The single most common question in the clinic is also the one with no clean answer: how long will it last? Patients want a number, ideally a round one, ideally larger than their remaining lifespan. The instinct is to oblige with the famous “about twenty years” and move on. The truthful reply is more interesting and only slightly longer: a hip replacement does not have a use-by date stamped on the box. It has a survival curve, and the curve has a different answer depending on whose hip it is.
A survival curve — a Kaplan–Meier curve, if you want the surname — plots the percentage of replacements still in place over the years, counting a return to theatre for revision as the “event”. It starts at 100% on the left and only ever steps downward, because no revised hip un-revises itself. The headline figure everyone quotes is simply one point read off that curve, usually at ten or fifteen years, and quoting a single point off a whole curve is like describing a film by its middle frame. Below you can reshape the curve with the two things that move it most: the patient’s age and how hard they intend to use the joint.
Two patterns jump out the moment you start dragging. First, the curve for a sedentary eighty-year-old barely droops — they will, with no unkindness intended, run out of patient before the implant runs out of life. Second, the curve for an active fifty-year-old falls visibly faster, because a younger, busier joint simply clocks up more steps, more load and more wear per year. This is the awkward paradox of the field: the patients most desperate for a new hip and most able to enjoy it are precisely the ones whose replacements have to work hardest and revise soonest. The implant does not know your birthday; it only counts your strides.
When the curve does step down, why?
A revision is never just “it wore out”. The reasons cluster, and they arrive on different timetables — some in the first weeks, some only after a decade of faithful service. Knowing which is which is half of consenting a patient honestly, because the early risks and the late risks ask for completely different conversations.
| Reason for revision | Typically when | Plain-English version |
|---|---|---|
| Infection | Early, or any time | Bacteria get a foothold on the metal; the hardest problem in the book |
| Dislocation / instability | First weeks to months | The ball keeps popping out of the socket |
| Periprosthetic fracture | Any time, often after a fall | The bone around the implant breaks, not the implant itself |
| Aseptic loosening | Late, 10+ years | The bond to bone quietly fails with no infection to blame |
| Bearing wear | Late, 15+ years | The sliding surfaces grind thin over many millions of steps |
Notice that exactly one of those rows — bearing wear — is the slow, dignified “used it up” story patients picture when they ask the question. The others are events, not expiry dates, and an event can land in year one as easily as year twenty. This is why a registry can report 95% survival at ten years and an individual surgeon can still have a patient back in theatre at ten weeks: the average is made of curves, and a curve is made of people who each had their own way of needing a second operation.
“How long will it last?” is really three questions wearing one coat: how long until it wears out, how likely something goes wrong, and how long will I last. Only the patient can answer the third, and it is usually the one that decides the matter. — What I actually say in clinic
So the honest answer to the famous question is a curve and a shrug delivered with confidence: for most people having the operation at the usual age, the great majority of hips are still going strong at fifteen and twenty years, and modern bearings have pushed that further. For a young, active patient the number is lower and the conversation is longer — not because the surgery is worse but because they will simply ask more of it. A good clinic does not quote the curve. It works out which point on the curve is theirs, and then it tells them the truth about that point.