48 / 232
Plastic and Reconstructive Surgery
• April 2016
In 104 cases (97 percent), an implant was
placed (i.e., porous polyethylene, porous polyeth-
ylene with titanium, or preformed anatomical tita-
nium mesh). Of the three remaining cases, one
involved a superomedial defect, one involved a
very small floor defect with entrapped muscle that
was felt to not require an implant, and one patient
developed intraoperative mydriasis during poste-
rior dissection and thus only a reduction after lim-
ited dissection was performed. The three patients
with nasoorbitoethmoid fractures also had reduc-
tion of the medial canthal tendon–bearing bone
fragment(s) with a titanium miniplate.
Nonoperative Cohort
The remaining 43 patients were managed
nonoperatively because they did not have indica-
tions for surgery. Younger age was strongly cor-
related with a lower operative intervention rate
(
p
< 0.001; OR, 1.15 per year; 95 percent CI, 1.06 to
1.25). Follow-up completion rates were compara-
ble between surgical and conservatively managed
patients (81 percent versus 76 percent;
p
= 0.5).
Demographically, there were no significant differ-
ences compared with the operative group except
for age (Table 2). There were no cases of globe
malposition in conservatively managed patients.
Outcomes
Two patients had 20/40 vision or worse on the
side of the fracture at the conclusion of follow-
up. One patient had suffered a significant retinal
injury and the other had suffered traumatic optic
neuropathy caused by an orbital foreign body. In
both cases, the final visual acuity was better than
the preoperative visual acuity.
A total of three patients (2.8 percent) had
complications requiring reoperation. Two cases
involved plate infection that resolved after implant
removal and antibiotics. One patient underwent
plate removal in a successful attempt to correct
hyperglobus and diplopia. Three patients experi-
enced postoperative globe malposition, with one
patient demonstrating mild residual enophthal-
mos and two patients (including the patient who
underwent hardware removal) demonstrating
mild hyperglobus. None of these patients demon-
strated a functional impairment as a result of the
globe malposition. There were no cases of eyelid
malposition or unacceptable scarring. The overall
complication rate was 4.7 percent.
DISCUSSION
Orbital fractures are among the most com-
mon facial fractures in the pediatric population.
8,9
Fig. 1.
Management algorithm for surgical decision-making in pediatric orbital fractures.
Table 2. Between-Cohort Differences
Overall
Surgical (
n
= 107)
Nonoperative (
n
= 43)
Age, yr
12.6 ± 4.2
13.3 ± 3.4
10.7 ± 5.4
12 yr or younger, %
31
25
44
Male sex, %
73
75
70
Race, %
Caucasian
52
46
69
African American
41
47
24
Other
7
7
7
Follow-up, mo
10.2
10.1
10.2
Diplopia, %*
51.5
62.5
22.2
Gaze restriction, %*
58.7
74.3
20.9
Initial VA worse than 20/40, %
16.8
20.6
7.5
Final VA worse than 20/40, %
1.4
2.0
0.0
VA, visual acuity.
*At the time of presentation.
26