*3.1. Test–Retest*

Although 212 responses were received, only 37 individuals responded to the questionnaire twice as required. After checking the response times, we found four respondents that although responded twice to the questionnaire, didn't wait the necessary minimum two weeks period between the responses. After excluding these other four respondents, calculations were made using 33 participants: 24/33 were women (72%) and their average age was 50.9 (SD: 8.87) years. Among them, 24/33 (72%) were family physicians, 2/33 were nurses (6%), 1/33 was a dermatologist (3%) and 4/33 (12%) had other specialties.

The internal consistency measured with the Cronbach's alpha coefficient showed none of the items significantly altered the consistency of the instrument. The overall alpha coefficient was 0.84 (95%, CI: 0.79–0.84). Table 1 shows the alpha coefficient for each of the eight items in the questionnaire. With respect to temporal stability, the intraclass correlation coefficient value stands at 0.93 (95% CI: 0.852–0.964), thus showing excellent reliability of the test.


**Table 1.** Psychometric validation.

1 IDI: Item Discrimination Index, a Items scored on a 5-point response scale ranging from 1 "very dissatisfied" to 5 "very satisfied"; b Items scored on a 3-point response scale.

#### *3.2. Descriptive Analysis of the Items*

All discriminative rates were above zero and, for the first four questions, they were above one. The highest discrimination item was Item 3, "Clinical quality" (discrimination index = 1.40), and the lowest discrimination item was Item 8, "Future use" (0.51). Table 1 describes the average score for each group and the discrimination index of each of the eight items in the questionnaire.

The lowest ceiling effect score was for Item 3 "Clinical quality" and the highest score was for Item 5, "Health e ffects". Furthermore, two out of five items were above the ceiling-e ffect criterion of 15%.

#### *3.3. Exploratory Factor Analysis*

The Kaiser–Meyer–Olkin test of sampling was adequate (KMO = 0.818) and the Bartlett test of sphericity was significant (Chi-square 424.188; gl = 28; *p* < 0.001), indicating that the items were appropriate for a factor analysis. Two factors emerged with an eigenvalue greater than one (Table 2). The factor with questions about the quality of telemedicine technology (Items 1–5) was named Quality. The other factor contained items relating to technical di fficulties in telemedicine (Items 6–8) and was named Di fficulties. The two constructs together accounted for 61.2% of total variance, all factor loadings being higher than 0.40. Figure 1 shows the scree plot representing the number of dimensions extracted.

**Table 2.** Exploratory factor analysis (EFA): data on commonalities of items, item loadings in Factor 1 and Factor 2.


Extraction method, principal component analysis, rotation method, varimax with Kaiser normalization; rotation converged in three iterations. Factors: a Quality, b Di fficulties.

**Figure 1.** Scree plot.

These results show that the Catalan version of the Health Optimum questionnaire to assess practitioner's perceptions of telemedicine tools is statistically robust.
