**4. Discussion**

In France, nurses work in hospitals or care institutions or have private practices [31]. Both groups of practitioners experienced many changes during the COVID-19 pandemic [4]. However, to our knowledge, no study has investigated which protective factors were mobilized by nurses in these different contexts to preserve their QoL during the pandemic. Thus, the goals of the present study were to identify protective factors used by nurses and assess if the protective factors used differed as a function of their type of practice (NHCI or NPP). The results sugges<sup>t</sup> that, for both groups of nurses, high levels of perceived stress and problem-avoidant coping strategies were associated with poor QoL, whereas high social support and solution-oriented coping strategies were associated with good QoL. Surprisingly, resilience was not significantly associated with QoL in either practice.

Nurses' stress has been investigated in many studies [45]. Researchers have shown that even in ordinary times (i.e., without a pandemic) nurses face many sources of stress

resulting from their workload or patient-related issues, which affects their QoL [11,46,47]. The COVID-19 pandemic has exacerbated occupational stress because of the pressure put on care services and the central role nurses play in these services [48]. Prior research has shown that social support played an important role in protecting nurses from negative outcomes (e.g., burnout) caused by their exposure to multiple stress sources and helped them to preserve their QoL during the pandemic [49,50]. Our results confirmed these findings. The stress caused by high levels of uncertainty during the current crises may have increased the need for social support [51]. Undoubtedly, social support (providing assistance and information as needed) allows nurses to cope with both feelings of uncertainty and perceived stress [52,53]. The provision of social support at work can be problematic for NPPs because they cannot receive direct support from their institutions or colleagues. Their professional interactions are restricted to patients, their families, and the patients' physician [31], thus limiting the possibility of sharing experiences and asking for advice. On the contrary, NHCIs, who are included in teams, have more possibilities to receive support in the regular course of their work compared to NPPs. Our results support this assumption since NHCIs reported higher scores on social support than NPPs, though the effect size was very small. Nonetheless, the protective factors associated with QoL differed very little between the two types of nurses despite drastic differences in the nature of their practice. To our knowledge, no study has specifically compared NHCIs and NPPs on determinants of QoL. It is thus difficult to explain that, despite the difference in perception of social support, perception of QoL did not differ between nurses working in these different practice contexts. Nurses may have adapted differently in each work setting to the pandemic by relying on coping strategies appropriate to their own setting to maintain their QoL.

Our analyses confirmed the importance of coping strategies to preserve QoL, which is consistent with findings from previous research [29]. However, our study showed that NHCIs more frequently used coping strategies involving social relationships, such as seeking emotional and instrumental support or venting. NPPs used more individualistic coping strategies, such as active coping, planning, positive reframing, and acceptance. This indicates that both groups of nurses used some strategies to protect their QoL, but these strategies differed according to the professional context. This is what was proposed by Moos in his conceptual framework linking context and coping [54]. To him, cognitive appraisal of the stressful situation and coping responses are interdependent with, among others, an individual's environmental system (i.e., supra-personal and social climate factors, such as pressures arising from professional activities that threaten health). Although developed to understand coping patterns in everyday life, this observation could be valid in professional settings where individuals are exposed to numerous stress sources. Moreover, it is noteworthy that the two coping strategies that were the most strongly associated with a satisfying QoL were positive reframing and acceptance. These strategies focus on finding the positive aspects of stressful situations while still acknowledging the existence of negative events [55,56]. Focusing on positive affect generates positive emotions that, according to Fredrickson [57], lead to enhancing one's social, intellectual, and physical resources, all of which are durable and can be mobilized in other stressful situations. Hence, developing and encouraging the use of these strategies could help increase nurses' QoL. Additionally, negative coping strategies such as denial, blaming self, substance use, and behavioral disengagement were associated with low QoL. The use of these strategies reduces negative affect generated by stressful situations without addressing them [58]. Thus, the continued use of negative coping strategies is likely to lead to an accumulation of problems that might eventually seem insurmountable, which in turn, can elicit more avoidant behaviors, creating a vicious circle [14,59]. Our finding that both NHCIs and NPPs seldom used these strategies indicates that most of them adapted well to their contexts by using healthy coping strategies.

Surprisingly, resilience was not significantly associated with QoL, a finding contrary to our expectations based on previous research [14,23,26]. It is possible that the inclusion of the coping strategies in the analyses masked the link between resilience and QoL since

resilience and coping tend to be closely associated [60,61]. The correlations we observed between resilience and some coping strategies, such as active coping, planning, positive reframing, acceptance, and humor, support this hypothesis. To assess this, a mediation analysis should be conducted. However, the present design using the Brief-COPE inventory, which assesses 14 coping strategies, does not lend itself well to such analysis. Another explanation could be that resilience has an indirect effect on QoL. Indeed, the relationship between resilience and QoL is not precisely defined [62], and QoL is a broad concept that includes both physical and psychological aspects [63]. Thus, resilience might not affect QoL when considered as a composite concept, but it might moderate the effect that crises may have on some of the facets of QoL.

QoL can be improved by working on strengthening social support and promoting problem-oriented coping strategies while discouraging problem-avoidant behaviors. Given the similarity of the associations between these variables and QoL for NHCIs and NPPs, such strategies could benefit both groups. Our findings support the development of primary prevention interventions aimed at all nurses. They also advocate applying salutogenic models in the healthcare system, such as Neuman's Systems Model, which considers protective factors and depicts their role as buffers against the stressors that individuals encounter [64,65]. Lastly, our results support the use of managemen<sup>t</sup> policies and practices that foster social support to encourage and help nurses face the sources of stress they encounter rather than avoid them.

The main limitation of the present study was its cross-sectional nature, which did not allow for causal explanations. Moreover, some of the differences and regression coefficients we reported were of small magnitude, although statistically significant. Furthermore, NPPs were older on average and more experienced than NHCIs because, to work in private practice, nurses must have worked for at least two years at a hospital or care institution. Additionally, NPPs do not exist in all countries or might take different forms depending on local culture and legislation. Comparisons between the present study results and those found in other countries with different healthcare systems and organizations thus might be limited. As this study used self-reported measures and volunteer samples, our findings do not generalize to reflect the characteristics of non-respondents. Finally, the present study only investigated individual protective factors, but organizational and environmental factors might also play an important role in nurses' QoL [66,67]. It should not be inferred from the results of our study that nurses are solely responsible for protecting themselves from professional stress sources via the use of protective factors and coping strategies.
