These include patient - reported outcome
measures on fatigue (Chalder Fatigue Scale), 10 physical function (SF - 36), 11
mood (Hospital
Anxiety and Depression Scale; HADS), 12 pain (visual analogue pain rating scale), sleepiness (Epworth Sleepiness Scale) 13
and quality of life (EQ - 5D).14 Other services used one or more of the NOD outcome
measures listed above, plus additional outcome
measures including the Work
and Social Adjustment Scale.15
Many of the scales demonstrated weak psychometrics in at least one of the following ways: (a) lack of psychometric data [i.e., reliability
and / or validity; e.g., HFQ, MASC, PBS, Social Adjustment Scale - Self - Report (SAS - SR)
and all perceived self - esteem
and self - concept scales], (b) items that fall on more than one subscale (e.g., CBCL - 1991 version), (c) low alpha coefficients (e.g., below.60) for some subscales, which calls into question the utility of using these subscales in research
and clinical work (e.g., HFQ, MMPI - A, CBCL - 1991 version, BASC, PSPCSAYC), (d) high correlations between subscales (e.g., PANAS - C), (e) lack of clarity regarding clinically - relevant cut - off scores, yielding high false positive
and false negative rates (e.g., CES - D, CDI)
and an inability to distinguish between minor (i.e., subclinical)
and major (i.e., clinical) «cases» of a disorder (e.g., depression; CDI, BDI), (f) lack of correspondence between items
and DSM criteria (e.g., CBCL - 1991 version, CDI, BDI, CES - D, (g) a factor structure that lacks clarity across studies (e.g., PSPCSAYC, CASI; although the factor structure is often difficult to assess in studies of pediatric populations, given the small sample sizes), (h) low inter-rater reliability for interview
and observational methods (e.g., CGAS), (i) low correlations between respondents such as child, parent, teacher [e.g., BASC, PSPCSAYC, CSI, FSSC - R, SCARED, Connors Ratings Scales - Revised (CRS - R)-RSB-, (j) the inclusion of somatic or physical symptom items on mental health subscales (e.g., CBCL), which is a problem when conducting studies of children with pediatric physical conditions because physical symptoms may be a feature of the condition rather than an indicator of a mental health problem, (k) high correlations with
measures of social desirability, which is particularly problematic for the self - related rating scales
and for child - report scales more generally,
and (l) content validity problems (e.g., the RCMAS is a
measure of
anxiety, but contains items that tap
mood, attention, peer interactions,
and impulsivity).