Session Information
Session Type: ACR Poster Session C
Session Time: 9:00AM-11:00AM
Background/Purpose: A brief patient-reported measure of
physical function, the PROMIS PF-10a, is sensitive to changes in disease
activity among patients with rheumatoid arthritis (RA). However, less is known
about which components of composite disease activity measures–swollen and
tender joint counts, patient and provider global assessments, pain, or
inflammatory markers–are most strongly associated with reported changes in
physical function. Better understanding the impact of each of these domains on
physical function is critical to interpreting changes in PF-10a and
understanding its role in the clinical setting. The current study evaluates the
relative contribution of patient and provider assessments, pain, and
inflammatory markers, on self-reported physical function.
Methods: Clinical and demographic data were abstracted from
the electronic health record for all patients seen at a university-based
rheumatology clinic between February 2013 and February 2015 with at least two
ICD-9 codes for RA. All patients had PF-10a and Clinical Disease Activity Index
(CDAI) scores recorded on at least two occasions. Pain was assessed at each
visit using a visual analog scale (0-10cm). The inflammatory marker C-reactive
protein (CRP) was assessed at least every 3 months. Mixed effects linear
regression was used to model the relationship between change in CDAI score over
baseline and change in PF-10a score, and then to model the relationship between
changes in each component of the CDAI (patient global assessment (PGA),
evaluator global assessment (EGA), swollen (SJC) and tender joint count (TJC))
and changes in PF-10a score over time. The model was adjusted for age, gender,
pain, and CRP.
Results: Of 326 patients, 82% were women with a mean (SD) age
of 59 (14). The group was racially/ethnically diverse. About half of patients
had moderate-severe CDAI scores (≥ 10) at
baseline. In an unadjusted model, changes in CDAI and PF-10a were significantly
associated (p<0.001); a 12-point increase in CDAI, the Minimal Clinically
Important Improvement threshold, was associated with an average 3.3 point
decrease in PF-10a, 95% CI (2.7-3.8). In the multivariate model, changes in CDAI
remained significantly associated with changes in PF-10a; changes in pain but
not CRP were associated with changes in PF-10a score (Table). In a model isolating
CDAI components, changes in PGA and pain level were associated with changes in
PF-10a score, while changes in EGA, SJC, TJC, and CRP were not.
Conclusion: Changes in the patient global assessment and in pain
levels were significantly associated with reported changes in physical function
as measured by the PROMIS-PF10a, while changes in the evaluator global
assessment, tender and swollen joint count, and CRP level were not. These
findings suggest that PF-10a measures a distinct construct, self-reported physical
function, that may reflect patients’ global assessment and pain level but is
unrelated to provider assessments.
|
Table. Mixed effects linear regression model coefficients (95% CI) representing the association between average change in physical function score (PF-10a) and change in CDAI components, pain, and CRP for 326 RA clinic patients. |
||
|
|
Adjusted Models* |
|
|
|
Composite CDAI Model |
Component CDAI Model |
|
CDAI score (0-76) |
-0.1 (-0.2 — -0.04) |
|
|
Patient Global Assessment (0-10) |
|
-0.8 (-1.2 — -0.5) |
|
Provider Global Assessment (0-10) |
|
-0.4 (-0.8 — -0.0004) |
|
Tender Joint Count (0-28) |
|
0.05 (-0.1 — 0.2) |
|
Swollen Joint Count (0-28) |
|
-0.05 (-0.3 — 0.2) |
|
Pain score, VAS (0-10) |
-1.2 (-1.5 — -1.0) |
-0.6 (-0.9 — -0.3) |
|
CRP, mg/L |
-0.01 (-0.04 — 0.01) |
-0.01 (-0.04 — .01) |
|
*Models adjusted for age and gender. Results from the unadjusted model were similar. Higher PF-10a scores reflect better function and lower CDAI scores reflect less disease activity. Bolded covariates reached statistical significance, p<0.05 |
||
To cite this abstract in AMA style:
Wahl ER, Gross A, Chernitskiy V, Katz PP, Yazdany J. Changes in a Patient-Reported Measure of Physical Function, PF-10a, Are Most Strongly Associated with Changes in the Patient Global Assessment Portion of a Composite Measure of RA Disease Activity [abstract]. Arthritis Rheumatol. 2015; 67 (suppl 10). https://acrabstracts.org/abstract/changes-in-a-patient-reported-measure-of-physical-function-pf-10a-are-most-strongly-associated-with-changes-in-the-patient-global-assessment-portion-of-a-composite-measure-of-ra-disease-activity/. Accessed .« Back to 2015 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/changes-in-a-patient-reported-measure-of-physical-function-pf-10a-are-most-strongly-associated-with-changes-in-the-patient-global-assessment-portion-of-a-composite-measure-of-ra-disease-activity/