IVF success rates in Toronto: what to actually look at

A 'success rate' is a fraction, and clinics get to choose the bottom of it. Per embryo transfer counts only patients who got far enough to have an embryo to transfer — it quietly excludes cancelled cycles and cycles that produced no viable embryos. Per egg retrieval includes those harder outcomes, so it's a lower, more honest number. Cumulative live birth rate per retrieval — the chance of a baby from one retrieval, counting all fresh and frozen transfers from it — is the number that best matches what patients actually care about.

The same clinic, same year, same patients can honestly report figures tens of percentage points apart depending on which fraction it publishes. When you see a rate, your first question should never be 'is it high?' — it should be 'per what?'

The single strongest predictor of IVF success is the age of the person providing the eggs. Outcomes decline gradually through the mid-30s and steeply after about 38 — which means any rate not broken out by age band is close to meaningless. A clinic whose patient base skews younger will post better blended numbers than an identical clinic serving older patients, with zero difference in quality.

This cuts both ways when you evaluate clinics. A clinic that takes on many patients over 40, low-responders, and complex cases will show worse blended numbers precisely because it accepts harder cases. Some of the best programs for difficult diagnoses look mediocre in undifferentiated statistics. Ask for outcomes in your age band, for patients with your diagnosis — anything else is noise.

Canadian IVF outcomes are collected in CARTR Plus, the national assisted-reproduction registry operated in partnership between the Canadian Fertility and Andrology Society (CFAS) and BORN Ontario. Nearly all Canadian IVF clinics submit cycle-level data to it, and it's the basis for the national outcome statistics you'll see cited at conferences and in the press.

What CARTR Plus is not: a public, clinic-by-clinic league table. Unlike the United States — where clinic-level data is published and, predictably, gamed through patient selection — Canada reports outcomes nationally. That frustrates comparison shoppers, but it also means Canadian clinics face less pressure to turn away hard cases to protect their statistics. When a clinic tells you its results are benchmarked against CARTR Plus, that's a legitimate claim worth probing in detail.

Beyond denominator choice, watch for these patterns. Selected subgroups: a rate computed only on chromosomally tested (euploid) embryo transfers will look outstanding anywhere, because the testing did the filtering. Vintage effects: a chart from several years ago may predate current lab staff and protocols. Survivorship framing: 'per transfer' figures ignore everyone whose cycle ended before transfer. And patient selection: strict intake criteria inflate outcomes before a single procedure happens.

None of this requires dishonesty — each number can be technically true. The distortion is in what's compared against what. Two clinics' published rates almost never share a denominator, a time window, and a patient mix, which is why we don't reprint them in our directory.

In consults, ask: 'For patients my age with my diagnosis, what is your cumulative live birth rate per retrieval, and how many such patients did you treat last year?' The second half matters — a rate computed on a handful of patients is statistical noise. Ask how their results are reported to CARTR Plus and whether their lab holds external accreditation; embryology lab quality is one of the few clinic-level factors with real outcome leverage.

Then weigh the things that are observable: clinic volume in cases like yours, lab accreditation, protocol individualization, and how straight the answers were. A clinic that explains its numbers carefully — including their limitations — is showing you exactly the trait you want in the team running your cycle.

Red flags, for completeness: any language resembling a guarantee, rates quoted without a denominator or age band, refusal to discuss how numbers were calculated, and heavy upselling of add-ons framed as success boosters. None of these prove a clinic is bad — but each one means the marketing is doing work the evidence isn't.