HSC Section 3 - Trauma, Critical Care and Sleep Medicine

Annals of Plastic Surgery • Volume 83, Number 1, July 2019

Postoperative CT Use in Orbital Floor Fractures

(94.1%) had significant physical examination findings (eg, diplopia, extraocular movement restriction, enophthalmos/exophthalmos). There was no statistically significant difference in the timing for return to the operating room for the imaging based population compared with the physical examination population ( P = 0.53). With respect to financial burden, the standard cost of a face and sinuses CT scan without contrast at our institution in 2017 was $143.84, and the standard cost for treatment of orbital fracture repair in 2017 ranged from $4661 to $6668. Almost $11,000 was spent to perform 76 CT scans, of which a maximum of 17 were used to alter medical management for 6 patients (Table 3). In other words, approximately $2013.76 is needed to be spent obtaining CT scans to reveal 1 addi- tional patient who needed reoperation (number needed to treat = 14). Initial univariate analyses revealed that a patient's postoperative physical examination was significantly associated with whether the CT scan affected clinical care ( χ 2 < 0.01). A multivariate regression model was performed to reveal no association between postoperative CT scans and change in management ( P = 0.995). When adding poten- tial confounders (covariates) to this model, the association remained in- significant (all P > 0.5) (Table 4). DISCUSSION It has been estimated that since the 1980s the use of CT imaging has almost doubled every 2 years. 15 In this era of exponential imaging use, CT scans are routinely obtained postoperatively for patients under- going orbital floor fracture repairs. This practice is commonly justified for various reasons including the assessment of the accuracy of bone

The patient demographics and injury characterizations are docu- mented in Table 1. Our cohort was composed of a majority of white pa- tients (68%) and patients younger than 60 years (85%). When considering mechanism of injury, our group was almost evenly split be- tween motor vehicle accidents, blunt trauma, assaults, and falls. The laterality of the fracture was also split equally between the left orbital floor (47%) and right orbital floor (46%). The clinical management details of the entire patient cohort and of those receiving change in care following postoperative imaging are distinguished in Table 2. Plastic surgeons and ophthalmologists were the operating physicians for 89% of our patients. Transconjunctival ap- proaches were preferred (64%), followed by midlid incisions (16%). To ensure there was no selection bias among those who received postoperative CT scans and those who did not, the demographics, injury characterizations, and clinical characteristics of these 6 patients were compared with the study population. The only statistically significant difference between these 2 groups was calculated for the percentage of black patients ( P = 0.01). However, because of the small sample size of the change in management group (n = 6), a bigger sample size is needed to draw reasonable statistical conclusions. For the 83 patients who received postoperative imaging, surgeons ordered 76 CT scans, 4 MRI scans, and 32 x-ray scans. The average time to obtain a CT scan after repair was 87 days. In our patient population, the positive predictive value of CT scans was 10.3%, whereas the positive predictive value of clinical examination was 17.6%, supporting the ad- vantage of clinical examination in postoperative management. Further- more, of the 34 patients who returned to the operating room, 32 patients

TABLE 1. Patient Demographics and Injury Characterization

Patients With Change of Care Following Postoperative CT †

Entire Patient Cohort*

P

Age, y

1 – 20

48 (22%) 82 (38%) 54 (25%) 31 (14%)

1 (17%) 2 (33%) 1 (17%) 2 (33%) 0 (0%) 45 ± 22 ‡ 4 (67%) 2 (33%) 0 (0%) 0 (0%) 1 (17%) 1 (17%) 3 (50%) 2 (33%) 4 (67%) 1 (17%) 1 (17%) 1 (17%) 2 (33%) 1 (17%) 0 (0%)

0.99 1.00 0.98 0.64 1.00 0.61 1.00 0.01 0.82 0.79 0.99 0.54 1.00 0.15 1.00 0.98 1.00 0.94 0.94 0.61 1.00

21 – 40 41 – 60 61 – 80

≥ 81

2 (1%)

Mean age, y

38 ± 19 ‡

Race

White Black Asian Other

147 (68%)

9 (4%)

29 (13%) 32 (15%) 48 (22%)

Hispanic

Type of orbital fracture Left orbital floor

102 (47%) 99 (46%)

Right orbital floor

Bilateral orbital floors

16 (7%)

Concomitant orbital injuries

145 (67%)

Mechanism of injury

Motor vehicle accident

51 (24%) 43 (20%) 62 (29%) 49 (23%) 10 (5%)

Blunt trauma

Assault

Fall

Other

Unknown

2 (1%)

*Percentage of total 217 patients. † Percentage of total 6 patients with change of treatment after postoperative CT scan. ‡ SD of sample.

www.annalsplasticsurgery.com

© 2019 Wolters Kluwer Health, Inc. All rights reserved.

39