Reliability of the Physician Global Assessment Scores for Determination of Disease Activity Status within the Pediatric Rheumatology Care and Outcomes Improvement Network

Brandt Groh¹, Ottar Kristinsson², Lisabeth V. Scalzi³, C. April Bingham⁴, Ronald Laxer⁵, Cagri Yildirim-Toruner⁶, Esi Morgan⁷, Michelle Batthish⁸, Beth Gottlieb⁹, Julia G. Harris¹⁰, Murray Passo¹¹, Michael Shishov¹² and Sheetal S. Vora¹³, ¹Pediatrics, Penn State Milton S. Hershey Medical Center, Hershey, PA, ²Pediatrics, Penn State Children's Hospital, Hershey, PA, ³Department of Rheumatology, Penn State Hershey Children’s Hospital, Hershey, PA, ⁴Penn State Health Children's Hospital, Hershey, PA, ⁵Div of Rheumatology, The Hospital for Sick Children, Toronto, ON, Canada, ⁶Rheumatology, Nationwide Children's Hospital, Columbus, OH, ⁷Pediatric rheumatology, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, ⁸Division of Pediatric Rheumatology, McMaster Children's Hospital, Hamilton, ON, Canada, ⁹Pediatric Rheumatology, Cohen Children's Medical Center of New York, New Hyde Park, NY, ¹⁰Children's Mercy - Kansas City, Kansas City, MO, ¹¹Division of Rheumatology PTD, Medical University of South Carolina, Charleston, SC, ¹²Pediatric Rheumatology, Phoenix Children's Hospital, Phoenix, AZ, ¹³Pediatric Rheumatology, Levine Children's Hospital, Charlotte, NC

Meeting: Pediatric Rheumatology Symposium 2017

Keywords: juvenile idiopathic arthritis (JIA), measure, physician data, polyarthritis and remission

Session Information

Date: Thursday, May 18, 2017

Title: Quality, Health Services and Education Research Poster Breakout I

Session Type: Abstract Submissions

Session Time: 4:45PM-5:15PM

Background/Purpose: Within the Pediatric Rheumatology Care and Outcomes Improvement Network (PR-COIN), the Physician Global Assessment (PhGA) metric is a key determinant of Òclinically inactiveÓ juvenile idiopathic arthritis (JIA)—the target to which all participating centers are encouraged to treat. While reliability has been documented within a single center, it has not been investigated whether or not there may be significant variation in how PhGA is assigned among centers across this network.

This study aims to determine if there is significant PhGA variability among centers across the PR-COIN network.

Methods: 15 PR-COIN centers were included with a total sample size of 5,888 patient entered in the database as of May 1, 2016. Utilizing the most recent PhGA score entered for each patient, we calculated the mean PhGA scores for the following JIA subsets: RF negative poly JIA; RF positive poly JIA; persistent oligo JIA; extended oligo JIA; psoriatic arthritis; enthesitis-related arthritis; and undifferentiated arthritis. To compare PhGA scores among centers, we used either center-specific mean data combining scores for all subtypes or center-specific mean data for a single subtype of interest. One-way ANOVA statistics were calculated to compare means among centers.

Results: Among participating centers, the mean PhGA values for all included ILAR JIA subtypes as a composite were significantly different (p = 0.04). PhGA remained significantly different with a restriction of the analysis to the five centers with total enrollments over 500 patients (n = 3,644) (p < 0.0001) and also with the restriction to a relatively homogeneous subtype, RF negative poly JIA (n = 1,903) (p < 0.0001). Limiting the analysis both to the larger centers and to the RF negative poly JIA subtype (n = 1,086; Table 1) again resulted in a significant difference between PhGA means (p < 0.0001). This analysis indicates significant variability of mean PhGA scores among centers. While it remains possible that mean PhGA differences among centers reflect true differences in patient outcomes, it is also possible that the subjective nature of this metric accounts for this variability.

Conclusion: To continue use of PhGA scores in the determination of disease activity in JIA, efforts should be made to better standardize this metric both within and among participating PR-COIN centers. Alternatively, a better defined composite metric could be substituted for the PhGA, or the determination of disease activity could be limited to joint counts. Further study is needed to determine if mean joint count scores correlate well with mean PhGA scores.

Table 1: Physician Global Assessment scores for RF-negative poly JIA patients at PR-COIN centers with total enrollments > 500 patients

Centers

294

182

154

199

257

Mean

4.95

3.71

3.12

3.58

2.65

3.80

3.13

3.41

3.58

3.31

Range

0 – 10

0 – 7

0 – 8

0 – 10

Disclosure: B. Groh, None; O. Kristinsson, None; L. V. Scalzi, None; C. A. Bingham, None; R. Laxer, 5; C. Yildirim-Toruner, None; E. Morgan, None; M. Batthish, None; B. Gottlieb, 5; J. G. Harris, None; M. Passo, 5; M. Shishov, 5; S. S. Vora, None.

To cite this abstract in AMA style:

Groh B, Kristinsson O, Scalzi LV, Bingham CA, Laxer R, Yildirim-Toruner C, Morgan E, Batthish M, Gottlieb B, Harris JG, Passo M, Shishov M, Vora SS. Reliability of the Physician Global Assessment Scores for Determination of Disease Activity Status within the Pediatric Rheumatology Care and Outcomes Improvement Network [abstract]. Arthritis Rheumatol. 2017; 69 (suppl 4). https://acrabstracts.org/abstract/reliability-of-the-physician-global-assessment-scores-for-determination-of-disease-activity-status-within-the-pediatric-rheumatology-care-and-outcomes-improvement-network/. Accessed .

ACR Meeting Abstracts - https://acrabstracts.org/abstract/reliability-of-the-physician-global-assessment-scores-for-determination-of-disease-activity-status-within-the-pediatric-rheumatology-care-and-outcomes-improvement-network/