Membership information +44 113 241 0727
Medicolegal advice +44 113 243 6436

Penetrating the eyeball

01 September 2013

Ms J, a 36-year-old banker with myopia, consulted Mr R, an ophthalmologist, with a one-week history of pain and blurring of vision in the left eye. Mr R diagnosed anterior uveitis and prescribed corticosteroid eye drops, and proceeded to give a sub-tenon’s injection of 0.5ml depomedrol under local anesthesia in the lower outer corner of the left eye. The patient felt minor pain with the local anaesthetic injection but felt excruciating pain with the depomedrol injection. Within seconds a black spot blocked the central vision in the left eye. The spot expanded rapidly until the vision was completely lost. Mr R continued injecting till the full dose was given. On examining the left eye Mr R found that the eye was filled with fluid – he arranged a follow-up consultation the next day.

Ms J called later that afternoon to ask if she could see Mr R immediately but was advised to return the next day. Ms J chose to see another ophthalmologist who diagnosed a localised retinal detachment and referred her to a retinal surgeon, who performed surgery eight hours later. The retinal detachment was caused by two needle punctures penetrating the eyeball and injecting depomedrol into the eye instead of the intended sub-tenon’s space. She underwent surgery to repair the retinal detachment and remove the intraocular drug but complete removal of the steroid was not possible.

Postoperatively, the retina was flat, but scattered retinal hemorrhage and macular nerve fibre layer oedema was noted. About three weeks later, Ms J developed an inferior retinal detachment, epiretinal membrane and retinal necrosis. She underwent further surgery to remove the epiretinal scar membrane and correct the retinal detachment. Her intraocular pressure was raised postoperatively but was controlled with medical treatment.

The iritis subsided, the intraocular pressure normalised and the remaining subretinal steroid dissipated completely within three months. Her final visual acuity was hand movement in the left eye and 6/6 in the right eye. The left eye remained painful and uncomfortable. Ms J had difficulty with near work and computer work, suffered eye strain and easy fatigue in the right eye and experienced frequent headaches and imbalance when walking downstairs.

She was assessed as having 20% impairment of vision and 20% impairment of the whole person, with 50% loss of capacity. She also developed depression and was under the care of a psychiatrist. She returned to work six months later but, due to mental distress and intense eye pain, she had to work part-time in a less intense position, and with a lower salary.

Ms J made a complaint and a civil claim. The claim was indefensible and was settled for a substantial sum.

Learning points

  • Ample guidance is available through professional bodies and the scientific literature on the management of common eye conditions. Periocular corticosteroid is not indicated for uncomplicated anterior uveitis. Where topical corticosteroids are ineffective, a sub-conjunctival injection of a short acting corticosteroid may be considered. Mr R chose the wrong primary method of treatment, the wrong injectable drug and the wrong route of injecting the drug.
  • Periocular injections carry a risk of globe penetration that is much higher in myopic eyes. The records showed no evidence of discussion of indication, risks or alternatives. No written consent was taken. When a non-standard treatment is offered, a thorough discussion of the indications, risks and alternatives is mandatory and written consent is advisable. Guidance on the principles of taking informed consent is available in a number of different countries.
  • Mr R failed to discontinue the injection when the patient had severe pain and loss of vision. Even though the globe had been injured, the extent of damage may have been reduced had he stopped immediately. Immediate exclusion of a penetration either by ultrasound or by clinical examination is mandatory when patient symptoms suggest globe penetration. Failure to do this established a breach in the duty of care. Early diagnosis and referral for emergency intervention may have reduced the extent of the irreversible damage.
  • Adverse outcomes and complications are part of a doctor’s working life. Responding to these events in a timely manner, showing respect, being open and communicating honestly help to reduce the impact of these events on both the patient’s wellbeing as well as the doctor’s professionalism.
  • A patient can withdraw consent at any time during the procedure. When pain is not what you expect, it is good practice to stop and reconsider your treatment.