An ever-increasing proportion of ophthalmic surgery, especially cataract surgery, is performed under local anesthesia (LA). Of all injection LA techniques, sub-Tenon’s block (STB) has gained increasing popularity. During the late 1990’s, the proportion of patients who received STB as their LA technique for cataract surgery was 6.7% in the United Kingdom, 7% in Japan, and 8% in the United States (1–3). This proportion is believed to have increased significantly since then. In an attempt to reduce the incidence of complications from the use of sharp needles, Stevens (4) described the modern STB technique in 1992. He also described the blunt-tipped cannula for this technique (5). This block has become the block of choice providing akinesia as well as anesthesia during intraocular surgery in many hospitals throughout the world. There have only been a few descriptions of serious complications attributable to STB. In 1997, retrobulbar hemorrhage after STB was reported by Olitsky and Juneja (6). Globe perforation upon dissecting Tenon’s capsule with scissors was recently described by Frieman and Friedberg (7). Severe orbital cellulitis after STB in an immunocompromised patient has been reported by Redmill et al (8). Recently, three cases of postoperative diplopia resulting from rectus muscle trauma during STB were reported (9). We add another five cases with four different complications to this emerging spectrum of adverse outcomes. All complications were observed over the course of 18 mo, during which an estimated 7250 patients received STB in our institution. A realistic assessment of the incidence of complications can only be derived from a prospective trial because significant underreporting is usually associated with retrospective case reviews. All cases were obtained from the critical incident database or personal reports by surgeons and anesthesiologists. Case notes were reviewed when complications were attributed to the use of STB. Two patients suffered intraocular hemorrhage or hyphema. A third patient suffered permanent loss of vision, possibly from central retinal artery occlusion (CRAO). The fourth case describes a globe perforation as a consequence of ineffective STB. In the fifth case, central nervous system (CNS) spread of LA caused prolonged loss of consciousness. The aim of this report is to raise awareness of complications from the use of STB and stimulate a debate about their prevention.
A blunt sub-Tenon’s anesthesia cannula 1.10 × 25 mm or 19-gauge × 1 in. (BD-Visitec, Sarasota, FL) was used in all cases. The technique for STB is briefly described as follows. After topical anesthesia and asepsis of the eye, a suitable lid speculum is inserted. The patient is then asked to gaze upwards and laterally. The conjunctiva and Tenon’s fascia is lifted with forceps in a place 5 mm away from the corneal limbus in the infero-nasal quadrant, creating a small tent of conjunctiva and adherent Tenon’s capsule. A small nick incision into the side of the conjunctival tent is made with blunt-tipped Westcott scissors. The closed scissors are then slid through the nick in the conjunctiva and used to create a path in Tenon’s capsule and the intermuscular fascia. The blunt and curved anesthesia cannula is then inserted onto bare sclera and glided along the path created by the Westcott scissors, following the contour of the globe, until posterior to the equator. The forceps should maintain fixation of the eye during dissection and insertion of the needle. The needle can be advanced to a depth of approximately 2 cm depending upon globe size. A total volume of 3–4 mL of LA is then injected slowly. It effectively irrigates the retrobulbar space. Digital massage or an orbital pressure device is sometimes applied to disperse the LA.
Sharp needle techniques for eye anesthesia are associated with rare complications such as globe perforation, hemorrhage, optic nerve injury, and brainstem anesthesia. With blunt needle STB, the incidence of those complications has been reduced even further (10). The following case reports serve as a reminder that those serious complications can still occur.
A 65-yr-old woman presented for day-case cataract surgery under LA. Her preoperative best-corrected visual acuity was 20/200. The axial length was 23 mm. She had a normal intraocular pressure (IOP). Her only regular medications were atenolol and bendrofluazide for hypertension. Topical anesthesia was established using amethocaine 1% eye drops. A STB of 5 mL of LA solution was given in the infero-nasal quadrant (2.5 mL of 2% lidocaine and 2.5 mL of 0.5% bupivacaine with 30 U/mL of hyaluronidase). Digital pressure and massage was applied to aid the spread of the block. The patient was then transferred onto the operating table at which stage blood was found in the anterior chamber. There was no evidence of globe perforation, and the IOP was normal on digital palpation. The operation was initially postponed. B-scan ultrasound showed a posterior vitreous detachment but no evidence of retinal or choroidal detachment. The patient underwent uneventful phacoemulsification and intraocular lens implantation later that day under general anesthesia. On follow-up 1 mo later, the vision was 20/20 in that eye.
An 86-yr-old woman presented to another hospital for day-case cataract surgery under LA. The preoperative visual acuity was recorded as 20/80. She suffered from hypertension and mild left ventricular failure. Her medication was 75 mg of aspirin per day. The anesthesiologist administered a STB with 4 mL of LA (2% lidocaine with 30 U/mL of hyaluronidase). The eye was then massaged through closed lids to disperse the anesthetic fluid. During the ocular massage, the patient reported feeling pain. A hyphema was found when the eye was examined. The IOP was measured at 6 mm Hg. A globe perforation was suspected. The operation was canceled, and the patient referred to the vitreo-retinal service at our institution for further management. On arrival, her visual acuity was hand movements only. There was an organizing clot in the anterior chamber. The IOP was 26 mm Hg. B-scan ultrasound showed the hyphema and a posterior vitreous detachment. There was no evidence of globe rupture, retinal, or choroidal detachment. A routine coagulation screen was normal. The patient was treated conservatively and discharged home. Cataract surgery has now been rescheduled at her local hospital.
A 33-yr-old man was admitted for excision of a pterygium with a free conjunctival autograft under LA. He was in good general health. His preoperative visual acuity was 20/15 in both eyes. Preoperatively, he was given 1 mg IV of midazolam for sedation. Proxymetacaine 0.5%, amethocaine 1%, and epinephrine 0.01% eye drops were applied. A STB with 5 mL of LA (0.5% bupivacaine with 30 U/mL of hyaluronidase) was then administered in the infero-nasal quadrant followed by ocular massage. The surgery was uneventful, and the patient was discharged home that afternoon. Ten days later, the patient attended the ophthalmic casualty department complaining of poor vision in the operated eye. He had noticed this on the day after surgery as soon as he removed the eye pad. The visual acuity was now counting fingers with poor color perception. There was a central scotoma and a relative afferent pupil defect. The operative site was healing well. Funduscopy revealed retinal thickening at the macula and punctate refractile bodies at the level of the internal limiting membrane. Ocular coherence tomography confirmed thickening at the right macula. Fluorescein angiography failed to demonstrate edema or ischemia, and retinal perfusion was clinically normal. Multifocal electro retinography showed normal a waves but attenuated b waves, suggesting inner retinal posttransductional block with preserved photoreceptors. The visual evoked potentials were normal. Ocular blood flow studies using ultrasound showed normal ophthalmic artery flow but reduced velocities in both retinal arteries. Computed tomography imaging of the orbits and brain was normal. All blood tests, including coagulation studies, thrombophilia screen, anticardiolipin antibodies, and inflammatory markers, were normal. During the work-up after the unexpected loss of vision, the patient reported that he had had an episode of painless transient visual loss in the peripheral field of both eyes 6 mo previously. It had resolved completely over 20 min with no recurrence, and the matter was never investigated further. No systemic condition predisposing to retinal vasculopathy was found during a work-up at a cardiovascular hospital. There has been no improvement in vision. The optic disk has become pale over time, and the eye is now blind.
A 66-yr-old woman was admitted with a history of cataract and mild ptosis in the left eye. She suffered from hypothyroidism and hypertension for which she took thyroxine, nifedipine, atenolol, and bendrofluazide. She was scheduled for cataract extraction. The operating surgeon gave a STB immediately before the start of surgery. After topical anesthesia with 1% amethocaine, 3 mL of LA (0.5% bupivacaine) was injected into the infero-nasal quadrant. There was no anesthesiologist present, and no sedation was given. During the later stages of the operation, the patient was in discomfort and started squeezing her eyelids. This made the operation increasingly difficult. A posterior capsular tear developed, and anterior vitrectomy had to be performed. Towards the end of the operation there was a hemorrhage from angle vessels. The implanted lens had to be removed to stabilize the bleeding. Postoperatively, the patient was prescribed a dose of 250 mg of acetazolamide. She was discharged home on the same day. On the fifth postoperative day, an ultrasound examination showed a posterior vitreous detachment and intra-gel opacities. There was a retro-hyaloid hemorrhage, as well as nasal and temporal choroidal hemorrhages. During follow-up examination 5 wk later, her visual acuity was counting fingers only, and the IOP in the operated eye was 30 mm Hg. The persisting hemorrhage obscured the fundus view. She was referred to the vitreoretinal service and underwent a vitrectomy and intraocular lens implantation 2 wk later. During the vitrectomy, an infero-nasal retinal break was noted. It was interpreted to be the scarred entry site of an accidental needle perforation during the previous operation. The further recovery of the patient was complicated by increased IOP and bacterial keratitis. The final visual acuity in the operated eye was 20/200, and the IOP was normal.
A 67-yr-old man was admitted for cataract surgery in his left eye under LA. Six months earlier he had undergone uneventful cataract surgery in his other eye under STB. He had a history of myocardial infarction followed by three-vessel bypass surgery 4 yr before. He also suffered from systemic hypertension. His regular medication included small-dose aspirin, lisinopril, and furosemide. The preoperative visual acuity was 20/80 with an axial length of 25.4 mm. The orbital examination was unremarkable. Eye drops of oxybuprocaine 0.4% and amethocaine 1% were instilled before STB with 5 mL of LA (3 mL of 2% lidocaine, 2 mL of 0.5% bupivacaine, and 30 U/mL of hyaluronidase) into the infero-nasal quadrant. The surgery was uneventful and lasted 30 min. During the operation, the patient was cardio-vascularly stable. The oxygen saturation was 100%, and the heart rate (HR) was 95 bpm. When the drapes were removed at the end of surgery, he was unresponsive to verbal command. He was transferred to the recovery unit. His cardiovascular variables were blood pressure (BP) of 138/78 mm Hg, HR of 94 bpm, oxygen saturation of 97%, and blood sugar of 80 mg/dL. He was still unresponsive 90 min after the administration of the STB. His eyes were moving under the closed lids in a manner similar to rapid eye movement sleep and continued to move when the eyes were opened. He showed no response to verbal or painful stimuli. The pupil in the unoperated eye was normal in size with a preserved light reflex. Both plantar responses were diminishing. His cardiorespiratory variables remained stable: HR of 100 bpm, BP of 120/56 mm Hg, respiratory rate of 18 breaths/min, and the blood glucose level was 85 mg/dL. Fifty minutes later, the patient showed some signs of recovery but was unable to speak for another half hour. When questioned, he could remember being given the anesthetic block but had no memory of the cataract surgery or subsequent events. Postoperatively, he complained of a headache, for which he was given acetaminophen. Three weeks later, his visual acuity was 20/30. He made a full recovery with no neurological deficits.
In the first two cases, cataract surgery had to be postponed because of hyphema developing between administration of the anesthetic block and the start of the procedure. In both cases, STB was followed by ocular massage, a manipulation that can produce peak IOP of up to 400 mm Hg, with considerable globe distortion and shearing forces on the long posterior ciliary arteries and the anterior ciliary arteries in the anterior segment (11). None of the known risk factors for hyphema, such as iris vascular abnormalities or Fuchs’ heterochromic iridocyclitis, were identified in the two patients (12,13). Both patients had preexisting but well controlled systemic hypertension. A short-lived increase of BP during the administration of the block or during ocular massage may have caused or exacerbated the intraocular hemorrhage, but such a pressure spike was not detected during any of the two cases. The preoperative exposure to aspirin in Case 2 may have increased the risk of hemorrhage, but a large prospective trial did not confirm this risk (14). Ocular massage is sometimes performed in an attempt to reduce the chemosis caused by subconjunctival placement of LA. Careful placement of the needle tip into the sub-Tenon’s space behind the equator of the globe helps to prevent ballooning of the conjunctiva. It is safer for the surgeon to manage the problem under control of a microscope using micro swabs or a wick, or by opening the conjunctiva, rather than by digital massage through closed lids.
In Case 3, the patient suffered substantial and irreversible visual loss after pterygium surgery. The precise pathogenesis remains unclear, but the fundal appearance, ocular coherence tomography, and electro retinography results suggest inner retinal dysfunction at the macula. Bupivacaine toxicity to the retina was excluded because there is now considerable experience using intracameral bupivacaine during cataract surgery, with no evidence of any damage to vision (15).
The lack of recovery over time is also inconsistent with a toxic pathogenesis of the maculopathy. Inner retinal ischemia seems the most likely explanation for the irreversible changes found. The flow velocities in the retinal vessels were reduced on ultrasound examination, despite a normal fluorescein angiogram. No systemic condition has been identified in this patient, which might predispose to retinal ischemia. However, a number of local causes need to be considered. Spikes in IOP, either because of the volume of anesthetic or to manipulation of the globe by the surgeon or anesthesiologist, may compromise retinal perfusion. Even without changes in IOP, retrobulbar block, peribulbar block, and STB all cause a marked decrease in pulsatile ocular blood flow, persisting at least for 10 minutes after the injection (16,17). The differential diagnosis of CRAO relevant to this case includes atherosclerotic disease, valvular heart disease, carotid artery disease, temporal arteritis, vasculitis, or hypercoagulable states. All of which have been excluded. Some rare causes of CRAO are compression from retrobulbar hemorrhage (18), constriction of the retinal artery associated with use of nasal topical epinephrine (19), or cocaine (20). Furthermore, retinal infarction is a recognized complication of ocular migraine (21). None of the above causes were implied in this case, but in the absence of embolic disease, the history of transient visual loss six months earlier favors the possibility of vasospasm. The risk of retinal ischemia after LA is very small but deserves consideration in cases where the history or examination suggests a degree of preexisting compromise to retinal perfusion.
In case 4, cataract extraction was performed with ineffective STB without sedation and without anesthetic staff being present. The patient felt discomfort during the later stage of the operation and was squeezing her eyes. This increased the IOP and made the operation very difficult and may have caused a hemorrhage from ciliary angle vessels. During a later operation, a scarred retinal break was found and interpreted to be the old entry site of a needle. We find it inconceivable that a blunt STB cannula would have penetrated the sclera, although this possibility cannot be excluded. This type of blunt needle has been in use for almost 10 years, and a perforation has never been reported. The likely sequence of events was failure of the STB and residual ocular muscle function, allowing unexpected globe movement during surgery. This in turn may have led to the accidental perforation of the globe with a sharp needle that was used during the surgical procedure. Enough time should be allowed for the block to be applied and assessed before surgery begins. Especially during high turnover cataract surgery, this is best achieved by the presence of a dedicated anesthesiologist.
In the fifth case, the patient became unresponsive for nearly three hours after STB. CNS depression can be a consequence of intravascular absorption of LA or direct spread into the cerebrospinal fluid. As the milligram dose (mass) of LA drug used for effective clinical results in STB is relatively small, IV injection of the LA would be unlikely to cause problems other than the mildest of systemic symptoms (22). We assume that these central effects were caused by direct spread of the LA into the subarachnoid space. The incidence of CNS depression from intrathecal anesthetic injection has been reported as between 1 in 350 and 1 in 500 when sharp needles are used (23). A case of CNS depression after blunt-needle STB has not been described in detail previously, but the national survey of LA for ocular surgery gives some indication of the incidence of this complication (10). One of 4380 patients who received STB became unresponsive, representing an incidence of 2.28 per 10,000 blocks. With ultrasound techniques, it can be demonstrated that the LA fluid tracks around the globe into the retrobulbar space after sub-Tenon’s injection and pools around the optic nerve sheath (24). In this case of CNS depression, it is not clear how the drug then crossed the dural sheath, unless it is postulated that there was a defect. It is possible that the tip of the Westcott scissors may have perforated the dural sheath upon dissection of the sub-Tenon’s space. That this is possible is shown (Fig. 1) during an optic nerve fenestration operation on a patient similar to Case 5. The optic nerve sheath is exposed and is within reach of the scissor tip. This simulation indicates that during dissection of sub-Tenon’s space, the optic nerve sheath can be injured. CNS depression after STB is rare, and the extent of brainstem anesthesia is exceedingly variable. This case illustrates that anesthetic cover remains an essential prerequisite where LA eye blocks are performed.
The above described complications range from self-limiting to irreversible loss of vision and to CNS depression. Irrespective of the introduction of the blunt needle, STB is still associated with small but significant risks, both to sight and to life.
1. Eke T, Thompson JR. The National Survey of Local Anaesthesia for Ocular Surgery. I. Survey methodology and current practice. Eye 1999; 13: 189–95.
2. Oshika T, Amano S, Araie M, et al. Current trends in cataract and refractive surgery in Japan: 1998 survey. Jpn J Ophthalmol 2000; 44: 268–76.
3. Kollarits CR, Jaweed S, Kollarits FJ. Comparison of pain, motility, and preoperative sedation in cataract phacoemulsification patients receiving peribulbar or sub-Tenon’s anesthesia. Ophthalmic Surg Lasers 1998; 29: 462–5.
4. Stevens JD. A new local anaesthesia technique for cataract extraction by one quadrant sub-Tenon’s infiltration. Br J Ophthalmol 1992; 76: 670–4.
5. Stevens JD. Curved, sub-tenon cannula for local anesthesia. Ophthalmic Surg 1993; 24: 121–2.
6. Olitsky SE, Juneja RG. Orbital hemorrhage after the administration of sub-Tenon’s infusion anesthesia. Ophthalmic Surg Lasers 1997; 28: 145–6.
7. Frieman BJ, Friedberg MA. Globe perforation associated with subtenon’s anesthesia. Am J Ophthalmol 2001; 131: 520–1.
8. Redmill B, Sandy C, Rose GE. Severe orbital cellulitis following sub-Tenon’s local anaesthesia. Eye 2001; 15: 554–6.
9. Jaycock PD, Mather CM, Ferris JD, Kirkpatrick JN. Rectus muscle trauma complicating sub-Tenon’s local anaesthesia. Eye 2001; 15: 583–6.
10. Eke T, Thompson JR. The National Survey of Local Anaesthesia for Ocular Surgery. II. Safety profiles of local anaesthesia techniques. Eye 1999; 13: 196–204.
11. Ernest JT, Goldstick TK, Stein MA, Zheutlin JD. Ocular massage before cataract surgery. Trans Am Ophthalmol Soc 1985; 83: 205–17.
12. Blanksma LJ, Hooijmans JM. Vascular tufts of the pupillary border causing a spontaneous hyphaema. Ophthalmologica 1979; 178: 297–302.
13. Feldman ST, Deutsch TA. Hyphaema following Honan balloon use in Fuchs’ heterochromic iridocyclitis. Arch Ophthalmol 1986; 104: 967.
14. Kallio H, Paloheimo M, Maunuksela EL. Haemorrhage and risk factors associated with retrobulbar/peribulbar block: a prospective study in 1383 patients. Br J Anaesth 2000; 85: 708–11.
15. Anderson NJ, Nath R, Anderson CJ, Edelhauser HF. Comparison of preservative free bupivacaine vs. lidocaine for intracameral anesthesia: a randomized clinical trial and in vitro
analysis. Am J Ophthalmol 1999; 127: 393–402.
16. Hulbert MF, Yang YC, Pennefather PM, Moore JK. Pulsatile ocular blood flow and intraocular pressure during retrobulbar injection of lignocaine: influence of additives. J Glaucoma 1998; 7: 413–6.
17. Pianka P, Weintraub-Padova H, Lazar M, Geyer O. Effect of sub-Tenon’s and peribulbar anesthesia on intraocular pressure and ocular pulse amplitude. J Cataract Refract Surg 2001; 27: 1221–6.
18. Kraushaar MF, Seelenfreund MH, Freilich DB. Central retinal artery closure during orbital haemorrhage from retrobulbar injections. Trans Am Acad Ophthalmol Otolaryngol 1974; 78: 65–70.
19. Maaranen TH, Mantyjarvi MI. Central retinal artery occlusion after a local anaesthetic with adrenaline on nasal mucosa. J Neuroophthalmol 2000; 20: 234–5.
20. Wallace RT, Brown GC, Benson W. Sudden retinal manifestations of intranasal cocaine and methamphetamine abuse. Am J Ophthalmol 1992; 2: 158–60.
21. Safran AB. Vascular complications of ocular migraine. Klin Monatsbl Augenheilkd 1985; 186: 491–4.
22. Hamilton RC. Complications of ophthalmic regional anaesthesia. In: Finucane BT, ed. Complications of regional anesthesia: complications of ophthalmic regional anesthesia. Philadelphia: Churchill Livingstone, 1999: 47–8.
23. Hamilton RC. Brain-stem anesthesia as a complication of regional anesthesia for ophthalmic surgery. Can J Ophthalmol 1992; 27: 323-5.
24. Winder S, Walker SB, Atta HR. Ultrasonic localization of anesthetic fluid in sub-Tenon’s, peribulbar, and retrobulbar techniques. J Cataract Refract Surg 1999; 25: 56–9.