A swollen, painful eye socket: when an infection needs surgery

The question I am asked most often about a red, swollen eye is not "what is this" but "does this need an operation". Fair enough. Nobody wants surgery, and nobody wants to be told they needed it a day too late. The honest answer sits on one thin sheet of tissue behind the eyelid, and once I know which side of it the infection has reached, most of the rest of the decision makes itself.

I trained where this presented weekly, and I still see it regularly in my own clinic and on call. Here is how I actually work through it, in the order I work through it at the bedside.

What is the orbital septum, and why does it decide everything?

The orbital septum is a thin, fibrous sheet running behind the eyelid skin and muscle, attached to the bony rim of the socket, and it is the wall between two very different problems. In front of it sits the eyelid itself. Behind it sits the orbit proper, the bony socket holding the eyeball, the muscles that move it, the fat around it, and the optic nerve carrying vision to the brain.

An infection in front of the septum, preseptal cellulitis, is confined to skin and soft tissue that has no direct route to the eye or nerve. An infection that crosses to the other side, orbital cellulitis, has reached a compartment that does not tolerate swelling well, because it is a closed bony box with only one soft exit at the front. That single anatomical fact, which side of a membrane the infection sits on, is why one case goes home on oral antibiotics and another is admitted the same afternoon.

How do preseptal and orbital cellulitis actually look different?

Preseptal cellulitis causes a red, swollen, often tender eyelid, sometimes with a visible source such as a stye, an insect bite, or a small skin wound nearby. The eye underneath moves freely in every direction, vision stays normal, and there is no pain when the eye itself is moved, only tenderness of the lid on touch. Patients often feel otherwise well, or mildly unwell with a low fever.

Orbital cellulitis looks more dramatic and behaves differently on examination. The eye itself may be pushed forward, a sign called proptosis, and moving it in any direction hurts and is restricted, because the swollen tissue behind it is now in the way of the muscles that steer it. Vision can blur or dim. Some patients develop double vision. Fever tends to be higher and the patient generally looks more unwell, not just the eyelid. Recognising these features on examination, rather than judging by how red or puffy the lid looks from across the room, is what actually separates the two conditions.

Which specific findings tell me the infection is behind the septum?

Four findings, checked in every patient with a swollen eye, are what push my thinking from preseptal toward orbital cellulitis. The first is proptosis, the eye sitting further forward than its fellow, which happens because swollen tissue inside a rigid bony box has nowhere else to go. The second is restricted or painful eye movement, since the muscles that turn the eye run directly through the inflamed orbital fat and cannot glide freely through it.

The third is double vision, a direct consequence of that restricted movement, since the two eyes are no longer aiming at exactly the same point. The fourth, and the one that changes the tempo of everything, is reduced vision, sometimes accompanied by a relative afferent pupillary defect, a specific pupil response found by shining a light alternately between the two eyes that indicates the optic nerve on the affected side is not conducting signals normally. Any one of these four findings, on its own, is enough to treat the case as orbital rather than preseptal, and a reduced vision or a relative afferent pupillary defect specifically is what makes me move fastest, because it means the optic nerve itself is under threat.

Why does an abscess need draining rather than just more antibiotics?

Once bacteria have established a pocket of pus, a subperiosteal abscess sitting between the bone of the orbit and its lining, or a true orbital abscess within the orbital fat itself, antibiotics reaching that pocket through the bloodstream become far less effective. A collection under pressure also compresses the surrounding tissue and the optic nerve directly, in a way that simply is not going to resolve by waiting for medication to catch up.

Surgical drainage removes that source of pressure and infection at once, rather than asking the body and the antibiotics to clear something that has already walled itself off. It is not the first step for every patient with orbital cellulitis. Many respond fully to intravenous antibiotics without ever needing an operation. But once imaging confirms a collection, or a patient is not improving on antibiotics after a reasonable trial, or vision is genuinely threatened, surgery moves from an option to the next necessary step, and delaying it further does not buy anything.

How does a CT scan actually settle the question?

Clinical findings tell me which direction to worry in. A CT scan of the orbit and sinuses is what confirms it. It shows the sinuses, which are the source of infection in most adult cases, it shows whether the septum has genuinely been crossed, and it shows whether a discrete collection, an abscess, has formed and exactly where it sits. This is not a scan I order to satisfy paperwork. It is the single investigation that changes management, because "orbital cellulitis without a collection" and "orbital cellulitis with a subperiosteal abscess" are treated differently from that point onward, and no amount of careful examination alone substitutes for seeing it directly.

What actually happens in hospital, and when does surgery come in?

Orbital cellulitis is admitted to hospital, imaged, and started on intravenous antibiotics chosen to cover the likely bacteria, usually in conjunction with ear, nose and throat colleagues given how often the sinuses are the source. Vision, pupil response, eye movement and the degree of proptosis are checked repeatedly, often several times a day in the early stages, because this is a condition that can change within hours rather than settle predictably overnight.

Surgery is reserved for three situations. A confirmed abscess on scan is one. A patient who is not improving, or is worsening, despite appropriate intravenous antibiotics after a reasonable interval is the second. Any sign that vision is under threat, whether that is deteriorating acuity, a worsening relative afferent pupillary defect, or dimming colour vision, is the third, and in that situation surgery is not delayed to see whether antibiotics might still work. The reason this whole pathway exists, admission, imaging, intravenous treatment, a low threshold for surgery, is that vision loss from orbital cellulitis is preventable when it is caught and acted on in time, and is not reliably reversible once the optic nerve has been compressed for too long.

The anatomy behind all of this is identical in children, but the pattern of illness and the pace at which it moves can differ, since children develop orbital cellulitis more often than adults, usually from sinus infection, and can deteriorate quickly. I see and treat adults; I do not manage this in children, and I would rather hand that specific question to someone who does every day. My colleague Dr Chan Li Yen has written a companion piece on recognising when a child's swollen eyelid is a genuine emergency, and paediatric assessment belongs with a paediatric ophthalmologist rather than an extrapolation from adult signs.

What should I actually do if my eye is swelling right now?

If the swelling is spreading over hours, if moving the eye hurts, if vision is changing, or if there is fever alongside it, that is a same-day eye emergency assessment, not an appointment to book for later in the week. Do not wait to see whether it settles overnight, because the findings that separate a straightforward eyelid infection from a sight-threatening orbital one are exactly the findings I have described above, and they are worth having checked properly rather than guessed at from a mirror.

This sits within the wider group of orbital conditions I see regularly, alongside causes such as thyroid eye disease that can also push an eye forward, though far more slowly and without infection. A red, swollen eyelid is sometimes something far simpler, such as a chalazion, a blocked eyelid gland, and telling that apart from a spreading infection is part of the same assessment. If you are trying to work out whether what you are seeing is ordinary tiredness or something that needs a proper look, I have written separately about what actually makes eyes look tired. You can also read about the full group of orbital and thyroid eye conditions I look after, or bring your questions to a consultation.