gms | German Medical Science

Objective: Although being a straightforward neurosurgical procedure, cranioplasty after decompressive craniotomy remains associated with a high complication rate. This retrospective observational study analyses the most common determinates of complications after cranioplasty.

Methods: A retrospective analysis was performed of all patients who underwent cranioplasty following hemicraniectomy from January 2010 till December 2015 at a single institution. Predictors of surgical site infection, post-operative hemorrhage, shunt-dependent hydrocephalus, post-operative epilepsy, osteonecrosis and the need for allogenic cranioplasty were evaluated in a multivariate logistic regression model.

Results: One hundred twenty-four patients met the inclusion criteria. The diagnoses leading to decompressive hemicraniectomy were as follows: 49.6% (62/124) ischemic stroke, 29.6% (37/124) traumatic brain injury, 15.2% (19/124) subarachnoid hemorrhage and 5.6% (7/124) intracerebral hemorrhage. The overall complication rate was 33.0% (41/124). Surgical site infection occurred in 12.9% of cases (16/124), hemorrhage in 9.68% (12/124), shunt dependency in 29% (36/124), epilepsy in 29.8% (37/124), osteonecrosis in 14.5% (18/124) and 22.6% (28/124) of patients eventually needed an allogenic cranioplasty. Neither patient-specific nor surgery-specific-factors proved to be significant predictors of infection. Shunt dependent hydrocephalus was predicted by prior dependency on an external ventricular drainage (p<0.005), prior invasive neuromonitoring (p=0.036), and the presence of a hygroma after decompressive hemicraniectomy (p=0.025). Epilepsy was more common in patients who suffered a traumatic brain injury (p=0.007), a subarachnoid hemorrhage (p=0.03) or had an external ventricular drainage (p=0.049). The need for an allogenic cranioplasty was directly correlated with patient’s age mainly due to the higher prevalence of osteonecrosis in younger patients (p=0.038).

Conclusion: The primary goal of this retrospective cohort analysis was to identify adjustable risk factors for post-cranioplasty complications. We have shown that surgery-specific factors such as duration of surgery, incision type or lag between hemicraniectomy and cranioplasty do not play a significant role in predicting complications. Patient-specific risk factors for surgical site infection play a more pronounced role in the development of complications and if possible, they should be adjusted accordingly.