Background/Purpose: Despite clinicians commonly using the bulge sign test to assess presence of knee effusion, prior studies have questioned the validity of the test to detect the prevalence of knee effusion based on ultrasound imaging. It remains unknown if the bulge sign test over time is a sensitive test to detect changes in knee effusion. Hence, we evaluated the association between the bulge sign test and magnetic resonance (MR) imaging-based knee effusion volume at a single visit and change over 2 years.

Methods: We selected individuals from the Osteoarthritis Initiative (OAI) who had no radiographic knee osteoarthritis at baseline and had MR images and a clinical exam at baseline and 24-month OAI visits. We enriched the sample to include 250 individuals who had an increase in radiographic severity (Kellgren-Lawrence grade). We selected the 1 knee per person. Two readers used a semi-automated software to quantify MR-based knee effusion volume (includes synovitis and effusion; intra-tester reliability: 0.73 to 0.97). Reader 1 reviewed all segmented images to ensure consistency between readers, knees, and time. The bulge sign was performed during a standardized clinical exam. We defined an incident bulge sign as a person with a positive bulge sign at the 24-month visit but not at baseline. We used an independent-sample t-test to compare baseline effusion volume between those with and without a positive bulge sign at baseline. A second independent-test was used to compare change in effusion volume between those with an incident bulge sign and those who never had a positive bulge sign. We also used logistic regression with ROC curves to explore the optimal cutpoint for baseline effusion volume and change in effusion volume for baseline positive bulge sign and incident bulge sign, respectively.

Results are reported as mean (standard deviation). Results: At baseline of OAI visit 352 participants were eligible participants with a mean age of 59.6 (8.6) years, mean BMI of 28.4 (4.5) kg/m2, and 62% female. The 47 participants with a positive bulge sign (15.38 [12.83] cm³) had greater knee effusion (t=-3.45, p=0.001) than those without a positive bulge sign (8.81 [6.14] cm³). Based on an ROC curve, knee effusion volume had an area under the curve of 0.67 with an optimal cutoff between those with and without a positive bulge sign of 8.39 cm³. At the 24-month OAI visit, the 40 (14%) participants with an incident positive bulge sign (9.85 [11.90] cm³) had greater increases in knee effusion volume(t = -4.21, p = 0.0001) than the 251 participants without an incident positive bulge sign (1.65 [7.69] cm³). Based on an ROC curve, change in knee effusion volume had an area under the curve of 0.70 with an optimal cutoff between those with and without a positive bulge sign of 2.06 cm³.

Conclusion:  The bulge sign test is associated with effusion volume and changes in effusion volume but may misclassify individuals with and without larger volumes of effusion or those with larger increases in knee effusion volume. Clinicians and researchers need to be aware of these properties when relying on the bulge sign test.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/the-clinical-utility-of-the-bulge-sign-in-evaluating-knee-osteoarthritis/