Where does the data come from?

Primarily from patients submitting what they actually paid out of pocket after a procedure, via our submission form. We supplement this with MBS schedule fee data from Services Australia and publicly available price references. Provider-supplied pricing is not accepted without independent patient verification.

How do you calculate medians and ranges?

We use the median (P50) as our primary figure and show the P25 to P75 interquartile range alongside it. Reports that appear clearly erroneous are flagged for manual review before aggregation. The median is resistant to extreme outliers; a single unusually high or low submission cannot move it far from the centre of the distribution.

What do you exclude from the dataset?

We do not accept placement fees or payments from providers, and provider-supplied pricing is excluded unless independently confirmed by a patient submission. We also exclude clinical reviews, before-and-after imagery, and any content that could constitute a recommendation about a specific provider's quality of care.

How often is the data updated?

Each submission is processed individually and rolled into city/procedure aggregates on a daily refresh cycle. The signal date shown on each procedure page reflects when the underlying dataset was last refreshed.

Why do costs vary so much for the same procedure?

Several factors drive variation: private health insurance cover (tier, extras, waiting periods), whether you were treated as a public or private patient, the surgeon's gap fee structure, anaesthetist billing, and hospital charges. Inclusions also vary: one quote might bundle anaesthetist fees; another might not.

What are the key limitations?

We do not verify submissions against actual invoices: data is self-reported. Small sample sizes, especially in regional areas or for less common procedures, mean some figures carry low confidence. Historical data may not reflect current fee levels. Always get a written estimate from your provider before any treatment.

How the numbers work

Show the answer, then show the workings.

Every figure on CostDoc comes with a source label, a sample size, and a confidence band. If you want to understand exactly where the number comes from and what it can’t tell you, this page is the full explanation.

Last updated April 2026

Step 01

Collect

Patient submissions are the primary input. MBS schedule data and publicly available price references supplement the core dataset.

Step 02

Label

Every record carries a source label: patient self-report, clinic list price, modelled estimate, or in-depth evidence. Source is never hidden.

Step 03

Aggregate

Median (P50) and interquartile range (P25-P75), not averages. Outlier-flagging before aggregation. Sample count shown on every figure.

Step 04

Publish

Each figure shows sample size, confidence tier, data source, and the date the benchmark was last refreshed.

Where the data comes from

Three independent inputs. The core is always the patient submission; everything else is supplementary. No source is treated as authoritative on its own.

Patient-reported costs

The backbone of the dataset. Patients submit what they actually paid out of pocket for a procedure in a specific city, including whether they held private health insurance and whether they received a written quote beforehand. Submissions are anonymous. Optionally, patients can attach a de-identified invoice image; those are stored privately for internal review and never published. Provider-supplied pricing is not accepted unless a matching patient submission exists independently.

MBS schedule data

Where a real MBS item number and schedule fee are available from Services Australia, we display them with clear attribution. For procedure-level cost breakdowns, such as how much Medicare paid, how much the insurer paid, and what the patient paid, we use modelled estimates derived from typical specialist billing patterns. These are always labelled as estimates, never presented as government data. For the current Medicare schedule fee on a specific item, visit health.gov.au.

Public pricing references

We review publicly available price disclosures, procedure cost guides, and health fund comparison data to identify pricing patterns and cross-check patient-supplied reports. Individual sources are not quoted or attributed. This reference layer is used as a validation check, particularly for procedures where direct submissions are thin, and is not a primary data source.

What we do not accept

Clinics cannot pay to appear, pay to rank, or supply their own pricing without independent patient verification. We do not accept promotional content, sponsored placement, or provider-authored cost figures. Any clinic that contacts us to suggest a number be changed without a supporting patient submission is politely declined.

How we calculate medians and ranges

The goal is to give patients a useful budgeting range, not a single reassuring number. That means being honest about spread, not just central tendency.

Median (P50), not mean

We use the median as the primary cost figure. If you ranked every submission for a procedure in a city from lowest to highest, the median is the value sitting in the middle. A single highly subsidised public patient or a premium international patient can pull a mean well away from where most people actually land. The median does not move much because of one extreme submission. When you see a cost figure on CostDoc, half of reporters paid less than that amount and half paid more.

Interquartile range (P25 to P75)

Alongside the median, we display the 25th percentile (P25) and the 75th percentile (P75). This interquartile range shows where the middle 50% of reported costs cluster. If the P25 is $2,000 and the P75 is $4,500 for a procedure in Sydney, most patients reporting that procedure paid somewhere in that band. The spread itself is informative: a wide P25 to P75 gap signals that costs are genuinely variable for that procedure, and you should budget conservatively and seek multiple quotes.

Outlier rejection before aggregation

Before rolling submissions into aggregates, we flag records that sit well outside the expected range for that procedure and city. Flagged records are held for manual review. If a submission looks clearly erroneous on review, it is removed from the dataset entirely. If it looks plausible but extreme, it stays and is reflected in the published min/max values. We do not silently drop high submissions to make medians look lower.

Out-of-pocket definition

All costs represent the patient’s total out-of-pocket expense: the final amount paid after subtracting any Medicare rebate and any private health insurance contribution. This is the number that matters when deciding whether you can afford a procedure. Where a procedure is fully bulk-billed or covered, the out-of-pocket is recorded as $0.

When the data updates

Each submission is processed individually as it arrives. City and procedure aggregates are refreshed on a daily cycle once new submissions clear the review queue. The signal date shown beside each benchmark reflects when that specific aggregate was last recalculated.

We do not batch submissions and release them quarterly. If twenty people submit their colonoscopy costs in Brisbane this week, those reports flow into the Brisbane aggregate within about 24 hours. The practical implication is that benchmarks for popular procedures in major cities are relatively fresh; benchmarks for rare procedures in regional areas may reflect submissions from months ago. That’s why we show the signal date: so you can judge currency before acting on a figure.

Data quality

Confidence scoring

Every cost figure carries a confidence rating of high, medium, or low. This is driven primarily by three factors: how many submissions underpin the aggregate, how tightly those submissions cluster (a wide P25 to P75 spread reduces confidence), and whether the figure is drawn from direct patient-reported city data or a broader modelled estimate. We always show the submission count behind every statistic so you can weigh it yourself.

Outliers are visible, not hidden

We publish the median, the minimum, the maximum, and the P25 to P75 range so the spread is visible. A wide gap between the published minimum and maximum is a signal to dig into the distribution before assuming the median applies to you.

Source labels on every record

Records carry one of four labels: patient self-report, clinic list price, modelled estimate, or in-depth evidence. A verification label is only applied after a private manual review of uploaded supporting evidence. Upload presence alone does not mark a record as verified.

Modelled estimates are labelled clearly

Where a per-city sample is too thin to publish a meaningful city median, we show a modelled estimate derived from comparable Australia-wide reports. These carry an amber “Estimated” badge and are never presented as verified patient-reported city costs.

Provider information

Where provider data comes from

Provider details are sourced from the AHPRA public register and publicly available professional directories. This includes practitioner names, registration status, practice locations, and specialisations as listed in official registers. We do not purchase provider data from third parties.

Google ratings

Where available, we display Google Business profile ratings and review counts alongside provider listings. These come from public Google Business data and are not generated, moderated, or influenced by CostDoc.

AHPRA compliance

CostDoc does not make clinical outcome claims, quality rankings, or provider superiority lists. We are strictly a billing and cost transparency platform. We do not rank providers by quality, recommend one provider over another, or publish testimonials about medical outcomes. This is consistent with AHPRA advertising guidelines.

Corrections and takedowns

Providers can claim their profile to update practice details or correct inaccurate information. Takedown requests are reviewed within seven days. Contact hello@costdoc.com.au for any correction request.

Cosmetic injectables and prescription-only treatments

Treatments involving prescription-only medicines, including botulinum-toxin-based wrinkle treatments, dermal fillers, GLP-1 weight-loss medicines, and fat-dissolving injections, are governed by the Therapeutic Goods Act 1989(Cth). CostDoc's pricing figures for these treatment areas are aggregate, city-level ranges only, derived from anonymised patient-reported costs. We do not publish per-provider pricing for these treatments, do not recommend any specific medicine, and hold no commercial relationship with any medicine manufacturer or distributor. Individual patient costs vary with units used, medicine chosen, practitioner, and consultation fees. Anyone considering these treatments should consult a qualified practitioner who can assess suitability and disclose the specific medicine and dose.

Modelled specialist fee breakdowns

Where real MBS schedule fees are available, we display them alongside patient-reported data with clear attribution. For the component-level cost breakdown shown on many procedure pages, covering what Medicare paid, what the insurer covered, and what the patient paid out of pocket, we use modelled estimates based on typical specialist billing patterns. These are always labelled as estimates so patients can compare them with real-world submissions.

Attribution language used on every modelled figure

Estimated breakdown based on typical Australian specialist billing patterns. These are modelled figures, not sourced from government data. For actual Medicare rebates, visit health.gov.au.

What we exclude

To keep the focus on billing transparency and avoid AHPRA compliance issues, we explicitly do not collect or publish:

Provider-paid placement
No clinic pays to appear in results or to improve its position. If we are ever approached with that offer, we decline.
Clinical reviews
No testimonials about outcomes, success rates, or treatment quality. That is for review platforms. Our scope is billing.
Before-and-after imagery
No identifiable patient photos or visual outcome claims.
Provider-only pricing
Providers cannot submit their own prices without a matching independent patient submission. The patient experience is the primary record.
Surgeon rankings
No leaderboards based on non-financial metrics. We do not rank providers by quality of care.

Honest limitations

A dataset that is honest about its limits is more useful than one that papers over them. Here is what our data cannot tell you.

We do not verify against invoices

Submissions are self-reported. We cannot confirm that the amount a patient says they paid matches their actual invoice. We ask for receipts voluntarily, and some patients provide them, but verification is not mandatory. Treat figures as informed estimates, not audited costs.

Inclusions vary between quotes

One patient's submission might include the surgeon fee, anaesthetist, and hospital gap all-in. Another might be surgeon-only. We ask submitters to specify what is included, but inconsistency between submissions is a real source of spread in the data. Where inclusions are known they are noted; where they are not, the range itself communicates the uncertainty.

Small samples mean less certainty

Some procedures and locations have hundreds of reports. Others have a handful. We display the sample count and confidence rating beside every figure so you can judge how much weight to put on it. A low-confidence figure with n=4 should not be treated the same way as a high-confidence figure with n=120.

Regional coverage is patchy

Our strongest data is for capital cities and common procedures. Regional and rural areas, and specialist procedures, often have limited or no patient-reported data. When city data is unavailable we fall back to a modelled estimate, which is labelled clearly.

Costs change over time

Fee schedules change. Specialist billing gaps have been rising faster than inflation for many procedure categories. A submission from eighteen months ago reflects what that patient paid at that time, not necessarily what you would pay today. Always confirm current pricing directly with your provider.

Your situation may differ

Your out-of-pocket cost will depend on your private health insurance cover (tier, extras, waiting periods), whether you are treated as a public or private patient, case complexity, and your surgeon's specific fee structure. CostDoc figures are a guide for setting expectations, not a guarantee.

Data governance questions

Unusual reports are flagged for review before aggregation. We use the median (resistant to outliers) and publish the interquartile range and min/max so the spread is visible. If a submission looks clearly fraudulent on review, it is removed from the dataset entirely. If it is plausible but extreme, it stays and is reflected in the published range.

Invoice files are stored privately for manual review only. They are never published on the public site and are accessed only through time-limited reviewer links. No invoice image is used for anything other than internal verification.

Where real MBS schedule fees exist, they are displayed with attribution. Procedure-level cost breakdowns for Medicare paid, insurer paid, and out-of-pocket amounts are modelled estimates based on typical billing patterns and are clearly labelled as such. This preserves the distinction between modelled baselines and verified patient-reported costs.

The median is resistant to outliers. A single highly subsidised public patient or a premium international patient can shift a mean significantly, but it does not move the median far from where most patients actually land. We also show the P25 to P75 range so you can see how tightly the data clusters.

Each procedure page labels whether the benchmark comes from direct patient-reported city data or a broader modelled estimate. The active label and sample size are shown beside the median so the origin of the figure is always visible. Pages show an amber badge when no patient-reported city data is available for that procedure.

Use the takedown and correction workflow linked from every provider profile, or email hello@costdoc.com.au. We follow a seven-day review process for all takedown requests.

Contribute

The data gets better every time someone submits.

Every anonymous report strengthens the signal for the next patient researching the same procedure. It takes about 60 seconds. You don’t need to upload an invoice.

Share your cost Read the guides

Reviewed by the CostDoc Editorial Team. Last reviewed April 2026. Not a medical professional. For clinical questions, consult a qualified healthcare provider.

Modelled specialist fee breakdowns

Attribution language used on every modelled figure

Estimated breakdown based on typical Australian specialist billing patterns. These are modelled figures, not sourced from government data. For actual Medicare rebates, visit health.gov.au.

What we exclude

To keep the focus on billing transparency and avoid AHPRA compliance issues, we explicitly do not collect or publish:

Provider-paid placement

No clinic pays to appear in results or to improve its position. If we are ever approached with that offer, we decline.

Clinical reviews

No testimonials about outcomes, success rates, or treatment quality. That is for review platforms. Our scope is billing.

Before-and-after imagery

No identifiable patient photos or visual outcome claims.

Provider-only pricing

Providers cannot submit their own prices without a matching independent patient submission. The patient experience is the primary record.

Surgeon rankings

No leaderboards based on non-financial metrics. We do not rank providers by quality of care.