What the MBS actually pays for in eye care, and why billing is hard.

Ask what Medicare pays for an eye clinic's work and the honest answer is: it depends on the item number, the setting, the referral, the eye, the date, and how many times the same test has already been claimed in the past twelve months. The Medicare Benefits Schedule presents as a price list - item, descriptor, fee - but it behaves like a rulebook, and ophthalmology sits at the hard end of it because eye care is procedural, imaging-heavy and repetitive in exactly the ways the rules constrain. Billing done well is not data entry after the consult. It is a rules-engine problem, and a practice that treats it as typing pays for the difference in lost revenue, rework, or compliance risk.

The schedule fee is not the price, and the benefit is not the fee

Start with the three numbers that sit behind every claim, because conflating them is where confusion usually begins. The fee a practice charges is its own decision. The schedule fee is what the MBS assigns to the item. The benefit - the amount Medicare actually pays - is a percentage of the schedule fee that depends on where the service happened: 85 per cent of the schedule fee for out-of-hospital services, 75 per cent for professional services that form part of an episode of hospital treatment for a private patient, and 100 per cent only for non-referred GP attendances.¹ The difference between the fee charged and the benefit paid is the patient's gap, and bulk billing is simply the practice accepting the benefit as full payment. So the same clinical act can produce a different Medicare payment depending on whether it happened in consulting rooms or a day surgery, which means the question "what does this item pay" has no single answer - and that is before any of the item's own conditions apply.

An item number is a small contract, and every word in it is a condition

An MBS item looks like a code with a dollar value, but the descriptor attached to it is a set of conditions that all have to be true before the benefit is payable. Take the most ordinary line in a specialist clinic's day: item 104, the initial specialist attendance, with a schedule fee of $103.95 and an out-of-hospital benefit of $88.40. The descriptor does not say "first visit"; it says "initial attendance in a single course of treatment", after referral.² "Single course of treatment" is a defined concept that decides when a patient goes back to attracting the higher initial rate rather than the subsequent one, and a new referral letter does not by itself restart it. Every word in that descriptor - "after referral", "initial", "single course of treatment" - is load-bearing, and the same is true across the schedule: the descriptor is the contract, and the fee is only what you get when the whole contract is satisfied.

The referral is part of the claim

Specialist benefits exist only downstream of a valid referral, which makes the referral a live variable in every claim rather than paperwork settled at the first visit. A GP referral is valid for twelve months unless the GP specifies otherwise, a referral from another specialist for only three, and the validity period runs from the date of the specialist's first service under it, not from the date on the letter.³ For an episodic surgical patient that is rarely a problem. For the patients who define modern ophthalmology it is a constant one, because a macular degeneration or glaucoma patient attends for years, and somewhere in that run of visits every referral quietly expires. A clinic that only discovers the lapse at the front desk has an unclaimable consultation and an awkward conversation; and when a new referral does arrive, whether the next visit is billed at initial or subsequent rates depends on the course-of-treatment rules rather than on the fresh piece of paper.³ Referral state - who referred, when the clock started, when it runs out - has to be tracked per patient, per referrer, continuously.

Frequency caps make billing stateful

The rules so far are at least visible on the day. The harder class of rule is the one that depends on history. Computerised perimetry - the visual-field test that glaucoma monitoring is built on - is capped at a maximum of three examinations in any twelve-month period, counting related perimetry items together.⁴ Optical coherence tomography, the scan at the centre of macular disease care, is narrower still: the ophthalmology OCT item is payable only for diagnosis of a condition treatable with a PBS-listed intraocular medication, and only once in any twelve-month period.⁵ A retina clinic scans treated patients at almost every visit because the scan drives the treatment decision, yet at most one of those scans a year attracts that rebate - the rest are simply part of what the practice provides. The operational consequence is easy to miss: whether today's test is claimable cannot be determined from today's encounter. It depends on what was claimed for this patient across the preceding twelve months, which makes correct billing stateful, and state is exactly what a busy front desk cannot hold in its head.

Procedures price each other, and the schedule moves underneath you

Procedural billing adds arithmetic between items. Under the multiple operation rule, operations performed on the same occasion are not each paid in full: the most expensive attracts 100 per cent of its schedule fee, the next 50 per cent, and each remaining procedure 25 per cent, with the aggregate becoming the schedule fee the benefit is calculated from.⁶ In eye care that rule is routine rather than exotic, because so much of the work comes in pairs - one procedure per eye, often on the same day.

And the item numbers themselves do not stay still. On 1 July 2025 the ophthalmology schedule was restructured to implement recommendations of the MBS Review Taskforce - the body that reviewed how more than 5,700 MBS items align with contemporary clinical practice - along with advice from the Medical Services Advisory Committee.⁷ The intravitreal injection item that anti-VEGF care ran on for years, 42738, no longer exists: injections are now claimed under eye-specific items, 43030 for the left eye and 43032 for the right, each with a schedule fee of $359.95 and each payable only as an independent procedure.⁸ Overnight, every template, favourite list and fee schedule that mentioned the old number was wrong. A restructure like that is not a rare event to weather once; it is how the schedule works, with items amended, split and deleted every July and November, which means a practice's billing knowledge has a shelf life measured in months.

Getting it wrong costs in both directions

The costs of imperfect billing are asymmetric in visibility but real in both directions. Undercharging is silent: the perimetry that sat within the cap but was never claimed, the attendance billed at subsequent rates when the course-of-treatment rules supported initial, the second-eye procedure never entered. No letter arrives about revenue you failed to claim, so most practices never learn the size of that leak. Overcharging is the opposite: it arrives with a signature required. Medicare claiming is monitored, and a practitioner whose claiming cannot be supported can be referred to the Professional Services Review, where "inappropriate practice" is judged against what the general body of the profession would find acceptable, and outcomes include reprimand, repayment of benefits and full or partial disqualification from Medicare.⁹ The provider number on the claim belongs to the practitioner, not to the software or the receptionist, so the person carrying that risk is the doctor - which is precisely why billing accuracy is a clinical-governance issue and not an administrative nicety.

Billing is a rules engine, and it belongs inside the workflow

Put the pieces together and the shape of the problem is clear. Correct billing in an eye clinic requires evaluating descriptor conditions, setting-dependent benefit levels, per-patient referral clocks, twelve-month frequency windows, per-eye item selection and the multiple operation rule - at the moment of care, against the patient's actual history, using this month's schedule. That is not a task for memory or for a laminated cheat sheet at the front desk. It is a rules engine, and it belongs inside the clinical workflow, where the system already knows which eye was injected, when the referral's first service was, and what was claimed in the last twelve months. This is the approach we are building Trenthos Vision around: MBS logic encoded as rules the software checks at the point of billing - frequency caps tested against claim history, referral expiry surfaced before the appointment rather than after it, per-eye and same-occasion arithmetic applied automatically, and every decision logged so the claim can be explained later. None of that replaces judgement about what care the patient needs; it makes the payment for that care land where the rules already say it should. The MBS is not going to get simpler. The practical response is to stop treating its rules as folklore held by whoever has been at the desk longest, and start treating them as what they are: logic, executable by software, checked every time.