Clinician's biases toward surgery in cerebellar hematomas: an analysis of decision-making in 94 patients

No studies have examined clinical decision-making in cerebellar hemorrhages. Clinical and CT features may influence surgery in patients with a spontaneous cerebellar hematoma. One commonly accepted adage is to remove a clot when 3 cm or larger in axial diameter on presentation CT scan. It is possible that certain preferences impact on outcome.

We analyzed 94 patients with spontaneous cerebellar hematomas between the years of 1973-1993. Thirty-one patients underwent suboccipital craniectomy and clot removal with or without ventriculostomy. Deterioration denoted worsening of consciousness, new brainstem signs, or presentation in coma. Statistical analysis was performed utilizing a tree-based model fitted by binary recursive partitioning. Ninety-five percent confidence intervals (CI) were calculated.

The overall probability of surgical intervention was 33% (CI 23-43%). The chance of surgery in stable patients was 7% (CI 2-20%). Neurologic deterioration was seen in 54 patients (57%) and increased the prospects of a surgical procedure (52%, CI 38-66%). Surgery was performed in all deteriorating patients with small hematomas (size <3 cm), but large clots (size >3 cm) did not substantially influence surgical decision-making (45%, CI 30-60%) except in patients younger than 70 years old (57%, CI 41-82%).

Clinicians at our institution usually wait for clinical deterioration to unfold prior to operating on patients with cerebellar hematomas. After deterioration occurs, they prefer small hematomas but will operate on large hematomas when patients are younger than 70, generally withholding surgery from older patients. These attitudes may impact on outcome and should be considered in future treatment trials.