Neuraxial Anesthesia Reduces Joint Surgery Complications

May 10, 2013

By James E. Barone MD

NEW YORK (Reuters Health) May 10 - In a large retrospective study, knee and hip arthroplasty patients had better perioperative outcomes with spinal or epidural anesthesia rather than general anesthesia.

Which anesthetic technique is better has been debated for many years. Hip and knee arthroplasties are being done more frequently, so evidence-based guidance for anesthetic management is needed, but randomized trials are difficult to do because complications after these procedures are generally so infrequent.

The new conclusions are drawn from data on hundreds of thousands of cases, analyzed by investigators based at New York City's Hospital for Special Surgery and Weill Medical College of Cornell University. They reported their findings this month in Anesthesiology.

In an email to Reuters Health, lead author Dr. Stavros G. Memtsoudis said, "While at our hospital the use of neuraxial anesthesia for total knee and total hip arthroplasties represents standard practice, it clearly does not in many other institutions."

The authors looked at records from Premier Perspective, Inc., which oversees an administrative database comprising patients from nearly 400 hospitals, mostly non-teaching and urban institutions.

The type of anesthesia used for hip and knee arthroplasty was included in 382,236 of 528,495 available records covering a four-year period, with 11.1% having neuraxial anesthesia, 74.8% having general anesthesia, and 14.2% receiving a combination of the two.

While the number of 30-day deaths was small after all three types of anesthesia, it was significantly fewer for neuraxial and combined vs. pure general anesthesia--0.10%, 0.10% and 0.18% respectively (p < 0.001).

With the exceptions of cardiac and gastrointestinal complications, which occurred at similar rates in all three groups, all other adverse outcomes (pulmonary embolism, respiratory failure, pneumonia, stroke, infections, acute renal failure, and need for blood products) occurred significantly more often after general anesthesia.

Like all retrospective studies, this one has a few limitations, such as the well-known issues associated with using administrative databases to answer clinical questions and the lack of information about possible readmissions and specific anesthetic-related complications, such as spinal headaches and intubation mishaps.

However, Dr. Memtsoudis noted, "One of the main benefits of using large databases is the ability to study low incidence outcomes that are otherwise difficult to evaluate in prospective studies."

He noted that the incidence of pulmonary embolism after these procedures has become almost trivial over the years, which is in stark contrast to the cumulative amount of research published and effort expended to treat and prevent it. While differences may seem relatively small, the effect sizes seen in the study were "similar or larger than those achieved by many medications used to prevent adverse events in the perioperative period, for example, statins for the prevention of cardiac events."

Dr. Mark D. Neuman, an assistant professor of anesthesiology and critical care at the University of Pennsylvania, was not involved with this research. He said, "Any observational data set will have its own strengths and weaknesses, but the Premier data set has been used in a number of papers including several in high-profile journals."

Another assistant professor of anesthesiology and critical care at Penn, Dr. Nabil M. Elkassabany, said the paper "adds to our already existing belief that regional anesthesia is a better way to go in case of primary total joint arthroplasty."

At his institution, he estimates the ratio is 30% spinal to 70% general anesthesia, adding, "We are evaluating our own practice and educating our staff to get everyone on board."

"Over 90% of patients at the Hospital for Special Surgery receive neuraxial anesthesia," Dr. Memtsoudis said.

Dr. Neuman thinks there is still some uncertainty. He said, "This paper is important as an additional piece of evidence that anesthesiologists can incorporate into their decision-making process. Nonetheless, studies like this should not be over-interpreted. There is still a lot of work to be done before we can say definitively that neuraxial anesthesia is better."

"I believe that future research will have to investigate possible confounding factors not measured in databases," Dr. Memtsoudis said. "We have to confirm that the use of neuraxial anesthesia is not just a surrogate marker of better overall care, which may explain some of the outcomes."

"For some of the more common outcomes in the study, such as infectious complications, it is possible that a randomized controlled trial could be designed with enough power to show a difference," said Dr. Neuman.

Dr. Elkassabany said, "Anesthesiologists should have a firm belief that their role in the perioperative area is not limited to the intraoperative period, and what we do on every day basis will have an impact on overall patient surgical outcomes."


Anesthesiology 2013.