Table 1.

Measurement properties of type 2 diabetes-specific (subscales of) PROMs with sufficient content validity

Name of PROM	Number of items	Content validity [25]1	Structural validity	Internal consistency (Cronbach’s alpha)	Construct validity2	Test–retest reliability and measurement error	Cross-cultural validity	Responsiveness
Description		Relevance, comprehensiveness and comprehensibility of the items in a PROM	Extent to which the items in a (sub)scale measure only one construct	Interrelatedness among the items of a (sub)scale	Correlations to similar PROMs as expected	Consistency of results when the same PROM is completed at a different point in time	Same construct is being measured across countries	Ability to measure change over time
Diabetes-specific symptoms
DSSCI [78]	38	⊕⊕	?	0.92	R=0.57 with Illness Perception Questionnaire Identity subscale	?	?	?
DSC-R Hypoglycemic symptoms [79]	3	⊕	8-factor model with CFI 0.90 RMSEA 0.05	0.77	HbA1c levels	?	?	Global rating of change
DSC-R Hyperglycemic symptoms [79]	4	⊕	8-factor model with CFI 0.90 RMSEA 0.05	0.79	HbA1c levels	?	?	Global rating of change
DSC-R Cardiovascular symptoms [79]	3	⊕	8-factor model with CFI 0.90 RMSEA 0.05	0.69	BMI	?	?	Global rating of change
DSC-R Neuropathic pain [79]	4	⊕	8-factor model with CFI 0.90 RMSEA 0.05	0.76	R=0.38 SF-36 Bodily Pain	?	?	Global rating of change
DSC-R Neuropathic sensoric [79]	6	⊕	8-factor model with CFI 0.90 RMSEA 0.05	0.84	R=0.37 SF-36 Bodily Pain	?	?	Global rating of change
DSC-R Ophthalmologic symptoms [79]	5	⊕	8-factor model with CFI 0.90 RMSEA 0.05	0.85	?	?	?	Global rating of change
DQLCTQ/ DQLCTQ-R Frequency of symptoms [80]	7	⊕	?	0.77	HbA1c levels and type of diabetes (T1D vs T2D)	?	?	Metabolic control
DQLCTQ/ DQLCTQ-R Bothersomeness of symptoms [80]	7	⊕	?	0.80	HbA1c levels and type of diabetes (T1D vs T2D)	?	?	Metabolic control
Fatigue
DQLCTQ/DQLCTQ-R Energy/Fatigue [18, 80]	5	⊕	?	0.85	HbA1c levels and type of diabetes (T1D vs T2D)	0.85	?	Metabolic control
W-BQ12 Energy [81, 82]	4	⊕	Unidimensional model with CFI 0.93	0.77−0.87	R=0.85 with SF-36 Vitality	0.80	?	?
Diabetes distress
DDS Emotional burden [15, 83–86]	5	⊕	?	>0.80	R=0.51 with SF-36 Mental Health, R=0.48 with SF-36 MCS, R=0.52 & 0.44 with HADS Anxiety & Depression, R=0.55 with CESD	0.76−0.78 (whole DDS)	?	?
DDS Physician-related distress [15, 83–86]	4	⊕	?	>0.80	?	0.76−0.78 (whole DDS)	?	?
DDS Interpersonal distress [15, 83–86]	3	⊕	?	>0.80	R=0.33 with SF-36 Social Functioning, R=0.48 with CESD	0.76−0.78 (whole DDS)	?	?
PAID-20 Diabetes-related emotional problems [18, 87–91]	12	⊕	Inconsistent results (different models proposed)	0.93−0.95	R=0.53 with HFS, R=0.60 with STAI	0.80−0.83	?	Change after interventions
Anxiety
Diabetes Questionnaire Worries [92, 93]	3	⊕⊕⊕	?	?	R=0.41 with SF-36 Mental Health	weighted κ 0.49−0.60 T1D, 0.46−0.52 T2D	?	?
DQLCTQ Worry (HFS) [18, 80]	17	⊕	?	0.94	Self-perceived control	0.73	?	Metabolic control
DQOL Diabetes-related worry [51, 94]	3−7	⊕	?	0.67	R=0.46 with PAIS psychological distress, R=0.46 with Diabetes-39 Anxiety and Worry scale	0.80	?	?
Physical function
IWADL [24, 95, 96]	7	⊕⊕	?	0.94	± High	0.91 Measurement error: low	?	Weight loss
Sexual function
Diabetes-39 Sexual Functioning [97, 98]	3	⊕	⊕	⊕ >0.80, 0.92	?	?	?	?
Emotional function
DQLCTQ/DQLCT-R Mental health [18, 80]	5	⊕	?	0.82	Self-perceived control	0.83	?	Metabolic control
Social function
Diabetes Questionnaire Barriers [92, 93]	5	⊕⊕⊕	?	?	R=0.42 with SF-36 Social Functioning	Weighted κ 0.41−0.67 T1D; 0.32−0.64 T2D	?	?
Overall self-rated health
Diabetes Questionnaire How you feel [92, 93]	5	⊕⊕⊕	?	?	R=0.56 with SF-36 General Health Perceptions	Weighted κ 0.50−0.65 T1D; 0.40−0.63 T2D	?	?
DQOL Impact [51, 94]	18−27	⊕	?	0.67	R=0.50 with Symptom Checklist-90-R	0.89	?	?
SPH Feel Healthy [99]	5	⊕	?	?	?	?	?	?

Data was extracted from 16 systematic reviews [11–26] and some additional validation studies. This is not a comprehensive systematic review, but provides the most relevant evidence on the measurement properties of these PROM scales. No information on criterion validity was found

COSMIN criteria for sufficient measurement properties were used [44]: structural validity consistent factor model with CFI >0.95, RMSEA <0.08; internal consistency (Cronbach’s alpha) >0.70; test–retest reliability (ICC) >0.70; construct validity/responsiveness (correlations with PROMs measuring similar constructs) ≥0.50

¹ ⊕ Very low, ⊕⊕ low and ⊕⊕⊕ moderate refer to the quality of the evidence (higher quality studies or more available studies lead to higher quality evidence)[46]

² Construct validity: only correlations with PROMs measuring similar constructs are presented

?, no information available; CESD, Center for Epidemiologic Studies Depression scale; CFI, comparative fit index; DDS, Diabetes Distress Screening; DSC-R, Diabetes Symptom Checklist-Revised; DSSCI, Diabetes Symptom Self-Care Inventory; DQLCTQ, Diabetes Quality of Life Clinical Trial Questionnaire; DCLCTQ-R, Diabetes Quality of Life Clinical Trial Questionnaire -Revised; DQOL, Diabetes Quality of Life; HADS, Hospital Anxiety and Depression Scale; HFS, Hypoglycemia Fear Survey; ICC, intraclass correlation coeffient; MCS, Mental Component Summary; PAID, Problem Areas in Diabetes; PAIS, Psychosocial Adjustment of Illness Scale; RMSEA, root mean square error of approximation; SPH, self-perception of health; STAI, State-Trait Anxiety Index; T1D, type 1 diabetes; T2D, type 2 diabetes; W-BQ12, Well-being and Treatment Satisfaction scales