Content validity of the PROMIS® pediatric family relationships measure for children with chronic illness

Recommendations for Patient-Reported Outcomes Measurement Information System pediatric measures in youth with chronic pain: a COnsensus-based Standards for the selection of health Measurement INstruments systematic review of measurement properties

ABSTRACT

The Patient-Reported Outcome Measurement Information System (PROMIS) pediatric measures assess physical, emotional, and social health among children and adolescents. However, their measurement properties have not been systematically examined in youth with chronic pain. A systematic review applying the COnsensus based Standards for the selection of health Measurement INstruments (COSMIN) methodology was conducted to evaluate self-reported PROMIS pediatric measures in youth with chronic pain, assessing 8 measurement properties across all versions (item bank, short form, and computer adaptive testing) from 63 studies covering 25 measures. Moderate or high-quality evidence was most available for content validity, structural validity, internal consistency (measurement precision), and construct validity. Four short-form PROMIS pediatric measures-mobility, anxiety, depressive symptoms, and physical stress experiences-achieved recommendation for the use in chronic pain clinical trials; 7 approached recommendation and 14, including the commonly used PROMIS Pediatric Pain Interference Scale, would be recommended with further evidence. Recommendations were also provided for the use of each measure in observational studies. Overall, based on the existing evidence, a total of 11 self-reported PROMIS pediatric short-form measures, including pain intensity, pain behavior, mobility, sleep disturbance, sleep-related impairment, anxiety, depressive symptoms, psychological stress experiences, physical stress experiences, family relationships, and positive effect, are recommended or approaching recommendation for use in youth ages 8 to 19 years with chronic pain. Research is needed to further establish test-retest reliability, measurement errors, cross-cultural validity, and responsiveness. Future work should expand the evaluation of PROMIS pediatric measures in subpopulations of youth with chronic pain, particularly young children and those with neurodevelopmental disabilities.

Influences of health and environmental deprivation on family relationships among children with chronic disease

PURPOSE

Families play a key role in managing chronic illness. Among chronically ill children, we describe the Patient-Reported Outcomes Measurement Information System (PROMIS) Family Relationships measure over time and its associations with sociodemographics, environmental deprivation, and health.

METHODS

Parents of children aged 8-18 years with asthma (n = 171), type 1 diabetes (n = 199), or sickle cell disease (n = 135), recruited in pediatric clinics and emergency departments (ED), completed demographic surveys. Every six months for up to three years, children completed PROMIS Family Relationships, Anxiety, and Depressive Symptoms short forms (T-scores; mean 50, SD = 10), and a 5-level health status item. Linear mixed models were fit to estimate associations.

RESULTS

Older baseline age was associated with weaker family relationships. For example, for each 3-year higher baseline age, relationships were 3 points weaker for males (- 3.0; 95%CI - 5.7 to - .0.2) and females (- 3.1; 95%CI - 6.0 to - 0.3) with asthma recruited in the ED. For each 1-unit higher mean overall health, relationships were 4.6 points (95%CI 3.2-6.1) stronger for children with diabetes and about 2 points stronger for children with asthma (2.3; 0.7-3.9) and sickle cell disease (2.1; 0.3-3.9). Family relationships were 0.3-0.5 points weaker for each 1-unit increment in mean anxiety or depressive symptoms across all three diseases. Relationships were not significantly associated with environmental deprivation and generally stable over time.

CONCLUSIONS

Family relationships were weaker among older children and generally stable over time, yet fluctuated with physical and mental health. Monitoring PROMIS Family Relationships scores may facilitate referrals for chronically ill children who need support.

Reliability and validity of PROMIS® pediatric family relationships short form in children 8-17 years of age with chronic disease

PURPOSE

Families play a key role in managing pediatric chronic illness. The PROMIS® pediatric family relationships measure was developed primarily within the general pediatric population. We evaluated the Family Relationships short form in the context of pediatric chronic diseases.

METHODS

Children aged 8-17 years with asthma (n = 73), type 1 diabetes (n = 122), or sickle cell disease (n = 80) completed the Family Relationships 8a short form and the PROMIS Pediatric Profile-25's six domains representing physical, mental, and social health. Parents (N = 275) of these children completed the parent versions of the same measures. We evaluated reliability of the Family Relationships measure using Cronbach's alpha and IRT-based marginal reliability, and the standard error of measurement (SEM). Convergent/discriminant validity were assessed from correlations between the Family Relationships domain and the PROMIS-25 domains.

RESULTS

SEM increased for scores above the normative mean of 50. Cronbach's alpha and IRT-estimated marginal reliabilities exceeded 0.80 for children and parents across diseases, except in asthma, where marginal reliability was 0.75 for parents. Scores displayed small to large correlations in the expected directions with social and mental health domains. The largest correlations occurred with parents' proxy reports of children's depressive symptoms in sickle cell disease and asthma, r = - 0.60 (95% CI - 0.74, - 0.48) and r = - 0.58 (95% CI - 0.68, - 0.48) respectively.

CONCLUSIONS

The Family Relationships 8-item short form demonstrated adequate reliability and convergent/discriminant validity for use in pediatric chronic conditions, though scores above the mean displayed greater uncertainty. Evidence of the measure's reliability and validity in multiple contexts furthers the case for its use.