Secondary Logo

Journal Logo

Ambulatory Anesthesia

Painless Needle Insertion in Regional Anesthesia of the Eye

Vaalamo, Mikko O. MD; Paloheimo, Markku P.J. MD, PhD; Nikki, Pertti H. MD, PhD

Author Information


The popularity of ambulatory surgery for cataract extraction and intraocular lens implantation procedures has increased greatly during the last few years. Patient comfort in the ambulatory setting is very important [1].

Premedication with benzodiazepines and/or opioids has been used before the block to relieve anxiety and pain associated with the injections [2]. These drugs can delay street fitness postoperatively. Patients presenting for eye surgery are often elderly and would benefit from painless eye block without the use of sedatives.

Painless needle insertion in retrobulbar or peribulbar block has not been studied widely. The transconjunctival route of needle insertion was found significantly less painful than the percutaneous route when amethocaine drops were used to anesthetize the conjunctiva [3]. However, amethocaine stings when applied to the eye [4].

The purpose of these two studies was to evaluate a new technique of applying topical anesthesia to the conjunctiva before the retrobulbar block. In Study 1, we examined this technique by comparing two topical anesthetics, oxybuprocaine and lidocaine, with balanced salt solution (BSS) in a clinical setting. Study 2 was made with volunteers serving as their own controls to compare oxybuprocaine applied by two different methods to the conjunctiva. Pain was assessed with a visual analog scale (VAS) and quantitative surface electromyography (qEMG). The ability of qEMG to reveal subtle painful mimics during insertion of the needles was compared with the VAS score while studying the effects of topical anesthesia.


Study 1

The study protocol was approved by the local ethics committee. Written, informed consent was obtained from each patient. Ninety patients scheduled for cataract day-case surgery under local anesthesia were enrolled in the study. The patients had been fasting but were allowed liquids and were asked to take their daily medication. No premedication was used. Patients who had received analgesic or sedative medication, or who did not understand the principle of VAS, were excluded from the study. The patients were randomly allocated into three groups, with 30 patients in each.

An intravenous catheter (20 gauge, Viggo, Sweden) was inserted in the dorsum of the hand. qEMG was obtained from adhesive skin electrodes placed over the lower third of both sternocleidomastoid muscles and platysma with the ground electrode on the chest. The anesthesia and brain activity monitor (ABM; Datex/Instrumentarium Corp., Finland) was used to graphically record qEMG during the injections on a strip recorder with paper speed 1 cm/s. An increase of 5 mm or larger in the qEMG activity within 10 s from the time of needle insertion was considered as a response to the insertion.

The study drugs were 4% lidocaine (Lidocain 40 mg/mL; Medipolar, Finland) (Group L), 0.4% oxybuprocaine (benoxinate) (Oftan-obucain 4 mg/mL; Leiras, Finland) (Group O), and balanced salt solution (BSS; Alcon, TX) as control drug (Group C). The trial drugs were drawn into 2-mL syringes and labeled with the patient's number by a nurse not participating in the study. Five minutes before the application of the study drug, a drop of oxybuprocaine was administered conjunctivally. Two cotton tip sticks were soaked in the trial drug for 1 min and then placed conjunctivally on the caruncle and on the inferotemporal fornix for 1 min. After 1 min, the first insertion was made using a 12-mm, 30-gauge needle (Microlance; Becton Dickinson, Fraga, Spain) into the medial corner of the caruncle. The needle was directed to avoid the medial rectus muscle [5]. The local anesthetic solution was a mixture containing 10 mL of lidocaine 2%, 10 mL of bupivacaine 0.75%, 0.1 mL of epinephrine 0.1%, and 0.5 mL (75 IU) of hyaluronidase. The injection (3 mL, 1 mL/15 s) was started 15 s after the needle insertion to distinguish the pain caused by the needle insertion alone. Gentle digital massage to enhance subcutaneous spread of the anesthetic was applied for 15 s after the injection.

A second injection (3 mL) was administered similarly with a 31-mm, 27-gauge needle (PrecisionGlide; Becton Dickinson, NJ). The entry point was as far lateral as possible in the inferotemporal fornix of the conjunctiva, directed tangentially to the globe, and aimed behind the globe within the muscle cone. The eye was in the primary gaze position. Ocular compression (Autopressor; Storz, Germany) was applied for 10 min at a pressure setting of 25 mm Hg. If akinesia was incomplete at 10 min, a supplemental injection was administered intraorbitally either inferolaterally or nasally.

A 50-cm (0.5 mr) VAS was used to rate the pain of needle insertions, with 0 representing no pain and 50 being the worst pain that the patient could imagine. First, the insertion of the intravenous catheter was rated (VAS cat) to confirm that VAS was understood, and to see if there would be a difference in pain scores of this quite standard pain between the study groups. The pain on catheter insertion was compared to pain on orbital insertion. The VAS scores of the eye block needle insertions were requested immediately after the injections (medial injection was VAS1). VAS scores more than 15/50 were considered to represent pain and were calculated.

All patients were monitored during the anesthesia and perioperatively with pulse oximetry for oxygen saturation, continuous electrocardiogram, and noninvasive arterial blood pressure measurements intermittently.

All data were expressed as median (range). Demographic data and the VAS scores from the needle insertions were compared with the Kruskal-Wallis one-way analysis of variance. The VAS scores between the control group and the two treatment groups were compared using the Mann-Whitney U-test. Differences between groups in terms of positive qEMG responses and proportions of VAS scores of more than 15 were compared using Fisher's exact test. A P-value less than 0.05 was considered significant.

Study 2

The second study with volunteers was made to compare our previous practice of dropping oxybuprocaine three times on the eye before injections with the new technique of using additional cotton tip sticks soaked in oxybuprocaine.

Seven healthy ASA grade I-II volunteers were studied. The qEMG was measured in the same way as in Study 1. The subjects were given three consecutive drops of oxybuprocaine into both eyes at 3-min intervals. A cotton tip stick soaked in oxybuprocaine was placed on the caruncle of one eye and another soaked in BSS was placed to the caruncle of the other eye for 1 min.

The subjects were not told which topical anesthetics had been used. After 1 min, the 30-gauge needle was inserted in the medial corner of the caruncle and held in place for 15 s. No anesthetic was injected. A second needle insertion was done immediately after the removal of the first one to the caruncle of the other eye. The VAS scores were assessed immediately after withdrawal of the needles.

Fisher's exact test was used to compare the proportion of subjects who had the VAS score more than 0 and the positive qEMG responses.


Study 1

The patients in the three groups had comparable characteristics except for the sex distribution Table 1. There was no difference in VAS scores or qEMG response between females and males. For the medial injection alone (VAS1), the difference between Group C and the treatment group was significant Table 2. The VAS scores for intravenous catheter insertion were similar in all groups Table 3. The proportion of VAS scores more than 15 from the orbital block insertions were significantly higher in Group C (13/30) than in Group O (3/30) or Group L (6/30) Table 3. The proportions of patients with a VAS score of more than 15 were calculated because analgesic techniques that produce VAS values in the range of 0-30 on a scale of 0-100 have been reported to represent adequate analgesia [6].

Table 1
Table 1:
Demographic Data of the Patients in Study 1
Table 2
Table 2:
VAS Scores of Needle Insertions in Study 1
Table 3
Table 3:
Number of Patients with Positive Quantative Surface Electromyography Response to the Needle Insertions in Study 1, Number of Patients with Visual Analog Scale (VAS) Score More Than 15, and Number of Patients with More Pain on Orbital Injections Than on Catheter Insertion (VAS > VAS cat)

Positive qEMG responses were most often found in the control group but the difference was not statistically significant Table 3. All operations were completed successfully and there were no complications.

Study 2

The use of an oxybuprocaine cotton tip stick in addition to three drops of oxybuprocaine made the needle insertion virtually pain free. Only one of seven subjects rated pain more than 0 on VAS when the cotton tip stick contained oxybuprocaine. In the other eye, the three drops of oxybuprocaine alone prevented pain totally only for one subject (P < 0.05). Only two subjects had a positive qEMG response in the eye that had had oxybuprocaine also on the cotton tip stick, whereas all seven had a positive qEMG response for the other eye (P < 0.05).


In these studies we have evaluated a new technique of topical conjunctival anesthesia with two topical anesthetics compared to BSS in reducing the pain of needle insertion in eye block. In Study 2 we compared the new technique with one previously used in our hospital.

We applied the medial injection first because some patients blink strongly [5]. The medial injection relaxes the orbicular muscle and may reduce the risk of needle penetration of the globe, especially if the patient has deeply situated eyes. On the other hand, if the injections are made in reverse order, i.e., the medial injection after the inferolateral one, the pain caused by the medial injection is probably reduced.

There have been studies comparing different premedications before regional anesthesia of the eye. Alfentanil, 12.5 micro gram/kg, injected into the deltoid muscle relieved pain associated with the peribulbar block, which was performed 15 min after premedication [2]. Even for the alfentanil group the VAS +/- SD was 35.0 +/- 21.9 (on a scale of 0-100) which is more than our control group scored. When premedication is not used, drug side effects such as delayed recovery, postural hypotension, and emetic sequelae can be avoided [1].

Oxybuprocaine was first introduced for opththalmic anesthesia in 1954 [7]. It is mostly used for corneal anesthesia. Lidocaine has been widely used for topical anesthesia of mucous membranes. It does not affect regeneration of the corneal epithelium and it produces adequate anesthesia [8]. Lidocaine has been used recently in a cataract operation as the sole topical anesthesia [9]. We did not find a significant difference between oxybuprocaine and lidocaine. Lidocaine spray on oral mucosa has been found to produce maximum analgesia as late as 4-5 min after application [10]. Therefore, a pause longer than 1 min between the application of lidocaine and the needle insertions might reduce pain even more.

Facial qEMG is used to monitor the depth of anesthesia and has been used to monitor pain also [11]. We used sternocleidomastoid muscles and platysma to monitor qEMG because monitoring the upper facial muscles of the contralateral side of the patient's face was too sensitive to distraction during the procedure. EMG reflects a startle-like reflex at needle insertion. We had hoped to get more objective information of the reaction to the painful topical stimulus but it did not correlate with the VAS scores in Study 1.

Applying topical lidocaine or oxybuprocaine with cotton tip sticks to the conjuctiva proved to reduce the pain of the needle insertions of the orbital block significantly when compared to placebo or to our previous practice of dropping a topical anesthetic on the eye. We recommend the use of both anesthetic drops and cotton tip sticks for optimal conjunctival anesthesia before regional anesthesia of the eye.


1. Hamilton RC, Gimbel HV, Strunin L. Regional anaesthesia for 12000 cataract extraction and intraocular lens implantation procedures. Can J Anaesth 1988;35:615-23.
2. Virkkila MEJ, Ali-Melkkila TM, Kanto JH. Premedication for outpatient cataract surgery: a comparative study of intramuscular alfentanil, midazolam and placebo. Acta Anaesthesiol Scand 1992;36:559-63.
3. Sweeney EJ, Barber K, Prosser JA. A comparison of percutaneous and perconjunctival routes of administration or peri-ocular anaesthesia for day case cataract surgery. Anaesthesia 1993;48:336-8.
4. Sobol WM, McCrary JA. Ocular anesthetic properties and adverse reactions. Int Ophthalmol Clin 1989;29:195-9.
5. Hustead RF, Hamilton RC, Loken RG, et al. Periocular local anesthesia: medial orbital as an alternative to superior nasal injection. J Cataract Refract Surg 1994;20:197-201.
6. Mantha S, Thisted R, Foss J, et al. A proposal to use confidence intervals for visual analog scale data for pain measurement to determine clinical significance. Anesth Analg 1993;77:1041-7.
7. Sclegel HE, Swan KC. Benoxinate (dorsacaine) for rapid corneal anesthesia. Arch Ophthalmol 1954;51:663-70.
8. Marr WG, Wood R, Senterfit L, et al. Effect of topical anesthesia on regeneration of corneal epithelium. Am Ophthalmol 1957;43:606-10.
9. Williamson CH. Clear corneal incision with topical anesthesia. In: Gills JP, Hustead RF, Sanders DR, eds. Ophthalmic anesthesia. Thorofare, NJ: SLACK Inc., 1993:176-83.
10. Schonemann NK, van der Burght M, Arendt-Nielsen L, et al. Onset and duration of hypoalgesia of lidocaine spray appliet to oral mucosa--a dose response study. Acta Anaesthesiol Scand 1992;36:733-5.
11. Paloheimo M. Quantitative surface electromyography (qEMG): applications in anaesthesiology and critical care. Acta Anaesthesiol Scand 1990;34(93 Suppl):1-50.
© 1995 International Anesthesia Research Society