The bulge sign – a simple physical examination for identifying progressive knee osteoarthritis: data from the Osteoarthritis Initiative

Abstract

Objective

To examine whether the presence of bulge sign or patellar tap was associated with frequent knee pain, progression of radiographic OA (ROA) and total knee replacement (TKR).

Methods

This study included 4344 Osteoarthritis Initiative participants examined at baseline for bulge sign and/or patellar tap. The clinical signs were categorized as no (none at baseline and 2 years), resolved (present at baseline only), developed (present at 2 years only) and persistent (present at both time points). Frequent knee pain and progression of ROA over 4 years and TKR over 6 years were assessed. Binary logistic regression was used to examine the associations.

Results

A total of 12.7% of participants had bulge sign only, 2.0% had patellar tap only and 3.3% had both. A positive baseline bulge sign was associated with an increased risk of frequent knee pain [OR 1.31 (95% CI 1.04, 1.64), P = 0.02] and TKR [OR 1.47 (95% CI 1.06, 2.05), P = 0.02]. Developed bulge sign was associated with an increased risk of frequent knee pain [OR 1.75 (95% CI 1.34, 2.29), P < 0.001] and progressive ROA [OR 1.67 (95% CI 1.11, 2.51), P = 0.01]. Persistent bulge sign was associated with an increased risk of frequent knee pain [OR 1.60 (95% CI 1.09, 2.35), P = 0.02], progressive ROA [OR 1.84 (95% CI 1.01, 3.33), P = 0.045] and TKR [OR 2.13 (95% CI 1.23, 3.68), P = 0.007]. Patellar tap was not examined for its association with joint outcomes due to its low prevalence.

Conclusion

The presence of bulge sign identifies individuals at increased risk of frequent knee pain, progression of ROA and TKR. This provides clinicians with a quick, simple, inexpensive method for identifying those at higher risk of progressive knee OA who should be targeted for therapy.

Rheumatology key messages

• Knee effusion detected using bulge sign was associated with adverse clinical and structural joint outcomes.

• Bulge sign provides a simple method for identifying individuals at risk of progressive knee OA.

Introduction

Knee OA is a major cause of pain and disability and a common reason for presentation to primary care [1]. It has a multifactorial aetiology and the natural history of knee OA varies across patients [2, 3]. It is important to identify individuals at higher risk of disease progression so that these people can be targeted for treatment in order to improve patient outcomes [4, 5]. Currently there is no simple method to determine which people are at most risk of onset and progression of knee OA. A simple clinical test that enables this has the potential to improve patient outcomes and reduce health care costs by targeting treatment to those patients at increased risk of adverse joint outcomes.

There is increasing evidence that knee effusion plays an important role in the pathogenesis of knee OA [6–14]. A synovial effusion assessed by MRI identifies individuals at risk of cartilage loss [6–9], development of knee OA [10, 11], worsening of knee pain [12, 13] and knee replacement surgery [14]. However, using MRI to detect synovial effusion adds significantly to health care costs and participant burden, and may be unnecessary if a routine physical examination of the knee could identify those at risk of knee OA progression. The presence of a knee effusion in OA indicates the presence of inflammation. A knee effusion can be detected using physical examination in the clinical setting using the bulge sign and patellar tap tests. These are used in routine clinical practice to detect the presence of a knee effusion and have shown adequate interrater reliability [15]. While the bulge sign is used to detect small effusions, a large effusion must be present for a patellar tap to be present [16, 17]. These clinical tests may offer the potential of a simple method for predicting clinical and structural joint outcomes in knee OA. In a large prospective cohort study of people with or at risk for knee OA, a positive bulge sign at baseline was associated with an increased risk of knee replacement within 3 years [18].

The aim of this study was to determine whether the presence of a bulge sign or patellar tap at baseline and their patterns over 2 years were associated with frequent knee pain and progression of radiographic OA (ROA) over 4 years and the risk of total knee replacement (TKR) over 6 years in a large cohort of individuals with or at risk for knee OA. We hypothesized that the presence of bulge sign or patellar tap at baseline and persistence of bulge sign or patellar tap over 2 years would be associated with increased risk of frequent knee pain, progression of ROA and TKR.

Methods

Osteoarthritis Initiative (OAI)

Data were extracted from the OAI database, a publicly available multicentre observational cohort study of knee OA (https://oai.nih.gov). The OAI comprises data of 4796 participants 45–79 years of age at baseline. OAI exclusion criteria were inflammatory arthritis, severe joint space narrowing in both knees, unilateral knee joint replacement and severe joint space narrowing in the contralateral knee, inability to undergo MRI or to provide a blood sample, required use of walking aids except a single straight cane ⩽50% of the time or unwilling to provide informed consent. Participants were recruited at four clinical sites and the study was approved by the institutional review boards at each of the sites. All participants gave informed consent.

Participants in the current study

Bilateral standing posteroanterior fixed-flexion knee radiographs [19] were assessed for baseline Kellgren–Lawrence (KL) grading (0–4) (n = 4369). If both knees had no evidence of ROA, the dominant knee was selected for analyses. If only one knee had evidence of ROA, this was the knee selected for analyses. If both knees had evidence of ROA, the most severe knee (i.e. highest KL grade) was selected for analyses. When the severity was equal between sides, the most painful knee was selected for analyses. In the case of equal pain in both knees, the dominant knee was selected for analyses. The current study included participants with KL grade, bulge sign and/or patellar tap data available at baseline (n = 4344).

Clinical assessment of knee effusion – the bulge sign and patellar tap

Physical examination of bulge sign and patellar tap was performed at baseline and the 2 year follow-up, with the participants lying supine on the examination table in a relaxed comfortable position with the knees in the extended position and all muscles relaxed [20, 21]. Examiners performed knee examination under supervision of physicians after receiving central training. At least one examination performed by each examiner per month was repeated or observed by the local investigator for quality assurance [20]. For bulge sign, the examiner stroked upwards with the edge of the hand on the medial side of the knee to drain the fluid proximal to the patella. The examiner then proceeded to push the fluid inferiorly into the lateral aspect of the knee. The test was deemed positive if the examiner observed fluid moving towards the medial recess of the knee. For patellar tap, the examiner emptied the suprapatellar pouch by pressing the area above the kneecap with the palm of one hand. This pushed the fluid under the patella and lifted it. While keeping the pressure on with the first hand, the examiner used the fingertips of his/her other hand to press down on the patella. The test was deemed positive if the patella was felt to move down and tap against the femur. The tests were repeated once by the same examiner to confirm findings (positive or negative) and were recorded as positive if confirmed in the second examination [20]. Data regarding the reliability of these examinations were not available but interrater reliability of bulge sign and patellar tap has been previously reported as adequate (intraclass correlation coefficient 0.97 for bulge sign and a prevalence-adjusted bias-adjusted κ of 0.78 for patellar tap) [15]. The patterns of bulge sign and patellar tap from baseline to the 2 year follow-up were categorized as no sign (present at neither baseline nor 2 year follow-up), resolved sign (present at baseline but not the 2 year follow-up), developed sign (present at the 2 year follow-up but not baseline) and persistent sign (present at both baseline and the 2 year follow-up).

Assessment of knee pain

Knee pain was assessed using the Likert scale version of the WOMAC pain subscale [22] at baseline and yearly follow-ups. It consists of five items, with the score ranging from 0 to 20, with 20 being the worst pain. ‘Symptomatic’ was defined as a WOMAC pain score ⩾5, based on the Low Intensity Symptom State-attainment Index cut-off [23]. This definition has been used in a previous OAI study in which a WOMAC pain score ⩾5 represented the upper tertile of all participants with any knee pain in the cohort [24]. Frequent knee pain from baseline to the 4 year follow-up was defined if a participant reported symptomatic knee pain at three or more of the five time points. Frequent knee pain from the 2 year to 4 year follow-up was defined if a participant reported symptomatic knee pain at two or more of three time points.

Assessment of progression of ROA

Progression of ROA from baseline to the 4 year follow-up was defined by an increase in KL grade of ⩾1 from baseline to the 4 year follow-up in participants with a baseline KL grade of 0–3. Progression of ROA from the 2 year to 4 year follow-up was defined by an increase in KL grade of ⩾1 from the 2 year to 4 year follow-up in participants with a KL grade of 0–3 at 2 years.

Assessment of TKR

At each available follow-up, participants indicated whether they had received TKR surgery. It was defined as the study knee with patient-reported TKR, which was confirmed on subsequent radiograph between baseline and the 6 year follow-up visit. For participants with missing data for TKR due to loss to follow-up at any study visit, it was assumed that the participant had not undergone knee replacement surgery.

Statistical analyses

Demographic, clinical and radiological data were systematically entered into a computerized database. Descriptive statistics of participant characteristics were tabulated. The χ² test and binary logistic regression were used to examine the relationship between KL grade and the prevalence of bulge sign/patellar tap at baseline. Binary logistic regression was used to examine the associations of prevalent bulge sign/patellar tap at baseline and patterns of bulge sign/patellar tap over 2 years with frequent knee pain and progression of ROA over 4 years and the risk of TKR over 6 years. As TKR is unlikely to be performed in people without ROA, supported by data in the current study that only eight participants with a KL grade of 0–1 at baseline had a TKR over 6 years, the risk of TKR was examined in participants with a KL grade of 2–4 at baseline. All the analyses were adjusted for gender, baseline age, BMI and KL grade, with analyses of ROA progression and risk of knee replacement further adjusted for baseline WOMAC pain score. Multiple imputation by chained equations [25] was used to impute missing data on BMI, WOMAC pain score, prevalence and patterns of bulge sign and patellar tap, frequent knee pain and progression of ROA over 4 years. All tests were two-sided and a P-value <0.05 was considered statistically significant. Statistical analyses were performed using Stata 14.0 (StataCorp, College Station, TX, USA).

Results

Participant characteristics at baseline, as well as the change in bulge sign, patellar tap and joint outcomes, are shown in Table 1. At baseline, 1051 (24.2%) participants had knee pain (WOMAC pain score ⩾5) and 2233 (51.4%) participants had ROA (KL grade ⩾2). The prevalence of bulge sign and patellar tap was 16.1% and 5.3%, respectively. Over 2 years, 429 (11.3%) participants had bulge sign resolved, 412 (10.9%) participants developed bulge sign and 177 (4.7%) had persistent bulge sign; the corresponding figures for patellar tap were 151 (4.0%), 218 (5.8%) and 34 (0.9%), respectively. Over 4 years, 626 (16.8%) participants reported frequent knee pain, and progression of ROA was observed in 474 (14.7%) participants with a KL grade of 0–3. Over 6 years, 221 (5.1%) participants had a TKR, mostly [n = 213 (96.4%)] in those with a KL grade of 2–4. There was no significant difference in age, gender, BMI or prevalence of bulge sign/patellar tap between participants who completed the 4 year follow-up (n = 3356) and those who did not (n = 988). The non-completers had a higher KL grade (1.7 vs 1.4; P < 0.001) and worse knee pain (3.6 vs 2.6; P < 0.001) and were more likely to have ROA (57.5% vs 49.6%; P < 0.001) than the completers.

Table 1.

Characteristics of study participants (N = 4344)

	Values
Baseline characteristics
Age, years, mean (s.d.)	61.3 (9.2)
Female, n (%)	2520 (58.0)
BMI, kg/m2, mean (s.d.)	28.7 (4.8)
WOMAC pain score (range 0–20), mean (s.d.)	2.8 (3.4)
WOMAC pain score ≥5, n (%)	1051 (24.2)
KL grade, n (%)
0	1426 (32.8)
1	685 (15.8)
2	1164 (26.8)
3	784 (18.0)
4	285 (6.6)
Positive bulge sign, n (%)	698 (16.1)
Positive patellar tap, n (%)	230 (5.3)
Change
Patterns of bulge sign over 2 years (n = 3791), n (%)
No bulge sign	2773 (73.1)
Resolved bulge sign	429 (11.3)
Developed bulge sign	412 (10.9)
Persistent bulge sign	177 (4.7)
Patterns of patellar tap over 2 years (n = 3781), n (%)
No patellar tap	3378 (89.3)
Resolved patellar tap	151 (4.0)
Developed patellar tap	218 (5.8)
Persistent patellar tap	34 (0.9)
Frequent knee pain from baseline to 4 years (n = 3733), n (%)	626 (16.8)
Frequent knee pain from 2 years to 4 years (n = 3797), n (%)	667 (17.6)
Progression of ROA from baseline to 4 years (n = 3222), n (%)	474 (14.7)
Progression of ROA from 2 years to 4 years (n = 2963), n (%)	217 (7.3)
TKR from baseline to 6 years (n = 4344), n (%)	221 (5.1)
TKR from baseline to 6 years in those with ROA at baseline (n = 2233), n (%)	213 (9.5)
TKR from 2 years to 6 years (n = 4296), n (%)	173 (4.0)
TKR from 2 years to 6 years in those with ROA at baseline (n = 2186), n (%)	166 (7.6)

Of the 4341 participants examined for both bulge sign and patellar tap at baseline, 551 (12.7%) participants had bulge sign only, 85 (2.0%) had patellar tap only and 145 (3.3%) had both (Fig. 1). The prevalence of bulge sign and patellar tap increased with a higher KL grade (P < 0.001) (Fig. 2).

Fig. 1.

Venn diagram for the prevalence of bulge sign and patellar tap at baseline

Fig. 2.

Prevalence of bulge sign and patellar tap according to KL grade at baseline

(A) Prevalence (standard error) of bulge sign according to KL grade (P < 0.001). (B) Prevalence (standard error) of patellar tap according to KL grade (P < 0.001).

Due to the low prevalence of patellar tap at baseline (5.3%), the low number of participants in most of the categories of patellar tap patterns over 2 years and the very low sensitivity of patellar tap for MRI-detected medium/large effusion synovitis (7.6–10.8%) [20, 21], we did not perform analyses on the association between patellar tap and joint outcomes.

The associations between bulge sign and joint outcomes are shown in Table 2. After adjustment for gender, baseline age, BMI and KL grade, the presence of bulge sign at baseline was associated with an increased risk of frequent knee pain [odds ratio (OR) 1.31 (95% CI 1.04, 1.64), P = 0.02]. After adjustment for gender, baseline age, BMI, KL grade and WOMAC pain score, the presence of baseline bulge sign was associated with an increased risk of TKR [OR 1.47 (95% CI 1.06, 2.05), P = 0.02] but not the progression of ROA [OR 1.20 (95% CI 0.92, 1.56), P = 0.17]. While resolved bulge sign was not significantly associated with the risk of frequent knee pain [OR 1.25 (95% CI 0.94, 1.65), P = 0.12], both developed [OR 1.75 (95% CI 1.34, 2.29), P < 0.001] and persistent [OR 1.60 (95% CI 1.09, 2.35), P = 0.02] bulge sign were associated with an increased risk of frequent knee pain (P for trend < 0.001). While resolved bulge sign was not significantly associated with the risk of progressive ROA [OR 1.02 (95% CI 0.64, 1.64), P = 0.92], both developed [OR 1.67 (95% CI 1.11, 2.51), P = 0.01] and persistent [OR 1.84 (95% CI 1.01, 3.33), P = 0.045] bulge sign was associated with an increased risk of progressive ROA (P for trend = 0.005). While resolved [OR 1.39 (95% CI 0.86, 2.25), P = 0.18] or developed [OR 1.35 (95% CI 0.82, 2.24), P = 0.24] bulge sign was not significantly associated with the risk of TKR, persistent bulge sign was associated with an increased risk of TKR [OR 2.13 (95% CI 1.23, 3.68), P = 0.007] (P for trend = 0.008). The results remained similar when the imputed dataset was analysed (Supplementary Table 1, available at Rheumatology online).

Table 2.

Associations of bulge sign with frequent knee pain, progressive ROA and TKR

	Univariable analysis, OR (95% CI)	P-value	Multivariable analysis, OR (95% CI)	P-value
Baseline bulge sign
Frequent knee pain from baseline to the 4-year follow-up (≥3 of 5 time points)a
Presence of bulge sign	1.70 (1.38, 2.10)	<0.001	1.31 (1.04, 1.64)	0.02
Progression of ROA from baseline to the 4-year follow-upb
Presence of bulge sign	1.40 (1.08, 1.81)	0.01	1.20 (0.92, 1.56)	0.17
TKR from baseline to the 6-year follow-upb,c
Presence of bulge sign	2.07 (1.53, 2.80)	<0.001	1.47 (1.06, 2.05)	0.02
Patterns of bulge sign
Frequent knee pain from the 2-year to 4-year follow-up (≥2 of 3 time points)a
No bulge sign	1.00		1.00
Resolved bulge sign	1.57 (1.21, 2.03)	0.001	1.25 (0.94, 1.65)	0.12
Developed bulge sign	2.07 (1.61, 2.66)	<0.001	1.75 (1.34, 2.29)	<0.001
Persistent bulge sign	2.22 (1.55, 3.17)	<0.001	1.60 (1.09, 2.35)	0.02
P for trend		<0.001		<0.001
Progression of ROA from the 2-year to 4-year follow-upb
No bulge sign	1.00		1.00
Resolved bulge sign	1.08 (0.68, 1.71)	0.76	1.02 (0.64, 1.64)	0.92
Developed bulge sign	1.72 (1.15, 2.58)	0.01	1.67 (1.11, 2.51)	0.01
Persistent bulge sign	2.00 (1.11, 3.58)	0.02	1.84 (1.01, 3.33)	0.045
P for trend		0.001		0.005
TKR from the 2-year to 6-year follow-upb,c
No bulge sign	1.00		1.00
Resolved bulge sign	1.91 (1.21, 3.01)	0.006	1.39 (0.86, 2.25)	0.18
Developed bulge sign	1.55 (0.96, 2.51)	0.07	1.35 (0.82, 2.24)	0.24
Persistent bulge sign	3.16 (1.89, 5.28)	<0.001	2.13 (1.23, 3.68)	0.007
P for trend		<0.001		0.008

^aAdjusted for gender, baseline age, BMI and KL grade.

^bAdjusted for gender, baseline age, BMI, KL grade and WOMAC pain score.

^cAnalyses performed in participants with KL grade ≥2 at baseline.

Discussion

This study demonstrated that a clinically detected knee effusion using the bulge sign was associated with adverse clinical and structural joint outcomes over the medium to long term, including frequent knee pain and progression of ROA over 4 years and TKR over 6 years. A positive bulge sign at baseline was associated with increased risk of frequent knee pain over 4 years and total knee replacement over 6 years. Furthermore, developed bulge sign from baseline to 2 years was associated with an increased risk of frequent knee pain and progressive ROA from 2 years to 4 years, and persistent bulge sign over 2 years was associated with an increased risk of frequent knee pain and progressive ROA from 2 years to 4 years and an increased risk of TKR from 2 years to 6 years. These results demonstrate that the persistent presence of a knee effusion at multiple time points on clinical examination identifies people at increased risk of disease progression.

In clinical practice, synovial effusion can be detected clinically, or more definitively using imaging. However, the use of MRI is costly and may not always be available. Clinical examination using a bulge sign can detect a knee effusion in a simple, inexpensive, non-invasive and highly reproducible manner [15]. Previous studies examined the diagnostic performance of bulge sign for MRI-detected effusion synovitis using the OAI data and showed a sensitivity of 37.6–40.3% and a specificity of 73.7–86.9% for medium/large effusion synovitis [20, 21]. The only previous study examining the clinical implications of bulge sign for joint outcomes showed an association between the presence of bulge sign at baseline and rapid progression to knee arthroplasty within 3 years in the OAI in people with or at risk for knee OA [18]. This previous study aimed to identify patient factors predicting rapid progression to knee arthroplasty by examining a multitude of baseline person-level and knee-level factors. Bulge sign was 1 of the 31 variables in the multivariable analyses [18]. Although the statistical method used to screen the predictor variables was valid and clearly presented, including 31 variables in multivariate logistic regression analyses simultaneously would be a problem due to the limited number of knee arthroplasties (128 knees, 116 persons) and the inadequate power of the study.

In this current study we aimed to examine the association between the presence and patterns of bulge sign (a physical examination of knee effusion/inflammation) and clinical and structural joint outcomes with a longer follow-up. We did not examine the other potential predictors for disease progression. We chose the potential confounders based on the literature, which are established risk factors for progression of knee OA that are also associated with knee effusion/inflammation, including age, gender, BMI, KL grade and WOMAC pain score. We did not perform Bonferroni correction, as there were only two exposures (presence and patterns of bulge sign) and the three outcomes (frequent knee pain, progression of ROA and TKR) were interrelated and represented different measures of the same disease process. The results from our study extend the findings of the previous study [18], showing the association of a positive bulge sign at baseline and a spectrum of clinically significant joint outcomes, including frequent knee pain over 4 years and TKR over 6 years, independent of the severity of knee OA. As the presence of bulge sign may fluctuate over time, our study also extended previous data by examining the patterns of bulge sign over 2 years and their associations with joint outcomes from 2 years onwards. We found developed bulge sign over 2 years was associated with an increased risk of frequent knee pain and progressive ROA from 2 years to 4 years and persistent bulge sign over 2 years was associated with an increased risk of frequent knee pain and progressive ROA from 2 years to 4 years and TKR from 2 years to 6 years. These data suggest that the presence of bulge sign, particularly when it is present on more than one occasion, may be used as a clinical marker to identify people at increased risk of disease progression with deteriorating clinical and structural joint outcomes over the medium to long term.

Knee OA is a heterogeneous chronic condition with varied natural history and multiple phenotypes [2, 3]. Currently clinicians do not have a simple method to identify patients who are most at risk of disease progression to be targeted for early and tailored therapy. Based on the findings of the current study, a clinically detected knee effusion using the bulge sign may provide a simple means of enabling clinicians in the primary care setting to predict disease trajectories for their patients. If a clinically detected knee effusion identifies those with an OA phenotype who are at increased risk of disease progression, this would enable personalized or precision medicine, targeting patients most likely to benefit from the treatment to optimize patient outcomes. As the presence of knee effusion reflects synovial inflammation [26, 27], therapies with potent anti-inflammatory effects may offer a theoretical construct for disease modification. Additionally, these people should also be targeted for treatments that modify other known risk factors contributing substantially to the progression of knee OA, such as obesity and muscle weakness [28]. Previous studies have examined the clinically detected knee effusion, including the bulge sign, as a predictor of response to intra-articular corticosteroids in clinical trials of patients with knee OA [29, 30], and patients with clinically detectable effusions showed the best response to treatment in terms of pain relief [29]. There is also evidence that chondroitin sulphate or combined chondroitin sulphate and glucosamine reduces joint swelling and effusion assessed by clinical examination in patients with painful knee OA [31, 32] and chondroitin sulphate reduces cartilage volume loss in knee OA patients with clinical signs of synovitis [33–35]. Taken together, these data support the importance of targeting patients with a clinically detected knee effusion for therapeutic intervention.

This study has limitations. Two clinical signs of effusion, i.e. bulge sign and patellar tap, were examined in the OAI. In the current study, patellar tap was not investigated for its relationship with joint outcomes due to its low prevalence and very low sensitivity in detecting knee effusion [20, 21]. In the OAI, physical examination of bulge sign was performed by examiners who had received central training and were supervised by physician examiners. Any misclassification is mostly likely to be non-differential and thus underestimates the magnitude of observed associations. Due to unavailability of the data, we were unable to examine the progression of patellofemoral OA. Although most of the data analysed in our study were available for both knees, the target knee (i.e. the most symptomatic knee) for each participant was chosen for the statistical analyses, which is the one most likely to have disease progression, thus increasing the power of the study. Therefore we do not believe it would change the conclusion of this study if we perform the statistical analyses on both knees. Despite the fact that the following is speculative, there was the probability that the contralateral knee had no OA symptoms and thus did not develop the disease in the time studied and, even if there was a slight sign of OA, it may not progress. Thus it will increase the background noise. We will, however, keep this option in mind for future studies using this cohort.

Participants who dropped out had more severe ROA and worse knee pain. As these participants were more likely to have persistent knee pain, progressive ROA and TKR, their dropout is likely to have reduced the power of the study to detect the associations between bulge sign and knee joint outcomes. For the incidence of TKR, we have assumed that participants with missing data due to loss to follow-up did not have TKR. It was likely that some of these participants may have undergone knee replacement surgery but were not reported or confirmed on subsequent radiographs. This is likely to result in non-differential misclassification of TKR and thus underestimates the magnitude of observed associations.

The present study also has a number of strengths. The OAI intentionally recruited participants with or at risk for knee OA. This enriches the population for the exposures and outcomes of interest, providing a unique opportunity to investigate the association between clinical signs and joint outcomes with adequate power. The long follow-up duration of this study, combined with the high retention rates, enabled several clinical and structural outcomes to be examined. The use of a clinically detected knee effusion is a strength of this study, as it represents a simple and non-invasive means of predicting disease trajectories, available to all clinicians, without requiring more sophisticated and costly use of imaging. The bulge sign was assessed at baseline and 2 years later by physical examination with high interobserver reliability [15]. From these, the patterns of bulge sign were investigated, with relationships observed between developed and persistent bulge sign over 2 years and adverse clinical and structural joint outcomes over 4 or 6 years. This is important, as knee effusion may fluctuate with time, and thus an isolated baseline assessment may provide limited insight about the effect of clinically detected effusion on joint outcomes many years later.

This study has shown that the presence of a clinically detected knee effusion, identified by the presence of a bulge sign, identifies individuals at increased risk of progression of knee OA across the spectrum of clinical and structural outcomes, including frequent knee pain, progression of ROA and TKR. The presence of a bulge sign, particularly when present on more than one occasion, provides clinicians with a quick, simple and inexpensive method for identifying those who should be targeted for therapy.

Supplementary data

Supplementary data are available at Rheumatology online.