Limbal anaesthesia versus topical anaesthesia for clear corneal phacoemulsification


Carlo Cagini
Policlinico Monteluce
Via Brunamonti
Perugia 06100
Tel: + 39 075 57 20 357
Fax: + 39 075 57 83 951


Purpose: To compare the safety and clinical efficacy provided by limbal anaesthesia with topical anaesthesia in cataract surgery.

Methods: A total of 117 consecutive patients undergoing routine cataract surgery were randomly assigned to receive limbal or topical anaesthesia. Limbal anaesthesia was administered with a cellulose ophthalmic sponge soaked in preservative-free lidocaine hydrochloride 4% applied to the temporal perilimbal area for 45 seconds immediately before surgery. For topical anaesthesia lidocaine 4% was instilled in each patient at 10-min intervals four times before surgery. We studied phaco time, perioperative pain, visual outcome and intraoperative complications. The level of intraoperative pain was scored on a scale of 1−10, where 1 = no pain and 10 = severe pain.

Results: 55 patients (91.6%) in the topical group and 54 patients (94.7%) in the limbal group tolerated the procedure well, giving pain scores of 1−3, with no statistical difference. No patients in either group required supplemental anaesthesia and no intraoperative complications were recorded. No eyes had epithelial defects at the end of surgery or at postoperative check-ups.

Conclusion: Limbal anaesthesia in cataract surgery is safe and the two anaesthesia techniques do not present differences in the degree of analgesia achieved.


In recent years the popularity of topical anaesthesia in cataract surgery has grown progressively (Leaming 2000) due to its advantages over injected local anaesthesia; it causes minimal pain during administration, has an absence of intraoperative or postoperative complications (ptosis, globe or optic nerve lesions), and allows for faster postoperative recovery. In topical anaesthesia, multiple administrations of eyedrops (3–5 times) are required to achieve analgesia, and topical anaesthetics have toxic effects on the corneal epithelium. This can decrease visibility for the surgeon and make surgery more difficult, cause discomfort during the early postoperative period, reduce lacrimation and, on rare occasions, lead to serious keratopathy (Liu et al. 1993).

Alternative techniques of surface anaesthesia have been proposed to avoid multiple administrations of anaesthetics and thus any possible toxic effects on the cornea (Bloomberg & Pellican 1995; Koch 1999; Lanzetta et al. 2000).

We proposed an alternative to topical anaesthesia, namely limbal anaesthesia (Cagini et al. 2004). The technique consists of applying a cellulose ophthalmic sponge, soaked in preservative-free lidocaine hydrochloride 4%, to the limbal area for 45 seconds immediately before starting surgery.

The advantage is the absence of epithelial involvement, leading to greater comfort for the patient and quick visual recovery. To verify the effectiveness and safety of limbal versus topical anaesthesia, we conducted a randomized study comparing these two techniques of anaesthesia in patients undergoing elective cataract surgery and intraocular lens (IOL) implantation.

Patients and Methods

The study involved 117 consecutive patients (49 males, 68 females) who underwent elective cataract surgery; 60 were randomly assigned to topical anaesthesia (group 1) and 57 to limbal anaesthesia (group 2).

There were no patients in either group undergoing anti-glaucoma treatment. In group 1, 13 patients were undergoing therapy for diabetes, while group 2 had 10 patients undergoing such treatment. To minimize bias, all operations were performed at the Eye Clinic Institute of Perugia by only one surgeon (CC). All types of cataracts were included in this study. Patients were excluded if they had inflammation or injury, had undergone previous ocular surgery, or demonstrated barriers to co-operation and communication during surgery (dementia, hearing impairment, etc.). All patients gave informed consent to the surgery.


On the day of surgery, the head nurse in day surgery randomly assigned, by dice rolling, eligible cases to either topical anaesthesia (group 1) or limbal anaesthesia (group 2). In the preoperative holding area, the head nurse of the surgery wing prepared two different series of 1-ml sterile syringes on a daily basis. The syringes in the first series contained balanced salt solution (solution A), whereas the syringes in the second series contained preservative-free lidocaine hydrochloride 4% (solution B). The syringes were identical and could be distinguished only by the identifying initials on them. At this point, in the preoperative holding area an eyedrop was instilled in each patient at 10-min intervals four times before surgery, using solution A for the patients in group 1 and solution B for those in group 2.

Before starting surgery, limbal anesthesia was performed in all patients by applying a cellulose ophthalmic sponge to the limbal area for 45 seconds. The sponge for group 1 patients was soaked in solution B, whereas for those in group 2 the sponge was soaked in solution A. Thus, the surgeon, the nursing staff and the patients themselves were masked with regard to the type of anaesthesia used on each patient.

Surgical procedure

All patients had identical preoperative topical preparation with tropicamide 0.5% (Visumidriatic 0.5%®, Visufarma Srl, Rome, Italy), cyclopentolate 1% (Ciclolux®, Allergan SpA, Rome, Italy), cyprofloxacine 0.3% Ofloxacin (Exocin®, Allergan), diclofenac 0.1% (Voltaren Ofta®, Novartis SpA, Origgio, Italy) drops at 10-min intervals four times before surgery. One drop of benoxinate 0.4% and povidone-iodine 5% was instilled as soon as the patient was stretched out on the operating table, before patient monitoring and preparation were completed. Skin preparation was then performed with povidone-iodine 10%, the patient was draped and the lid speculum was inserted. At this point, the sponge was applied at the corneal limbus for 45 seconds. Before starting surgery, the povidone-iodine 5% was rinsed from the conjunctiva. Surgery was identical in all eyes and comprised a three-plane, 3.2-mm temporal clear corneal tunnel, two side-port incisions, continuous curvilinear capsulorhexis, hydrodissection with balanced salt solution (BSS®, Industria Terapeutica Splendore, Casoria, Italy) with 0.2–0.4 m1 of preservative-free lidocaine 1%, bimanual phacoemulsifications and bimanual irrigation/aspiration of the residual lens cortex. A foldable IOL was implanted with the injector in the capsular bag. During all surgical procedures, we used high molecular weight sodium hyaluronate viscoelastic. None of the patients received systemic sedatives.

On completion of the operation, the surgeon recorded any verbal expression of pain made by the patient during surgery, as well as any complications or the need for supplemental anaesthesia. The surgeon also listed any pain or discomfort due to tissue handling demonstrated by the patient during surgery, rating it on a scale of 0–5.

In a postoperative area outside the operating theatre, at 30−60 mins post-surgery, the same observer asked the patients to grade the levels of pain felt during the delivery of anaesthesia and during surgery. This person was unaware of the type of anaesthesia used during surgery. Each patient was shown a 0−10 visual analogue pain scale with both numeric and symbolic indices. If the patient was unable to see the scale or read the text, the scale was described and a verbal response obtained.

Statistical analysis

Statistical analysis was performed using SAS Version 8.1. Comparisons between topical versus limbal anaesthesia were performed using Student's t-test, Mann–Whitney rank sum test or Fisher's exact test, depending on data type. A p-value <0.05 was considered significant.


No patients in either group required supplemental anaesthesia and no intraoperative complications were recorded. The mean phaco time was 49.1 ± 18.5  seconds(8–101 seconds) in the topical anaesthesia group and 50 ± 20.2 seconds (18–120 seconds) in the limbal anaesthesia group. The difference between the two groups was not statistically significant.

No epithelial defects were detected at the end of the surgery and at the time of medication.

In the topical group the surgeon recorded an expression of pain in 12 patients (20%), and in the limbal group an expression of pain was likewise recorded in 12 patients (21%); the mean score was 0.3 ± 0.69 in the topical group (group 1) and 0.26 ± 0.58 in the limbal group (group 2) (Fig. 1). Table 1 indicates the phases of the procedure in which the patient perceived pain, divided to reflect the two groups of patients. Surgeon discomfort was noted in 11 patients (18.3%) in the topical group, and in 9 patients (15.8%) in the limbal group, while the mean score was 0.3±0.72 and 0.29±0.78 respectively (Fig. 2). In these cases as well, Fisher's exact test underscored the fact that the pain levels perceived by the surgeon and surgeon discomfort were not related to the type of anaesthesia.

Figure 1.

Pain score distribution recorded by the surgeon during the surgery.

Table 1.  The phases of the procedure in which patients perceived pain, in the two groups of patients.
Scleral contact24
Corneal cut20
Iridal contact41
IOL insertion47
Figure 2.

Surgeon discomfort distribution.

With regard to the evaluation of pain indicated by the patient and recorded by an independent observer 30–60 mins after surgery, the mean score was 1.3 ± 0.9 in the topical group (group 1) and 1.33 ± 1.1 in the limbal group (group 2). Moreover, the percentage of patients who did not experience any pain whatsoever throughout the surgical procedure totalled 18.3% in group 1 (topical anaesthesia) and 19.3% in group 2 (limbal anaesthesia) (Fig. 3). In fact, in patients who received topical anaesthesia, 26 had a pain grade of 1 (43%), 18 (30%) grade 2, four (7%) grade 3, and one (2%) grade 4. In the group that received limbal anaesthesia, 23 patients (40%) had a pain grade of 1, 20 (35%) grade 2, two (4%) grade 3, and one (2%) of grade 7 (Table 2). Fisher's exact test indicated that the pain levels perceived by patients were unrelated to the type of anaesthesia.

Figure 3.

Pain score distribution recorded 30’-60’ after surgery.

Table 2.  Percentage of pain score distribution recorded 30-60’ after surgery
Grade 018%19%
Grade 143%40%
Grade 230%35%
Grade 37%4%
Grade 42% 0
Grade 5 0 0
Grade 6 0 0
Grade 7 02%


The goal of this study was to verify if the use of limbal anaesthesia in cataract surgery provides the same safety and effectiveness as the current gold standard in this field, namely topical anaesthesia. In order to do this, we performed a randomized, double-blind study using the same anaesthetic in both cases. This study design thus allowed us to avoid the bias arising from the use of two different anaesthetics.

Analysis of the data demonstrates that limbal anaesthesia is a technique that is safe and effective, on a par with the safety and effectiveness offered by topical anaesthesia. In fact, the levels of pain felt by patients and reported at the end of the operation were low in both groups, without any significant differences. Moreover, neither group required additional anaesthesia in order to complete surgery safely.

Analysis of the individual patients shows that 18.3% and 19.3% in the topical and limbal groups, respectively, reported that they experienced no pain whatsoever during the operation. Adding these patients (‘pain score 0’) to those who, at the end of surgery, reported pain scores of 1 and 3, (i.e. a very slight sensation of tissue handling), yields two groups, composed of 55 patients (91.6%) for the topical group and 54 patients (94.7%) for the limbal group. They represent the total number of subjects who, at the end of the procedure, reported either no pain or a feeling of ‘pressure’ on the eye, probably due to differences in intraocular pressure arising during the various phases of the procedure, or extremely brief piercing pain, probably due to certain phases of surgery such as scleral or iris contact. The data we collected are similar to the data reported by other authors.

Likewise, the percentage of patients in whom the surgeon recorded pain or discomfort at the end of the surgical procedure was low, and was similar in the two groups: 20% in group 1 and 21% in group 2. No patients demonstrated a pain level higher than 3; a score of 3 was recorded in only 3.3% and 1.7% of cases in the topical group and the limbal group, respectively.

Moreover, in our experience we did not report any intraoperative complications, nor did we observe any increased surgical difficulties due to intraoperative pain.

The parity in the phaco times of the two groups also suggests that there were no particular differences in the intraoperative management of the two groups of patients.

Therefore, based on these results the two anaesthesia techniques do not present differences in the degree of analgesia achieved, nor are there differences in terms of the difficulties encountered by the surgeon as a result of any discomfort or pain demonstrated by the patient.

In our opinion, limbal anaesthesia offers advantages over topical anaesthesia. The latter requires repeated instillation of the anaesthetic to ensure adequate analgesia. Insufficient administration of the agent, as well as patient preparation too far ahead of the operation, can make the surgical procedure painful or unsafe. Inversely, excessive instillation of the drug can jeopardize intraoperative visibility due to possible distress of the corneal epithelium.

With limbal anaesthesia, the surgeon is confident that the appropriate dose of anaesthetic has been administered at the right time, fully respecting the corneal epithelium. At the same time, the surgeon does not need to modify his or her normal modus operandi, and the time needed to administer anaesthesia can be used to accustom the patient to the light of the microscope and explain the various phases of the operation. In our experience, this practice has been very positive, as it augments the patient's peace of mind.

Limbal anaesthesia indubitably simplifies the preoperative preparation of the patient, which is reduced solely to inducing mydriasis. This aspect can be very important, especially in a busy outpatient practice.

How can such brief surface anaesthesia achieve adequate analgesia? As with other techniques involving perilimbal anaesthesia, the precise mechanism of this anaesthesia is not completely understood. In limbal anaesthesia, the drug-soaked sponge is applied at the limbus for a short period of time. Consequently, only a modest amount of lidocaine penetrates the anterior chamber. In topical anaesthesia, the drug must generally be administered 20–30 mins before surgery, with multiple administrations to permit the drug to penetrate the cornea and ensure that adequate concentrations reach the anterior chamber. This makes it possible to achieve adequate iris and ciliary body anaesthesia. Similar considerations can be made for anaesthesia with anaesthetic jelly, which is also applied about 20–30 mins before surgery.

Nevertheless, recent observations indicate that the intracameral concentration of anaesthetic agents is not correlated with the level of intraoperative analgesia (Crandall et al. 1999; Bardocci et al. 2003). In other words, a high concentration of lidocaine in the anterior chamber does not guarantee good analgesia. In fact, at times good analgesia is achieved in patients even with low anaesthetic concentrations in the anterior chamber.

In limbal anesthesia, the drop of anaesthetic that is instilled before applying the disinfectant makes it possible to anaesthetize the conjunctiva and thus apply the blepharostat without any problems. The analgesia achieved with the next application of the sponge can be justified with the block of the pericorneal and annular plexus: this would explain the analgesic effect on the entire circumference of the cornea. In fact the patient does not feel pain during contact with the main point of entry, during the side-port incisions or during accidental contact with other portions of the limbus or cornea. It is interesting to note that laser iridotomies can be accomplished with a single drop of topical anaesthetic, which is necessary strictly for permitting the application of the contact lens. In fact, in some cases the treatment can be performed even without applying the contact lens.

Thus, the extent to which the iris and ciliary body contribute to the sensation of pain during cataract surgery has yet to be clarified, and it is likely that intraocular tissues are not as sensitive as once thought.

Nevertheless, it must be noted that in limbal anesthesia the anaesthetic agents may also reach the iris and ciliary body by the conjunctival−scleral route with iris and ciliary body anaesthesia (Ahmed & Patton 1985; Schoenwald et al. 1997).

In conclusion, our experience demonstrates that limbal anesthesia can be considered an evolutionary and valid alternative to the multiple preoperative administration of anaesthesia, as it ensures optimum analgesia, permits safe surgery and does not have any toxic effects on the corneal epithelium. We are convinced that its simplicity makes it the anaesthesia of choice in clear corneal cataract phacoemulsification.