Table 4.
Content and construct validity for health-related quality of life (HRQL) measures in children
| Measure | Content validity | Construct validity |
|---|---|---|
| Differentiates health & disease | Factor analysis, correlation with instruments, preknown information, clinical symptoms, etc. | |
| CHIP-AE | Healthy ≠ illness | Most items correlate most highly with the sub-domain in which they had originally been placed (factor analysis). Differentiates in predicted direction between four groups of teenagers known to differ in their current health status |
| School population ≠ illness | ||
| CHIP-CE | Healthy ≠ major chronic illness | Basic conceptual framework is supported by factor analysis. Emotional discomfort correlates (r 0.63) with Baltimore How I Feel scale. Self-esteem correlated with CHQ satisfaction (r 0.58). Limitation of activity correlates (r 0.53) with CHQ physical functioning scale |
| CHQ-CF87 | Healthy ≠ illness | Presence of diabetes symptoms and concerns correlated with lower physical and psychosocial functioning of CHQ. In asthma 7 of 9 dimensions of the CHQ correlate with the Child Health Assessment Questionnaire. The CHQ correlated with the KINDL on the domains: physical, emotional, and self esteem |
| Norm ≠ diabetes | ||
| Healthy ≠ chronic health condition | ||
| CHQ-PF50 | Norm ≠ diabetes | CHQ scales loaded highest on their hypothesized vector (factor analysis). Strong correlations between CHQ bodily pain and HUI 2/HUI 3 pain scales (r 0.51–0.60), and between CHQ mental health and HUI 2/HUI 3 emotion (r 0.53–0.64). CHQ is correlated with symptom severity in Juvenile Chronic Arthritis (JCA) and symptom activity in asthma |
| Norm ≠ diabetes | ||
| Healthy ≠ chronic health condition | ||
| CHQ-PF28 | Healthy ≠ HIV | Severity of sickle cell disease correlates with mean physical summary score. All correlation coefficients between CHQ domains and the visual analog scale rating of the child’s health are positive and significant |
| Healthy ≠ chronic health condition | ||
| DISABKIDS | Healthy ≠ severe health condition | Differentiates between levels of severity of asthma and arthritis. Correlations were highest with dimensions of HRQL measures evaluating similar concepts |
| FS II(R) | Healthy ≠ medical problems/complaints | Moderate correlation with clinical rating. Negatively correlated with days hospitalized, days absent from school, and days in bed |
| HUI 2 | Off treatment ≠ on treatment | HUI 2/3 pain correlates with CHQ bodily pain (r 0.51–0.60). HUI 2 mobility correlates with CHQ physical functioning (r 0.45–0.58). HUI 2/3 emotion correlates with CHQ mental health (r 0.53–0.64). Important differences in HRQL scores between patients, parents, and physicians |
| KIDSCR-52 | ? | Strong correlation with KIDSCREEN-27 (r 0.63–0.96). High correlation with KINDL for dimensions assessing similar constructs (r 0.51–0.68) |
| KIDSCR-27 | Healthy ≠ physically ill | Strong correlation with KIDSCREEN-52 (r 0.63–0.96). Moderate to high correlation with other HRQL measures assessing similar constructs (r 0.36–0.63). Correlation with psychosomatic complaints (r 0.52) |
| Healthy ≠ mentally ill | ||
| KINDL | Healthy ≠ DM | KINDL physical correlates with CHQ physical scales (r −.38–0.55). KINDL emotional and self esteem correlates with CHQ mental health and self esteem (r 0.41–0.62). Factor solution in line with the original subscales. Strong correlation with the Short Form-36 for mental health and physical functioning (r 0.53–0.86) |
| Healthy = chronically ill | ||
| PedsQL4.0 | Healthy ≠ chronic health condition (CP, ADHD, headache, asthma, DM, cancer, rheumatic disease) | Most PedsQL items load most highly on their conceptually derived scale (factor analysis). Significant correlations in the expected direction with the PedsMIDAS total score (headache-specific measure of disability). Negatively correlated with the GMFCS score. Discriminates between children with extremity fractures and children with traumatic brain injury. Related to indicaters of morbidity and illness burden. Change over time as a result of clinical intervention |
| On ≠ off treatment | ||
| TACQOL | Healthy ≠ chronic diseases, medical treatment | Almost all items (93%) loaded higher on their own factors than on other factors. Correlation between TACQOL and KINDL (r 0.24–0.60). Mean correlation between contextual similar domains of the EuroQoL and the TACQOL was −0.55) |
| YQOL-R | Healthy ≠ chronic health condition | All scales correlate highly with the scales of the KINDL. Low correlation with two measures assessing different constructs: the Functional Disability Inventory and the Children’s Depression Inventory |