Arthritis Today

Diabetes May Not Raise Knee Replacement Risks

Study findings contradict previous research linking the disease with poor surgical outcomes.


Patients with diabetes, controlled or uncontrolled, who undergo total knee replacement are no more likely to experience infections or other postsurgical complications than patients without diabetes, according to researchers at Kaiser Permanente, a large integrated health care organization.

The findings, published in The Journal of Bone & Joint Surgery, contradict earlier studies showing that diabetes can lead to poor outcomes after knee replacement surgery. For instance, a 2009 Duke University study found that patients with uncontrolled Type 1 or Type 2 diabetes had a significantly higher risk of postoperative stroke and wound infection than patients who had controlled or no diabetes.

High levels of glucose (sugar) in the blood – the hallmark of diabetes – is thought to cause postsurgical complications because it adversely affects many organs and processes in the body. But the Kaiser researchers did not find such an association.

“We were surprised at the findings,” says lead author Annette L. Adams, PhD, research scientist at the Kaiser Permanente Southern California Department of Research and Evaluation in Pasadena. “We thought people with higher blood sugar would have more postoperative complications. At first we wondered if we’d made a mistake, because our results were so different from what other people found. It is counterintuitive.”

For the study, Adams and colleagues retrospectively reviewed the electronic health records of more than 40,000 Kaiser Permanente patients who had a first-time knee replacement between 2001 and 2009. Slightly more than 81 percent of the patients did not have diabetes, 12.5 percent had controlled diabetes and 6.2 percent had uncontrolled diabetes.

Definitions of controlled and uncontrolled were based on hemoglobin A1c levels (commonly known as HbA1c or A1c) – a measure of average blood sugar levels during the previous three months. Diabetes patients with an A1c level of 7 percent or higher were considered to have uncontrolled diabetes. Those with an A1c level less than 7 percent were considered to have controlled diabetes.

The investigators then looked at three main surgical outcomes in all knee replacement patients: deep infection, blood clots in the legs or lungs, and revision surgery (an operation to replace a failed knee implant). They also looked at the rates of heart attack and hospitalization for any reason within a year after surgery.

Adverse outcomes in general were rare. Fewer than 1 percent of patients developed a deep infection or blood clot, slightly more than 1 percent underwent revision surgery and another 1 percent experienced heart attack. Hospitalizations for any reason within a year of surgery were more common, experienced by 27.1 percent of patients.

After adjusting for age, sex, weight and other health problems, no differences in outcomes were seen in patients who had controlled or uncontrolled diabetes compared with those who did not have diabetes.

Adams says there are several possible reasons why the Kaiser results differ from earlier studies.  

One is a difference in the methods used to determine the glycemic status of patients with diabetes. Previous researchers relied on diagnostic codes, whereas the Kaiser team looked at actual laboratory tests as well as codes. They also used data from the large Kaiser Permanente Total Joint Replacement Registry to validate surgical outcomes.

Another possible reason for the different results is that the Kaiser patients with diabetes were healthier and their outcomes better overall.

“Except for rehospitalization, all the outcomes we studied were rare in our population,” Adams says. “Patients with poorly controlled diabetes were also rare – even when we tried changing the definition of glycemic control.” She notes that the rarity of both makes it harder to draw statistically significant conclusions.

Still, Adams says the evidence clearly suggests that factors other than blood sugar levels are responsible for postsurgical complications in patients with diabetes.

“I think multiple factors are involved. Diabetes can have a trickle-down effect on many body functions and organs, so it is not operating in isolation. There are other things [such as obesity] that are interacting with it,” she explains.

Kaiser researchers undertook the study because more than half of all diabetes patients also have osteoarthritis (OA), and may eventually need joint replacement surgery, Adams says.

“Diabetes matters, and doctors and patients need to be mindful of the presence of diabetes when considering major surgery,” she says. “But it is not the only factor in play, and glycemic control as we defined it does not appear to lead to poor surgical outcomes.”  

Osama Hamdy, MD, PhD,  medical director of the Obesity Clinical Program and clinical investigator at the Joslin Diabetes Center in Boston and an assistant professor of medicine at Harvard Medical School, is not convinced.

“Although this study suggests that you don’t have to postpone elective surgery for total knee replacement if diabetes is uncontrolled, it’s important to remember that this is a retrospective study that is missing [some] very crucial information, such as the type of diabetes treatment, duration of diabetes and the status of diabetes control during hospitalization when [insulin therapy] can rapidly improve blood glucose levels,” he says.

In addition, he says, the study didn’t look at more common complications among hospitalized diabetes patients, such as urinary-tract infection, hospital-acquired pneumonia and acute renal failure.

“Finally, the study didn’t look at length of hospital stay, which is frequently longer if diabetes is uncontrolled,” says Dr. Hamdy. “I think the study is not giving us the whole picture, and we should take its conclusions cautiously until we have a prospective, randomized controlled study.”