Secondary Logo

Journal Logo

Original Article

Evaluation of 19 cases of inadvertent globe perforation due to periocular injections

Gadkari, Salil S MS

Author Information
Indian Journal of Ophthalmology: Mar–Apr 2007 - Volume 55 - Issue 2 - p 103-107
doi: 10.4103/0301-4738.30702
  • Open


Reports of inadvertent globe perforation during periocular injection are not uncommon.1 Peribulbar anesthesia, though presumed to be safer than retrobulbar injection, is not without risk of this complication.2 Timely detection and management of this sight threatening complication is important, as the incidence of proliferative vitreoretinopathy (PVR) is high. The purpose of this study was to analyze the risk factors, management, and visual outcome of the cases referred with inadvertent globe perforation during periocular injection.

Materials and Methods

We retrospectively analyzed 19 eyes of 19 consecutive patients who took treatment for this condition at our center between January 1998 and December 2004. The number of patients have to be considered in light of the fact that we are a unit restricted to vitreoretina specialty. The period of follow up varied between 6 weeks and 2 years. All patients were treated by one surgeon (the author). Data were retrieved from our medical records by the author.

Information regarding circumstances at the time of the inadvertent perforation was sought verbally from the referring operating surgeon. This included time of detection, type of injection, person giving the injection, clinical setting, and any recognized risk factor. Data thus obtained were entered in a separate form in the patient's records. This practice was routinely followed in the case of patients referred with postoperative complications.

In the entire series, either a 25G (length 26 mm) or 26G (length 13 mm) disposable needles had been used for the injection. All blocks were for cataract surgery, while the sub-conjunctival injections were given for postoperative inflammation. Evaluation of the patient involved recording of visual acuity, complete ophthalmic examination, and B-scan Ultrasonography (USG). Patients with an A scan intraocular lens (IOL) power calculation of +16D or less was labeled as myopes for this study. Where media clarity permitted and retina was flat (4 eyes) laser/cryopexy was performed. Those with dense vitreous hemorrhage and/or retinal detachment (RD) were operated with a standard three-port pars plana vitrectomy (14 eyes) with silicone oil/gas injection within 2-8 weeks of the perforation. Intervention was performed within 10 days of presentation. One case with defective projection of light, (Case number 19) showing a large suprachoroidal hemorrhage was left alone. Postoperative visual acuity was recorded at 6 weeks.

Statistical methods

Categorical data were presented as proportions (%). The Chi-square test was used for testing the difference between proportions and association between the categorical variables. Differences were considered significant at P < 0.05; all P values were two-tailed. Data analyses were performed with the statistical program SPSS version 11.0 (SPSS Inc, Chicago).


Table 1 tabulates the relevant information of each case in the series. Perforation was suspected on the table in 10 (10/19) eyes. Table 2 shows the symptoms and signs in these patients. In patients who were detected later (9/19 eyes), the absence of visual improvement or decreased vision after surgery was the main cause for referral. Clinical features (confirmed with USG in presence of opaque media) including localized vitreous hemorrhage 2 (11%), dense vitreous hemorrhage 8 (42%), RD 8 (42%), hyphema 2 (11%), retinal whitening 1 (5%), and subretinal hemorrhage 1 (5%) were found in the entire series, when examined at our center. Table 3 shows the percentages of the variables studied. Peribulbar injections were given in 10, retrobulbar in six, and subconjunctival in three patients. Thirteen had been injected by ophthalmologists (residents-7, consultants-6), while anesthetists were responsible for 6 (P =0.10). Though 10 cases were detected on table prior to surgery, the surgeon proceeded with surgery prior to referral in 3 [Table 1; case numbers 2, 3, 9]. Anesthetists diagnosed significantly lesser percentage of mishaps at the time of injection compared to the ophthalmologists (17 vs. 69%, P =0.033). Our series also showed that 13/19 cases came from high volume community ophthalmology set ups as compared to 6/19 from private practices (P =0.10).

Table 1
Table 1:
Master chart
Table 2
Table 2:
Symptoms and signs at the time of injection in 10 patients with suspected perforation
Table 3
Table 3:
Percentages of the variables studied

Myopia was the most prevalent risk factor (n =10, P =0.008), compared to other risk factors described in literature. In all 18 operated eyes the break was localized either pre-operatively or during vitrectomy. Break localization was predominantly in the inferotemporal quadrants (P =0.046) compared to others. Double perforation was observed in one eye and no relation to the time of detection could be statistically shown. Vitreous surgery was performed in 14 cases while only four could be treated with laser/cryopexy. Visual acuity greater than 20/200 was achieved in 12 out of 19 patients. RD recurred in one eye. Fig. 1 ‘scatter figure’ shows the pre- and post-operative visual acuity of the entire series. It displays data in the laser/cryopexy group, vitrectomy with silicone oil group and vitrectomy with gas group.

Figure 1
Figure 1:
Pre- and post-operative visual acuity▪ Vitrectomy + Si Oil group ▴ Vitrectomy + Gas group lseer/cryo group 1= 20/20; 2= 20/30; 3= 20/40; 4=20/50;5=20/80; 6=20/100;7=20/200; 8=Finger counting (FC) 6 m - FC 3 m; 9=Less than 3 m; 10=Hand movements or light perception; 11=Defective light perception


Inadvertent globe perforation can occur during peribulbar, retrobulbar, sub-conjunctival injections, strabismus surgery,3 botulinum toxin injections for strabismus45 and has been reported even during chalazion surgery.6 Complications related to periocular injections, ranging from innocuous sub-conjunctival hemorrhage to intracranial diffusion have been described.7

Peribulbar injections are considered safer. However, some techniques need more than one injection, statistically increasing the chance of a mishap.8 Single use disposable fine gauge needles were used in this series. The decreased pressure required for globe perforation with such needles has been proven using a pressure transducer on fresh/preserved cadaver eyes.9 The risk factors for perforation described in literature10 and observed in our series were myopia in 10 (53%), block in superonasal quadrant in 3 (16%), previous buckling surgery in one (5%) and repeat injection in one (5%). Myopia has dual risk due to thinned out sclera and an elongated eyeball.11 Superonasal blocks require a sound technique due to restricted clearance, and studies have shown that inferior blocks are safer.12 Buckling surgery causes adhesions between the globe and orbital tissues. Some studies show a higher incidence of this complication during a second injection.13 In three patients, none of these predisposing factors could be found.

The first step in iatrogenic perforation is the penetration of the globe. At this stage, the damage is usually restricted to a retinal break. Injection of the medication into the globe results in a sudden rise of intraocular pressure. This could lead to a central retinal artery occlusion14 or pressure induced damage to the intraocular structures, called ocular explosion.151617 In such conditions, the effect of the medication on the retina is important. Lignocaine is relatively safe as shown by reversal of electroretinogram changes in animal models.18 Botulinum toxin is not retinotoxic but causes pupillary dilation and cycloplegia for a few months.5 Gentamycin however is known to cause macular ischemia.

The Royal College of Ophthalmologists carried out a postal survey to assess the prevalence of globe perforation during local anesthesia in cataract surgery. Anesthetist accounted for 75% of the cases.1 There was no difference between the ophthalmology residents and consultants. Our study did not show any significant difference between anesthetists and ophthalmologists but did show that anesthetists often missed the mishap on the table. Double perforation, is more likely to be missed preoperatively19 as the medication is injected into the retrobulbar space, however our series had an isolated case. A high index of suspicion is required in all cases of iatrogenic globe perforation as almost 50% are not recognized at the time of occurrence.202122 Nine cases (48%) in our series were likewise missed on the table.

Perforation signs noted by Berglin et al. in a series of 25 cases, included vitreous hemorrhage in 100%, sub-retinal hemorrhage in 76% and RD in 56%.23 Modarres et al. reported vitreous hemorrhage in all seven myopic cases, with RD in four of them.24 Gillow et al. reported six cases with vitreous hemorrhage in all and RD in five.25 Wearne et al. reported 20 cases where nine had RD.26 Localization of the retinal break in most series was posterior to the equator. The inferotemporal location was the commonest, as in our series.

Laser photocoagulation or cryopexy has been advocated for treatment of breaks when visible and not obscured by vitreous hemorrhage.26 Laser is easier in posterior lesions while cryopexy is easier in peripheral ones.

Early vitrectomy with silicone oil/gas tamponade was done in dense vitreous hemorrhage and/or RD. Early vitrectomy helps to treat the retinal breaks and clear vitreous hemorrhage. It serves to remove any medication, which may have found its way into the globe, like depot steroids.27 Increased duration from injury to surgery increases chances of PVR as in any posterior segment perforating injury. Vitrectomy helps to remove the vitreous scaffolding as also the tract along which traction can occur. Silicone oil was preferred as a tamponade in this series, as a high rate of PVR and poor outcome in RD cases has been reported in literature. A series of six cases by Gillow et al. had poor outcome where RD was present.25 Berglin et al. reported 25 eyes where retinal detachment was present in 14 eyes and 11 eyes developed PVR.23 In Wearne et al. series of 20 eyes, 17 were managed by surgery and nine patients with RD had a poor visual outcome.26 In this series, eight out of 19 patients who had a RD did reasonably well, with vision of better than 20/200 in five eyes. The early intervention and use of silicone oil as tamponade could be responsible for these results. Comparison between the silicon oil and gas group could not be drawn because only three underwent gas tamponade. While visual rehabilitation was variable in different series, generally eyes without a RD did better than eyes with RD. Gillow et al. had a visual acuity of better than 20/200 in 33.3% of cases.25 Duker reported vision of greater than 20/200 in 25% of cases.28 Puri et al. reported a visual acuity of better than 20/40 in six patients, four of these were treatable by cryopexy or laser.29

The Study retrospectively evaluated a series of cases referred with globe perforation, treated at a tertiary center. The study was unable to ascertain the incidence of this complication, as we were not privy to confidential information of the total blocks given or surgeries done by referring hospitals and whether some cases were referred elsewhere. A larger multi-centre study would help to increase our understanding of this complication.


There was no significant higher risk of this complication in the hands of the anesthetists or in community settings; however, anesthetists were more likely to miss the complication. Myopia was a significant risk factor. Breaks were usually located inferotemporally. Awareness of this complication amongst doctors is a key factor in management as it allows early intervention. The visual rehabilitation was comparable with other international series.

1. Gillow JT, Aggarwal RK, Kirby GR. A survey of ocular perforation during ophthalmic local anesthesia in the United Kingdom Eye. 1996;10:537–8
2. Stevens JD. A new local anesthesia technique for cataract extraction by one quadrant sub tenons infiltration Br J Ophthalmol. 1992;76:670–4
3. Morris RJ, Rosen PH, Fells P. Incidence of globe perforation during strabismus surgery Br J Ophthalmol. 1990;74:490–3
4. Liu M, Lee HC, Hertle RW, Ho AC. Retinal detachment from inadvertent intraocular injection of Botulinum toxin A Am J Ophthalmol. 2004;137:201–2
5. Mohan M, Fleck BW. Globe perforation during botulinum toxin injection Br J Ophthalmol. 1999;83:503–4
6. Shiramizu KM, Kreiger AE, McCannel CA. Severe visual loss caused by ocular perforation during chalazion removal Am J Ophthalmol. 2004;137:204–5
7. Marques-Gonzalez A, Onrubia-Fuertes X, Bellver-Romero J, Seller Losada JM, Pertusa-Collado V, Barbera-Alacreu M. Intracranial diffusion. A complication of retrobulbar anesthesia Rev Esp Anestesiol Reanim. 1997;44:284–6
8. Stevens JD, Franks WA, Orr G, Leaver PK, Cooling RJ. Four-quadrant local anaesthesia technique for vitreoretinal surgery Eye. 1992;6:583–6
9. Walter SG, Taboada J, O’Connor P. Retrobulbar anesthesia risk. Do sharp needles really perforate the eye more easily than blunt surfaces? Ophthalmology. 1993;100:506–10
10. Schneider ME, Milstein DE, Oyakawa RT, Ober RR, Campo R. Ocular perforation from a retrobulbar injection Am J Ophthalmol. 1998;106:35–40
11. Gopal L, Badrinath SS, Parikh S, Chawla G. Retinal detachment secondary to ocular perforation secondary to retrobulbar anesthesia Indian J Ophthalmol. 1995;43:13–5
12. Ortiz M, Valls R, Valles J, Blanco , Vidal F. Topography of peribulbar anesthesia Reg Anaesthesia. 1995;20:337–42
13. Ball JL, Woon WH. Globe perforation by second peribulbar injection Eye. 2002;16:663–5
14. Parikh S, Shanmugam MP, Biswas J. Bisected macula following retrobulbar steroid injection Indian J Ophthalmol. 1999;47:247–8
15. Brar GS, Ram J, Dogra MR, Pandav SS, Sharma A, Kaushik S, et al Ocular explosion after peribulbar anesthesia J Cataract Refract Surg. 2002;28:556–61
16. Bullock JD, Warwar RE, Green WR. Ocular explosion from periocular anesthetic injection: A clinical histopathologic, experimental and biophysical study Ophthalmology. 1999;106:2341–53
17. Kumar CM. Orbital regional anesthesia: Complications and their prevention Indian J Ophthalmol. 2006;54:77–84
18. Liang C, Peyman GA, Sun G. Toxicity of intraocular lignocaine and bupivicaine Am J Ophhalmol. 1998;125:191–6
19. Berg P, Kroll P, Kuchle HJ. Iatrogenic eye perforation in para and retrobulbar injections Klin Monatsbl Augenheilkd. 1986;189:170–2
20. Ramsay RC, Knobloch WH. Ocular perforation following retrobulbar anesthesia for retinal detachment surgery Am J Ophthalmol. 1978;86:61–4
21. Hay A, Flynn HW Jr, Hoffmann JI, Rivera AH. Needle perforation of the globe during retrobulbar and peribulbar injections Ophthalmology. 1991;98:1017–24
22. Grizzard WS, Kirk NM, Pavan PR, Antworth MV, Hammer ME, Roseman RL. Perforating ocular injuries caused by anesthesia personnel Ophthalmology. 1991;98:1011–6
23. Berglin L, Stenkula , Algvere PV. Ocular perforation during retrobulbar and peribulbar injections Ophthalmic Surg Lasers. 1995;26:429–34
24. Modarres M, Parvaresh MM, Hashemi M, Peyman GA. Inadvertent globe perforation during retrobulbar injection in high myopes Int Ophthalmol. 1997;21:179–85
25. Gillow JT, Aggarwal RK, Kirby GR. Ocular perforation during peribulbar anesthesia Eye. 1996;10:533–6
26. Wearne MJ, Flaxel CJ, Gray P, Sullivan PM, Cooling RJ. Vitreoretinal surgery after inadvertent globe perforation during ocular local anesthesia Ophthalmology. 1998;105:371–6
27. Gopal L, Bhende M, Sharma T. Vitrectomy for accidental intraocular steroid injection Retina. 1995;15:295–9
28. Duker JS, Belmont JB, Benson WE, Brooks HL Jr, Brown GC, Federman JL, et al Inadvertent globe perforation during retrobulbar and peribulbar anesthesia: Patient characteristics, surgical management and visual outcome Ophthalmology. 1991;98:519–26
29. Puri P, Verma D, Mc Kibbin. Management of ocular perforation resulting from peribulbar anesthesia Indian J Ophthalmol. 1999;47:181–3

Source of Support: Nil,

Conflict of Interest: None declared.


Globe perforation; inadvertent intraocular injection; periocular injection

© 2007 Indian Journal of Ophthalmology | Published by Wolters Kluwer – Medknow