Traditionally, penetrating eye injuries have been operated upon under general anaesthesia (GA) in order to diminish local manipulation, which could aggravate the amount of injury and lead to further complications . During induction of GA, the anaesthesiologist should avoid drugs and procedures that could increase intraocular pressure (IOP) such as an increase of central venous pressure, increasing choroidal blood volume (coughing, emesis, Valsalva manoeuvre, respiratory acidosis, hypercapnia) and increase of extraocular muscle tone (succinylcholine, ketamine). Laryngoscopy and endotracheal intubation are the events most likely to cause a significant increase in IOP [2,3]. Pretreatments, including lidocaine intravenously (i.v.) , propofol  or clonidine by mouth  have been suggested to blunt these sympathetic responses. Patients having an open eye injury often have a full stomach that complicates the management of GA. On the other hand, the patient's general condition may be poor and thus ill-fitted for GA.
Injecting a large volume (8–10 mL) of fluid into the orbit (e.g. peribulbar block) may significantly increase IOP  and aggravate a penetrating eye injury. On the other hand, extraocular muscle relaxation due to peribulbar block can reduce IOP . Our hospital has a long tradition of performing major ophthalmic surgery under regional anaesthesia and conscious sedation . After having successfully treated an ASA IV patient with a penetrating eye injury under regional anaesthesia, we decided to study this method further.
The protocol was approved by the head of Ophthalmic Surgery and the Institutional Ethics Committee. After informed consent, adult ASA I-IV patients were studied. The surgeon selected the patients suitable for regional anaesthesia. The ophthalmological inclusion criteria for this study were a penetrating eye injury with a maximum wound length of 8 mm, reaching up to 4 mm posteriorly from the limbus. It was not possible to randomize the patients, but 18 adult patients operated on under GA served as controls. Three experienced staff anaesthesiologists skilled in eye blocks performed the regional anaesthetic blocks. Seven others gave conventional GA. All the patients whose eye injuries did not fulfil the above-mentioned inclusion criteria received GA. All patients fasted for at least 4 h before the induction of anaesthesia.
Patients were given fentanyl 50 μg i.v. and topical anaesthesia to the eye with amethocaine drops (Amethocaine®). A double-injection technique  was used for the combined retrobulbar peribulbar block. The first transconjunctival injection (3–4 mL 1:1 mixture of bupivacaine 0.75% (Marcain®; Astra, Sodertalje, Sweden) and lidocaine 2% (Xylocain®; Astra, Sodertalje, Sweden) with hyaluronidase 0.5 mL (Wydase®; Wyeth, Philadelphia, PA, USA), and epinephrine 1 mgmL−1 0.1 mL) was made at the medial corner of the caruncle with a sharp, disposable 12 mm 30-G needle (Microlance®; Becton Dickinson, Drogheda, Ireland). The second intraconal injection was administered inferolaterally through the conjunctiva with a 31 mm 27-G sharp needle (PrecisionGlide®; Becton Dickinson, Franklin Lakes, NJ, USA). The axial length of the bulbus was not known. Echography could not be carried out owing to the eye injury. A total of 3–4 mL of the same local anaesthetic mixture as above, epinephrine excluded, was injected. The local anaesthetic mixture used in this study was studied in our department . The volume of the local anaesthetic was determined according to the lean body weight of the patient (3 mL for the patients weighing <70kg, 3.5 mL for those weighing 70 kg and 4mL for the patients weighing >80kg). However, if the orbita seemed full, the lids seemed tight or the bulbus began to protrude, the injection was interrupted to avoid an increase in the IOP. No pressure was applied on the globe after finishing the injection. The ocular akinesia was estimated after 10 min and an additional dose of local anaesthetic was injected if necessary.
Before the sedation was begun, all patients received i.v. glycopyrrolate 0.2 mg, and lidocaine 40 mg was administered through the i.v. cannula to diminish burning caused by propofol. A loading dose of propofol (Diprivan®; Milano, AstraZeneca Pharmaceuticals, Italy) 0.15 mg kg−1 i.v. was given followed by a propofol infusion 1.5 mgkg−1 h−1 i.v. The objective was to achieve a light level of conscious sedation so that the patients could understand and remember instructions given and communicate during the operation. The patients were observed and continuously monitored by Cardiocap®, Cardiomac® (Datex, Helsinki, Finland) (electrocardiography, non-invasive blood pressure, end-tidal CO2 and SPO2). If the patient felt pain in the eye during the operation, the time was recorded and local anaesthetic was added by an anaesthesiologist or the surgeon using a retrobulbar or sub-tenon technique, respectively. Immediately after the operation, both the patient and the surgeon were interviewed on their opinion about the quality of the anaesthesia provided. The opinions were recorded on a verbal rating scale of 0–10 (0 = completely unsatisfactory, 10 = excellent). The patients were observed in the postanaesthesia care unit (PACU) for at least 30 min after cessation of the propofol infusion. The criteria for discharge from the PACU were those defined by Aldrete and Krovulik , i.e. the patients were cardiovascularly stable, breathing normally and fully orientated.
After glycopyrrolate 0.2 mg i.v., patients in the control group received routine GA with fentanyl, thiopental or propofol i.v.; vecuronium or rocuronium i.v. was given for muscle relaxation. Anaesthesia was maintained with isoflurane in O2/N2O (30%/70%), and intermittent doses of fentanyl given as needed. Neuromuscular blockade was antagonized with neostigmine mixed with glycopyrrolate.
For postoperative pain, all patients received i.v. ketoprofen or acetaminophen. If needed for postoperative pain, the patients received oxycodone 0.05 mg kg−1 i.v. in the PACU and 0.1 mg kg−1 intramuscularly (i.m.) in the ward. Nausea and emesis were treated with droperidol 1.25 mg i.v. or i.m., or metoclopramide 25 mg rectally. In the ward, the patients were observed according to our clinical routine. In the first 24 h (from the induction of anaesthesia) we recorded the occurrence of postoperative pain and nausea, and analgesic and antiemetic medication.
There were more males than females in the study group (Table 1). Typically, there were two subgroups of patients having a penetrating eye injury: healthy working males who suffered accidents and elderly patients who had fallen at home.
For induction of regional anaesthesia, the mean (range) volume of the local anaesthetic was 7.4 (6.5–8) mL of a lidocaine/bupivacaine mixture. Before surgery, six patients needed an additional retrobul-bar local anaesthetic injection (2.9, range 2–4 mL). During the operation, one patient needed an additional dose (sub-tenon's block 0.2 mL). Six patients fell asleep during the operation and the propofol infusion was interrupted until they awoke. For sedation, the patients received propofol, on average (range) 12.3 (7–20) mg i.v. as a bolus and 1.7 (0.40–3.9) mg−1 kg−1 h as an infusion. All patients were haemodynamically stable and no oculocardiac reflexes were observed. Both the patients and surgeons were very satisfied with this particular anaesthetic method. The patients gave a median (range) score of 10 (8–10) on the verbal rating scale and the median (range) of the surgeon's score was 8 (8–10). No problems or surgical complications related to local anaesthesia were reported by the surgeons.
Patients receiving regional anaesthesia spent 93.4 (35–225) min in the operating room and those receiving GA 135.5 (45–240) min. The operation time was 57.2 (20–115) and 96.7 (25–165) min, respectively. At the end of the operation, all regional anaesthesia patients fulfilled the criteria to be transferred to the ward, but, according to the protocol, they were observed in the PACU for 30 min (mean 33.7, range 15–180 min). Patients receiving GA needed longer observation in the PACU compared with the regional anaesthesia patients (mean 86.7 min, range 30–165 min, P < 0.0001). Their wounds were also more extensive compared with those in the regional anaesthesia group. The regional anaesthesia patients were discharged significantly sooner to the ward than those patients who had received GA (P < 0.0001). Only one patient needed analgesia in the PACU (one dose of oxycodone) after operation under regional anaesthesia compared with four patients in the GA group (seven doses). During the first 24 h postoperatively, one patient in the ward, in the regional anaesthesia group, needed oxycodone compared with two patients in the GA group. No patient had post-operative nausea or vomiting.
An ophthalmologist carefully examined the patients on the first postoperative day. They received a further ophthalmologic examination one month after they had been discharged from hospital. Nine regional anaesthesia patients and seven GA patients had no complications and they had no complaints after the first postoperative month. Four regional anaesthesia and six GA patients underwent vitrectomy within one month. Four regional anaesthesia patients and three GA patients developed cataracts that required operation and one eye was enucleated or eviscerated in both groups. One GA patient lost sight in the eye and one regional anaesthesia patient lost sight from the wounded eye before the accident. One regional anaesthesia patient needed a corneal transplant. One regional anaesthesia patient developed postoperative endophthalmitis. There was no reason to believe that the anaesthesia method used in this study had aggravated the injury.
Local anaesthesia has become increasingly popular for a variety of intraocular procedures . Our study showed that under certain conditions and when given by experienced ophthalmic anaesthesiologists, combined peri- and retrobulbar block also is suitable in penetrating eye injuries. The main techniques used to administer local anaesthesia are topical anaesthesia for cataract surgery, peri- and retrobulbar anaesthesia, or, as in our clinic, a combination of these two techniques  for retinal surgery.
Both peri- and retrobulbar anaesthesia are known to increase intraocular pressure. Stevens and colleagues  found a greater rise in IOP using peribulbar anaesthesia compared with retrobulbar anaesthesia and it was attributed to the greater volume of solution used in the peribulbar technique. Jay and colleagues  reported mean increases of 4.7–6.1 mmHg immediately after retrobulbar injection volumes of 3, 4 and 5 mL. Increases in the IOP did not differ significantly between these three groups.
O'Donoghue and colleagues  compared IOP after retrobulbar (mean 3.5 mL) and peribulbar injection (mean 9 mL) of bupivacaine/lidocaine/hyaluronidase. They described a patient undergoing penetrating keratoplasty for spontaneous corneal perforation associated with bullous keratopathy. The perforation was sealed preoperatively with corneal glue and a contact lens. Their peribulbar injection caused an immediate increase in IOP sufficient to result in displacement of the glue and of the contact lens with a flattening of the anterior chamber.
There is no universally accepted and clinically reliable end-point marker of the local anaesthetic volume needed to provide akinesia and anaesthesia. We usually estimate the volume according to the patient's lean body weight as described above. Frow and colleagues  showed that total upper eyelid droop represents a reliable, reproducible clinical marker for the optimal volume of local anaesthetic injection required to produce satisfactory conditions without unduly increasing intraocular pressure. The mean (range) volume injected in their study was 9.1 (4–15) mL and the increase in IOP was 6.9 (3.2–11.4) mmHg. A slight increase in IOP in the patients studied probably also occurred because the total volume injected for the combined peri- and retrobulbar block was 7.4 (range 6–8) mL, on average.
On the other hand, extraocular muscle relaxation due to peribulbar block can reduce IOP. Nociti and colleagues  found that the IOP increased significantly in the first 1 min and then decreased at 5 and 15 min after the peribulbar block induced with bupivacaine. An experienced anaesthesiologist can interrupt the injection if the orbita seems full, the lids tight or the globe begins to protrude. Note that individuals vary greatly in their IOP response to local anaesthetic injection.
The regional anaesthesia patients reported high satisfaction with the chosen anaesthetic method. Previously, patient-controlled sedation with propofol has been associated with high patient satisfaction during day case cataract surgery [8,15,16] and during extractions of molar teeth . The relevance of the satisfaction rating in our study can be influenced by the sedative dose of propofol. However, lack of adequate analgesia cannot be corrected with sedation and the patient is the only competent assessor of his or her general experience during the surgical procedure.
In the hands of experienced anaesthesiologists, regional anaesthesia appears to be a suitable method for small penetrating eye injuries. It is important that the total volume of the block solution is adjusted to the anatomical conditions of the orbit, and a total volume of 8–10 mL should not be exceeded. The number of patients included in this study was limited and the risk of adverse events cannot yet be excluded. Based on the results here, the eye injuries in co-operative adult patients, with a maximum wound length of 8 mm reaching up to 4 mm posteriorly from the limbus, are now operated upon using combined peri- and retrobulbar anaesthesia in our clinic.
The authors thank Professor P. H. Rosenberg for fruitful discussions during the preparation of the manuscript.
1. Donlon JV. Anesthesia and eye, ear, nose and throat surgery. In: Miller RD, ed. Anesthesia.
New York, USA: Churchill Livingstone, 1994: 2175-2196.
2. Cunningham AJ, Barry P. Intraocular pressure: physiology and implications for anaesthetic management. Can J Anaesth
3. Murphy DF. Anesthesia and intraocular pressure. Anesth Analg
4. Vanacker B, Dekegel D, Dionys J, et al.
Changes in intraocular pressure associated with the administration of propofol. Br J Anaesth
5. Ghignone M, Noe C, Calvillo O, et al.
The effects of clonidine on intraocular pressure. Anesthesiology
6. Frow MW, Miranda-Caballo JI, Akhtar TM, Hugkulstone CE. Single injection peribulbar anaesthesia
: Total upper eyelid drop as an end-point marker. Anaesthesia
7. Nociti JR, Serzedo PSM, Zuccolotto EB, Nunes AMM, Ferreira SB. Intraocular pressure and ropivacaine in peribulbar block
: a comparative study with bupivacaine
. Acta Anaesthesiol Scand
8. Kallio H, Uusitalo R, Maunuksela EL. Randomised, prospective trial: topical anesthesia with or without propofol sedation and retrobulbar/peribulbar anesthesia for cataract extraction. J Cataract Refract Surg
9. Kallio H, Paloheimo M, Maunuksela EL. Hyaluronidase as an adjuvant in bupivacaine
-lidocaine mixture for retrobulbar/peribulbar block
. Anesth Analg
10. Aldrete JA, Krovulik D. A postanesthetic recovery score. Anesth Analg
11. McGoldrick KE, Feitl M, Krupin T. Neural blockade in ophthalmological surgery. In: Cousins MJ, Bridenbaugh PO, eds. Neural Blockade in Clinical Anaesthesia and Management of Pain.
Philadelphia, USA: Lippincott-Raven, 1998: 533-556.
12. Stevens J, Giubilei M, Lanigan L, Hykin P. Sub-tenon, retrobulbar and peribulbar local anaesthesia
: the effect upon intraocular pressure. Eur J Implant Ref Surg
13. Jay WM, Carter H, Williams B, Green K. Effect of applying the Honan intraocular pressure reducer before cataract surgery. Am J Ophthalmol
14. O'Donoghue EO, Batterbury M, Lavy T. Effect on intraocular pressure of local anaesthesia
in eyes undergoing intraocular surgery. Br J Ophthalmol
15. Herrick IA, Gelb AW, Nichols B, Kirby J. Patient-controlled propofol sedation for elderly patients: safety and patient attitude toward control. Can J Anaesth
16. Janzen PRM, Chistys A, Vucevic M. Patient-controlled sedation using propofol in elderly patients in day-case cataract surgery. Br J Anaesth
17. Osborne GA, Rudkin GE, Curtis NJ, et al.
Intra-operative patient-controlled sedation. Anaesthesia