Pediatric PROMIS Methodology

A total of 88 cognitive interviews were conducted with 77 children and adolescents across two sites on 318 items. From this initial set of items, 25 items were deleted and 35 were revised and underwent a second round of cognitive interviews. A total of 293 items were retained for field testing.

Pediatric Wave 1 Testing

Field testing took place in North Carolina and Texas among 4,129 children ages 8–17. Participants visiting pediatric clinics were recruited for enrollment. We enrolled a subsample of children from schools and after school programs in Chapel Hill, NC. Psychometric analysis of each proposed item bank included confirmatory factor analysis, exploratory factor analysis (as needed), and item response theory analyses including assessment of differential item functioning by age and sex. Detailed results of the analyses are available or under review in peer reviewed journals as of April 2010. At this time, we have item banks and short forms for each of the following domains: Anger, Anxiety, Asthma, Depressive Symptoms, Fatigue, Pain Interference, Peer Relationships, Physical Function (Mobility), and Physical Function (Upper Extremity).

Parent Proxy Testing

Six major research hospitals across the U.S. participated in Parent Proxy testing. Pediatric items were modified for use with parents/guardians and cognitively tested prior to administration. Approximately 1935 parent–child dyads (children ages 5–17) were enrolled in item bank testing. For children ages 5–7, only the parent–proxy forms were collected. (PROMIS pediatric self–report items are only tested for use among children ages 8–17.) Short form versions of the parent proxy–report scales are available at Assessment Center.

Pediatric Chronic Illness Testing

UNC partnered with more than two dozen pediatric hospitals across the U.S. to administer PROMIS pediatric items to children with sickle cell disease, obesity, cancer, rheumatic disease, chronic kidney disease, and rehabilitative needs. Researchers collected responses from more than 1400 children. Data analysis is ongoing with publications in preparation.

Pediatric Wave 2 Testing

UNC and Texas A&M (with Scott and White Clinic) performed a cross–sectional study to evaluate the use of computerized adaptive testing (CAT) with the PROMIS items in children. Approximately 334 children were enrolled from general and subspecialty pediatrics clinics. Assessment of construct validity with the PedsQL Generic module is being performed.

Additionally, in the subgroup of children with asthma, the Pediatric Asthma Quality of Life Questionnaire was assessed for construct validity. Lastly, a subset of children was sampled for test–retest reliability. Data analysis is ongoing with publications in preparation.