Skip to main content

Advertisement

Table 2 Measurement properties of the short form QTAC-PREM post-acute care measure

From: Short form version of the Quality of Trauma Care Patient-Reported Experience Measure (SF QTAC-PREM)

Subscales Item numbers and descriptionsa Sub-scale factor loadings (n = 117)b Subscale Cronbach’s alpha and corrected item-subscale correlations, (n = 117)b Subscale to global rating item correlationsd, (n = 207)c Reliability coefficient (95% CI), (n = 76)e
Post-discharge information and communication   0.64 0.61g  
1. Received adequate written instructions (2/3) 0.40 0.27 0.25g 0.78 (0.67–0.86)
7. Explained next steps in recovery (9) 0.82 0.58 0.52g 0.71 (0.52–0.91)h
8. Described recovery timeline (10) 0.55 0.34 0.29g 0.55 (0.35–0.76)h
9. Understandable explanations (12) 0.74 0.57 0.43g 0.69 (0.55–0.80)
12. Post-discharge recovery guidance (17) 0.68 0.55 0.79g 0.86 (0.79–0.91)
Stand-alone items
2. Pain management (4)f 0.25g 0.90 (0.77–1.00)h
5. Difficulty scheduling appointments (8a–c)i
10. Family physician informed (16)f 0.40g 0.87 (0.78–0.92)
11. Perceived safety of care (15)j 0.10 0k
13. Global rating item (18) 0.90 (0.83–0.94)
  1. aCorresponding long form item number in (brackets)
  2. bn = 117 complete cases available out of total sample of 207 completed surveys
  3. cThe available sample size varied from n = 118 to n = 207 depending on the number of complete cases for the subscale/individual items
  4. dSpearman’s correlations
  5. eIntraclass correlation coefficient unless otherwise indicated
  6. fItem did not load onto a sub-scale
  7. gSignificant at the p < 0.01 level
  8. hCohen’s Kappa coefficient
  9. iThis item underwent major revision for inclusion on the short form and therefore could not be assessed using data from the long form validation study. However, we recommend it be reported as a stand-alone item as it does not conceptually match the construct of the subscale
  10. jItem excluded from factor analysis due to limited variation among complete cases
  11. kNo reliability estimate because of limited response variance among the sample of test–retest respondents