How scoring works
The GCR Rating is a single, comparable number that tells patients which hospitals and clinics perform best within a specific medical specialty. This page sets out exactly what the Rating measures today, the model behind it, and how patient-reported outcomes will reshape it — written to be read, checked, and cited.
The GCR Rating is a single number — out of 5, shown to three decimals, for example 4.972 — given to a hospital or clinic for one medical specialty it offers.
A facility does not receive a single whole-unit score. It earns a separate Rating for each specialty it treats, because a hospital can be excellent in one area and weaker in another. The Rating exists to answer one question fairly: among facilities treating the same specialty, which perform best? The ranking is simply the sorted list those Ratings produce, highest first.
The three decimals are not false precision for its own sake — they break ties between closely matched facilities so the ranking stays stable and unambiguous.
Today, the Rating is calculated from verified public review data — the feedback patients share publicly after they have been treated. We do not ask facilities to supply data about themselves. Three signals drive the current Rating:
The average score patients give a facility for a specialty — the core measure of how well it is judged to perform.
How many verified reviews exist. More reviews mean more evidence, and more confidence that an average reflects reality.
When the reviews were left. Recent feedback weighs more, so the Rating tracks how a facility performs now — not years ago.
Review volume and recency are deliberately not shown as columns in the ranking. They shape the Rating behind the scenes; the patient sees one fair number, not a wall of metrics.
Raw averages mislead when review counts differ. A facility with three five-star reviews is not more proven than one with four hundred reviews averaging 4.8 — but a naïve average would rank it higher. GCR avoids this with a Bayesian model.
Each facility's average is pulled toward the specialty-wide mean in proportion to how few reviews it has. This is called shrinkage: with little evidence, a facility sits close to the typical score for its specialty; as verified reviews accumulate, the Rating relies more on the facility's own data and less on that prior.
The Bayesian average
R — the facility's own average for the specialty, after recency weighting up-weights recent reviews.
v — the facility's number of verified reviews.
C — the mean Rating across all facilities in that specialty (the prior).
m — a confidence weight: how many reviews are needed before a facility's own average dominates.
The result is normalised to a 0–5 scale and reported to three decimals. Specialty means are computed per specialty, so a facility is only ever compared with its true peers.
Because the prior is the specialty mean — not zero and not five — a brand-new facility starts near the middle of its field and earns its way up or down on real, verified evidence.
No competitor rates a hospital as a single entity and calls it fair. GCR computes a distinct Rating for every specialty a facility offers, each on its own specialty-specific scale. A hospital can rank in the top ten for dental implants and mid-table for orthopaedics — and the rankings show exactly that. A facility's headline asset is therefore its position within a specialty and geography, such as “#1 in Dental — Hungary” or “Top 10 worldwide in Hair Transplant”.
Public reviews capture experience; they do not always capture results. A patient can have a pleasant visit and a poor clinical outcome, or vice versa. The next stage of the GCR Rating closes that gap.
GCR is introducing patient-reported outcome measures (PROMs) — structured reports from patients on recovery, side effects and long-term results, collected over time, from two days to five years after treatment. Combined with clinical outcome data where it is available, PROMs are being introduced as the future primary signal behind the Rating, with verified public reviews moving to a supporting role as coverage grows.
Today · live
The Bayesian model on rating, volume and recency, described above.
Rolling out
Structured follow-up surveys at set intervals after treatment — 2 days, then weeks, months and years — tracking recovery and long-term results.
The roadmap
As PROMs reach sufficient coverage in a specialty, patient-reported outcomes and clinical results become the primary input, with public reviews as a supporting signal.
GCR ranks hospitals and clinics across 25 medical specialties — including dental, hair transplant, plastic and cosmetic surgery, IVF and fertility, ophthalmology, orthopaedics, cardiology and oncology — worldwide, and scoped to each country and city. Every specialty has its own fixed ranking; every location ranking is its own page.
This page supersedes earlier descriptions that referenced 16 specialisations: GCR now ranks 25. Browse all specialties →
Questions about the methodology or a specific Rating? Contact the GCR team →
