Keywords
anophthalmic surgery - ocular evisceration - ophthalmology resident performance -
oculoplastic training - ophthalmology resident outcomes
Evisceration is the removal of the intraocular contents of the eye, leaving behind
a scleral shell with preserved extraocular tissues. Evisceration is performed for
several reasons, including trauma, phthisis bulbi, cosmesis, and blind, painful eyes.
Compared with eye removal procedures like enucleation and exenteration, evisceration
is considered technically easier, preserves more tissue, offers improved implant motility,
and boasts a better cosmetic result.[1]
[2]
[3]
[4]
[5]
[6]
[7] Recent studies also suggest that eviscerations have a lower complication rate regarding
postoperative infection and implant extrusion.[2]
[8]
[9] Historically, evisceration was avoided due to concerns for an increased risk of
sympathetic ophthalmia; however, recent literature suggests this risk is no greater
than other eye removal procedures like enucleation.[4]
[8]
[10] For these reasons, the frequency of ocular evisceration has increased over the years.[11]
[12] Typical complications of evisceration include infection and wound dehiscence with
implant exposure or extrusion, as well as minor surgical sequelae such as eyelid malposition,
superior sulcus deformity, enophthalmos, and cyst formation.[7]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21] Current literature explores various surgical techniques, efforts to minimize complications,
and long-term outcomes of anophthalmic sockets. Many of these published works arise
from training institutions. However, there is a dearth of information regarding ophthalmology
trainee performance and long-term outcomes in cases of ocular evisceration. As an
increasingly common, technically easier procedure in a non-seeing eye, eviscerations
are ideal cases for trainees. The aim of this study is to use patient outcomes and
complications to evaluate the appropriateness of ophthalmology residents performing
ocular evisceration, as well as using ocular evisceration procedures to enhance surgical
training and evaluate resident surgical performance.
Methods
Data Acquisition
A retrospective chart review of eviscerations performed between October 2011 and May
2017 by ophthalmology residents as the primary or assistant surgeon under the guidance
of a single oculofacial plastic surgeon was completed. Data collected included reason
for evisceration, resident's role in the case and resident's month of oculoplastic
training, surgical technique, size of implant, subsequent complications, and duration
of follow-up.
Surgical Technique
On the day of surgery, the patient was placed under general anesthesia. Phenylephrine
ophthalmic drops (10%) were applied to the operative eye, and the nonoperative eye
was shielded. A 360-degree peritomy was performed, followed by conjunctival and Tenon's
dissection into the quadrants at the 2, 4, 8, and 10 o'clock positions. The cornea
was excised and the intraocular contents were removed. In all cases, the cornea and
intraocular contents were submitted for histopathologic analysis. The internal scleral
surface was scrubbed repeatedly with 100% alcohol to remove any uveal antigens, followed
by copious irrigation with antibiotic solution. Four scleral petals were created from
the limbus to the optic nerve at the 2, 4, 8, and 10 o'clock positions. An orbital
sizer was used to determine an implant size that allowed for filling of the anophthalmic
socket without placing undue tension when closing the scleral petals around the implant.
The implant—a porous polyethylene spherical orbital implant (Medpor: Stryker, Kalamazoo,
Michigan)—was soaked in a combination antibiotic and anesthetic solution, and was
then placed into the scleral cavity. The superior and inferior scleral petals were
closed with a combination of running and interrupted slow-absorbing sutures. The horizontal
petals were similarly closed. Copious irrigation was again performed with antibiotic
solution to allow for visualization of Tenon's. In a portion of cases, slow-absorbing
suture was used to close Tenon's in a running manner, followed by a separate closure
of the conjunctiva with running absorbable suture. In the remainder of cases, Tenon's
and conjunctiva were closed simultaneously with running absorbable suture. A combination
antibiotic and steroid ointment was applied, an appropriately sized conformer was
inserted, and a temporary tarsorrhaphy was placed using nonabsorbable suture. At the
conclusion of the case, a retrobulbar injection of 0.75% bupivacaine with epinephrine
(1:200,000) was administered, and a pressure patch and eye shield were placed. The
tarsorrhaphy, pressure patch, and shield remained in place for 1 week, after which
a combination antibiotic and steroid ointment was applied to the eye three times per
day for the next 2 weeks.
Teaching Methodology
A single attending oculoplastic surgeon oversaw these cases. Early in the chart series,
third-year residents were intermittently assigned to oculoplastic surgeries and taught
in a modified “see one, do one” manner under direct supervision of the attending oculoplastic
physician. Later in the chart review period, first-year residents were assigned to
a 3-month intensive block of oculoplastic education, including time in clinic, the
minor procedure room, and the operating room. In each learning environment, a baseline
level of competency was established through previous interaction and direct evaluation
of surgical technique in both wet laboratory and the operating room by the attending
oculoplastic surgeon. Intraoperatively, a graduated, step-wise approach to mastering
the procedure was employed. If a step was completed incorrectly, the resident was
informed and tasked with repeating the step correctly on the spot under demonstrative
guidance of the attending surgeon until the resident was able to perform more than
50% of the key steps of the procedure, at which point the case became a primary resident
case.
Statistical Analysis
Data were collected and cross-checked for errors. Percentage, range, and average mean
were calculated. Statistical analysis included chi-square test, Pearson's product–moment
correlation coefficient, and odds ratio (OR) calculations. Significance was determined
by p < 0.05.
Results
A total of 104 eviscerations were completed: 57 (54.8%) were completed with a resident
as the primary surgeon, while 47 (45.2%) were completed with a resident as the assistant
surgeon. Patient demographics were wide ranging ([Table 1]). There were many similarities between attending and resident-led case demographics;
however, attending cases had nearly twice the amount of patients with penetrating
ocular trauma compared with resident cases. Reasons for evisceration included blind,
painful eyes from multiple etiologies (such as end-stage glaucoma; 58.7%), endophthalmitis
(12.5%), corneal perforation (10.6%), phthisis (9.6%), and trauma (7.7%; [Table 2]). There were nearly double the number of etiologic cases of endophthalmitis in attending
cases (17.0% compared with 8.8% in resident cases), and chart review of attending
trauma cases revealed more globally obliterative traumas than resident cases.
Table 1
Patient demographics
|
Demographic
|
Total (%)
|
Attending
|
Resident
|
|
Sex
|
|
Female
|
37 (35.6%)
|
16 (34.0%)
|
21 (36.8%)
|
|
Male
|
67 (64.4%)
|
31 (66.0%)
|
36 (63.2%)
|
|
Age
|
|
Range
|
10–97 y
|
14–92 y
|
10–97 y
|
|
Average
|
60.2 y
|
44.8 y
|
58.4 y
|
|
Laterality
|
|
Right eye
|
42 (40.4%)
|
24 (51.1%)
|
18 (31.6%)
|
|
Left eye
|
62 (59.6%)
|
23 (48.9%)
|
39 (68.4%)
|
|
Follow-up
|
|
Shortest
|
9 d
|
9 d
|
9 d
|
|
Longest
|
53 mo
|
53 mo
|
52 mo
|
|
Average
|
11.3 mo
|
14.2 mo
|
9.5 mo
|
|
Previous ocular surgery
|
|
Patients (no.)
|
81 (77.9%)
|
36 (76.6%)
|
45 (78.9%)
|
|
Surgeries (no.)
|
138
|
61
|
77
|
|
History of penetrating trauma
|
28 (26.9%)
|
17 (36.2%)
|
11 (19.3%)
|
|
History of intraocular infection
|
29 (27.9%)
|
13 (27.8%)
|
16 (28.1%)
|
Notes: There were more male patients undergoing ocular evisceration. The average patient
age was 60 years, and there was an incidental predominance of left eye evisceration.
Follow-up was widely variable, with an average follow-up period of 11.3 months. Follow-up
was slightly longer at 14.2 months in attending-led cases compared with 9.5 months
in resident cases. The majority of patients had previous ocular surgery, which ranged
from cataract extraction and glaucoma procedures to retinal detachment repairs, posterior
pole vitrectomies, and repair of ruptured globes. There was nearly two times the number
of patients with penetrating trauma in attending-led cases compared with resident
cases.
Table 2
Reasons for ocular evisceration
|
Reason for evisceration
|
% (n)
|
Attending
|
Resident
|
|
Blind, painful eye
|
58.7 (61)
|
48.9 (23)
|
66.7 (38)
|
|
Endophthalmitis
|
12.5 (13)
|
17.0 (8)
|
8.8 (5)
|
|
Corneal perforation
|
10.6 (11)
|
10.6 (5)
|
10.5 (6)
|
|
Phthisis
|
9.6 (10)
|
12.8 (6)
|
7.0 (4)
|
|
Trauma
|
7.7 (8)
|
8.5 (4)
|
7.0 (4)
|
|
Other
|
0.9 (1)
|
2.1 (1)
|
0.0 (0)
|
Notes: Trauma occurred within the preceding 3 months to qualify as the reason for
evisceration. Attending trauma cases tended to have more globally disruptive trauma
than resident cases. Blind, painful eyes occurred from several pathologies, including
end-stage glaucoma and keratitis. Evisceration is absolutely contraindicated in cases
of intraocular malignancy, and there was no intraocular malignancy present on histopathologic
analysis of eviscerated tissue.
Duration of follow-up ranged from 9 days to nearly 5 years, with slightly longer follow-up
for attending-led cases compared with resident cases ([Table 1]). During this time, a total of six complications occurred. Complications were defined
as wound dehiscence, implant exposure or extrusion, surgical site infection, and sympathetic
ophthalmia. Of the six complications, one occurred in resident primary cases (1.8%
of resident primary cases) and five in resident-assist cases (10.6% of resident-assist
cases). The complication rate for all cases in total was 5.77%. There were no cases
of implant extrusion or sympathetic ophthalmia. Minor adverse events and long-term
surgical sequelae included eyelid malposition, conjunctival cysts, pyogenic granuloma,
implant migration, superior sulcus defect, and enophthalmos. Sequelae occurred with
a near-equal incidence in both resident-led and attending-led cases ([Table 3]). There was no incidence of implant migration, superior sulcus defect, or enophthalmos.
There was no significant difference in the rate of complications (χ
2(1, N = 104) = 3.74, p = 0.05), long-term sequelae (χ
2(1, N = 104) = 1.53, p = 0.22), or total adverse events (complications of surgery and minor or long-term
sequelae (χ
2(1, N = 104) = 0.01, p = 0.91) between resident- and attending-led cases. An infectious etiology for evisceration
was associated with a slightly increased risk of experiencing a major complication
in both attending- and resident-led cases (OR = 1.23). There was a larger risk for
complications in patients who had a history of previous ocular surgery (OR = 3.35
in attending cases; the single complication in resident cases was insufficient to
calculate this for resident-led surgeries).
Table 3
Complications, sequelae, and surgical variables after ocular evisceration
|
Number of eviscerations
(n = 104)
|
Resident role
|
|
Primary surgeon
(n = 57)
|
Assistant surgeon
(n = 47)
|
|
Major complications
|
1
|
5
|
|
Implant exposure
|
0
|
3
|
|
Wound dehiscence
|
1
|
1
|
|
Infection
|
0
|
1
|
|
Minor surgical sequelae
|
21
|
18
|
|
Ptosis
|
6
|
5
|
|
Other eyelid malposition
|
6
|
5
|
|
Conjunctival cyst
|
5
|
4
|
|
Other
|
4
|
4
|
|
Surgical variables
|
|
|
|
Implant size (average)
|
19.6 mm
|
20.1 mm
|
|
Separate or combined closure of Tenon's and conjunctiva
|
40 separate
16 combined
|
44 separate
3 combined
|
Notes: As a case series of 104 ocular eviscerations, the overall complication rate
was 5.77%. The major complication rate was 1.8% in resident primary cases and 10.6%
in resident-assist cases. Minor surgical sequelae occurred with a near-equal incidence
of 36.8% in resident primary and 38.3% in resident-assist cases. Other sequelae included
one case each of fornix shortening (in the setting of an initial alkali burn), persistent
pain (in a patient with neuropathic pain syndrome), symblepharon formation, pedunculated
benign conjunctival lesion, pyogenic granuloma (occurring twice in the same patient),
and chronic discharge. No significant differences were observed among surgical variables.
The orbital implant size ranged from 16 to 22 mm, with the average size being a 19.9-mm
implant. There was no statistical difference in choice of implant size (χ
2(1, N = 104) = 0.86, p = 0.35) or adverse events (χ
2(1, N = 104) = 0.09, p = 0.76) between these two groups. Additionally, there was no correlation between
the type of conjunctiva and Tenon's closure (whether closed separately or simultaneously)
with postoperative complications (χ
2(1, N = 103, due to exclusion of one patient who lacked conjunctiva) = 1.11, p = 0. 292; Fisher's exact test yielded p = 0.2857 in resident surgeries, and p = 0.2920 in attending-led surgeries), similar to a recently published report.[21]
A Pearson product–moment correlation coefficient revealed a slight negative correlation
between the resident's month of training and the presence of postoperative complications,
though this was not significant (r = −0.23, n = 40, p [one-tailed] = 0.08). There was a small but significant positive correlation between
the occurrence of all adverse events (complications and long-term sequelae) and duration
of follow-up (r = 0.029, n = 104, p [one-tailed] = 0.001).
Discussion
Ocular evisceration is employed in patients with cosmetic, painful, infectious, or
other pathologic processes that compromise ocular integrity. Historically, enucleation
was the preferred anophthalmic procedure in these patients, as it was believed that
enucleation had a lower risk of sympathetic ophthalmia and better postoperative pain
control.[22]
[23]
[24]
[25]
[26] The frequency of eviscerations has increased in the past few decades,[11]
[12] in part because evisceration is considered a technically easier procedure and now
has a known low rate of sympathetic ophthalmia.[4]
[8]
[10]
[27] As such, ocular evisceration is an ideal procedure for ophthalmology trainees. However,
review of the current literature reveals a lack of information regarding ophthalmology
resident performance and outcomes in cases of ocular evisceration. In this study,
we found that not only were trainees capable of safely performing ocular eviscerations
with a complication rate similar to published reports but also that a trainee's surgical
skills could be both taught and evaluated by performance of ocular evisceration procedures
and assessing postoperative complications.
Our rate of complications (1.8% in resident primary cases and 10.6% in resident-assist
cases) falls within published reports, suggesting that evisceration surgery is a safe
procedure for ophthalmology residents to learn and practice surgical skills. A review
of the literature shows complication rates after evisceration range from 0 to 53.8%.[2]
[3]
[6]
[7]
[9]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21] The wide range of complication rates reported is likely due to variability in patient
demographics, underlying ocular pathology, surgical technique (including type and
size of implant, wrapping material, and presence or absence of sclerotomies), and
duration of follow-up (with longer follow-up being associated with a higher incidence
of complications).[6]
[9]
[13]
[14]
[16]
[21] Collectively, recent reports have an average incidence of 8.6% for implant exposure,
1.0% for surgical site infection, and 2.5% for superior sulcus deformity.[2]
[3]
[6]
[7]
[9]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21] Our study found that primary resident cases had a below-average incidence of complications.
There were no superior sulcus defects or enophthalmos noted in our study. This is
likely due to follow-up of less than 5 years in the majority of resident cases, and
longer follow-up may reveal the development of sulcus or socket deformities. Similar
to published reports, our study did find complication rates directly correlated to
the length of follow-up. Of note, the power of our study to compare resident to attending
outcomes is low (∼60%, confidence interval [CI] = 95%), and expanding the number of
cases reviewed in the future will increase the applicability of our findings.
Minor and surgical sequelae occurred at nearly equal rates between resident-led and
resident-assist surgeries. Of note, our rate of these minor and surgical sequelae
seems higher than previously published reports. In the current work, 10.5% of resident
primary cases exhibited eyelid malposition (entropion, ectropion, or blepharoptosis),
and 8.8% developed conjunctival cysts. Similarly, 10.6 and 8.5% of resident-assist
cases displayed similar findings, respectively. Published data show an average incidence
of 7.4% for eyelid malposition and 2.2% for conjunctival cysts.[7]
[9]
[14]
[16]
[17]
[19]
[21] Not all reports include these sequelae in their follow-up, and it is possible that
current publications do not accurately represent the overall incidence of these sequelae.
It is also possible that variations in surgical technique, underlying pathology, or
presence of trauma account for the discrepancy, or that our sample is too small to
reveal representative rates.
Our work is the first to propose that not only can resident surgeons safely perform
ocular evisceration, but also that evisceration outcomes can be used as a marker for
evaluating resident performance. For example, three of the five conjunctival cysts
in resident primary cases occurred in cases performed by the same resident. Conjunctival
cysts occur from erroneous implantation of conjunctiva in posterior layers like Tenon's
capsule. These cases had no history of ocular trauma; as such, the occurrence of conjunctival
cysts in these patients is likely a direct reflection of surgical technique and can
be used to assess resident surgical performance. It is acknowledged that this assessment
has more utility for short-term complications when surgical assessment and teaching
occurs during residency.
This study is also the first to demonstrate that complications in resident evisceration
cases are inversely correlated with the number of months the resident spent on the
oculoplastic service. While a nonsignificant trend in this study, this finding is
consistent with previously published work in ophthalmology surgical training. Surgical
training—including length of training—predicts performance and outcomes, though not
necessarily transferability, across subspecialties.[28]
[29]
[30]
[31]
[32]
[33]
[34]
Limitations of this study include the small number of residents assessed (n = 12) and differences in oculoplastic training. Early in the years of this retrospective
chart review, residents did not have an immersive oculoplastic training period. Later,
residents underwent an immersive 3-month-long oculoplastic rotation. This allowed
for better evaluation of skills and interpretation of those skills in patient outcomes.
Some residents continued to perform oculoplastic cases after completion of their 3-month
block, and those with contiguous months of oculoplastic performance were included
in this study's data analysis. Those without a known block of contiguous months of
oculoplastic training were omitted from those portions of the data analysis. This
results in less statistical power, and is a source of improvement in future work.
Several tools have been developed to measure resident physician technical skills,[35]
[36]
[37] including a proposed curriculum for oculoplastic training,[38] and perhaps ocular eviscerations should be considered as a safe, formalized teaching
and evaluation tool for resident surgical skills.
Future directions would also include increasing the sample size and duration of follow-up
studied. With a confidence interval of 5%, the current work gives us a power of approximately
60% when comparing resident to attending outcomes. A larger sample size is needed
to improve the power of this work. In addition, while some patients were followed
up up to 5 years, the majority have been followed up only for approximately 2 years
to date. This is insufficient time to see complications like superior sulcus deformity
and enophthalmos. Thus, larger studies with longer follow-up and an increased number
of resident surgeons may be necessary to further support the findings of this study.
The goals of this project were multiple: (1) to compare resident evisceration outcomes
to those in published reports; (2) to assess outcomes after ocular evisceration; and
(3) to look for correlation between surgical outcomes and duration of resident's oculoplastic
training. This study supports the ACGME goal of enhancing resident physician education
in surgical competency, assessment of resident performance, and utilizing outcomes
for improving resident education.[39] This study shows that not only are ocular eviscerations safe for resident surgeons
to perform, but they are also a procedure that is ideal for training and evaluating
resident surgical performance.
