dramatically lower neurotoxicity (pro-convulsive activity = 11, compared to cefepime's 160) 1. Multiple case reports demonstrate complete neurological recovery within 24-48 hours after switching from cefepime to piperacillin-tazobactam 5, 2. The guidelines explicitly recommend this switch when neurotoxicity occurs 1.
While piperacillin-tazobactam can also cause neurotoxicity at very high concentrations (>157 mg/L steady-state), the threshold is much higher and the risk substantially lower than with cefepime 1.
Why NOT the Other Options
Stopping vancomycin alone would be incorrect because:
- Vancomycin does not cause this EEG pattern
- The patient needs continued Gram-positive coverage given her severe pneumonia
- The temporal relationship and EEG findings point specifically to cefepime
Administering Keppra (levetiracetam) would be inappropriate because:
- This is not primary seizure activity—it's drug-induced encephalopathy 3, 6
- The FDA label for Keppra shows it can cause confusion and psychotic symptoms itself 7
- Treating the symptom rather than removing the cause delays recovery and adds unnecessary medication burden
- Guidelines explicitly state to discontinue the offending beta-lactam, not add antiepileptics 1
Starting risperidone would be dangerous because:
- This is not primary psychosis—it's toxic encephalopathy
- Risperidone has no role in treating drug-induced encephalopathy 8
- Adding an antipsychotic would further cloud the clinical picture and potentially worsen mental status
Critical Clinical Pitfall
The most dangerous error here would be diagnostic anchoring—attributing the encephalopathy to sepsis, metabolic derangement, or structural brain pathology when it's actually iatrogenic and completely reversible 3. The case reports emphasize that CIN can occur even with appropriate renal dose adjustments 2, and that delayed recognition leads to worse outcomes.
Expected Clinical Course
Based on the evidence, neurological symptoms should resolve within 24-48 hours of cefepime discontinuation 4, 5, 2. If symptoms persist beyond 72 hours after switching antibiotics, reassess for alternative diagnoses.
The guidelines recommend therapeutic drug monitoring when available, but immediate discontinuation should not be delayed waiting for levels 1. The clinical diagnosis is sufficient given the classic presentation.