Table 1 presents Means, Standard Deviations and Pearson correlation coefficients among DTS, STAI-S, and MHC-SF. In general, health-care professionals showed high levels of anxiety and traumatic intensity. In fact, 80 health-care professionals (34.48%) reported a DTS score greater than 40 which is the optimal cut-point for accurate classification of those with or without PTSD (efficiency = 0.83) according to Davidson and colleagues (1997). Moreover 125 participants (53.88%) had a score greater than 32, a cut-point score with an efficiency of 0.94 according to McDonald and colleagues (2009) to correctly classify the presence or absence of any anxiety disorder. Although there is not a consensus regarding STAI-S cut-off points, we used the criterion proposed by  that sets a cut-off point of 35 with an efficiency of 0.87. According to this criterion, sixty-four health-care professionals (70.69%) presented a STAI-S score greater than 35, which could indicate the presence of anxiety disorders. Regarding the presence of positive health, only 52 health-care professionals (22,415) showed a flourishing mental health diagnosis based on Keyes  criteria (participants experienced at least 1 of the 3 HWB symptoms and 6 of the 11 PWB/SoWB symptoms ‘every day’ or ‘almost every day’ in the past month).
As expected, the two pathology measures were significantly correlated with each other. Moreover, in line with the Complete State Model of Health, both measures of pathology were also negatively correlated with the indicators of presence of positive health (i.e., MHC-SF). Considering the MHC-SF subscales, social well-being shows the weakest correlations with pathology measures.
To test the two-continua model of positive mental health (i.e., MHC-SF) and mental illness (i.e., STAI-S; DTS), we first conducted a PA. Only the first and second eigenvalues of the real dataset (i.e., 2.67, 1.19) exceeded mean random values (i.e., 1.09, 1.07). Five variables were introduced into the EFA analysis to test factor loadings of all MHC-SF subscales, HWB, PWB, SoWB, STAI-S and DTS. The N:p ratio was 46.4, higher than those generally recommended in the literature to yield factors’ good recovery (e.g., , and communalities were relatively high (all greater than 0.55 except for SoWB that was 0.45), indicating a good factor recovery . All the sub-scales of MHC-SF essentially loaded on the first factor (53.43% of variance explained) and the STAI-S and DTS loaded on the second factor (21.86% of variance explained) (Table 2). These results support the two-continua model of mental health: positive mental health (MHC-SF; factor 1) and mental illness (STAI-S and DTS; factor 2). The correlation between factors was − 0.34, which is a first indicator of the relationship between mental illness and positive mental health.
Regarding the importance of PPE accessibility for professionals’ mental health, seventy health-care professionals (30.2%) indicated that they had access to the PPE, and 162 (69.8%) indicated that they had not. Participants who had access to PPE reported lower levels of state anxiety and traumatic intensity than those who did not have access to the equipment. Concerning the presence of positive mental health, professionals with access to PPE informed of greater well-being compared to those without access (Table 3).
Finally, we expected PPE availability to moderate the relationship between mental illness and positive mental health. Specifically, PPE availability should moderate the relationship between anxiety and well-being. To test our hypothesis, MHC-SF was subjected to a hierarchical regression analysis. We introduced PPE availability and STAI-S (centered score) as predictor variables at the first step and added a computed interaction term at the second step. Gender and Profession were entered as covariables and both were no significant, B = 0.01, t (226) = 0.15, p = 0.88, 95% CI [− 0.11, 0.13], B = 0.09, t (226) = 1.49, p = 0.14, 95% CI [− 0.03, 0.20]. As expected, this analysis revealed that the main effect of PPE availability, B = −0.56, t (226) = −4.34, p < 0 0.01, 95% CI [− 0.82, − 0.31] and the main effect of STAI-S, B = −0.41, t (226) = −6.52, p < 0.01, 95% CI [− 0.54, − 0.29] were significant. Most relevant for the purposes of the present analysis, the data revealed a significant PPE availability × STAI-S interaction, B = −0.36, t (226) = −2.99, p < 0.01, 95% CI [− 0.60, − 0.12]. As depicted in Fig. 1, this interaction revealed that among participants without PPE availability, STAI-S were strongly related with MHC-SF, B = −0.52, t (226) = −6.47, p < 0.01, 95% CI [− 0.68, − 0.36]. This relationship was not significant among participants who reported PPE availability, B = −0.16, t (226) = −1.75, p = 0.08, 95% CI [− 0.34, 0.02]. Also, we expected the relationship between trauma intensity and well-being to be moderated by PPE availability. Similarly, MHC-SF was subjected to a hierarchical regression, with PPE availability and DTS as predictor variables. Again, the covariables Gender and Profession were not significant, B = 0.06, t (226) = 1.00, p = 0.32, 95% CI [− 0.06, 0.17], B = 0.05, t (226) = 0.80, p = 0.42, 95% CI [− 0.07, 0.16]. The main effects of PPE availability, B = −0.55, t (226) = −4.17, p < 0.01, 95% CI [− 0.80, − 0.29] and DTS, B = −0.32, t (226) = −5.44, p < 0.01, 95% CI [− 0.44, − 0.21] were significant. According to our hypothesis, the PPE availability × DTS interaction was also significant, B = −0.29, t (226) = −2.18, p = 0.03, 95% CI [−0.55, − 0.03]. This interaction reveled that among participants without PPE availability, DTS was strongly related with MHC-SF, B = −0.41, t (226) = −5.78, p < 0.01, 95% CI [− 0.55, − 0.27]. However, this relationship was not significant for participants who reported having access to PPE, B = −0.12, t (226) = −1.12, p = 0.27, 95% CI [− 0.34, 0.09] (Fig. 2).