Since the first confirmed case has been announced in Yemen on 10th April 2020, in Ash Shihr, which is a port city in the Hadhramout province southern Yemen, extreme fear and anxiety extended to the other provinces from the possibility of COVID-19 outbreak. The HCPs as a front line, are the most vulnerable to catch COVID-19 than other people, particularly with a critical shortage in PPE due to the current war and civil conflict in the country . At this critical moment, it is crucial to understand the HCPs' preparedness to cope with COVID-19 outbreak. Accordingly, the current study aimed to explore the level of knowledge, attitudes, anxiety, and preventive behaviors among the Yemeni HCPs towards the COVID-19 pandemic.
Based on the findings, although the majority (60.0%) of respondents had never attended training courses of COVID-19, most (69.80%) of them had an adequate level of knowledge about COVID-19. However, these four statements on the importance of wearing masks in the society, having to wear N95 mask only during intubation, suction, bronchoscopy and cardiopulmonary resuscitation, the possibility to treat the disease by usual antiviral drugs and antibiotics as the first-line treatment scored the lowest correct answers (69.9%, 68.8% 28.47% & 27.3%), respectively. This result probably highlights the need to focus more on some parts in future educational courses related to COVID-19. The adequate level of knowledge among the respondents could be attributed to their educational level as most of them (73.0%) had a bachelor degree or higher like Master Degrees. Such an educated professional group could act to obtain knowledge about COVID-19 from different sources of information. In this regard, the result showed that only (20.0%) of HCPs obtained their information about COVID-19 from the official websites of the Ministry of Public Health and Population and the WHO. This indicates that health authorities should pay more attention to encourage HCPs to use official websites as an essential and credible source of information. Likewise, 57.1% of the HCPs appeared to use social media and news media as the main source of their information, which is a considerable concern. This is because using such media can misguide the HCPs by spreading numerous fabricate and unverified information.
It is worth noting that the respondents' level of knowledge was only statistically significantly different according to their age, occupation and educational level. Our results are consistent with the result of a previous study  which reported that level of knowledge toward COVID-19 differs significantly across different age groups, educational levels and levels of profession. The results are also in line with the results of Giao et al.  and Saqlain et al.  in relation to the difference in the level of respondents’ anxiety based on their profession. Concerning the level of respondents’ attitude, it significantly differed based on participants' occupations. This corroborates with a study of Giao et al. , which reported a significant association between respondents’ attitude and their occupations. In contrast, the result seems in disagreement with the results of Saqlain et al. and Rahman and Sathi , who stated that a positive attitude toward COVID-19 did not significantly vary across the different occupations. Likewise, the results revealed that the respondents' level of anxiety was significantly different based on their gender and educational levels. These results support results reported by Al-Hanawi et al.  that respondents' level of worry due to COVID-19 differs significantly across gender and educational level. The result also in line with a previous studies [23, 24] in China, which indicated that females have high levels of anxiety compared to males. Similarly, the respondents' level of self-reported preventive behaviour significantly differed according to their sex, occupation, years of working experience and educational level. These results are in agreement with the results of Rahman and Sathi  on the variation of respondents' preventive behavior according to the different age groups, Al-Hanawi et al.  in regard to the respondents' gender, Saqlain et al.  regarding the respondents' years of working experience and Khasawneh et al.  in relation to the respondents' educational level.
Regarding the respondents’ attitude, the result showed that 85.10% of the respondents had optimistic attitudes towards COVID-19. Unfortunately, the findings showed that 75.1 % of them think that they will not get infected by the virus, and almost 29.4% of them are willing to move to other places inside the homeland to protect himself from the COVID-19 outbreak. This result means that most of the respondents are either confident to protect themself or are not aware that COVID-19 is very contagious. Likewise, one-third of the respondents probably leave their work and go to other places for fear of infection, which resulted in extra healthcare providers’ shortage and made the situation more serious. Accordingly, this information should be corrected as soon as possible via holding training courses about COVID-19 or disseminating such information through official websites. The high optimistic attitude in the current study could be explained by the limited cases reported in Yemen so far and the adequate level of knowledge that they obtained since the outbreak has started until this study was conducted. The result of Roy et al.  indicated that adequate awareness often leads to optimistic attitudes, which could positively affect the preparedness of the HCPs to meet pandemic issues. In addition, our study result showed a positive correlation between the respondents' knowledge and their attitude, which could support this speculation. Our findings are consistent with a study by Giao et al. , who found that healthcare workers had a high level of knowledge and a positive attitude towards the COVID-19 pandemic. These findings are also in line with the result of a cross-sectional study conducted among Saudi health college students . The results revealed that more than half of the students had a positive attitude toward MERS-CoV.
Concerning the respondents' anxiety, the result indicated that nearly half (51.00%) of the respondents had a moderate level of anxiety and 27.70% of them had a high level of anxiety about the COVID-19 outbreak. According to Roy et al. , fear and anxiety among a given population are usually expected because of the intense impact of the pandemic, which could affect the mental well-being and influence their behavior in the community. In this study, only 27.7% of the respondents exhibited a high level of anxiety about COVID-19. The low level of anxiety among the respondents could be attributed to the knowledge level that they have and because they are still in the first stage of the battle against the COVID-19 pandemic. Our result showed lower anxiety reported in studies conducted during the COVID-19 outbreak by Huang and Zhao  on Chinese healthcare workers and Nemati et al.  on Iranian nurses. Their results showed that the anxiety among healthcare workers was higher than that in other people. The high anxiety among the HCPs could be attributed to the uncontrolled nature of the pandemic and worries about being infected, particularly with the shortage of healthcare institutions and PPE.
Concerning the self-reported preventive behaviors, it was found that the majority (87.70%) of the respondents had a high-performance level of preventive behaviors toward COVID-19. Such high-proactive behavior toward COVID-19 could be attributed to the adequate level of knowledge among the respondents towards the COVID-19 outbreak. As shown by a previous study, those who had adequate knowledge exhibited optimistic attitudes and appropriate proactive practice toward COVID-19 . Another study demonstrated that the level of good knowledge in a given population about COVID-19 is significantly reflected in their behavior and attitude . Our finding seems to be relatively lower than a study conducted during COVID-19 by Taghrir et al.  on Iranians' medical students as the researchers found that 94.2% of respondents had high performance in preventive behaviors toward COVID-19. According to the results of this study, it was found that females exhibited a higher-performance-level in preventive behaviors as compared to males. This probably indicates that the women in this study were more compliance in preventive behaviors toward COVID-19. This result is consistent with result of Taghrir et al. , that female demonstrated more precautionary behaviors than males.
Another key result was the positive linear correlation between knowledge-attitude, knowledge-anxiety, knowledge-preventive behaviors, attitude-anxiety, attitude-preventive behaviors and anxiety-preventive behaviors. This result reaffirms the relationship between respondents' level of knowledge and their anxiety, attitude and preventive measures toward COVID-19 pandemic. Such correlation could be explained by the Theory of Reasoned Action , which stated that a person's intention to carry out a specific behavior is determined by their attitude toward this behavior. Our findings are in line with other studies result [20, 30, 32] showing that having a good level of knowledge about COVID-19 correlated with optimistic attitudes and proper practices towards COVID-19. In contrast, our results are in disagreement with the result of Nemati et al.  who found that most Iranian nurses exhibited their anxiety about themselves and their families as a result of COVID-19 though their knowledge that they had about COVID-19 seemed sufficient. Lin et al.  found that the anxiety levels were not influenced by the level of knowledge of COVID-19. However, they found that higher levels of attitudes were highly associated with high levels of anxiety.
In a study conducted in Hong Kong by Leung et al, , the results revealed that the level of anxiety during SARS outbreak was highly associated with behavioural responses such as facemask wearing. Another study by Roy et al.  revealed that people anxiety was correlated with their behavior. The results showed that under the effect of rumors, people modify their behavior to an undesirable one. Reuben et al.  also reported a relationship between respondents’ attitudes and their preventive behaviors.
Regarding the association between the respondents’ attitudes and their preventive behavior, Rubin et al.  conducted study during the swine flu outbreak. They reported a significant association between the respondents’ attitude and their behavior change (e.g. doing one or more avoidance behavior).
The current study had several limitations that should be addressed for future research. First, the data was collected through a web-based survey because it was not possible to conduct a face-to-face survey among the Yemeni healthcare providers during this exceptional period. Therefore, the data obtained might less reliable, since there are less accountability and a lack of a trained interviewer. In addition, the data collection was challenging owing to limited respondent availability and cooperation. Another limitation of this study was its exclusiveness to healthcare providers. Therefore, further research should involve different community-population, and when possible used a community-based studies design are recommended.