Table 4 Synthesized evidence
PROM (refs) | Measurement property | Summarized result | Overall rating* | Quality of evidence§ |
|---|---|---|---|---|
Structural validity | INFIT mean square range 0.66–1.18; OUTFIT mean square range 0.45–1.55 | – | Moderate | |
| Â | Internal consistency | Person separation reliability coefficient 0.94 | ? | Â |
|  | Reliability | ICC range = 0.81–0.91 |  +  | Moderate |
|  | Measurement error | SEM = 1.7; SDD95 = 6.7; SDD95/range = 0.16; SEM = 1.9; SDD90 = 4.8; SDD/range = 0.11; LOA = -2.06–1.40 | ? |  |
|  | Construct validity | 9 out of 20 hypotheses confirmed |  ±  |  |
|  | Responsiveness | RM ANOVA F = 29.89, p < 0.001; Effect size T1vsT2 = 0.916, T2vsT3 = 0.158; Correlation changes measured by PEDI and ABILHAND-Kids Spearman r = 0.430, p = 0.003; Correlation changes measured by AHA and ABILHAND-Kids Pearson r = –0.104, p = 0.493 | ? |  |
ABILHAND-Kids (Ukrainian version) [27] | Structural validity | Standardized residuals range = -2.19–1.58 | – | Moderate |
|  | Internal consistency | Person separation index = 0.95 | ? |  |
|  | Cross-cultural validity | 3 major DIF’s were observed across countries (Ukrainian versus Belgian cohort) | – | Moderate |
ABILHAND-Kids (Danish version) [28] | Structural validity | TLI = 0.98; CFI = 0.98; RMSEA = 0.07; SRMR = 0.07 Fit residuals (z) range = -2.178–2.170 | – | Moderate |
|  | Internal consistency | Cronbach’s alpha = 0.96 | ? |  |
|  | Measurement invariance | 1 non-uniform DIF was observed across age groups | – | Moderate |
|  | Reliability | ICC2.1 = 0.97 (95% CI 0.95–0.98) |  +  | High |
|  | Measurement error | SE = 0.5; LOAs range: –4.8–5.5; SDC = 5.15 points | ? |  |
ABILHAND-Kids (Turkish version) [29] | Structural validity | Residual (z) range = -1.636–1.934 | ? |  |
|  | Internal consistency | Cronbach’s alpha = 0.94 | ? |  |
|  | Measurement invariance | No DIF was observed |  +  | Very low |
|  | Reliability | ICC = 0.98 (95% CI 0.98–1.00) |  +  | Very low |
|  | Construct validity | 2 out of 2 hypotheses confirmed |  +  | High |
ABILHAND-Kids (Arabic version) [30] | Structural validity | Unidimensionality T-Tests (CI): 6.08% significant tests (lower limit of 95% CI = 2.60); Fit residual range = -2.06–2.01 | – | Moderate |
|  | Internal consistency | Person separation index = 0.93 | ? |  |
|  | Measurement invariance | No DIF was observed |  +  | Very low |
|  | Reliability | ICCagreement = 0.98 (95% CI 0.97–0.99) |  +  | Very low |
|  | Measurement error | SEMagreement = 0.24; MDC95 = 0.68 | ? |  |
|  | Construct validity | 6 out of 7 hypotheses confirmed |  +  | Moderate |
ABILHAND-Kids (Persian version) [31] | Structural validity | χ.2 probability = 0.40; PCA on the residuals, first residual factor accounts for 13% of the observed variance; Standardized residuals range = -1.34–1.60 |  +  | Low |
|  | Internal consistency | Cronbach’s alpha = 0.963 |  +  | Moderate |
|  | Cross-cultural validity | 2 major DIF’s were observed across countries | – | Very low |
|  | Measurement invariance | No DIF was observed |  +  | Very low |
|  | Reliability | ICCagreement = 0.7 (CI 95% 0.33–0.85) |  +  | Very low |
|  | Measurement error | SEM for CP measure = 11.21% (1.16 logits, raw score of 2.21); SDC for CP measure = 31.07% (3.21 logits, raw score of 6.13) | ? |  |
|  | Construct validity | 1 out of 1 hypothesis confirmed |  +  | Very low |
ChARM [36] | Structural validity | Unidimensionality T-Tests (CI): 8% significant tests, lower limit of 95% CI = 4.6; Fit residuals range = -1.603–1.484 | – | Moderate |
|  | Internal consistency | Cronbach’s alpha = 0.95 | ? |  |
|  | Construct validity | 1 out of 1 hypothesis confirmed |  +  | Low |
Structural validity | Rasch analyses showed misfits (INFIT mean square > 1.5 and/or Z-standardized values < -2 or > 2) for several items of all three subscales | ? |  | |
|  | Internal consistency | Three CHEQ subscales: Person separation reliability coefficient range = 0.89–0.94 | ? |  |
|  | Reliability | Opening questions: ‘performing the activity independently’ average κ = 0.63, ‘using the affected hand as support or to grasp’ average κ = 0.57; Three CHEQ subscales: average ICC 0.87–0.91 |  +  | Very low |
|  | Construct validity | 2 out of 2 hypotheses confirmed |  +  | Very low |
CHQ [50] | Construct validity | No hypotheses were defined a priori | ? | Â |
CHSQ (Original version) [38] | Structural validity | ‘Leisure and play domain’: INFIT mean square range = 0.8–1.5, INFIT Zstd range = -1.6–2.8; OUTFIT mean square range = 0.7–1.5, OUTFIT Zstd range = -1.7–1.8 ‘School/education domain’: INFIT mean square range = 0.7–1.2, INFIT Zstd range = -2.6–1.1; OUTFIT mean square range = 0.6–1.1, OUTFIT Zstd range = -2.1–0.4 ‘Activities of daily living domain’: INFIT mean square range = 0.7–1.2, INFIT Zstd range = -1.6–1.3; OUTFIT mean square range = 0.5–1.4, OUTFIT Zstd range = -1.4–0.8 | ? |  |
|  | Internal consistency | Three CHSQ domains: Person reliability coefficient range = 0.67–0.75 | ? |  |
|  | Cross-cultural validity | 7 items with DIF by cultural difference (Australian versus Taiwanese cohort) | – | Very low |
|  | Construct validity | 5 out of 7 hypotheses confirmed |  ±  |  |
CHSQ (Turkish version) [32] | Internal consistency | Three CHSQ-TR subscales: Cronbach’s alpha range = 0.83–0.86 | ? |  |
|  | Reliability | Three CHSQ-TR subscales; ICC range = 0.98–0.99 |  +  | Low |
|  | Construct validity | 1 out of 1 hypothesis confirmed |  +  | Moderate |
DHI [43] | Internal consistency | Cronbach’s alpha range = 0.83–0.94 | ? |  |
|  | Reliability | ICC range = 0.84–0.93 |  +  | Very low |
| Â | Construct validity | No hypotheses were defined a priori | ? | Â |
Structural validity | INFIT mean square range = 0.78–1.39; OUTFIT mean square range = 0.71–1.36 | ? |  | |
|  | Internal consistency | Cronbach’s alpha = 0.941 | ? |  |
|  | Reliability | ICC = 0.89 (95% IC 0.81–0.93) |  +  | Very low |
|  | Measurement error | SEM (logits) = 0.599; SDCindividual (logits) = 1.66; SDCgroup (logits) = 0.22 | ? |  |
|  | Construct validity | 7 out of 7 hypotheses confirmed |  +  | High |
IMAL [51] | Internal consistency | Two IMAL subscales: Cronbach’s alpha range = 0.94–0.95 | ? |  |
|  | Reliability | Two IMAL subscales: Spearman’s correlation range = 0.64–0.70 | ? |  |
|  | Measurement error | ‘How Often’ scale: SEM = 0.66 ‘How Well scale: SEM = 0.61 | ? |  |
| Â | Construct validity | No hypotheses were defined a priori | ? | Â |
PEDI self-care domain [52] | Construct validity | 1 out of 2 hypotheses confirmed |  ±  |  |
Construct validity | 11 out of 11 predefined hypotheses confirmed; for several analyses hypotheses could not be defined a priori |  ±  |  | |
|  | Responsiveness | 4 out of 6 hypotheses confirmed |  ±  |  |
PODCI (v2.0; Original version) [25] | Internal consistency | Cronbach’s alpha range = 0.82–0.93 | ? |  |
| Â | Construct validity | No hypotheses were defined a priori | ? | Â |
|  | Responsiveness | Moderate-large SRM (0.38–1.27)/effect size (0.32–1.37) for UE function, mobility, pain/comfort, happiness, global function; SRM 0.12/effect size 0.14 for sports/physical | ? |  |
PODCI (v2.0; Dutch version) [26] | Internal consistency | Cronbach’s alpha range = 0.161–0.928 | ? |  |
|  | Reliability | 4 subscales and total score: ICC = 0.636–0.972 (p < 0.025) ‘Pain and comfort’-subscale: ICC = 0.022 (p = 0.476) | – | Very low |
|  | Construct validity | 2 out of 2 hypotheses confirmed |  +  | Very low |
| Â | Responsiveness | No hypotheses were defined a priori | ? | Â |
PROMIS – Upper Extremity item bank (short form) [56] | Construct validity | 3 out of 3 hypotheses confirmed |  +  | Very low |
PROMIS – Upper Extremity item bank (CAT) [56] | Construct validity | 3 out of 3 hypotheses confirmed |  +  | Very low |
QuickDASH [57] | Internal consistency | Cronbach’s alpha = 0.91 | ? |  |
|  | Construct validity | Results in line with 1 hypothesis |  +  | Low |
Revised PMAL [39] | Structural validity | ‘How Often’ scale: EU associated with the first PCA contrast = 2.6 ‘How Well’ scale: EU associated with the first PCA contrast = 2.5 | ? |  |
|  | Internal consistency | Two rPMAL subscales: Person reliability index range = 0.89–0.90 | ? |  |
|  | Reliability | Two rPMAL subscales: ICC range = 0.93–0.94 |  +  | Very low |
|  | Construct validity | 2 out of 2 hypotheses confirmed |  +  | Very low |