Background/Purpose:  Child/parent
perspectives on their own/their child’s health are highly pertinent to disease
management. PROMIS is a
collection of patient-reported outcome measures that were developed to be non-disease
specific. For each PROMIS domain there are “item banks” (13-29 questions) and
associated “short forms” (8-10 questions) that generate a score representing
the level of that trait in comparison to a sample of US children (inclusive of
children with common chronic diseases). This study examines the applicability
of PROMIS tools to children with juvenile idiopathic arthritis (JIA) and their
parents.

Methods: A convenience sample of
patients with JIA aged 8-17 years (all JIA categories) and parent proxies (5-17
years) were recruited from a single center and completed 1 to 4 PROMIS item
banks. In all, 8 PROMIS items banks were administered: anger, anxiety,
depression, fatigue, pain interference, mobility, upper extremity, and peer
relationships. We used item response theory scoring to generate T-scores for
each item bank and 8-item short form. Responsiveness of the item banks and short forms over
disease activity states defined by the Juvenile Disease Activity Score
(JADAS-3) was evaluated using ANOVA tests for equal variances. Standard error
plots were generated for each full bank and corresponding short form.  The association between child and proxy
dyad responses was assessed using Spearman pairwise correlation coefficients.

Results:  229 patients (67.3% female, median age=13yrs)
and 224 parents (70.5% female, median age=12yrs) completed at least one PROMIS
form, resulting in 71-78 responses per item bank. Fatigue, mobility, and pain
interference full item banks and short forms showed significant responsiveness to
JIA disease activity (Figure; all p<0.05) for both patient and parent proxy
report. Peer relationships, anxiety, depression, and anger had significant
responsiveness (all p<0.05) over the disease activity levels for parent proxy
but not patient report. The upper extremity instrument did not show significant
responsiveness for either patient or parent proxy report. Patient and proxy
PROMIS short forms exhibited comparable standard error across the range of
T-scores to that seen in the full item banks. The pairwise correlations between
patient and proxy dyads for the item bank scores were: anger
(r=0.38), anxiety (r=0.36), depression (r= 0.45), fatigue (r=0.65), pain
interference (r=0.70), mobility (r=0.55), upper extremity (r=0.49), and peer
relationships (r=0.32).

Conclusion:   Our results demonstrate that the
8-item PROMIS pediatric short forms are feasible to complete, responsive to JIA
disease activity, and have comparable error to the full item banks for child
and proxy report. The moderate associations between patient and proxy report
are consistent with prior reports and suggest that administration of both forms
should be considered.