Four factors predict likelihood of future knee replacement

PHILADELPHIA – Knowledge of a patient’s age, gender, quality of life score, and radiographic knee osteoarthritis severity helps predict the odds of a patient undergoing total knee arthroplasty in the next 5 years.

Few studies have looked at the predictors of total knee arthroplasty (TKA) in community-based cohorts, despite the heavy burden of TKA in the United States, said Dr. Marc C. Hochberg, head of rheumatology and clinical immunology, University of Maryland, Baltimore. An estimated 4 million Americans already live with TKA, and more than half of adults diagnosed with knee osteoarthritis (OA) will undergo TKA (J. Bone Joint Surg. Am. 2013;95:385-92).

For the current analysis, Dr. Hochberg and his associates obtained 48 months of annual clinical follow-up data on 6,406 patients in the Osteoarthritis Initiative who had symptomatic, radiographic knee OA or who were at risk for the condition. Follow-up at 60 months involved only questionnaires. Consensus readings of knee radiographs were used for the analysis, along with knee-specific multiple variable regression models. The best models were selected based on Chi-square values and area under the receiver operating characteristic curve (AUC). Radiographic knee OA Kellgren-Lawrence (KL) severity grade 0, 1, 2, 3, and 4 was present in the right knee of 36%, 18%, 28%, 14%, and 3.5% of patients and in the left knee of 38%, 18%, 26%, 14.4%, and 3.3% of patients.

A little more than half of participants were women (58.4%), mean body mass index was 28.7 kg/m², mean Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain score was 2.47 (0-20 scale), mean WOMAC function score was 8.33 (0-68 scale), and mean Knee injury and Osteoarthritis Outcome Score (KOOS) (quality of life) was 66.10 (0-100 scale). Their average age was 62.7, he said at the World Congress on Osteoarthritis, sponsored by the Osteoarthritis Research Society International.

During the 60 months of follow-up, there were 91 right and 102 left TKAs. The AUC for age, sex, and BMI was 0.64 for the right knee and 0.67 for the left knee.

The best TKA prediction models included age, sex, BMI, and KOOS quality-of-life score, improving the AUC to 0.78 for the right knee and 0.80 for the left knee, Dr. Hochberg said. Addition of the WOMAC pain score was associated with a marginal, but significant improvement in the AUC (0.79 and 0.80, respectively). When KL grade was added to these models, the AUC reached 0.89 and 0.91 for the right and left knees, respectively, he said. Sensitivity analyses failed to demonstrate additional effects of other baseline variables when added.

Limitations of the study included the relatively few TKAs, that most patients had KL grade 2 at baseline, proportional hazards analyses were not used, and the study did not consider a correlation between knees or the change in independent variables over time, he said.

In a prospective Canadian study, willingness to undergo surgery was the strongest predictor of the time to first hip or knee total joint replacement (TJR) (Arthritis Rheum. 2006;54:3212-20), while an international OARSI Task Force reported that pain and disability was higher in patients recommended for hip or knee TJR (Osteoarthritis Cartilage 2011;19:147-54). The Task Force, which included Dr. Hochberg, could not, however, identify cut points for pain and disability that would discriminate between those who did or did not get the nod for surgery.

The Osteoarthritis Initiative is sponsored by the National Institutes of Health. Dr. Hochberg reported serving as a consultant for Allergan, Bioiberica, Iroko, Merck, and Pfizer, and as a scientific/medical advisory board member for Eli Lilly and Merck.

[email protected]

PHILADELPHIA – Knowledge of a patient’s age, gender, quality of life score, and radiographic knee osteoarthritis severity helps predict the odds of a patient undergoing total knee arthroplasty in the next 5 years.

Few studies have looked at the predictors of total knee arthroplasty (TKA) in community-based cohorts, despite the heavy burden of TKA in the United States, said Dr. Marc C. Hochberg, head of rheumatology and clinical immunology, University of Maryland, Baltimore. An estimated 4 million Americans already live with TKA, and more than half of adults diagnosed with knee osteoarthritis (OA) will undergo TKA (J. Bone Joint Surg. Am. 2013;95:385-92).

For the current analysis, Dr. Hochberg and his associates obtained 48 months of annual clinical follow-up data on 6,406 patients in the Osteoarthritis Initiative who had symptomatic, radiographic knee OA or who were at risk for the condition. Follow-up at 60 months involved only questionnaires. Consensus readings of knee radiographs were used for the analysis, along with knee-specific multiple variable regression models. The best models were selected based on Chi-square values and area under the receiver operating characteristic curve (AUC). Radiographic knee OA Kellgren-Lawrence (KL) severity grade 0, 1, 2, 3, and 4 was present in the right knee of 36%, 18%, 28%, 14%, and 3.5% of patients and in the left knee of 38%, 18%, 26%, 14.4%, and 3.3% of patients.

A little more than half of participants were women (58.4%), mean body mass index was 28.7 kg/m², mean Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain score was 2.47 (0-20 scale), mean WOMAC function score was 8.33 (0-68 scale), and mean Knee injury and Osteoarthritis Outcome Score (KOOS) (quality of life) was 66.10 (0-100 scale). Their average age was 62.7, he said at the World Congress on Osteoarthritis, sponsored by the Osteoarthritis Research Society International.

During the 60 months of follow-up, there were 91 right and 102 left TKAs. The AUC for age, sex, and BMI was 0.64 for the right knee and 0.67 for the left knee.

The best TKA prediction models included age, sex, BMI, and KOOS quality-of-life score, improving the AUC to 0.78 for the right knee and 0.80 for the left knee, Dr. Hochberg said. Addition of the WOMAC pain score was associated with a marginal, but significant improvement in the AUC (0.79 and 0.80, respectively). When KL grade was added to these models, the AUC reached 0.89 and 0.91 for the right and left knees, respectively, he said. Sensitivity analyses failed to demonstrate additional effects of other baseline variables when added.

Limitations of the study included the relatively few TKAs, that most patients had KL grade 2 at baseline, proportional hazards analyses were not used, and the study did not consider a correlation between knees or the change in independent variables over time, he said.

In a prospective Canadian study, willingness to undergo surgery was the strongest predictor of the time to first hip or knee total joint replacement (TJR) (Arthritis Rheum. 2006;54:3212-20), while an international OARSI Task Force reported that pain and disability was higher in patients recommended for hip or knee TJR (Osteoarthritis Cartilage 2011;19:147-54). The Task Force, which included Dr. Hochberg, could not, however, identify cut points for pain and disability that would discriminate between those who did or did not get the nod for surgery.

The Osteoarthritis Initiative is sponsored by the National Institutes of Health. Dr. Hochberg reported serving as a consultant for Allergan, Bioiberica, Iroko, Merck, and Pfizer, and as a scientific/medical advisory board member for Eli Lilly and Merck.

[email protected]

PHILADELPHIA – Knowledge of a patient’s age, gender, quality of life score, and radiographic knee osteoarthritis severity helps predict the odds of a patient undergoing total knee arthroplasty in the next 5 years.

Few studies have looked at the predictors of total knee arthroplasty (TKA) in community-based cohorts, despite the heavy burden of TKA in the United States, said Dr. Marc C. Hochberg, head of rheumatology and clinical immunology, University of Maryland, Baltimore. An estimated 4 million Americans already live with TKA, and more than half of adults diagnosed with knee osteoarthritis (OA) will undergo TKA (J. Bone Joint Surg. Am. 2013;95:385-92).

For the current analysis, Dr. Hochberg and his associates obtained 48 months of annual clinical follow-up data on 6,406 patients in the Osteoarthritis Initiative who had symptomatic, radiographic knee OA or who were at risk for the condition. Follow-up at 60 months involved only questionnaires. Consensus readings of knee radiographs were used for the analysis, along with knee-specific multiple variable regression models. The best models were selected based on Chi-square values and area under the receiver operating characteristic curve (AUC). Radiographic knee OA Kellgren-Lawrence (KL) severity grade 0, 1, 2, 3, and 4 was present in the right knee of 36%, 18%, 28%, 14%, and 3.5% of patients and in the left knee of 38%, 18%, 26%, 14.4%, and 3.3% of patients.

A little more than half of participants were women (58.4%), mean body mass index was 28.7 kg/m², mean Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain score was 2.47 (0-20 scale), mean WOMAC function score was 8.33 (0-68 scale), and mean Knee injury and Osteoarthritis Outcome Score (KOOS) (quality of life) was 66.10 (0-100 scale). Their average age was 62.7, he said at the World Congress on Osteoarthritis, sponsored by the Osteoarthritis Research Society International.

During the 60 months of follow-up, there were 91 right and 102 left TKAs. The AUC for age, sex, and BMI was 0.64 for the right knee and 0.67 for the left knee.

The best TKA prediction models included age, sex, BMI, and KOOS quality-of-life score, improving the AUC to 0.78 for the right knee and 0.80 for the left knee, Dr. Hochberg said. Addition of the WOMAC pain score was associated with a marginal, but significant improvement in the AUC (0.79 and 0.80, respectively). When KL grade was added to these models, the AUC reached 0.89 and 0.91 for the right and left knees, respectively, he said. Sensitivity analyses failed to demonstrate additional effects of other baseline variables when added.

Limitations of the study included the relatively few TKAs, that most patients had KL grade 2 at baseline, proportional hazards analyses were not used, and the study did not consider a correlation between knees or the change in independent variables over time, he said.

In a prospective Canadian study, willingness to undergo surgery was the strongest predictor of the time to first hip or knee total joint replacement (TJR) (Arthritis Rheum. 2006;54:3212-20), while an international OARSI Task Force reported that pain and disability was higher in patients recommended for hip or knee TJR (Osteoarthritis Cartilage 2011;19:147-54). The Task Force, which included Dr. Hochberg, could not, however, identify cut points for pain and disability that would discriminate between those who did or did not get the nod for surgery.

The Osteoarthritis Initiative is sponsored by the National Institutes of Health. Dr. Hochberg reported serving as a consultant for Allergan, Bioiberica, Iroko, Merck, and Pfizer, and as a scientific/medical advisory board member for Eli Lilly and Merck.

[email protected]