LONDON (Reuters) – New techniques used for hip and knee replacements appear to need reworking at a higher rate than older methods, British researchers said on Tuesday.

About one in 75 people needed to have their hip and knee replacements redone in the three years following the original procedure, according to the study published in the journal PLoS Medicine.

That figure is low but the revision rates were higher for people who had new surgical techniques called hip resurfacing and unicondylar knee replacements, a finding the researchers said raises concern about the procedures.

“On the basis of our data, consideration should be given to using hip resurfacing only in male patients and unicondylar knee replacement in elderly patients,” Jan van der Meulen of the London School of Hygiene and Tropical Medicine and colleagues wrote.

The researchers analyzed nearly 170,000 procedures carried out between 2003 and 2006 — about half of all such surgeries performed in England in this period — and found an overall revision rate of 1.4 percent, or about one in 75 people.

Yet hip resurfacing — a newer technique in which doctors replace just the surface of the femur instead of the whole joint — had a revision rate of 2.6 percent, the study found.

Unicondylar knee replacement when doctors only replace one side of the knee joint had a revision rate of 2.8 percent, van der Meulen said.

The results are in line with findings from other countries showing higher revision rates for the new procedures, which offer benefits because recovery time can be shorter as only part of the joint is replaced, he added in a telephone interview.

“That is an observation that is seen around the world,” he said. “Our registry is the largest in the world, which allows us to look at the most recent results.

Hip and knee replacements are some of the most frequently performed surgeries, and medical device makers like Britain’s Smith & Nephew and U.S.-based Stryker Corp have looked to new techniques to profit from an aging population.

The researcher did not look at brands or what types of products had the lowest revision rates, though that is something they hope to do in the future, van der Meulen added.

(Reporting by Michael Kahn; editing by Sharon Lindores)