Routine postoperative computed tomography scan after craniotomy: systematic review and evidence-based recommendations.

Over the last few years, the role of early postoperative computed tomography (EPOCT) after cranial surgery has been repeatedly questioned, but there is yet no consensus on the practice. We conducted a systematic review to address the usefulness of EPOCT in association with neurological examination after elective craniotomies compared to the neurological examination alone. Studies were eligible if they provided information about the number of patients scanned, how many were asymptomatic or presented neurological deterioration before the scan and how many of each of those groups had their management changed due to imaging findings. CTs had to be performed in the first 48 h following surgery to be considered early. Eight studies were included. The retrospective studies enrolled a total of 3639 patients, with 3737 imaging examinations. Out of the 3696 CT scans performed in asymptomatic patients, less than 0.8% prompted an intervention, while 100% of patients with neurological deterioration were submitted to emergency surgery. Positive predictive values of altered scans were 0.584 for symptomatic patients and 0.125 for the asymptomatic. The number of altered scans necessary to predict (NNP) one change in management for the asymptomatic patients was 8, while for the clinically evident cases, it was 1.71. The number of scans needed to diagnose one clinically silent alteration is 134.75, and postoperative imaging of neurologically intact patients is 132 times less likely to issue an emergency intervention than an altered neurological examination alone. EPOCT following elective craniotomy in neurologically preserved patients is not supported by current evidence, and CT scanning should be performed only in particular conditions. The authors have developed an algorithm to help the judgment of each patient by the surgeon in a resource-limited context.