Among 108 patients who underwent repair of a blowout fracture, the same number, 54 patients, were allocated to each group: group I (mannitol use group) and group II (mannitol non-use group) in randomized fashion. A mean age of 34.20 years, a proportion of males to females of 34 to 20, and fracture type distribution ratio of medial wall fractures to floor fractures to combined fractures as 28:12:14 were noted in group I. Preoperatively, group I included 30 symptomatic patients (20 enophthalmic patients, eight diplopia patients, and two patients with limitation of ocular movement). The mean age of 32.68 years, proportion of males to females 47 vs. 7, and fracture type distribution ratio of medial wall fractures to floor fractures to combined fractures as 31:13:10 were noted in group II. In group II, a total of 23 symptomatic patients (16 enophthalmic patients, five diplopia patients, and two patients with limitation of ocular movement) were recorded preoperatively. Therefore, no significant differences in these parameters were observed between the two groups, except for the sex distribution ratio (
Table 1). In the mannitol use group, there were seven cases (13.0%) of postoperative complications: one case of diplopia (1.9%) and six remnant enophthalmos (11.1%). Reoperation was performed in one patient (1.9%) in whom enophthalmos and diplopia occurred concurrently. That patient had diplopia in upward gaze and enophthalmos of 2 mm at 18 days after the initial operation. The previous implant was removed and replaced with a new implant after further reduction. The symptoms showed improvement; however, diplopia remained at the left upper extreme gaze at three months after reoperation. In the mannitol non-use group, there were seven cases (13.0%) of postoperative complications: one of retrobulbar hematoma (1.9%), three of diplopia (5.6%), two remnant enophthalmos (3.7%), and one case of operative scar problem (1.9%). Among them, five patients (9.3%) underwent reoperation. One patient had lateral dystopia and limitation of ocular motility at postoperative day 1 and diffuse retrobulbar hemorrhage without muscle entrapment was observed on CT. After reoperation for the decompression of intraorbital pressure, extraocular muscle movement showed improvement; however, diplopia remained at the lateral and inferior gaze, and impaired visual acuity continued at 1 year postoperatively. The overall incidence of preoperative and postoperative complications was not significantly different; however, preoperative symptomatic patients (30 vs. 23) had a tendency to be more frequent in the mannitol group and some serious immediate postoperative complications requiring reoperation (1 vs. 5) were more frequent in the non-mannitol group (P=0.093). The operative time was shorter in the mannitol use group (88.5±32.1 minutes) compared to the mannitol non-use group (90.7±38.9 minutes) (
Table 2). The fracture type distribution ratio of medial wall to floor of combined fracture as 11:11:3, as noted in the video recording of the Mannitol group, and 8:2:15 were noted in a video recording of Hartman's solution group. No significant differences in each fracture type were observed between before and after the visual analogue scale (VAS) score, because the number of each case was too small and the numbers of each case through classification were not the same. Therefore, we did not distinguish among these details, but evaluated the cases before and after scoring together. The visual analogue scale score before and after mannitol administration was shown to be markedly different, and this had statistical significance. However, no significant differences in Hartman's solution group were observed between before and after the VAS score (
Table 3). There was no occurrence of immediate or delayed recipient complications of infection, extrusion, displacement, or foreign body reaction related to the implant. There was no occurrence of major drug related complications in the central nervous, cardiovascular, pulmonary, or renal systems. Other minor drug related side effects such as headache, nausea, vomiting, and abdominal cramps were also not present. Some representative cases are shown in
Fig. 3.