Decompressive Craniectomy for Diffuse Cerebral Edema
Decompressive craniectomy (DC) should be performed as a tier 3 treatment option for diffuse cerebral edema when intracranial pressure (ICP) remains refractory to maximal medical management, with the understanding that while it reduces mortality and ICP effectively, it carries significant risk of severe disability. 1, 2
Evidence-Based Approach
When to Perform DC
The 2020 Brain Trauma Foundation guidelines position DC as a tier 3 intervention after failure of sedation, osmotic therapy, external ventricular drainage, and barbiturate coma 1. The RESCUE-ICP trial definitively showed that DC reduces mortality from 48.9% to 26.9% in patients with refractory intracranial hypertension, though this comes at the cost of increased severe disability (8.5% vs 2.1% vegetative state) 2.
Specific triggers for DC include:
- ICP persistently >20-25 mmHg despite maximal medical therapy
- Clinical signs of herniation with pupillary abnormalities
- Progressive midline shift on imaging
- Failure of external ventricular drainage and barbiturate coma 2
Surgical Technique Considerations
Large unilateral (temporal) craniectomy (>100 cm²) with duraplasty is the preferred approach for most cases of diffuse cerebral edema 2. The evidence shows:
- Unilateral DC achieved good outcomes (GOS 4-5) in 40-57% of patients versus 28-32% in medical management 2
- Bifrontal DC showed worse outcomes in the DECRA trial, with 70% poor outcomes versus 51% in controls 2
- The guidelines explicitly note insufficient evidence to discriminate between lateral and bifrontal approaches, but the weight of evidence favors unilateral decompression 1
Critical Timing Factors
Earlier intervention appears superior to delayed surgery. Studies show inverse correlation between timing and injury severity, with mean surgical timing around 30-42 hours post-admission 3. However, 85% of patients achieved ICP control <20 mmHg post-operatively, with mean ICP dropping from 23.9 to 14.4 mmHg 4.
Outcome Expectations and Prognostication
The CRASH prediction model demonstrates excellent discrimination (AUC 0.905) for predicting outcomes and should guide shared decision-making with families 3. Among 30-day survivors:
- Good outcomes occurred in 17% with GCS 3-5,67% with GCS 6-8, and 67% with GCS 9-15 4
- Overall good outcomes (GOS 4-5) range from 20-40% across studies 4, 5
- Mortality ranges from 22-27% 2, 5
Preoperative CT findings of transtentorial or uncal herniation predict mortality and poor outcomes and may indicate when DC should not be performed 6.
Common Pitfalls to Avoid
- Insufficient bone flap size: Craniectomies <100 cm² are associated with poor outcomes 1
- Performing bifrontal DC without clear indication: DECRA trial showed worse outcomes with this approach 2
- Delaying surgery beyond medical management failure: Earlier intervention correlates with better outcomes 3
- Inadequate family counseling: 30-40% of survivors remain vegetative or severely disabled; families must understand this risk 2
Expected Complications
- Subdural hygroma (50% of cases) 4
- Hemorrhagic swelling ipsilateral to craniectomy (16%) 4
- Hydrocephalus (10%) 4
- Need for subsequent cranioplasty
Special Populations
For non-traumatic causes (stroke, diabetic ketoacidosis), limited case reports suggest DC may be life-saving, but outcomes data are sparse 7. The same principles of refractory ICP and herniation apply, though evidence is predominantly from traumatic brain injury cohorts.
Age considerations: Most trials excluded patients >60-70 years, and outcomes worsen with advancing age, though this should not be an absolute contraindication 2.