Technique

Preoperative templating for total hip arthroplasty: how accurate is it, really?

The orthobro · 2026-08-26 · 7 min read

Templating gets treated as a formality — something the fellow did, printed, and stuck on the board. Then somebody asks you what size you would use and you find out you have no opinion.

Worth knowing what the exercise is actually good for, because it is neither as precise nor as pointless as the two camps claim.

How accurate is it

The honest number is: roughly half, exactly.

In one series, digital templating was exact in 52% of stems and 51% of cups, and accurate within one size in 90% of stems and 85% of cups.

Another study reported the exact size predicted in 42% of femoral components and 37% of acetabular components, with 87% of femoral and 78% of acetabular within one size.

Those two studies disagree by about ten points on exactness and agree closely on the thing that matters. Read them together and the picture is:

Templating tells you the neighbourhood, not the address.

Which is precisely how it should be used. The point was never to eliminate the trial. It is to walk in knowing that this is a size 4 or 5 hip and not a size 2 or a size 8 — so that the right implants are in the room, the right instruments are open, and a surprise is a surprise rather than a delay.

Why bother if you are going to trial anyway

Because it changes the outcome that is hardest to fix afterward.

Patients who had preoperative templating had a mean postoperative leg length discrepancy of 6.7 mm, against 11.5 mm in a control group without templating (p = 0.023).

Leg length is the classic. It is the thing a patient notices, the thing that generates complaints, and the thing that is very difficult to revise for. Templating is where the offset and length plan gets made, and there is no equivalent moment later in the case where you can make it with the same information and no time pressure.

The other reasons are more mundane and just as real: the correct implants and instruments get pulled, the rep knows what to bring, and the theatre is not waiting for a tray from another site.

The experience question, and the good news for residents

Two findings sit next to each other here and they are both worth knowing.

Experience helps. Digital templating of femoral implant size was significantly more reliable when performed by a senior surgeon. In one analysis, high experience produced adequate planning in 95% of cases, moderate experience 88%, and low experience 82%.

But the learning curve is short. In a study using mediCAD software where templating was done by someone new to hip arthroplasty — a PGY-2 orthopaedic resident or a hip fellow — there was no significant difference between accuracy in the first months of the arthroplasty course and the second. Whatever the skill is, it does not take a year to acquire.

Put those together and the practical reading is that templating is worth learning early because you get most of the available accuracy quickly, and the last 10 percentage points come from case volume you will accumulate anyway.

There is also evidence that a structured video teaching tool is effective for training residents in hip arthroplasty templating — which mostly says that this is a teachable, procedural skill rather than an intuition you develop by osmosis. So it is reasonable to expect to be taught it, and reasonable to go and learn it deliberately if you have not been.

How to actually get the practice

The obstacle is not difficulty. It is that nobody assigns it to you.

The fix is unglamorous: template the case yourself, before the case, even when someone else has already templated it. Then compare. You need 50 primary total hip arthroplasties to graduate, so there is no shortage of chances to practise.

Three things to record each time:

• What you templated: cup size, stem size, offset, neck length, and the plan for leg length.
• What actually went in.
• Why they differed, if they did.

That third line is where the learning is. The difference between your template and the implant is almost never random — it is a consistent bias in how you are reading the film, or a preference of that surgeon's you did not know about, and either one is fixable once you have four or five of them written down next to each other.

A run of ten of those, kept somewhere you can look back at, is worth more than any amount of reading about templating. It is also a genuinely good thing to be able to show an attending who asks what you have been working on.

Caveats worth carrying

Magnification. Two-dimensional templating depends on knowing the magnification of the radiograph. A marker of known size in the field is what makes that possible. Templating off a film with unknown magnification is guessing with extra steps.

Positioning of the film. Rotation of the pelvis and of the limb changes what you measure. A poorly positioned AP pelvis will produce a confident, wrong answer.

Abnormal anatomy. Accuracy figures above come mostly from primary, reasonably normal hips. Dysplastic hips have their own literature and their own worse numbers. Do not carry a 90% figure into a case that does not resemble the series it came from.

It is a plan, not a decision. The trial is still the decision. Templating that makes you reluctant to change your mind intraoperatively is worse than no templating at all.

The one-line version

Template every hip you are going to scrub, write down what you predicted and what went in, and look at the gap. Half the time you will be exactly right, most of the rest you will be one size away, and the times you are further off are the cases that will teach you the most.