Cannula Dislodgement: The Russian Roulette of Harm

The use of Luer lock syringes is standard practice in cataract surgery, with the specific purpose of avoiding cannula dislodgement. Despite their widespread adoption, ocular harm is still occurring—often due to the cannula shooting off the end of the syringe into the eye, damaging ocular structures and causing visual loss or even blindness in what should be a routine procedure.

An anonymous survey sent to members of the United Kingdom and Ireland Society of Cataract and Refractive Surgeons (UKISCRS), the ESCRS, and the readership of Eye News was designed to ascertain whether surgeons had experienced dislocation of a cannula during cataract surgery and whether any ocular damage was sustained.¹

Eighty-four per cent (84.00%) of respondents had experienced dislocation of the cannula during an intraocular procedure. Seventy-eight per cent (78.04%) of respondents observed episodes of harm due to this complication, while 50.37% of respondents indicated ocular damage occurred the last time a cannula dislocated. Another 16.08% indicated they never experienced a cannula dislocation. Some 22.95% indicated that a cannula dislocation occurred on average once per year; 38.43% reported it happened twice per year; 15.66% reported three times per year; and 6.92% stated it happened four times or more annually.

Most correspondents, 86.90%, felt that a better cannula design or a safety device was required to prevent further occurrences. The next significant group, 45.76%, suggested checking to ensure the cannula was tight each time, while 25.46% and 23.80%, respectively, suggested better scrub practitioner training and surgical practice.

The problem has been reported extensively in the literature and has been the subject of two recent safety alerts by the Royal College of Ophthalmologists (England) and a UKISCRS alert.^2–10

Cannula detachment is a well-recognised, unaddressed safety issue that can result in blindness or even loss of the eye.^2–10 Estimates of incidence vary between 0.009% and 0.07%.^2–3 The consequences can be blinding, and it is completely avoidable and indefensible.

Data from the European Union of Medical Specialists (UEMS) show that there are more than 40,000 ophthalmologists in Europe.¹² If 22.95% of these surgeons experience a cannula dislocation once per year, then it occurs 9,180 times annually. When including the numbers for those experiencing it even more frequently, the occurrence increases to an estimated 69,788 times per year. Accepting that in 50% of cases harm is seen, then avoidable ocular harm is occurring in almost 35,000 cases per year across Europe. For the majority, the harm will be minimal and potentially not even documented in the clinical record. For example, a small iris or angle bleed may settle rapidly and not cause any morbidity. However, the harm can be severe, and a Russian roulette of avoidable harm clearly unfolds daily in our work.

A paper published in Eye suggested “[s]imple statements that Luer locks never fail will not avoid future harm if human performance shortfalls cannot be eliminated from every possible step. The safest way to do this is at the manufacturing stage by moulding one-piece cannula-syringe devices, where syringe and cannula hub are moulded together—eliminating the possibility of detachment by force or inadequate tightening.”⁷

This recommendation is not financially feasible, as it would be costly and require accommodating every syringe size.

The issue has been highlighted by a paper as far back as 2012 in the Canadian Journal of Ophthalmology.⁸ They found “[d]espite the use of Luer locks, 60 cases of cannula detachment were reported; 196 respondents experienced this complication, and the most common cause of cannula detachment was stromal hydration (50%). Hydrodissection and viscoelastic were experienced by 18% and 17%, respectively. No severe damage resulted in most cases (76 cases), but some serious complications were reported: retinal damage (9%) and vitreous loss (17%).”

For 13 years, there has been no solution available, and from a medicolegal perspective, it is indefensible. I still have one case every two months or so that results in visual loss, which is heartbreaking; a solution needs to be enacted. As shown in the Canadian Journal of Ophthalmology, this most commonly occurs during stromal hydration—which is likely at the end of an otherwise uncomplicated procedure—taking a patient from a perfect outcome to potential blindness in an instant.

Some viscoelastic suppliers have adopted a screw cap that goes over the cannula and screws to the body of the syringe. While this is a welcome acknowledgement of the issue, it is not feasible for all the other cannulas used during the procedure. As shown above, the viscoelastic is not the main culprit but rather stromal hydration from a Luer lock syringe of volume ranging between 2 mL and 5 mL.

Furthermore, the screw cap solution potentially has inherent safety issues. It is well established that surgeons should not re-sheath a needle or a cannula, and that is what the screw cap encourages.

The Aseptic Non-Touch Technique (ANTT) is a widely used international healthcare method that prevents infection during clinical procedures by ensuring key aseptic sites and parts are protected. It clearly states that these parts, which include the tip of the cannula introduced into the eye, should not be touched—even with a gloved hand. Touch carries the risk of a ‘cannula-stick’ injury but also the risk of endophthalmitis and Toxic Anterior Segment Syndrome (TASS). As Figure 1 demonstrates, passing the cap over the cannula brings the cannula tip into proximity of the user’s fingers. 

The use of the screw cap is a step forwards, and in the absence of other safety methods, its use is encouraged. However, there must be a better solution that can be applied to all the syringes in use, particularly for the stromal hydration step, in which the viscoelastic is not involved.

As surgeons, we currently use syringes and cannulas with a known risk, and we need a solution that is cheap, fast, easy to apply to all syringes, and effective. We and our hospitals have a duty of care to minimise that risk.

In the interim, it is vital that surgeons are aware of this complication and do everything they can to minimise the risk, including checking that the cannula is tight each time.

An article featuring this research is expected to appear in the JCRS in November 2025.

Amar Alwitry BMedSci, BM, BS, MRCS, MRCSEd, MRCOphth, FRCOphth, MMedLaw PgD is a Consultant Ophthalmologist at Nuffield Health, England, UK.

Financial Declaration: The author is actively working with production partners to try to develop a safety device that will resolve the issue of cannula dislocation.

1. Alwitry A. J Cataract Refract Surg, 2025 June 17. doi:10.1097/j.jcrs.0000000000001724.

2. Ting DS, Vaideanu-Collins D, Ellerton C. Eye, 2016; 30: 764–765.

3. Rumelt S, Kassif Y, Koropov M, et al. Arch Ophthalmol, 2007; 125: 889–892.

4. Ting DSJ, Dees C, Ellerton C. Middle East Afr J Ophthalmol, 2017; 24: 54–56.

5. Kahawita S, Cugati S, Casson R. J Cataract Refract Surg, 2015; 41: 1107–1108.

6. Prenner JL, Tolentino MJ, Maguire AM. Arch Ophthalmol, 2003; 121: 128–129.

7. Wilde C, Ross A, Orr G, Dua H. Eye, 2019; 33: 341–342.

8. Pandey P, Kirkby G. Canadian Journal of Ophthalmology, 2012; 47(3): 280–283.

9. Dinakaran S, Kayarkar VV. J Cataract Refract Surg, 1999; 25: 720–721.

10. Yadav S, Tandon R, Singh R, Pujari A. BMJ Case Rep, 2019 Aug 15; 12(8).

11. Maubon L, Nderitu P, O’Brart DPS. Eye, 2022; 36: 1761–1766.

12. European Union of Medical Specialists, https://www.uems.eu/.