Discordance between physician s and parent s global assessments in juvenile idiopathic arthritis

Transcription

1 Rheumatology 2007;46: Advance Access publication 16 June 2006 Discordance between physician s and parent s global assessments in juvenile idiopathic arthritis doi: /rheumatology/kel201 F. Sztajnbok 1, D. L. Coronel-Martinez 1, A. Diaz-Maldonado 1, C. Novarini 1, A. Pistorio 2, S. Viola 1, N. Ruperto 1, A. Buoncompagni 1, A. Martini 1,3 and A. Ravelli 1 Objective. To investigate the discrepancy between physician s and parent s global assessments of disease status and the factors explaining discordance in patients with juvenile idiopathic arthritis (JIA). Methods. The mothers of 197 patients with JIA rated the child s overall well-being on a 10 cm visual analogue scale (VAS) and the attending physician rated the child s overall disease activity on a 10 cm VAS. A discordance score was calculated by subtracting the physician s global assessment from that of the parent s, leading to the definition of three patient groups: (1) no discordance, when physician s and parent s assessments were within 1 cm of each other; (2) negative discordance, when parent s assessment was underrated relative to the physician; and (3) positive discordance, when parent s assessment was over-rated relative to the physician. Negative and positive discordance was defined as marked when the difference between the two assessments was greater than 3 cm. Results. No discordance was found in 40.6% of the patients. Negative discordance was found in 51.3% of the patients, with 34% showing marked discordance. Positive discordance was found in 8.1% of the patients, with 2% showing marked discordance. Significant differences between groups included a shorter disease duration among patients with a markedly positive discordance (P ¼ 0.02) and a greater frequency of ongoing second-line drug therapy among patients with no discordance or with positive discordance (P ¼ 0.008). Patients with no discordance or with marked positive discordance had a significantly lower joint counts (P ¼ ). Conclusion. Parents and physicians often perceive the health status of children with JIA differently, with parents providing most frequently lower rating. KEY WORDS: Juvenile idiopathic arthritis, Physician s global assessment, Parent s global assessment, Discordance, Parent physician agreement. Introduction The physician s and parent s global assessments of the disease status are among the most important quantitative measures used to assess patients with juvenile idiopathic arthritis (JIA). The physician s global assessment usually consists in a rating of the overall level of the child s disease activity on a 10 cm visual analogue scale (VAS), with anchors of 0 ¼ no activity and 10 ¼ maximum activity ; the parent is generally asked to make a global assessment of the child s overall well-being on a 10 cm VAS, with anchors of 0 ¼ very good and 10 ¼ very poor, which is located at the bottom of the Childhood Health Assessment Questionnaire (C-HAQ) questionnaire [1]. Both measures have been shown to possess good measurement properties, including responsiveness to clinically important change [2 5] and have been selected for inclusion in the ACR Pediatric 30 core set of outcome variables [6]. Recently, a good agreement among independent physicians in assessing the overall level of disease activity was reported [7] and the mother and father were shown to be equally reliable proxy reporters of the child s pain [8]. Although both measures yield a subjective judgement on how the arthritis is doing, they address partially different constructs. In rating the child s current disease activity, the physician integrates the information obtained from clinical history, particularly regarding the intensity of pain and the duration of morning stiffness, with the findings of physical examination, specifically focusing on joint swelling and pain on pressure or passive motion, and the results of laboratory tests. Although joint complaints also have a major influence on the parent s global assessment, the parent perception of the child s health status is largely determined by the ability of the child to cope with arthritis symptoms, namely pain, and by the broad impact of the illness on the child s physical and psychosocial functioning [9]. In a recent study of the relationship of JIA severity measures in patients with different disease duration, the physician s global assessment was found to be a reliable indicator of disease activity in all stages of the illness, whereas in patients with longstanding disease the parent s global assessment proved to be a mixed variable that combine the effects of both the disease process and functional damage [10]. Studies of the relationship between the physician s and parent s global assessments in patients with JIA have shown that they are moderately correlated but not redundant [10, 11]. However, the degree of concordance between the two measures has never been investigated. In the present study, we examined the discrepancy between the physician s and parent s global assessments of disease 1 Pediatria II, Istituto di Ricovero e Cura a Carattere Scientifico G. Gaslini, 2 Servizio di Epidemiologia e Biostatistica, Istituto di Ricovero e Cura a Carattere Scientifico G. Gaslini and 3 Dipartimento di Pediatria, Universita` di Genova, Genova, Italy. Submitted 3 February 2006; revised version accepted 9 May Correspondence to: Angelo Ravelli, MD, Pediatria II, Istituto G. Gaslini, Largo G. Gaslini 5, Genova, Italy. angeloravelli@ospedale-gaslini.ge.it 141 ß The Author Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 142 F. Sztajnbok et al. status and the factors explaining discordance in a cohort of patients with JIA. Patients and methods Study population A cohort of 197 unselected patients who met the International League of Associations for Rheumatology (ILAR) criteria for JIA [12] and were seen at the study unit between February, 2002 and October 2004 was studied. To be included in the study, patients had to have both a physician s and a parent s global assessment of the disease status made at the time of the study visit. To ensure homogeneity of study analyses, only patients who had the parent s global assessment completed by the mother were included. Patients were in regular follow-up and all mothers gave informed consent to participate. Clinical assessment The medical charts of each patient were reviewed for the following information: sex, age at disease onset, JIA category, antinuclear antibody (ANA) and rheumatoid factor (RF) status, medications received at the time of the study visit, age and disease duration from disease onset to study visit and the mother s age and employment status (employed or housewife). Prior to the study visit, the mother of each patient made a global assessment of the child s overall well-being on a 10 cm VAS (0 ¼ very good; 10 ¼ very poor). The mother was asked the following question: Considering all the ways that arthritis affects your child, rate how your child is doing on the following scale by placing a mark on the line. The mother also rated the intensity of the child s pain on a 10 cm VAS (0 ¼ no pain; 10 ¼ very severe pain) and completed the C-HAQ, Italian version [13] (0 ¼ best; 3 ¼ worst). At the time of the study visit, the attending paediatric rheumatologist (A.R. or S.V.) performed a complete clinical assessment, including a detailed joint examination, and recorded for each joint the following articular indices: number of swollen joints, number of joints with pain upon movement/tenderness, number of joints with limited range of motion and number of joints with active disease (defined as the number of joints with swelling or, if no swelling was present, with limitation of movement with either pain upon movement or tenderness). The articular indices were assessed in a total of 67 joints (those that are included in the standard articular examination). The same physician who performed the joint examination made a global rating of the child s overall disease activity on a 10 cm VAS (0 ¼ no activity; 10 ¼ maximum activity), being blinded to all parent s assessments. Both physicians who completed the study evaluations have >15 yrs of experience in assessing patients with JIA, and followed the same methodology throughout the study. The laboratory parameters of JIA activity included the erythrocyte sedimentation rate (ESR), determined with the Westergren method, and the C-reactive protein (CRP), determined with nephelometry. Calculation of the discordance score A discordance score was calculated by subtracting the physician s global assessment from the parent s global assessment. Then, patients were separated in 3 groups: (1) no discordance ¼ physician s global assessment and parent s global assessment within 1 cm of each other; (2) negative discordance ¼ parent s global assessment underrated by more than 1 cm relative to physician s global assessment and (3) positive discordance ¼ parent s global assessment over-rated by more than 1 cm relative to physician s global assessment. Negative/positive discordance was defined as marked when the difference between the two assessments was greater than 3 cm. In summary, a negative discordance indicates that the physician perceived the child s disease status as worse than that assessed by the parent, and a positive discordance indicates that the parent perceived the child s disease status as worse than that assessed by the physician. Statistics Descriptive statistics were reported in terms of medians and ranges for the continuous variables and in terms of absolute frequencies and percentages for the categorical variables. The comparison of quantitative and qualitative variables between discordance categories was performed by means of Kruskal Wallis analysis of variance and chi-square or Fisher s exact test (if expected frequency was <5), respectively. The physician parent agreement for each of the two physicians was measured with the intraclass correlation coefficient (ICC). For the interpretation of ICC values, the following classification was used: <0.4 ¼ poor agreement; 0.4 and <0.75 ¼ moderate agreement; 0.75 ¼ good agreement. The correlation between physician s and parent s global assessments and JIA severity measures was compared using Spearman s rank correlation. For the purpose of this analysis, correlations >0.7 were considered high, correlations ranging from 0.4 to 0.7 were considered moderate and correlations <0.4 were considered low. The statistical package used for all analyses was the Statistica (StatSoft Corp., Tulsa, OK, USA). Results The socio-demographic and clinical features of the 197 study patients are presented in Table 1. Table 2 shows the results of the assessment of clinical measures of JIA severity, including the physician s and parent s global assessments. The study cohort is representative of the whole spectrum of activity and severity of the JIA patients who are usually seen in a paediatric rheumatology clinic; there is, however, an over-representation of patients with ANA positive disease that reflects the high prevalence of this form of JIA in our country. On average, the global rating provided by the physician was higher (worse) than that provided by the parents. The difference between physician s and parent s global assessments ranged from 9.4 to 4.5 (mean 2 2.8, median 1.3). No discordance in the perception of child s disease status was found in 80 (40.6%) patients. Discordance between the physician s and the parent s perception of the child s disease status, either negative or positive, was found in 117 (59.4%) patients. Negative discordance (the parent scored lower than the physician) was found in 101 (51.3%) patients, with 67 (34%) showing marked (>3 cm) discordance. Positive discordance (the parent scored higher than the physician) was found in 16 (8.1%) patients, with 4 (2%) showing marked (>3 cm) discordance. Table 3 depicts the comparison of socio-demographic and clinical features in patients divided by discordance category. There was no association between the direction and amount of discordance and gender ratio, age at disease onset and at study visit, disease duration at study visit, JIA subtype, ANA and RF status and mother s age and employment status. Significant differences between groups included a shorter disease duration among patients with markedly positive discordance (P ¼ 0.02) and a greater frequency of ongoing second-line drug therapy among patients with no discordance or with positive discordance within 3cm (P ¼ 0.008). The comparison of JIA severity measures among discordance categories is reported in Table 4. The parent s rating of the intensity of the child s pain, the C-HAQ score and the laboratory indicators of systemic inflammation were comparable across groups. Patients with no discordance or with marked positive discordance had a significantly lower extension and severity of arthritis, as expressed by joint counts (P ¼ ).

3 Discordance in juvenile idiopathic arthritis 143 TABLE 1. Socio-demographic and clinical features of the 197 study patients N % Mean (S.D.) Median Minimum Maximum Female Male Age at disease onset (yrs) 4.6 (3.3) Age at study visit (yrs) 8.4 (4.5) Disease duration (yrs) 3.9 (3.7) JIA category Oligoarthritis Polyarthritis Systemic arthritis Psoriatic arthritis Rheumatoid factor positive Antinuclear antibody positive Medication at the study visit None NSAIDs Second-line drugs Mother s age, yrs 38.2 (6.1) Mother s employment status a Employed Housewife a Assessed in 191 patients. NSAIDs, non-steroidal anti-inflammatory drugs. TABLE 2. Results of assessment of JIA severity measures in the 197 study patients Mean (S.D.) Median Minimum Maximum Physician s global assessment a (cm) 4.8 (3.3) Parent s global assessment a (cm) 2.8 (2.6) Parent s pain assessment a (cm) 3.0 (2.9) C-HAQ score b (units) 0.6 (0.7) No. swollen joints 2.8 (3.0) No. joints with pain on motion/tenderness 3.9 (5.8) No. joints with limited range of motion 3.4 (4.9) No. active joints 3.9 (4.5) Erythrocyte sedimentation rate c (mm/h) 28.8 (24.4) C-reactive protein d (mg/dl) 1.7 (3.2) a On a 0 10 cm visual analogue scale (0 ¼ best; 10 ¼ worst); b C-HAQ: Childhood Health Assessment Questionnaire (0 ¼ best, 3 ¼ worst); c normal <15 mm/h; d normal <0.3 mg/dl. TABLE 3. Socio-demographic and clinical features of the study patients by discordance category No discordance Negative discordance Positive discordance (n ¼ 80) >1cm (n ¼34) >3cm (n ¼ 67) >1cm (n ¼ 12) >3cm (n ¼ 4) P-value Male [n (%)] 22 (27.5) 9 (26.5) 15 (22.4) 3 (25.0) 2 (50.0) 0.75 Female [n (%)] 58 (72.5) 25 (73.5) 52 (77.6) 9 (75.0) 2 (50.0) Age at disease onset [yrs (mean S.D.)] Age at study visit [yrs (mean S.D.)] Disease duration [yrs (mean S.D.)] JIA category [n (%)] Oligoarthritis 54 (67.5) 25 (73.5) 50 (74.6) 7 (58.4) 3 (75.0) 0.16 Polyarthritis 13 (16.3) 8 (23.5) 14 (20.9) 4 (33.3) 1 (25.0) Systemic arthritis 10 (12.5) 1 (3.0) 3 (4.5) 1 (8.3) 0 (0) Psoriatic arthritis 3 (3.7) 0 (0) 0 (0) 0 (0) 0 (0) Medication at the study visit [n (%)] None 16 (44.4) 3 (8.3) 17 (47.2) 0 (0) 0 (0) 0.11 NSAIDs 39 (48.7) 22 (64.7) 39 (58.2) 9 (75.0) 3 (75.0) 0.28 Second line drugs 43 (53.7) 13 (38.2) 19 (28.4) 8 (66.7) 1 (25.0) Rheumatoid factor positive [n (%)] 6 (7.5) 0 (0.0) 7 (10.4) 1 (8.3) 0 (0) 0.32 Antinuclear antibody positive [n (%)] 59 (73.7) 27 (79.4) 57 (85.1) 9 (75.0) 3 (75.0) 0.48 Mother s age [yrs (mean S.D.)] Mother s employment status [n (%)] Employed 47 (61.0) 20 (62.5) 33 (50.0) 5 (41.7) 2 (50.0) 0.49 Housewife 30 (39.0) 12 (37.5) 33 (50.0) 7 (58.3) 2 (50.0) NSAIDs, non-steroidal anti-inflammatory drugs.

4 144 F. Sztajnbok et al. TABLE 4. Values of JIA severity measures of the study patients by discordance category No discordance Negative discordance Positive discordance (n ¼ 80) >1cm (n ¼ 34) >3cm (n ¼ 67) >1cm (n ¼ 12) >3cm (n ¼ 4) P-value Physician s global assessment a (cm) Parent s global assessment a (cm) Parent s pain assessment a (cm) C-HAQ score b (units) No. swollen joints No. joints with pain on motion/tenderness No. joints with limited range of motion No. active joints Erythrocyte sedimentation rate c (mm/h) C-reactive protein d (mg/dl) a On a 0 10 cm visual analogue scale (0 ¼ best; 10 ¼ worst); b C-HAQ: Childhood Health Assessment Questionnaire (0 ¼ best, 3 ¼ worst); c normal <15 mm/h; d normal <0.3 mg/dl. TABLE 5. Spearman s correlations between physician s and parent s global assessments and JIA severity measures Physician s global assessment Spearman s r Parent s global assessment Parent s pain assessment C-HAQ score No. swollen joints No. joints with pain on motion/tenderness No. joints with limited range of motion No. active joints Erythrocyte sedimentation rate C-reactive protein The Spearman s correlations between physician s and parent s global assessments and JIA severity measures are presented in Table 5. As expected, the parent s rating was better correlated with the parent-centred measures (parent s rating of the intensity of the child s pain and C-HAQ), whereas the physician s rating was better correlated with the physician-centred measures ( joint counts). Both the physician s and the parent s ratings were poorly correlated with the laboratory indicators of systemic inflammation. The physician parent agreement, measured through the ICC, was in the moderate range for both physicians (0.42 for A.R. and 0.58 for S.V.). Spearman s correlations between physician s ratings and JIA severity measures for physicians A.R. and S.V. were similar for joint counts and laboratory parameters, whereas physician S.V. showed better agreement with parent-centred measures (data not shown). Discussion The physician s and parent s global assessments of disease status are simple, easy to score and reliable clinical measures that enable translation of qualitative clinical impressions into quantitative data [14]. Although these measures are generally regarded as research tools, their regular application in daily practice may offer an important guide to follow the course of the disease over time and to monitor the effectiveness of therapeutic interventions. However, while parents make their proxy reports using this measure, it is the physicians who make judgements upon their reports, and it is the physicians judgements that are used to guide the therapeutic discussion with the parents. It is important, therefore, to ascertain whether parents and clinicians opinion converge or diverge and to identify the factors that may explain the discordance. Substantial disagreement between parents and physicians over disease severity can lead to difficulty in assessing the efficacy of treatments or in evaluating the need for additional interventions. To give an example, it could result in situations in which the physician concludes that a particular treatment is working and the parent says it is not, or vice versa. Furthermore, poor physician parent concordance may lead to parent s dissatisfaction and decreased compliance with therapeutic regimens. In JIA, there is an increasing body of evidence that the parents opinions do not coincide with those of the professionals caring for their children. We previously found that a sizable proportion of parents either under- or overestimate the degree of their children s functional ability, as measured with the C-HAQ, when compared with the objective physician s assessment [15]. Recently, only a moderate agreement between the parents and the physicians in rating of the intensity of children s pain was reported [8]. Our results support the notion that parents and physicians may perceive the disease status of JIA patients differently. Parents appeared to rate, on average, their child s health status as better than the physicians, suggesting that either the parents judged their child as coping well with their disease or that the physicians weighted the objective disease manifestations as more relevant when compared with the parents assessment of child s subjective complaints or poor physical and psychosocial functioning. In keeping with this hypothesis, we found that the physician s rating was better correlated with joint counts, whereas the parent s ratings showed a greater relationship with the parent s assessment of the intensity of the child s pain and the C-HAQ. We found evidence of a clinically relevant discordance (difference in VAS scores >1 cm) in parent s and physician s assessments in over half (60%) of the patients. Discordance occurred much more frequently in the negative direction (the parent scored lower than the physician) than in the positive direction (the parent scored higher than the physician), which is in keeping with the general tendency of the parents to judge the child s health status as better than that perceived by the physicians. This phenomenon was strengthened by the finding that most of the assessments in the negative direction were markedly discordant (difference in VAS scores >3 cm). It is of interest that both concordance and marked positive discordance were significantly associated with lower joint counts. A tendency toward a better agreement between physician s and parent s health assessments at low levels of disease severity was found in previous studies [8, 15]. The relationship between positive discordance and involvement of fewer joints may be explained by the fact that this patient group had, on average, the shortest

5 Discordance in juvenile idiopathic arthritis 145 disease duration. It can be hypothesized that parents are more likely to over-rate the impact of arthritis on their children s health at the time of disease onset, when, although the number of joints involved is frequently low [16], they are emotionally distressed about the occurrence of a new illness. This finding should be regarded with some caution, however, because marked positive discordance was observed in only four patients. We must acknowledge that by describing the parent physician discordance in rating the disease status of children with JIA, we cannot imply that the clinician s assessment is the correct one. As discussed above, physicians and parents were likely to be measuring disease status using different parameters. Furthermore, it is well-known that parents and doctors may have widely different perspectives relating to their beliefs about health and illness, their expectations of medical care, their priorities for treatment and the ways in which they interpret information about their child s disease. We should also recognize that the two physicians revealed a substantial difference in their agreement with parents assessments. This inter-physician discrepancy may have affected the discordance between the parents and doctors ratings. We asked the mothers to rate the health status of their children, but did not obtain information on children s self-reporting. However, using only parent s proxy reports instead of both parent s and patient s selfreports would fail to capture that parents and children may differ in their perception of health [8, 17]. Another potential limitation of our analysis is the lack of inclusion of several potential determinants of discordance, such as the socioeconomic status of the family or the mother s education level. In conclusion, we found that physicians and parents may perceive the health status of children with JIA differently, with parents providing most frequently lower rating. Our results support the inclusion of both physician s and parent s global assessments in the core set of outcome variables for JIA [6] and suggest that both measures should be incorporated in standard paediatric rheumatology care. Rheumatology Key messages We found that parents and physicians often perceive the disease status of patients with juvenile idiopathic arthritis differently, with parents providing, most frequently, lower rating. Our results support the inclusion of both physician s and parent s global assessments of the disease status in standard paediatric rheumatology care. Acknowledgements Dr Sztajnbok is the recipient of an Alpha Scholarship from the European Union (contract no. AML/B7-311/970666/II-0246-FI). The authors have declared no conflicts of interest. References 1. Singh G, Athreya BH, Fries JF, Goldsmith DP. Measurement of health status in children with juvenile rheumatoid arthritis. Arthritis Rheum 1994;37: Ruperto N, Ravelli A, Falcini F et al. Responsiveness of outcome measures in juvenile chronic arthritis. Rheumatology 1999;38: Ruperto N, Ravelli A, Migliavacca D et al. Responsiveness of clinical measures in children with oligoarticular juvenile chronic arthritis. J Rheumatol 1999;6: Moretti C, Viola S, Pistorio A et al. Relative responsiveness of condition specific and generic health status measures in juvenile idiopathic arthritis. Ann Rheum Dis 2005;64: Magni-Manzoni S, Cugno C, Pistorio A et al. Responsiveness of clinical measures to flare of disease activity in juvenile idiopathic arthritis. Clin Exp Rheumatol 2005;23: Giannini EH, Ruperto N, Ravelli A, Lovell DJ, Felson DT, Martini A. Preliminary definition of improvement in juvenile arthritis. Arthritis Rheum 1997;40: Falcone A, Cassone R, Rossi E, Pistorio A, Martini A, Ravelli A. Inter-observer agreement of the physician s global assessment of disease activity in children with juvenile idiopathic arthritis. Clin Exp Rheumatol 2005;23: Garcia-Munitis P, Bandeira M, Pistorio A et al. Level of agreement between children, parents and physicians in rating pain intensity in juvenile idiopathic arthritis. Arthritis Rheum 2006;55: Oliveira S, Ravelli A, Castell E et al. Health related quality of life of patients with juvenile idiopathic arthritis. A multicenter multinational cohort study conducted by the Pediatric Rheumatology International Trials Organization. Arthritis Rheum (in press). 10. Palmisani E, Solari N, Magni-Manzoni S et al. Relationship between juvenile idiopathic arthritis severity measures in early versus late disease. Arthritis Rheum (in press). 11. Ravelli A, Viola S, Ruperto N, Corsi B, Ballardini G, Martini A. Correlation between conventional disease activity measures in juvenile chronic arthritis. Ann Rheum Dis 1997;56: Petty RE, Southwood TR, Baum J et al. Revision of the proposed classification criteria for juvenile idiopathic arthritis: Durban, J Rheumatol 1998;25: Ruperto N, Ravelli A, Pistorio A et al. The Italian version of the Childhood Health Assessment Questionnaire (CHAQ) and the Child Health Questionnaire (CHQ). Clin Exp Rheumatol 2001;19(Suppl. 23):S Pincus T, Sokka T. Complexities in the quantitative assessment of patients with rheumatic diseases in clinical trials and clinical care. Clin Exp Rheumatol 2005;23(5 Suppl 39):S Ravelli A, Viola S, Migliavacca D, Pistorio A, Ruperto N, Martini A. Discordance between proxy-reported and observed assessment of functional ability of children with juvenile idiopathic arthritis. Rheumatology 2001;40: Felici E, Novarini C, Magni-Manzoni S et al. Course of joint disease in patients with antinuclear antibody-positive juvenile idiopathic arthritis. J Rheumatol 2005;32: Brunner HI, Klein-Gitelman MS, Miller MJ et al. Health of children with chronic arthritis: relationship of different measures and the quality of parent proxy reporting. Arthritis Rheum 2004;51: