- Email: [email protected]
Validation of the KOFUS (Knee Osteoarthritis Flare-Ups Score)
Available online at
www.sciencedirect.com Joint Bone Spine 76 (2009) 268e272
Original article
Validation of the KOFUS (Knee Osteoarthritis Flare-Ups Score) Marc Marty a,*, Pascal Hilliquin b, Sylvie Rozenberg c, Jean Pierre Valat d, Eric Vignon e, Philippe Coste f, Bernard Savarieau g, Franc¸ois Andre´ Allaert h b
a Service de Rhumatologie, Centre hospitalier Henri Mondor, Cre´teil 94010, France Service de Rhumatologie, Centre hospitalier Sud Francilien, Corbeil-Essonnes, France c Service de Rhumatologie, Groupe hospitalier Pitie´-Salpe´trie`re, Paris, France d Service de Rhumatologie, Universite´ Franc¸ois-Rabelais, Tours, France e Service de Rhumatologie, CHU de Lyon sud, Pierre Benite, France f Laboratoires Expanscience, De´partement me´dical, Puteaux, France g Gestion de l’e´tude, Nukle´us, Paris, France h Service de Biome´trie, Cenbiotech, Dijon, France
Accepted 17 July 2008 Available online 17 March 2009
Abstract Objective: To develop a diagnostic score for knee osteoarthritis flare-ups and to evaluate its sensitivity and specificity. Methods: We used two epidemiological databases built using the same methodology. One database was from a general-practice study and served to develop the score, whereas the other was from a rheumatology study and served to validate the score. Physicians determined the flare-up status of each patient. The rheumatologist diagnosis was the reference standard. Logistic regression was performed to identify factors significantly associated with having a flare-up. Results: Of the 6085 patients in the general-practice database, 52.3% had a knee osteoarthritis flare-up. The score was built by assigning points to features that were present, with a weighting system based on the odds ratio of each feature for having a flare-up (0, feature absent; 1, morning stiffness for longer than 20 min; 2, pain causing nocturnal awakenings and knee effusion; 3, limping, joint swelling, and increased warmth over the knee). The score could range from 0 to 14. The receiver-operating characteristic curve showed that 7 was the best cutoff for diagnosing a flare-up. In the rheumatologist database, the numbers of patients having a flare-up were 274 (46.4%) based on the score and 270 (45.7%) based on the rheumatologist diagnosis. Sensitivity of the score was 87.0%, specificity 87.9%, positive predictive value 85.8%, and negative predictive value 89.0%. The Youden index was 0.75. Conclusion: A score equal to or greater than 7 points correlated well with a rheumatologist diagnosis of flare-up. Our score may constitute a valid objective criterion for standardizing the diagnosis of knee osteoarthritis flare-up, most notably when screening patients for inclusion in therapeutic trials. Ó 2009 Socie´te´ Franc¸aise de Rhumatologie. Published by Elsevier Masson SAS. All rights reserved. Keywords: Knee osteoarthritis; Flare-up; Diagnostic score; Sensitivity; Specificity
Pain and functional impairment are the main clinical manifestations of osteoarthritis. Symptom intensity varies over time, and patients experience intermittent flare-ups during which the pain and functional impairment worsen [1]. The existence of flare-ups probably contributes to explain the
* Corresponding author. E-mail address: [email protected] (M. Marty).
inconsistent correlation of clinical symptoms and radiological changes, most notably at the knee [2]. The pathological concomitant of osteoarthritis flare-ups is the development of inflammatory lesions within the synovial membrane [3]. The pain becomes more severe and responds less well to rest, tending to become continuous. A joint effusion and warmth may be noted. Cytological examination of fluid aspirated from the knee shows mechanical properties. The synovial membrane releases substances that damage the cartilage, such as
1297-319X/$ - see front matter Ó 2009 Socie´te´ Franc¸aise de Rhumatologie. Published by Elsevier Masson SAS. All rights reserved. doi:10.1016/j.jbspin.2008.07.018
M. Marty et al. / Joint Bone Spine 76 (2009) 268e272
proinflammatory cytokines and metalloproteases, supporting the hypothesis that flare-ups increase the risk of cartilage destruction [4]. In patients with knee osteoarthritis, arthroscopic evidence of synovitis may predict faster cartilage loss over the next year [5]. Another hypothesis, however, is that flare-ups are merely markers for cartilage destruction. Variations in the frequency of flare-ups may explain the variability of the rate of cartilage destruction across patients. Accurate detection of flare-ups should allow early treatment. Nevertheless, the definition of osteoarthritis flare-up is not agreed on. An accurate definition, together with improved knowledge of the natural history of osteoarthritis and/or of the factors associated with flare-ups would help to develop recommendations aimed at limiting the number and adverse impact of flare-ups. The knee is the most appropriate site for studying osteoarthritis flare-ups, as its superficial location facilitates the detection and aspiration of joint effusions. The primary objective of this study was to define knee osteoarthritis flareups based on the clinical and radiological features. The secondary objectives were to describe patients who present with knee osteoarthritis flare-ups and to describe management modalities. 1. Methods Two observational cross-sectional studies were conducted in France, one among general practitioners and the other among rheumatologists, to identify the criteria that define knee osteoarthritis flare-ups. The data obtained from the general practitioners were used to develop a diagnostic score for patients seeking primary care. The data from the rheumatologists served only to validate the score. All the general practitioners and rheumatologists who participated in the study received a detailed research protocol that contained explanations about painful knee osteoarthritis flare-ups. 1.1. Patient selection for the study Patients were eligible for study inclusion if they met the clinical and radiological criteria for knee osteoarthritis established by the American College of Rheumatology [6]. They either had stable disease or were experiencing a flare-up according to the physician’s assessment. Each general practitioner was asked to include 3 patients in each category and each rheumatologist 2 patients in each category. Patients were not eligible if they had secondary knee osteoarthritis (inflammatory joint disease, crystal deposition disease, septic arthritis, or trauma) or a history of surgery on the study knee other than arthroscopy for meniscal disease. 1.2. Evaluation criteria For each patient, we recorded the following data: age, sex, and occupation; duration of the knee osteoarthritis, radiological findings, and current treatments; physical findings on the study day; pain severity over the last 48 h as assessed on a 100-mm visual analog scale (VAS); functional impairment
269
over the last week on a VAS; and treatment. Patients completed self-questionnaires on their perception of the flareup, its characteristics and triggers, and pain severity over the last 48 h. The study complied with French law about research involving patients. 1.3. Statistical analysis All available data were included in the analysis. The results are shown using the number of patients with available data as the denominator. Univariate analyses were performed, and a logistic regression model was then built to identify factors independently associated with flare-ups [7]. To build the score, each factor was assigned a value weighted for the odds ratio for having a flare-up. No points were assigned if the factor was absent. Receiver-operating characteristic (ROC) curves were used to determine the best cutoff for diagnosing flare-ups. The score was then validated using the rheumatologist database. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were computed. Youden’s index was computed by subtracting 1 from the sum of sensitivity and specificity. Statistical analyses were performed using SAS software version 8 for Windows (SAS Institute, Cary, NC). 2. Results The studies were conducted in 2005 and 2006. The general practitioner cohort (development cohort) comprised 6085 patients, of whom 3182 had a flare-up and 2903 were in stable condition. The rheumatologist cohort (validation cohort) was composed of 641 patients, of whom 285 had a flare-up and 356 were in stable condition. Detailed data are provided only on the general practitioner cohort. 2.1. Profile of the general practitioners who participated in the study The 1312 participating general practitioners had a mean age of 50.1 years. Nearly 91% of them were men. Mean time since first licensing was 20 years (Table 1). Table 1 Characteristics of the 1312 general practitioners who collected the data in the development cohort. Variables
N ¼ 1312
Age (years), mean (SD) Male gender, n (%) Time since first licensing (years), mean (SD) Place of practice, n (%) Urban Semi-rural Rural Type of practice, n (%) Single-physician office Group practice Number of patients seen per week Number of patients followed-up for knee osteoarthritis
50.1 (6.6) 1184 (90.8) 20.8 (7.6) 651 (50.1) 420 (32.3) 228 (17.6) 575 725 143.7 43.4
(44.2) (55.8) (36.7) (43.7)
270
M. Marty et al. / Joint Bone Spine 76 (2009) 268e272
2.2. Clinical and radiological features in the patients (general practitioner cohort) Data were available for 3182 patients experiencing flareups and 2903 patients in stable condition. We compared the features of the patients in these two categories (Table 2). Age and disease duration were not significantly different. The mean body mass index was slightly higher in the flare-up group, which also had higher rates of nocturnal awakenings caused by pain (P < 0.0001), limping (P < 0.0001), joint swelling (P < 0.0001), joint effusion (P < 0.0001), and increased warmth (P < 0.0001). The flare-up group had a longer mean morning stiffness duration (P < 0.0001) and higher mean scores for pain (P < 0.0001) and functional impairment (P < 0.0001). Radiological alterations were slightly more severe in the flare-up group (Table 2).
2.3. Treatments used before the study day Patients in the flare-up group used more analgesics (P < 0.0001) and nonsteroidal antiinflammatory drugs (NSAIDs) (P < 0.0001) and less symptomatic slow-acting drugs for osteoarthritis (P < 0.0001), compared to the patients in stable condition (Table 3).
Table 3 Previous treatments in the development cohort (general practitioners).
Analgesics, n of patients (%) Duration of analgesic therapy (days), mean (SD) NSAIDs, n of patients (%) Duration of NSAID therapy (days), mean (SD) SySADOA, n of patients (%) Duration of SySADOA therapy (days), mean (SD)
Flare-up, N ¼ 3182
No flare-up, N ¼ 2903
P-value
2660 (88.6) 30.0 (34.8)
2102 (79.0) 36.2 (37.0)
<0.0001 <0.0001
955 (31.8) 19.1 (23.2)
539 (20.3) 26.7 (33.6)
<0.0001 <0.0001
1672 (55.7) 85.2 (74.5)
1688 (63.41) 87.8 (72.5)
<0.0001 0.34
NSAIDs, nonsteroidal antiinflammatory drugs; and SySADOA, symptomatic slow-acting drugs for osteoarthritis.
2.4. Flare-up triggers Among patients experiencing a flare-up, the main identified triggers were unusual physical activity (62.4% of cases), injury (21.8%), a lifestyle change (15.8%), and treatment discontinuation (12.7%). 2.5. Medications prescribed at the study visit We compared the medications prescribed to patients with and without flares. The proportions of patients to whom
Table 2 Main characteristics of the 6085 patients seen by general practitioners (development cohort). Variables
Flare-up, N ¼ 3182
No flare-up, N ¼ 2903
P-valuea
Age (years), mean (SD) Male gender, n (%) Body mass index Knee OA duration (years), mean (SD) Morning stiffness (minutes), mean (SD) Pain severity over the last 48 h (100-point VAS, patient self-questionnaire), mean (SD) Functional impairment over the last week (100-point VAS, patient self-questionnaire), mean (SD) Nocturnal awakenings due to knee pain, n (%) Limp, n (%) Swelling, n (%) Warmth, n (%) Effusion, n (%)
66.4 1368 27.7 7.4 21.5 65.4
66.2 1337 27.0 7.4 16.2 39.9
(10.2) (46.6) (4.1) (5.9) (14.4) (21.6)
0.40 0.02 <0.0001 0.89 <0.0001 <0.0001
63.5 (16.2)
38.7 (22.4)
<0.0001
1829 2594 2434 2034 1453
770 1480 979 614 427
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001
Distribution of knee OA, n (%) One knee Both knees Joint narrowing, n (%) None Moderate Severe Subchondral sclerosis, n (%) None Moderate Severe Osteophytosis, n (%) Mild Moderate Severe OA, osteoarthritis; and VAS, visual analog scale. a Univariate analysis.
(10.9) (43.4) (4.5) (5.9) (16.4) (14.2)
(59.0) (83.4) (78.3) (65.8) (47.6)
(27.2) (52.1) (34.6) (21.8) (15.2)
1674 (53.18) 1474 (46.8)
1424 (49.7) 1441 (50.3)
0.01
103 (3.3) 1755 (56.6) 1241 (40.1)
140 (5.0) 1852 (65.9) 817 (29.1)
<0.0001
674 (22.9) 1760 (59.9) 503 (17.1)
762 (28.7) 1561 (58.7)) 336 (12.6)
<0.0001
900 (30.8) 1422 (48.6) 605 (20.7)
968 (36.9) 1208 (46.11) 444 (16.9)
<0.0001
M. Marty et al. / Joint Bone Spine 76 (2009) 268e272 Table 4 Treatments prescribed by the general practitioners at the study visit (development cohort).
Analgesics, n (%) NSAIDs, n (%) SySADOA, n (%) Topical agents Ice
Table 6 Factors associated with a physician diagnosis of flare-up in the development cohort (general practitioners), by logistic regression.
Flare-up, N ¼ 3182
No flare-up, N ¼ 2903
P-value
Association
2633 2383 2682 1881 963
2082 708 2563 1463 296
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001
Positively associated with a flare-up diagnosis
(86.0) (77.8) (87.6) (59.1) (30.3)
(74.0) (25.3) (91.6) (50.4) (10.2)
NSAIDs, nonsteroidal antiinflammatory drugs; and SySADOA, symptomatic slow-acting drugs for osteoarthritis.
NSAIDs and analgesics were prescribed were considerably larger in the flare-up group (Table 4). 2.6. Clinical characteristics of the study patients In the group of patients classified by their physicians as experiencing a flare-up, larger proportions of individuals reported worse pain than usual (P < 0.01), more nocturnal awakenings (P < 0.001), joint swelling (P < 0.001), warmth (P < 0.001), and trouble walking (P < 0.001) (Table 5).
271
Odds ratio [95%CI] P-value
Factors
Morning stiffness less than 10 min Morning stiffness 10e20 min Morning stiffness more than 20 min No nocturnal awakenings Nocturnal awakenings No limp Limp No knee swelling Knee swelling No warmth over the knee Warmth over the knee No knee effusion Knee effusion Negatively associated One knee involved with a flare-up Both knees involved
1
e
1.14 (0.97e1.34)
NS
1.31 (1.10e1.55)
<0.01
1
e
2.05 (1.78e2.3) 1 2.97 (2.54e3.47) 1 2.62 (2.26e3.05) 1
<0.0001 e <0.0001 e <0.0001 e
3.30 (2.86e3.81)
<0.0001
1 1.92 (1.63e2.25) 1 0.69 (0.06e0.79)
e <0.0001 e <0.0001
95%CI, 95% confidence interval.
2.7. Agreement between physicians and patients regarding presence of a flare-up Agreement was strong between physicians and patients regarding presence of a flare-up (kappa ¼ 0.78). Thus, among patients in whom the physicians diagnosed a flare-up, 96.5% reported that they were experiencing a flare-up.
2.9. Diagnostic score for knee osteoarthritis flare-up
2.8. Clinical features associated with flare-ups By multivariate analysis, the only factors significantly associated with having a flare-up were clinical features (Table 6). Features significantly associated with having a flare-up diagnosis by the general practitioner were morning stiffness for Table 5 Symptoms reported by the patients on a questionnaire in the development cohort (general practitioners).
Worse pain than usual, n (%) Continuous pain Pain causing nocturnal arousals Pain at night even when lying still Pain unresponsive to medications Morning stiffness and pain for at least 30 min Swelling Redness Warmth Trouble walking Some activities impossible Patient’s opinion about whether there is a flare-up Yes No
Flare-up, N ¼ 3182
No flare-up, N ¼ 2903
P-value
2300 892 1348 1085 887 847
(72.5) (28.1) (42.5) (34.1) (28.0) (26.7)
1994 779 1107 940 810 721
(68.8) (26.9) (38.2) (32.5) (28.0) (24.9)
0.01 0.29 <0.001 0.15 0.15 0.11
1668 730 889 2088 1224
(52.6) (23.0) (28.0) (65.8) (38.6)
1275 600 670 1811 1039
(44.0) (20.7) (23.1) (62.5) (35.9)
<0.001 0.03 <0.001 <0.01 0.03
3042 (96.5) 110 (3.5)
554 (19.3) 2320 (80.7)
longer than 20 min, nocturnal awakenings caused by knee pain, limping, knee swelling, increased warmth, and knee effusion. Furthermore, unilateral knee osteoarthritis was associated with a flare-up diagnosis, compared to bilateral involvement.
Kappa 0.78
The score was built by assigning 0 points to absence of each clinical feature and a number of points weighted for the odds ratio to each feature that was present. Weights were as follows: 1 for morning stiffness longer than 20 min; 2 each for nocturnal awakenings caused by knee pain and knee effusion; and 3 each for limping, knee swelling, and increased warmth. Thus, the score could range from 0 to 14 (Table 7). The ROC curve analysis showed that 7 was the best cutoff for separating patients with and without flare-ups. The score and the cutoff were validated using the data from the rheumatologists. The rheumatologist diagnosis was the
Table 7 Diagnostic score for knee osteoarthritis flare-up. Variables
Absent
Morning stiffness longer than 20 min Nocturnal awakenings Effusion Limp Swelling Warmth
0
1
0 0 0 0 0
2 2 3 3 3
Totala
0
14
a
A score 7 indicates a flare-up.
Present
M. Marty et al. / Joint Bone Spine 76 (2009) 268e272
272
Table 8 Flare-up status according to the score and to the rheumatologist diagnosis (validation cohort).a
Score 7 Score < 7 Total flare-ups (rheumatologist) % Total a
Flare-up, N ¼ 270
No flare-up, N ¼ 321
Total flare-ups (score)
% of total
235 35 270
39 282 321
274 317 591
46.4 53.6 100.00
45.7
54.3
100.00
These values were obtained based on the patients with no missing data.
reference standard. Of the 641 patients, 274 (46.4%) had score values greater than 7 points and 270 (45.7%) had a rheumatologist diagnosis of flare-up. Sensitivity of the score was 87.0%, specificity was 87.9%, PPV was 85.8%, and NPV was 89.0% (Table 8). The Youden index was 0.75. 3. Discussion The primary objective of this study was to identify and validate criteria for defining knee osteoarthritis flare-up, based on data from two large cross-sectional studies. We identified the clinical features used by general practitioners to define a knee osteoarthritis flare-up. These features were found to be signs of local inflammation, in keeping with recent evidence that inflammation contributes to the pathogenesis of osteoarthritis [8,9]. According to one hypothesis, flare-ups reflect episodes of chondrolysis responsible for a joint effusion [10]. Our study patients were evaluated only on the basis of clinical features; no blood tests or joint fluid studies were performed. Among the study parameters, those associated with a flare-up were morning stiffness for longer than 20 min, nocturnal awakenings due to knee pain, limping, swelling, warmth, and joint effusion. Some of these features, most notably, warmth over the joint, may be difficult to assess, although this does not seem to affect the ability of our score to identify patients experiencing a flare-up. We built a diagnostic score that could range from 0 to 14, and we found that values of 7 or more strongly indicated a flare-up. The metrological properties of the score were determined in a different population of patients, who were evaluated by rheumatologists. Content validity (the extent to which the score measures what is intended) may be open to debate. Nevertheless, given that there is no consensus about what constitutes a flare-up, the opinions of the 1312 general practitioners who participated in the study are clinically relevant. The good score performance in the validation cohort of patients evaluated by rheumatologists supports the relevance of our diagnostic score. The 86% PPV (probability of having a flare-up when the score is 7) and 89% NPV (probability of not having a flare-up when the score is <7) indicate good score performance. Overall, the patients underestimated the presence of the clinical features associated with flare-ups. Nevertheless, they
overestimated the presence of flare-ups, compared to their general practitioners (Table 5). The main limitations of our study are the absence of external validation criteria, such as blood tests or histological studies, and the cross-sectional design. Available recommendations for managing knee osteoarthritis fail to provide specific details on the treatment of knee osteoarthritis flare-ups [11e15]. We found that a physician diagnosis of flare-up was strongly associated with prescription of NSAIDs, analgesics, and ice. KOFUS could be used to standardize the diagnosis of knee osteoarthritis flare-up, most notably when selecting patients for therapeutic trials. References [1] Dieppe PA. Relationship between symptoms and structural change in osteoarthritis. What are important targets for osteoarthritis therapy? J Rheumatol 2004;70(31 Suppl.):50e3. [2] Dieppe P, Cushnaghan J, Tucker M, et al. The Bristol ‘OA500 study’ progression and impact of the disease after 8 years. Osteoarthritis Cartilage 2000;8:63e8. [3] Spector T, Hart DJ, Nandra D, et al. Low level increases in serum C-reactive protein are present in early osteoarthritis of the knee and predict progressive disease. Arthritis Rheum 1997;40:723e7. [4] Dougados M, Gueguen A, Nguyen M, et al. Longitudinal radiologic evaluation of osteoarthritis of the knee. J Rheumatol 1992;19:378e84. [5] Ayral X, Pickering EH, Woodworth TG, et al. Synovitis: a potential predictive factor of structural progression of medial tibiofemoral knee osteoarthritis e results of a 1 year longitudinal arthroscopic study in 422 patients. Osteoarthritis Cartilage 2005;13:361e7. [6] Altman R, Asch E, Bloch D, et al. Development of criteria for the classification and reporting of osteoarthritis: classification of osteoarthritis of the knee. Arthritis Rheum 1986;29:1039e49. [7] Lachin JM. Logistic regression model. In: Lachin JM, editor. Biostatistical methods. The assessment of relative risks. New York: John Wiley & Sons; 2000. p. 247e316. [8] Loyau G, Courtheoux F, Guaydier Souquie`res G, et al. La pousse´e congestive de l’arthrose est-elle en pousse´e inflammatoire? Rev Rhum 1988;55:440e3. [9] Bonnet CS, Walsh DA. Osteoarthritis, angiogenesis and inflammation. Rheumatology 2005;44:7e16. [10] Amor B. Pousse´e congestive d’arthrose. Chondrolyse et re´paration du cartilage. Rev Prat 1993;43:601e3. [11] Pendleton A, Arden N, Dougados M, et al. EULAR recommendations for the management of knee osteoarthritis: report of a task force of the Standing Committee for International Clinical Studies Including Therapeutic Trials (ESCISIT). Ann Rheum Dis 2000;59:936e44. [12] Altman RD, Lozada CJ. Practice guidelines in the management of osteoarthritis. Osteoarthritis Cartilage 1998;6(Suppl. A):22e4. [13] American College of Rheumatology Subcommittee on Osteoarthritis Guidelines. Recommendations for medical management of osteoarthritis of the hip and knee. Arthritis Rheum 2000;43:1905e15. [14] Jordan KM, Arden NK, Doherty M, et al. EULAR Recommendations 2003: an evidence-based approach to the management of knee osteoarthritis: report of a Task Force of the Standing Committee for International Clinical Studies Including Therapeutic Trials (ESCISIT). Ann Rheum Dis 2003;62:1145e55. [15] Zhang W, Moskowitz RW, Nuki G, et al. OARSI recommendations for the management of hip and knee osteoarthritis, Part II: OARSI evidencebased, expert consensus guidelines. Osteoarthritis Cartilage 2008;16: 137e62.
www.sciencedirect.com Joint Bone Spine 76 (2009) 268e272
Original article
Validation of the KOFUS (Knee Osteoarthritis Flare-Ups Score) Marc Marty a,*, Pascal Hilliquin b, Sylvie Rozenberg c, Jean Pierre Valat d, Eric Vignon e, Philippe Coste f, Bernard Savarieau g, Franc¸ois Andre´ Allaert h b
a Service de Rhumatologie, Centre hospitalier Henri Mondor, Cre´teil 94010, France Service de Rhumatologie, Centre hospitalier Sud Francilien, Corbeil-Essonnes, France c Service de Rhumatologie, Groupe hospitalier Pitie´-Salpe´trie`re, Paris, France d Service de Rhumatologie, Universite´ Franc¸ois-Rabelais, Tours, France e Service de Rhumatologie, CHU de Lyon sud, Pierre Benite, France f Laboratoires Expanscience, De´partement me´dical, Puteaux, France g Gestion de l’e´tude, Nukle´us, Paris, France h Service de Biome´trie, Cenbiotech, Dijon, France
Accepted 17 July 2008 Available online 17 March 2009
Abstract Objective: To develop a diagnostic score for knee osteoarthritis flare-ups and to evaluate its sensitivity and specificity. Methods: We used two epidemiological databases built using the same methodology. One database was from a general-practice study and served to develop the score, whereas the other was from a rheumatology study and served to validate the score. Physicians determined the flare-up status of each patient. The rheumatologist diagnosis was the reference standard. Logistic regression was performed to identify factors significantly associated with having a flare-up. Results: Of the 6085 patients in the general-practice database, 52.3% had a knee osteoarthritis flare-up. The score was built by assigning points to features that were present, with a weighting system based on the odds ratio of each feature for having a flare-up (0, feature absent; 1, morning stiffness for longer than 20 min; 2, pain causing nocturnal awakenings and knee effusion; 3, limping, joint swelling, and increased warmth over the knee). The score could range from 0 to 14. The receiver-operating characteristic curve showed that 7 was the best cutoff for diagnosing a flare-up. In the rheumatologist database, the numbers of patients having a flare-up were 274 (46.4%) based on the score and 270 (45.7%) based on the rheumatologist diagnosis. Sensitivity of the score was 87.0%, specificity 87.9%, positive predictive value 85.8%, and negative predictive value 89.0%. The Youden index was 0.75. Conclusion: A score equal to or greater than 7 points correlated well with a rheumatologist diagnosis of flare-up. Our score may constitute a valid objective criterion for standardizing the diagnosis of knee osteoarthritis flare-up, most notably when screening patients for inclusion in therapeutic trials. Ó 2009 Socie´te´ Franc¸aise de Rhumatologie. Published by Elsevier Masson SAS. All rights reserved. Keywords: Knee osteoarthritis; Flare-up; Diagnostic score; Sensitivity; Specificity
Pain and functional impairment are the main clinical manifestations of osteoarthritis. Symptom intensity varies over time, and patients experience intermittent flare-ups during which the pain and functional impairment worsen [1]. The existence of flare-ups probably contributes to explain the
* Corresponding author. E-mail address: [email protected] (M. Marty).
inconsistent correlation of clinical symptoms and radiological changes, most notably at the knee [2]. The pathological concomitant of osteoarthritis flare-ups is the development of inflammatory lesions within the synovial membrane [3]. The pain becomes more severe and responds less well to rest, tending to become continuous. A joint effusion and warmth may be noted. Cytological examination of fluid aspirated from the knee shows mechanical properties. The synovial membrane releases substances that damage the cartilage, such as
1297-319X/$ - see front matter Ó 2009 Socie´te´ Franc¸aise de Rhumatologie. Published by Elsevier Masson SAS. All rights reserved. doi:10.1016/j.jbspin.2008.07.018
M. Marty et al. / Joint Bone Spine 76 (2009) 268e272
proinflammatory cytokines and metalloproteases, supporting the hypothesis that flare-ups increase the risk of cartilage destruction [4]. In patients with knee osteoarthritis, arthroscopic evidence of synovitis may predict faster cartilage loss over the next year [5]. Another hypothesis, however, is that flare-ups are merely markers for cartilage destruction. Variations in the frequency of flare-ups may explain the variability of the rate of cartilage destruction across patients. Accurate detection of flare-ups should allow early treatment. Nevertheless, the definition of osteoarthritis flare-up is not agreed on. An accurate definition, together with improved knowledge of the natural history of osteoarthritis and/or of the factors associated with flare-ups would help to develop recommendations aimed at limiting the number and adverse impact of flare-ups. The knee is the most appropriate site for studying osteoarthritis flare-ups, as its superficial location facilitates the detection and aspiration of joint effusions. The primary objective of this study was to define knee osteoarthritis flareups based on the clinical and radiological features. The secondary objectives were to describe patients who present with knee osteoarthritis flare-ups and to describe management modalities. 1. Methods Two observational cross-sectional studies were conducted in France, one among general practitioners and the other among rheumatologists, to identify the criteria that define knee osteoarthritis flare-ups. The data obtained from the general practitioners were used to develop a diagnostic score for patients seeking primary care. The data from the rheumatologists served only to validate the score. All the general practitioners and rheumatologists who participated in the study received a detailed research protocol that contained explanations about painful knee osteoarthritis flare-ups. 1.1. Patient selection for the study Patients were eligible for study inclusion if they met the clinical and radiological criteria for knee osteoarthritis established by the American College of Rheumatology [6]. They either had stable disease or were experiencing a flare-up according to the physician’s assessment. Each general practitioner was asked to include 3 patients in each category and each rheumatologist 2 patients in each category. Patients were not eligible if they had secondary knee osteoarthritis (inflammatory joint disease, crystal deposition disease, septic arthritis, or trauma) or a history of surgery on the study knee other than arthroscopy for meniscal disease. 1.2. Evaluation criteria For each patient, we recorded the following data: age, sex, and occupation; duration of the knee osteoarthritis, radiological findings, and current treatments; physical findings on the study day; pain severity over the last 48 h as assessed on a 100-mm visual analog scale (VAS); functional impairment
269
over the last week on a VAS; and treatment. Patients completed self-questionnaires on their perception of the flareup, its characteristics and triggers, and pain severity over the last 48 h. The study complied with French law about research involving patients. 1.3. Statistical analysis All available data were included in the analysis. The results are shown using the number of patients with available data as the denominator. Univariate analyses were performed, and a logistic regression model was then built to identify factors independently associated with flare-ups [7]. To build the score, each factor was assigned a value weighted for the odds ratio for having a flare-up. No points were assigned if the factor was absent. Receiver-operating characteristic (ROC) curves were used to determine the best cutoff for diagnosing flare-ups. The score was then validated using the rheumatologist database. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were computed. Youden’s index was computed by subtracting 1 from the sum of sensitivity and specificity. Statistical analyses were performed using SAS software version 8 for Windows (SAS Institute, Cary, NC). 2. Results The studies were conducted in 2005 and 2006. The general practitioner cohort (development cohort) comprised 6085 patients, of whom 3182 had a flare-up and 2903 were in stable condition. The rheumatologist cohort (validation cohort) was composed of 641 patients, of whom 285 had a flare-up and 356 were in stable condition. Detailed data are provided only on the general practitioner cohort. 2.1. Profile of the general practitioners who participated in the study The 1312 participating general practitioners had a mean age of 50.1 years. Nearly 91% of them were men. Mean time since first licensing was 20 years (Table 1). Table 1 Characteristics of the 1312 general practitioners who collected the data in the development cohort. Variables
N ¼ 1312
Age (years), mean (SD) Male gender, n (%) Time since first licensing (years), mean (SD) Place of practice, n (%) Urban Semi-rural Rural Type of practice, n (%) Single-physician office Group practice Number of patients seen per week Number of patients followed-up for knee osteoarthritis
50.1 (6.6) 1184 (90.8) 20.8 (7.6) 651 (50.1) 420 (32.3) 228 (17.6) 575 725 143.7 43.4
(44.2) (55.8) (36.7) (43.7)
270
M. Marty et al. / Joint Bone Spine 76 (2009) 268e272
2.2. Clinical and radiological features in the patients (general practitioner cohort) Data were available for 3182 patients experiencing flareups and 2903 patients in stable condition. We compared the features of the patients in these two categories (Table 2). Age and disease duration were not significantly different. The mean body mass index was slightly higher in the flare-up group, which also had higher rates of nocturnal awakenings caused by pain (P < 0.0001), limping (P < 0.0001), joint swelling (P < 0.0001), joint effusion (P < 0.0001), and increased warmth (P < 0.0001). The flare-up group had a longer mean morning stiffness duration (P < 0.0001) and higher mean scores for pain (P < 0.0001) and functional impairment (P < 0.0001). Radiological alterations were slightly more severe in the flare-up group (Table 2).
2.3. Treatments used before the study day Patients in the flare-up group used more analgesics (P < 0.0001) and nonsteroidal antiinflammatory drugs (NSAIDs) (P < 0.0001) and less symptomatic slow-acting drugs for osteoarthritis (P < 0.0001), compared to the patients in stable condition (Table 3).
Table 3 Previous treatments in the development cohort (general practitioners).
Analgesics, n of patients (%) Duration of analgesic therapy (days), mean (SD) NSAIDs, n of patients (%) Duration of NSAID therapy (days), mean (SD) SySADOA, n of patients (%) Duration of SySADOA therapy (days), mean (SD)
Flare-up, N ¼ 3182
No flare-up, N ¼ 2903
P-value
2660 (88.6) 30.0 (34.8)
2102 (79.0) 36.2 (37.0)
<0.0001 <0.0001
955 (31.8) 19.1 (23.2)
539 (20.3) 26.7 (33.6)
<0.0001 <0.0001
1672 (55.7) 85.2 (74.5)
1688 (63.41) 87.8 (72.5)
<0.0001 0.34
NSAIDs, nonsteroidal antiinflammatory drugs; and SySADOA, symptomatic slow-acting drugs for osteoarthritis.
2.4. Flare-up triggers Among patients experiencing a flare-up, the main identified triggers were unusual physical activity (62.4% of cases), injury (21.8%), a lifestyle change (15.8%), and treatment discontinuation (12.7%). 2.5. Medications prescribed at the study visit We compared the medications prescribed to patients with and without flares. The proportions of patients to whom
Table 2 Main characteristics of the 6085 patients seen by general practitioners (development cohort). Variables
Flare-up, N ¼ 3182
No flare-up, N ¼ 2903
P-valuea
Age (years), mean (SD) Male gender, n (%) Body mass index Knee OA duration (years), mean (SD) Morning stiffness (minutes), mean (SD) Pain severity over the last 48 h (100-point VAS, patient self-questionnaire), mean (SD) Functional impairment over the last week (100-point VAS, patient self-questionnaire), mean (SD) Nocturnal awakenings due to knee pain, n (%) Limp, n (%) Swelling, n (%) Warmth, n (%) Effusion, n (%)
66.4 1368 27.7 7.4 21.5 65.4
66.2 1337 27.0 7.4 16.2 39.9
(10.2) (46.6) (4.1) (5.9) (14.4) (21.6)
0.40 0.02 <0.0001 0.89 <0.0001 <0.0001
63.5 (16.2)
38.7 (22.4)
<0.0001
1829 2594 2434 2034 1453
770 1480 979 614 427
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001
Distribution of knee OA, n (%) One knee Both knees Joint narrowing, n (%) None Moderate Severe Subchondral sclerosis, n (%) None Moderate Severe Osteophytosis, n (%) Mild Moderate Severe OA, osteoarthritis; and VAS, visual analog scale. a Univariate analysis.
(10.9) (43.4) (4.5) (5.9) (16.4) (14.2)
(59.0) (83.4) (78.3) (65.8) (47.6)
(27.2) (52.1) (34.6) (21.8) (15.2)
1674 (53.18) 1474 (46.8)
1424 (49.7) 1441 (50.3)
0.01
103 (3.3) 1755 (56.6) 1241 (40.1)
140 (5.0) 1852 (65.9) 817 (29.1)
<0.0001
674 (22.9) 1760 (59.9) 503 (17.1)
762 (28.7) 1561 (58.7)) 336 (12.6)
<0.0001
900 (30.8) 1422 (48.6) 605 (20.7)
968 (36.9) 1208 (46.11) 444 (16.9)
<0.0001
M. Marty et al. / Joint Bone Spine 76 (2009) 268e272 Table 4 Treatments prescribed by the general practitioners at the study visit (development cohort).
Analgesics, n (%) NSAIDs, n (%) SySADOA, n (%) Topical agents Ice
Table 6 Factors associated with a physician diagnosis of flare-up in the development cohort (general practitioners), by logistic regression.
Flare-up, N ¼ 3182
No flare-up, N ¼ 2903
P-value
Association
2633 2383 2682 1881 963
2082 708 2563 1463 296
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001
Positively associated with a flare-up diagnosis
(86.0) (77.8) (87.6) (59.1) (30.3)
(74.0) (25.3) (91.6) (50.4) (10.2)
NSAIDs, nonsteroidal antiinflammatory drugs; and SySADOA, symptomatic slow-acting drugs for osteoarthritis.
NSAIDs and analgesics were prescribed were considerably larger in the flare-up group (Table 4). 2.6. Clinical characteristics of the study patients In the group of patients classified by their physicians as experiencing a flare-up, larger proportions of individuals reported worse pain than usual (P < 0.01), more nocturnal awakenings (P < 0.001), joint swelling (P < 0.001), warmth (P < 0.001), and trouble walking (P < 0.001) (Table 5).
271
Odds ratio [95%CI] P-value
Factors
Morning stiffness less than 10 min Morning stiffness 10e20 min Morning stiffness more than 20 min No nocturnal awakenings Nocturnal awakenings No limp Limp No knee swelling Knee swelling No warmth over the knee Warmth over the knee No knee effusion Knee effusion Negatively associated One knee involved with a flare-up Both knees involved
1
e
1.14 (0.97e1.34)
NS
1.31 (1.10e1.55)
<0.01
1
e
2.05 (1.78e2.3) 1 2.97 (2.54e3.47) 1 2.62 (2.26e3.05) 1
<0.0001 e <0.0001 e <0.0001 e
3.30 (2.86e3.81)
<0.0001
1 1.92 (1.63e2.25) 1 0.69 (0.06e0.79)
e <0.0001 e <0.0001
95%CI, 95% confidence interval.
2.7. Agreement between physicians and patients regarding presence of a flare-up Agreement was strong between physicians and patients regarding presence of a flare-up (kappa ¼ 0.78). Thus, among patients in whom the physicians diagnosed a flare-up, 96.5% reported that they were experiencing a flare-up.
2.9. Diagnostic score for knee osteoarthritis flare-up
2.8. Clinical features associated with flare-ups By multivariate analysis, the only factors significantly associated with having a flare-up were clinical features (Table 6). Features significantly associated with having a flare-up diagnosis by the general practitioner were morning stiffness for Table 5 Symptoms reported by the patients on a questionnaire in the development cohort (general practitioners).
Worse pain than usual, n (%) Continuous pain Pain causing nocturnal arousals Pain at night even when lying still Pain unresponsive to medications Morning stiffness and pain for at least 30 min Swelling Redness Warmth Trouble walking Some activities impossible Patient’s opinion about whether there is a flare-up Yes No
Flare-up, N ¼ 3182
No flare-up, N ¼ 2903
P-value
2300 892 1348 1085 887 847
(72.5) (28.1) (42.5) (34.1) (28.0) (26.7)
1994 779 1107 940 810 721
(68.8) (26.9) (38.2) (32.5) (28.0) (24.9)
0.01 0.29 <0.001 0.15 0.15 0.11
1668 730 889 2088 1224
(52.6) (23.0) (28.0) (65.8) (38.6)
1275 600 670 1811 1039
(44.0) (20.7) (23.1) (62.5) (35.9)
<0.001 0.03 <0.001 <0.01 0.03
3042 (96.5) 110 (3.5)
554 (19.3) 2320 (80.7)
longer than 20 min, nocturnal awakenings caused by knee pain, limping, knee swelling, increased warmth, and knee effusion. Furthermore, unilateral knee osteoarthritis was associated with a flare-up diagnosis, compared to bilateral involvement.
Kappa 0.78
The score was built by assigning 0 points to absence of each clinical feature and a number of points weighted for the odds ratio to each feature that was present. Weights were as follows: 1 for morning stiffness longer than 20 min; 2 each for nocturnal awakenings caused by knee pain and knee effusion; and 3 each for limping, knee swelling, and increased warmth. Thus, the score could range from 0 to 14 (Table 7). The ROC curve analysis showed that 7 was the best cutoff for separating patients with and without flare-ups. The score and the cutoff were validated using the data from the rheumatologists. The rheumatologist diagnosis was the
Table 7 Diagnostic score for knee osteoarthritis flare-up. Variables
Absent
Morning stiffness longer than 20 min Nocturnal awakenings Effusion Limp Swelling Warmth
0
1
0 0 0 0 0
2 2 3 3 3
Totala
0
14
a
A score 7 indicates a flare-up.
Present
M. Marty et al. / Joint Bone Spine 76 (2009) 268e272
272
Table 8 Flare-up status according to the score and to the rheumatologist diagnosis (validation cohort).a
Score 7 Score < 7 Total flare-ups (rheumatologist) % Total a
Flare-up, N ¼ 270
No flare-up, N ¼ 321
Total flare-ups (score)
% of total
235 35 270
39 282 321
274 317 591
46.4 53.6 100.00
45.7
54.3
100.00
These values were obtained based on the patients with no missing data.
reference standard. Of the 641 patients, 274 (46.4%) had score values greater than 7 points and 270 (45.7%) had a rheumatologist diagnosis of flare-up. Sensitivity of the score was 87.0%, specificity was 87.9%, PPV was 85.8%, and NPV was 89.0% (Table 8). The Youden index was 0.75. 3. Discussion The primary objective of this study was to identify and validate criteria for defining knee osteoarthritis flare-up, based on data from two large cross-sectional studies. We identified the clinical features used by general practitioners to define a knee osteoarthritis flare-up. These features were found to be signs of local inflammation, in keeping with recent evidence that inflammation contributes to the pathogenesis of osteoarthritis [8,9]. According to one hypothesis, flare-ups reflect episodes of chondrolysis responsible for a joint effusion [10]. Our study patients were evaluated only on the basis of clinical features; no blood tests or joint fluid studies were performed. Among the study parameters, those associated with a flare-up were morning stiffness for longer than 20 min, nocturnal awakenings due to knee pain, limping, swelling, warmth, and joint effusion. Some of these features, most notably, warmth over the joint, may be difficult to assess, although this does not seem to affect the ability of our score to identify patients experiencing a flare-up. We built a diagnostic score that could range from 0 to 14, and we found that values of 7 or more strongly indicated a flare-up. The metrological properties of the score were determined in a different population of patients, who were evaluated by rheumatologists. Content validity (the extent to which the score measures what is intended) may be open to debate. Nevertheless, given that there is no consensus about what constitutes a flare-up, the opinions of the 1312 general practitioners who participated in the study are clinically relevant. The good score performance in the validation cohort of patients evaluated by rheumatologists supports the relevance of our diagnostic score. The 86% PPV (probability of having a flare-up when the score is 7) and 89% NPV (probability of not having a flare-up when the score is <7) indicate good score performance. Overall, the patients underestimated the presence of the clinical features associated with flare-ups. Nevertheless, they
overestimated the presence of flare-ups, compared to their general practitioners (Table 5). The main limitations of our study are the absence of external validation criteria, such as blood tests or histological studies, and the cross-sectional design. Available recommendations for managing knee osteoarthritis fail to provide specific details on the treatment of knee osteoarthritis flare-ups [11e15]. We found that a physician diagnosis of flare-up was strongly associated with prescription of NSAIDs, analgesics, and ice. KOFUS could be used to standardize the diagnosis of knee osteoarthritis flare-up, most notably when selecting patients for therapeutic trials. References [1] Dieppe PA. Relationship between symptoms and structural change in osteoarthritis. What are important targets for osteoarthritis therapy? J Rheumatol 2004;70(31 Suppl.):50e3. [2] Dieppe P, Cushnaghan J, Tucker M, et al. The Bristol ‘OA500 study’ progression and impact of the disease after 8 years. Osteoarthritis Cartilage 2000;8:63e8. [3] Spector T, Hart DJ, Nandra D, et al. Low level increases in serum C-reactive protein are present in early osteoarthritis of the knee and predict progressive disease. Arthritis Rheum 1997;40:723e7. [4] Dougados M, Gueguen A, Nguyen M, et al. Longitudinal radiologic evaluation of osteoarthritis of the knee. J Rheumatol 1992;19:378e84. [5] Ayral X, Pickering EH, Woodworth TG, et al. Synovitis: a potential predictive factor of structural progression of medial tibiofemoral knee osteoarthritis e results of a 1 year longitudinal arthroscopic study in 422 patients. Osteoarthritis Cartilage 2005;13:361e7. [6] Altman R, Asch E, Bloch D, et al. Development of criteria for the classification and reporting of osteoarthritis: classification of osteoarthritis of the knee. Arthritis Rheum 1986;29:1039e49. [7] Lachin JM. Logistic regression model. In: Lachin JM, editor. Biostatistical methods. The assessment of relative risks. New York: John Wiley & Sons; 2000. p. 247e316. [8] Loyau G, Courtheoux F, Guaydier Souquie`res G, et al. La pousse´e congestive de l’arthrose est-elle en pousse´e inflammatoire? Rev Rhum 1988;55:440e3. [9] Bonnet CS, Walsh DA. Osteoarthritis, angiogenesis and inflammation. Rheumatology 2005;44:7e16. [10] Amor B. Pousse´e congestive d’arthrose. Chondrolyse et re´paration du cartilage. Rev Prat 1993;43:601e3. [11] Pendleton A, Arden N, Dougados M, et al. EULAR recommendations for the management of knee osteoarthritis: report of a task force of the Standing Committee for International Clinical Studies Including Therapeutic Trials (ESCISIT). Ann Rheum Dis 2000;59:936e44. [12] Altman RD, Lozada CJ. Practice guidelines in the management of osteoarthritis. Osteoarthritis Cartilage 1998;6(Suppl. A):22e4. [13] American College of Rheumatology Subcommittee on Osteoarthritis Guidelines. Recommendations for medical management of osteoarthritis of the hip and knee. Arthritis Rheum 2000;43:1905e15. [14] Jordan KM, Arden NK, Doherty M, et al. EULAR Recommendations 2003: an evidence-based approach to the management of knee osteoarthritis: report of a Task Force of the Standing Committee for International Clinical Studies Including Therapeutic Trials (ESCISIT). Ann Rheum Dis 2003;62:1145e55. [15] Zhang W, Moskowitz RW, Nuki G, et al. OARSI recommendations for the management of hip and knee osteoarthritis, Part II: OARSI evidencebased, expert consensus guidelines. Osteoarthritis Cartilage 2008;16: 137e62.