Ophthalmic Regional Anesthesia: Medial Canthus Episcleral (Sub-Tenon) Single Injection Block

Following research committee approval from our four centers, patients scheduled for elective ophthalmic procedures with an expected duration of less than 60 min were eligible for this prospective observational study. Contraindications for regional anesthesia were clotting abnormalities, impaired mental status, uncontrolled glaucoma, recent surgical procedure on the same eye, and refusal to participate. During a period of 5 yr, anesthesiologists were asked to participate in this study. Participating anesthesiologists enrolled all eligible patients after the patient’s verbal consent and anonymously completed a standardized evaluation form.

Patients received oral premedication with 150 μg of clonidine 1 h before surgery. A peripheral intravenous catheter was inserted, and monitoring included continuous electrocardiography, pulse oxymetry, and automated noninvasive blood pressure measurement. Before induction of the blockade, oxybuprocaine or tetracaine drops were instilled on the eyeball and propofol was injected intravenously to obtain a brief period of sedation during eye puncture. Excessive sedation that could lead to apnea or loss of contact with the patient and result in possible movement during the puncture was avoided.

The technique used has been formally described. 6,7A 25-gauge, short-bevel needle was inserted to contact the conjunctiva between the eyeball and the semilunaris fold, at a depth of less than 1 mm (fig. 1). The bevel was directed toward the globe. The needle was then shifted slightly, medially displacing the semilunaris fold and caruncle away from the eyeball (fig. 2). The needle was then advanced in an anteroposterior direction, with the globe directed slightly medially by the needle until a “pop” was perceived, at a mean depth of 15–20 mm. At this point, the globe returned to the primary gaze position. This point is considered to be the depth marker, confirming the episcleral location of the needle tip. After an aspiration test, the local anesthetic mixture was slowly injected. A local anesthetic mixture, with hyaluronidase 25 U/ml, was chosen by the anesthesiologist as follows: lidocaine 2% in equal part with bupivacaine 0.5%, mepivacaine 2% alone, lidocaine 2% alone, or bupivacaine 0.5% alone. The injected volume was adjusted for each patient. The injection was continued until subconjunctival edema (chemosis), proptosis, and lid fullness appeared, which are considered predictive factors for success of the blockade. Ocular compression was then applied for 10–20 min using a Honan balloon set at 30 mmHg to lower intraocular pressure and resorb chemosis.

All data were collected prospectively in the four centers using a standardized form. Demographic data included age, gender, weight, and height. Surgical description included side, axial length of the globe (as measured preoperatively using echography), type of procedure, and duration. Anesthesiologist’s experience (inexperienced = resident and physicians having < 3 months’ experience with the technique, or < 50 ESA performed; experienced = staff physicians), propofol dosage, depth of needle insertion (at 5-mm intervals), mixture of anesthetic, and volume injected were noted.

All patients were followed at least until postoperative day 1. During the observation period, major complications (globe perforation, retro-bulbar or peribulbar hemorrhage, brainstem or optic nerve injury, or strabismus) and minor incidents (persistent chemosis after compression, subconjunctival edema, ocular hypertonia) were recorded. Discomfort and pain were also noted during the procedure.

Efficacy of the block was evaluated according to the level of akinesia (immobility) of the globe and eyelids and was scored by the anesthesiologist at 5, 10, 15, and 20 min after the end of injection. An 18-point scale was used in which each of the four rectus muscles, levator palpebra, and orbicularis of the eyelids were scored between 0 and 3 (0 = no block, 1 = partial akinesia unsuitable for surgery, 2 = partial but sufficient akinesia, 3 = total akinesia). The final score added the six subscores and ranged from 0 (no block) to 18 (total akinesia). The best akinesia score was defined as the highest score obtained for each block without or before supplemental injection when necessary. Time of onset was defined as the time elapsed from the end of injection to the best akinesia score. If akinesia was judged to be insufficient, the anesthesiologist was free to perform an additional injection at the same site to provide total akinesia. Any need for supplemental injection was noted.

Anesthesia–related scores of satisfaction were determined postoperatively by the surgeon, patient, and anesthesiologist, using a subjective verbal scale from 0 (total dissatisfaction) to 10 (total satisfaction).

Results are expressed as the median (minimum, maximum). To assess factors associated with complications, incidents, or pain, patients who reported one or more of these undesirable events were compared with those who did not. A Fisher exact test was used to compare categorical variables (gender, surgery side, type of surgical procedure, type of anesthetic mixture, supplemental injection, anesthesiologist’s experience, inclusion center). Ordinal and continuous variables (age, weight, height, biometry, propofol dosage, injected volume of local anesthetic mixture, depth of needle insertion) were compared using the Kruskal-Wallis test. Univariate logistic regression analysis was used to measure the odds ratio between undesirable events and relevant risk factors. All risk factors that had a P  value of less than 0.20 determined by univariate regression were entered into a multivariate logistic regression model. Using complications as the dependent variable, models were performed by the means of stepwise backward procedure. Odds ratios of complication occurrence are presented with 95% confidence intervals. All reported P  values are two-tailed, and a P  value of less than 0.05 was considered statistically significant. Statistical analyses were performed using SAS/STAT software, version 8.1 (SAS Institute, Cary, NC).

Of the 2,031 forms collected, 94% were fully completed. A total of 1,557 patients were from Caremeau Hospital, University Hospital of Nîmes; 474 patients were from the three other institutions. There were 1,176 women (58.5%) and 835 men (41.5%); median age, weight, and height were, respectively, 75 (range, 19–96) yr, 65 (range, 30–125) kg, and 165 (range, 140–190) cm. Surgery was conducted in 996 right eyes (50.4%) and 979 left eyes (49.6%). The median axial length of the globe was 23.2 (range, 19.5–34.4) mm. Most (1,747; 91.0%) of the surgical procedures were phacoemulsifications and posterior chamber artificial lens implantations. Also performed were 48 (2.5%) open globe procedures for lens extraction, 69 (3.6%) procedures to treat glaucoma, 44 (2.3%) procedures in the posterior segment (retinal surgery, vitrectomy), and 12 (0.6%) miscellaneous procedures. The median duration of surgery was 20 (range, 5–180) min. Anesthesia was performed in 1,603 cases (79.0%) by an experienced anesthesiologist and in 426 cases (21.0%) by an inexperienced anesthesiologist. The dose of propofol given at the time of the puncture was 0.50 (0.0–3.5) mg/kg. The median depth of needle insertion and injected volume were, respectively, 15 (5–35) mm and 10 (2–14) ml. Only one patient received Xylocaine; 7 (0.3%) received bupivacaine, 194 (9.6%) received mepivacaine, and 1,812 (90.0%) received a lidocaine-bupivacaine mixture. Hyaluronidase 25 U/ml was used in each case.

A total of 66 complications occurred in 60 of the patients. One patient had a retrobulbar hemorrhage; 59 had one or two minor incidents or pain and discomfort (table 1). In the patient with a retrobulbar hemorrhage, a resident had performed the ESA for a phacoemulsification, and the needle insertion depth was 15 mm. After compression, the procedure was canceled; no morbidity related to this complication occurred. Only one patient required general anesthesia for pain, and no complications due to unsuitable anesthesia occurred.

A comparison between patients who presented with a complication and those who did not is presented in table 2. Results of the univariate and multivariate analyses revealed that inexperience in the technique represented a complication risk factor (table 3).

The best akinesia score was 18 (range, 0–18) and was reached after 5 min (range, 5–35 min). A full score (18 out of 18) was achieved after the first injection for 1,756 (86.5%) patients. The anesthesiologist determined that reinjection was needed in 128 cases (6.3%). The anesthesia–related satisfaction scores were, respectively, for the surgeon, patient, and anesthesiologist: 10 (4–10), 10 (5–10), and 10 (2–10).

Based on this series of 2,031 cases, ESA provides relatively safe anesthesia. In regional ophthalmic anesthesia, the incidence of needlestick injuries is very low, but these injuries can be catastrophic. Incidence of and risk factors for complications have been reported in larger series 1,8–10for retrobulbar and peribulbar anesthesia, but not for ESA. To our knowledge, this prospective study on the safety of ESA is the largest published to date.

We observed 52 minor incidents, which, as expected, consisted of subconjunctival hemorrhages and chemosis and are related to the space of injection. In a large study, using a similar space of injection, Guise observed a similar incidence of these events. 3Transient and moderate spikes of ocular pressure are observed with ESA (J. Ripart, M.D., unpublished data, 1997). However, in this study the surgeons noted persistent ocular hypertonia in seven patients, all of whom were scheduled for phacoemulsification. In one case, the anesthesiologist indicated that compression time had been too short; in another case hypertonia occurred after a supplemental injection (total injected volume, 14 ml). The use of a large volume of drug for sub-Tenon anesthesia is questionable, as it can induce retinal damage. Pianka et al.  showed that sub-Tenon anesthesia causes a decrease in pulsatile ocular blood flow, even without a change in intraocular pressure. 5However, in posterior segment surgery, Li et al.  used a large volume of local anesthetic without complications. 11Moreover, we observed no complications among the 69 patients scheduled for a procedure to treat glaucoma.

One patient had a retrobulbar hemorrhage, which has not previously been reported with ESA. Theoretically, ESA has many advantages, such as a point of puncture in an avascular zone, reduced depth of needle insertion, and rare occurrence of myopic staphyloma in the medial aspect of the globe. 12Nevertheless, the complication we observed demonstrates that ESA does not avoid the sharp-needle complications previously described with retrobulbar or peribulbar block. Many other unpublished complications of ESA have been reported to our staff: three retrobulbar hemorrhages and one globe perforation. We are not aware of any other complications (brainstem, optic nerve injury, or retinal lesion), published or not. To avoid needle injuries, episcleral anesthesia may be performed using surgical blunt dissection and insertion of a cannula. Recently, Guise showed the excellent safety of this technique, 3observing no hemorrhage or perforation in a larger number of patients than in the current study. Nevertheless, the use of a cannula does not entirely avoid major complications. Retrobulbar hemorrhage, globe perforation, and central nervous system spread have been reported. 4,13,14 

To our knowledge, all complications reported here occurred when an inexperienced operator performed the ESA. This is in agreement with the strong association observed (odds ratio, 3) between the operator’s inexperience and the rate of complications in this survey. Performing ophthalmic regional anesthesia necessitates anatomical knowledge and technical skill. Lack of anatomical knowledge led one of our residents to puncture the lacrimal duct due to reluctance to insert the needle near the globe. However, blind needle insertion into the orbit carries its own hazards.

Secondary endpoints of this study were efficacy of the block based on akinesia and anesthesia-related satisfaction. Total akinesia was obtained in most of the patients (86.5%) after the first puncture and in more than 90% after a supplemental injection. These results are in agreement with other experiences with the use of episcleral anesthesia. 15,16The choice of an akinesia scale may be criticized regarding efficacy of the block. 6However, this scale is reproducible, and the efficacy of the anesthesia is in agreement with the good subjective anesthesia-related scores given by the surgeon or the patient.

This is the first survey of a large experience with ESA. ESA does not avoid sharp-needle complications and does require practice. Only one complication occurred among 2,031 patients; however, a larger patient group is needed to definitively evaluate the safety of ESA.