The good news, first. Losing a flap — the microsurgical tissue transplant at the heart of breast reconstruction — has become rare. When flap loss becomes uncommon enough that some surgeons never experience it, the profession has to ask what we should be measuring instead. The unit data from Queen Victoria Hospital, where I worked until December 2025, illustrates exactly where we stand: in a consecutive series of 247 patients during 2023, there were four total flap losses across three patients — a rate of 1.2% across nine consultant surgeons. By any historic measure, those are excellent results.
But excellent flap survival rates do not tell a patient what her recovery is going to look like. That is a different question — and one the field has been slower to answer.
My answer, developed with my colleagues Sarah Williams and Siri Gowda, and now published in the Journal of Plastic, Reconstructive & Aesthetic Surgery, is that it tells her very little — and that surgeons have been systematically undercounting the things that matter most to the patients sitting in front of them.
What We Were Missing
For most of the history of reconstructive breast surgery, success was measured in stark terms: did the reconstruction survive? Flap survival rates were the headline figure — the one quoted to patients in consultations and in published papers. Everything else — the seroma that needed draining three times, the wound that refused to close for six weeks, the infection that meant readmission and a course of intravenous antibiotics — was noted in the clinical record but rarely counted systematically.
This was not dishonesty. It reflected where the clinical challenge actually was. When preventing total flap loss demanded enormous concentration, measuring minor events felt like a distraction from what mattered.
But with microsurgery now well-established and most patients going home within two to three days, the challenge has shifted. The major complications are no longer the ones most patients face. The question is no longer only "will it survive?" It has become "what will my recovery actually be like?"
The answer to that second question lives almost entirely in the events surgeons have historically described as minor.
What We Actually See in Clinic
In the 247-patient series drawn from Queen Victoria Hospital’s 2023 practice — every consecutive autologous breast reconstruction with free tissue transfer that year — postoperative complications occurred in 33 patients (13.3%). Most were what surgeons classify as minor: seromas drained at clinic appointments, wound infections treated with oral antibiotics, areas of delayed wound healing managed with dressings. Eight patients required return to theatre, mostly for haematoma evacuation. There were no life-threatening events and no deaths.
By historic measures, those are excellent results. But looking more carefully at the 22 patients with Grade I or Grade II complications — managed without operative intervention — a different picture starts to form. These patients experienced extra clinic visits, courses of antibiotics, prolonged wound care, and in several cases, multiple appointments over many weeks. None of that shows up in a flap survival rate. None of it is captured by asking simply "were there any major complications?"
Although often considered minor, these complications can impose a disproportionate burden on patients through repeated clinic visits, prolonged wound care, anxiety and delayed return to normal activities.
— Williams, Gowda & Blackburn, JPRAS 2026
Why Counting Events Isn’t Enough Either
Here is a finding that genuinely surprised me, even though in retrospect it makes intuitive sense.
When we compared abdominal-based reconstructions (DIEP flaps) with thigh-based reconstructions (TUG and LUG flaps), the overall complication rates were almost identical — around 12 to 13% for both. If you stopped there, you would reasonably conclude that the two approaches carry similar risk.
But when we applied a weighted numerical score to each complication — reflecting not just whether it happened but how severe it was and what intervention it required — thigh-based reconstructions scored significantly higher (1.69 versus 0.43 for abdominal flaps). The same percentage of patients experienced complications, but complications associated with thigh donor sites were more likely to require operative intervention, prolonged management, or readmission. The event frequency was similar. The burden was not.
Complication rates, on their own, are hiding something important. Two procedures can look identical on paper while being meaningfully different in the recovery they impose.
The Scoring System
The tool we developed builds on work Siri Gowda and I published in 2024: a modified version of the Clavien–Dindo Classification — the surgical world's most widely used system for grading complications — adapted specifically for breast reconstruction. That classification grades complications from Grade I (minor, managed in the outpatient setting) through to Grade V (death), with Grade III subdivided into events requiring operative intervention under local or general anaesthetic.
The new addition is a numerical score assigned to each grade, reflecting the relative morbidity of that complication and the intervention required to manage it.
| Complication Grade | Examples | Score |
|---|---|---|
| Grade I | Seroma drained at clinic; wound managed with dressings; oral antibiotics | 1 |
| Grade II | Infection requiring IV antibiotics as an inpatient; seroma requiring admission | 3 |
| Grade IIIA | Wound debridement under local anaesthetic | 4 |
| Grade IIIB | Return to theatre under general anaesthetic — haematoma, washout, re-do anastomosis | 6 |
| Grade IIIC | Total flap loss | 10 |
| Grade IV | Life-threatening event requiring intensive care | 20 |
| Grade V | Death | 50 |
The scores are non-linear deliberately. The step from a minor wound issue to a return to theatre is not a small increment. It represents a qualitatively different experience for the patient — general anaesthesia, an extended hospital stay, anxiety for the family — and a very different demand on NHS resources. A score of 6 is not twice as bad as a score of 3; it is categorically worse in ways that a linear scale would not capture.
Across our entire cohort, the mean complication score was 0.49 — a reassuringly low overall morbidity burden, consistent with what high-volume ERAS centres report. But the mean score is only useful if it can be broken down meaningfully, and that is where the data became interesting.
What Age and BMI Actually Mean for Recovery
Surgeons have long known that older patients and patients with higher BMIs tend to have more complications. But "more complications" is not the whole story — and the scoring system allowed us to see the complete picture more clearly than raw rates alone.
Patients aged 60 or over had a mean complication score nearly three times higher than younger patients (0.98 versus 0.32, p = 0.023). This reflected not only a higher complication rate (26% versus 9%) but more severe events. Patients with a BMI of 30 or above showed a similar pattern — significantly higher mean scores than those with a BMI below 30 (0.90 versus 0.34, p = 0.047), and approximately three times the complication rate.
The combination of both risk factors produced the most striking findings. Patients aged 60 or over with a BMI of 30 or above had a mean complication score of 1.94, compared with 0.23 in younger patients with a BMI below 30. Their complication rate was 52.9% versus 6.8%.
I want to be clear about what these numbers do not mean. They do not mean that older patients, or patients with a higher BMI, should not have breast reconstruction. The operation remains transformative for the right patient, at any age and at any realistic BMI. What the numbers mean is that these patients deserve more specific conversations — not a generic reassurance that the complication rate is "around 13%", but an honest account of what that 13% is likely to look like for them specifically, and what they can do to shift it.
How This Changes the Consultation
Previously, I could say to a patient: "The chance of losing the reconstruction entirely is around 1%. The overall complication rate is about 13%." Both statements are accurate. But they do not distinguish between a single clinic visit for wound dressing and a readmission for intravenous antibiotics. They say nothing about whether her recovery is likely to be straightforward or drawn-out. And they give no basis for comparing the likely burden of a DIEP flap — which requires a longer operation and an abdominal scar — with a thigh-based reconstruction, which is technically simpler but, our data shows, carries a higher morbidity burden when things go wrong.
With a numerical morbidity score, the conversation becomes more specific and more honest. A patient in her late sixties with a BMI of 32 considering a thigh-based reconstruction should know — not as a reason to refuse her surgery, but as genuine informed consent — that the data suggest her recovery is more likely to be complex than that of a younger, slimmer patient having a DIEP flap. She should know what "complex" means in practice: not necessarily a disaster, but potentially multiple clinic visits, a longer recovery, a higher chance of needing to come back to hospital.
That is not a comfortable thing to say. But it is a useful one. Patients who go into surgery with accurate expectations cope better with setbacks. They are less likely to catastrophise a wound infection or a seroma because they understood the realistic range of outcomes before they consented.
The Bigger Picture
Beyond individual consultations, this kind of numerical scoring has value for the field as a whole. It allows fair comparison between units — adjusting for the fact that a centre taking on older and heavier patients should not look worse than one with a younger, lower-risk cohort, even if raw complication rates are higher. It supports training: a trainee surgeon whose cases are being assessed can be evaluated not just on whether anything went wrong, but on the morbidity burden associated with the complications that occurred. And it creates a platform for quality improvement, identifying which patient subgroups or operative strategies are associated with disproportionate burden before those patterns accumulate into a significant problem.
There is also a straightforward argument about honesty. The seroma that needs draining three times is not, individually, a disaster. But it means three additional clinic appointments during a period when a patient may already be managing chemotherapy, radiotherapy, and the emotional aftermath of a breast cancer diagnosis. The wound that takes ten weeks to heal instead of four is not a failure by any traditional measure — but for the patient, it is ten weeks of anxiety, of dressings, of not being able to shower normally.
These experiences matter. They deserve to be measured, not because we need another number in a journal paper, but because counting them properly is the first step to doing better — and because patients deserve to know, before they consent, what their recovery is most likely to actually involve.